Download Download attachment

Developing Inclusive and Sustainable Economic and Financial Systems
Volume 3
Islamic banking and finance – Essays on corporate
finance, efficiency and product development
Editorial Board
Dr. Hatem A. El-Karanshawy
Dr. Azmi Omar
Dr. Tariqullah Khan
Dr. Salman Syed Ali
Dr. Hylmun Izhar
Wijdan Tariq
Karim Ginena
Bahnaz Al Quradaghi
ISBN: 978-9927-118-23-4
Cover design: Natacha Fares
Copyright © 2015 The Authors
CONTENTS
Forewordv
Acknowledgmentsvii
Prefaceix
Introductionxi
PART 1: ISLAMIC CAPITAL MARKETS AND CORPORATE FINANCE
Chapter 1
Chapter 2
Chapter 3
Chapter 4
Chapter 5
Valuation of Islamic debt instruments, the Sukuk: Lessons
for market development
Mohamed Ariff and Meysam Safari
1
The impact of Islamic debt on company value
Fitriya Fauzi, Stuart Locke, Abdul Basyith and
Muhammad Idris
19
Islamic financing and bank characteristics in a dual banking
system: Evidence from Malaysia
Muhamed Zulkhibri
37
Leverage risk, financial crisis, and stock returns:
A comparison among Islamic, conventional, and
socially responsible stocks
Vaishnavi Bhatt and Jahangir Sultan
47
Is Shariah-compliant investment universally sustainable?
A comparative study
Mehdi Sadeghi
81
PART 2: I SLAMIC BANKING – EFFICIENCY, PROFITABILITY &
INTERNATIONAL PERSPECTIVES
Chapter 6
Chapter 7
Chapter 8
The nexus between economic freedom and Islamic bank
performance: Empirical evidence from the MENA banking
sectors
Fadzlan Sufian, Muhamed Zulkhibri and Abdul Majid
93
Efficiency of performance of banks in the Gulf region before,
during and after crises (financial and political)
Abdel Latef Anouze
111
The relationship between Islamic bank efficiency
and stock market performance: Evidence from GCC
countries
Samir Srairi, Imen Kouki and Nizar Harrathi
125
iii
Chapter 9
Conventional banks versus Islamic banks: What makes the
difference?
Huseyin Aytug and Huseyin Ozturk
137
PART 3: PRODUCT DEVELOPMENT IN ISLAMIC FINANCE
Chapter 10
Chapter 11
Chapter 12
Chapter 13
Estimating expected returns on Mudaraba time deposits
of Islamic banks
Zeynep Topaloglu Calkan
151
Self-adjusting profit sharing ratios for Musharakah financing
Volker Nienhaus
161
Indexing government debt to GDP: A risk sharing mechanism
for government financing in Muslim countries
Syed Aun R. Rizvi and Shaista Arshad
173
Concept and mathematics of Islamic valuation and financial
engineering
Nadi Serhan Aydin, FRM and Martin Rainer
181
Foreword
Hatem A. El-Karanshawy
Founding Dean, Qatar Faculty of Islamic Studies, Hamad bin Khalifa University, Qatar Foundation, Doha
The International Conference on Islamic Economics and
Finance (ICIEF) is the leading academic conference in the
discipline organized by the International Association for
Islamic Economics (IAIE) in collaboration with other key
stakeholders, including the Islamic Research and Training
Institute, Islamic Development Bank. It is the pioneering
international conference on Islamic economics organized
first in Makkah Al Mukaramah, Kingdom of Saudi Arabia,
in 1976 under the auspices of King Abdulaziz University
and has since been held in numerous locations around the
world. The conference as such has contributed immensely
to the promotion of Islamic economics and finance.
Since 2011, the Qatar Faculty of Islamic Studies (QFIS), of
Hamad bin Khalifa University, Qatar Foundation, has also
become a key partner in organizing the conference.
The global economy continues to face the perennial
problems of poverty, persistent youth unemployment,
excessive inequalities of income and wealth, high levels of
inflation, large macroeconomic and budgetary imbalances,
exorbitant debt-servicing burdens, inadequate and aging
public utilities and infrastructure, skyrocketing energy
prices, and growing food insecurity. The reoccurring
regional and global financial crises further intensify
and magnify these problems, particularly for the
underprivileged segments of the world population. As a
result, many countries are at the risk of failing to achieve by
2015 the Millennium Development Goals (MDGs) set by the
United Nations. Hence the achievement of an inclusive and
sustainable economic and financial system has remained
highly illusive.
The ICIEF presents an excellent opportunity for those
interested in Islamic economics and finance to present
their research and contribute to the development of an
inclusive and sustainable global economic and financial
system. It is through such a setting that thoughts can
be debated with the objective of advancing knowledge
creation, facilitating policymaking and promoting genuine
innovation for the industry and the markets. Disseminating
research presented at ICIEF to the greatest number of
researchers interested in the topic is important. It not only
advances the discourse, but also grants those who did not
have the privilege of attending the conference to partake in
the discussion.
To this end, this series of five volumes (two in Arabic to
follow) presents the proceedings of 8th and 9th conferences,
which were held in Doha and Istanbul respectively in 2011
and 2013. Each volume focuses on a particular sub-theme
within the broader theme of Developing Inclusive and
Sustainable Economic and Financial Systems.
The volumes are as follows:
Volume 1: Access to Finance – Essays on Zakah, Awqaf and
Microfinance
Volume 2: Islamic Economics and Social Justice – Essays
on Theory and Policy
Volume 3: Islamic Banking and Finance – Essays on
Corporate Finance, Efficiency, and Product
Development
Volume 4: Ethics, Governance, and Regulation in Islamic
Finance
Volume 5: Financial Stability and Risk Management in
Islamic Financial Institutions
We hope that this academic endeavor in partnership with
the Bloomsbury Qatar Foundation Publishing will benefit
the Islamic economics and finance community and policy
makers and that it will promote further academic study of
the discipline.
Cite this chapter as: El-Karanshawy H A (2015). Foreword. In H A El-Karanshawy et al. (Eds.), Islamic banking and finance –
Essays on corporate finance, efficiency and product development. Doha, Qatar: Bloomsbury Qatar Foundation
Developing Inclusive and Sustainable Economic and Financial Systems
Acknowledgements
Tariqullah Khan
President, International Association for Islamic Economics
At the International Association for Islamic Economics
(IAIE), we are grateful to acknowledge the unprecedented
success of the 8th and 9th International Conferences on
Islamic Economics and Finance, which were respectively
organized in the Qatar National Convention Centre, Doha,
December 19–21, 2011, and in the WoW Convention Centre
Istanbul, September 9–10, 2013. We greatly appreciate
the financial, academic and logistic support provided by
the Qatar Faculty of Islamic Studies, Hamad bin Khalifa
University at Qatar Foundation; Islamic Research and
Training Institute at the Islamic Development Bank; and
the Statistical, Economic and Social Research and Training
Centre for Islamic Countries.
We offer our sincere thanks to the sponsors of the 8th
International Conference on Islamic Economics and
Finance in Doha. Without their partnership and generous
contributions, the conference would not have been
possible. In addition to the Qatar Foundation and the
Islamic Development Bank, other sponsors included: Qatar
Central Bank (QCB), Qatar Financial Centre Authority
(QFCA), Qatar National Research Fund (QNRF), Qatar
National Bank, Qatar Islamic Bank, Qatar International
Islamic Bank, Masraf Al Rayan, and Qatar Airways.
We owe our deepest gratitude to the highly-esteemed panel
of reviewers who volunteered to dedicate their time and
energy in reviewing all the thousands of abstracts and papers
that were submitted to the conferences. The reviewers
of the English papers and abstracts included: Abdallah
Zouache, Abdel Latef Anouze, Abdelaziz Chazi, Abdul Azim
Islahi, Abdullah Turkistani, Abdulrahim AlSaati, Ahmet
Tabako lu, Anowar Zahid, Asad Zaman, Asyraf Dusuki,
Ercument Aksak, Evren Tok, Habib Ahmed, Hafas Furqani,
Hafsa Orhan Astrom, Haider Ala Hamoudi, Hossein Askari,
Humayon Dar, Ibrahim Warde, Iraj Toutounchian, Jahangir
Sultan, John Presley, Kabir Hassan, Karim Ginena, Kazem
Yavari, Kenan Bagci, Mabid Al-Jarhi, Maliah Sulaiman,
Marwan Izzeldin, Masooda Bano, Masudul Alam
Choudhury, Mehdi Sadeghi, Mehmet Asutay, Moazzam
Farooq, Mohamad Akram Laldin, Mohamad Aslam Haneef,
Mohamed Ariff Syed Mohamed, Mohammed Benbouziane,
Mohammed El-Komi, Monzer Kahf, Muhammad Syukri
Salleh, Murat Çizakça, Nabil Dabour, Nafis Alam, Nasim
Shirazi, Nazim Zaman, Necdet Sensoy, Nejatullah Siddiqi,
Rifki Ismal, Rodney Wilson, Ruhaya Atan, Sabur Mollah,
Salman Syed Ali, Savas Alpay, Sayyid Tahir, Serap Oguz
Gonulal, Shamim Siddiqui, Shinsuke Nagaoka, Simon
Archer, Tariqullah Khan, Toseef Azid, Turan Erol, Usamah
Ahmed Uthman, Volker Nienhaus, Wafica Ghoul, Wijdan
Tariq, Zamir Iqbal, Zarinah Hamid, Zeynep Topaloglu
Calkan, Zubair Hasan, and Zulkifli Hasan. The reviewers
of the Arabic papers and abstracts included Abdelrahman
Elzahi, Abdulazeem Abozaid, Abdullah Turkistani,
Abdulrahim Alsaati, Ahmed Belouafi, Ali Al-Quradaghi,
Aly Khorshid, Anas Zarqa, Bahnaz Al-Quradaghi, Layachi
Feddad, Mabid Al-Jarhi, Mohammed El-Gamal, Nabil
Dabour, Ridha Saadallah, Sami Al-Suwailem, Seif El-Din
Taj El-Din, Shehab Marzban and Usamah A. Uthman.
The primary objective of the conferences is to further
the frontiers of knowledge in the area of Islamic economics
and finance. Without the hard work and creativity of the
researchers who shared their work with us, the pool of
knowledge generated in the form of the conference papers
and presentations would not have been possible. We thank
all the authors who submitted their abstracts and papers to
the two conferences.
The IAIE has always endeavored to publish most of the
significant research papers contributed to its conferences.
Currently the selected papers of the 8th and 9th conference
are being published in five volumes under the common
theme of Developing Inclusive and Sustainable Economic
and Financial Systems. On behalf of the Editorial Board we
acknowledge that the partnership with the Bloomsbury
Qatar Foundation Publishing in this regard will be
highly beneficial in disseminating research output and in
promoting the academic cause.
Cite this chapter as: Khan T (2015). Acknowledgements. In H A El-Karanshawy et al. (Eds.), Islamic banking and
finance – Essays on corporate finance, efficiency and product development. Doha, Qatar: Bloomsbury Qatar Foundation
Developing Inclusive and Sustainable Economic and Financial Systems
Preface
Dr. Moh’d Azmi Omar
Director General, Islamic Research & Training Institute (IRTI)
Research in the area of Islamic banking and finance
has mushroomed since the new millennium. As the Islamic
banking industry continues to grow at impressive rates,
researchers have pondered on fundamental questions
relating to whether Islamic banks are unique from an
economic standpoint. Islamic banks have been compared
to their conventional counterparts on various measures
such as efficiency and profitability. Capital market
products such as Islamic mutual funds have also been the
subject of recent research. In order to achieve favorable
outcomes for the industry, it is important to empirically
explore and determine the economic characteristics of
Islamic banking and finance. This may inform how we
may go about determining the usefulness of a company to
issue Islamic debt, for example, and the impact that such
an issuance would have on the stock performance of the
company. In most parts of the world, Islamic banks operate
in a dual banking system together with conventional
banks. Researchers may be curious to explore whether
operating in a dual banking system affects the type of
Islamic banks that exist in the economy and the kind
of products that they offer. In light of the recent LIBOR
scandals, debates surrounding benchmarking of Islamic
products will recapture the attention of researchers; hence,
product development in Islamic banking will also remain
an important area. With the increasing availability of
data, empirical research in Islamic banking is expected to
continue to inform industry stakeholders. Understanding
the Islamic banking phenomenon from an economic
perspective with the support of empirical evidence may
help to align the practices of Islamic banks to its originally
intended objectives of offering a just and equitable form of
financing based on sound ethical principles.
This volume titled Islamic Banking and Finance – Essays
on Corporate Finance, Efficiency, and Product Development
aims to bring together papers on these themes. The volume
consists of selected papers from the 8th International
Conference on Islamic Economics and Finance held in
Doha during 19–21 December 2011 and from the 9th
International Conference on Islamic Economics and Finance
held in Istanbul during 9–11 September 2013. The papers
are presented here in their original form as presented at
the conferences, with changes limited to copyediting and
correcting typographical errors. The conferences were
organized by the Center for Islamic Economics and Finance,
Qatar Faculty of Islamic Studies (QFIS), Hamad bin Khalifa
University; Islamic Research and Training Institute (IRTI),
Islamic Development Bank (IDB); International Association
for Islamic Economics and Finance (IAIE); and Statistical,
Economic, and Social Research and Training Centre for
Islamic Countries (SESRIC).
We have selected the papers in this volume to reflect the
diversity of research in the field of Islamic banking and
finance. We hope that the papers that have been selected
here will help to inform and motivate researchers to
conduct rigorous empirical research to explore questions
that remained to be answered.
Cite this chapter as: Omar M A (2015). Preface. In H A El-Karanshawy et al. (Eds.), Islamic banking and finance – Essays
on corporate finance, efficiency and product development. Doha, Qatar: Bloomsbury Qatar Foundation
Developing Inclusive and Sustainable Economic and Financial Systems
Islamic banking and finance – Essays on corporate
finance, efficiency and product development:
An introduction to the issues and papers
Wijdan Tariq
Senior Researcher, Center for Islamic Economics and Finance, Qatar Faculty of Islamic Studies, Hamad bin Khalifa University
Academic interest in Islamic banking and finance has
grown over the years. The appeal of Islamic banking to
the Muslim world is decades old. However, curiosity from
the wider academic community has increased particularly
in the aftermath of the global financial crisis and ongoing
Eurozone crises. In recent years, prominent academics
from conventional economics have either written papers
on Islamic finance or commented in conferences on the
viability of the Islamic finance proposition (e.g., Beck,
et al., 2013; Abedifar, et al., 2013; Zaheer, et al., 2013;
Ongena and Şendeniz-Yüncü, 2011; Rogoff, 2011a; Rogoff,
2011b; Roubini, 2013, Baele, et al., 2014) This is a welcome
development and may help add more rigor and diversity to
the Islamic finance discourse.
Due to the increasing availability of data, empirical work
in Islamic finance is on the rise. The selection of papers in
this volume reflects this empirical trend. The majority of
the existing literature attempts to answer questions of a
comparative nature. Islamic banks are often compared to
conventional banks on various measures. Which banks are
more stable, more profitable, and more efficient? A more
fundamental question that some academics have asked is:
Are Islamic banks unique to begin with?
Efficiency and profitability comparisons are perhaps
the most common type of study in the field.1 Generally
speaking, the more robust studies seem to show that there
were few differences between Islamic and conventional
banks before the crisis. During the crisis, however, evidence
seems to suggest that Islamic banks were less profitable.
The size of the bank is an important control variable in such
comparative studies. Studies have shown that small Islamic
banks appear to perform better than small conventional
banks, but large conventional banks appear to perform
better than large Islamic banks. Furthermore, recent
anecdotal evidence appears to suggest that Islamic banks
may face challenges in liquidity management, hitherto an
area of lesser concern (Ali, 2013). Product development
will thus remain a key topic for research in Islamic banking
for reasons related to risk management as well as increasing
competitive advantage.
While organizing these conferences and selecting the
papers for this volume series, a deliberate attempt was
made at highlighting policy-relevance of research where
possible. The first volume covers issues related to access
to finance and human development including essays on
zakah, awqaf, and microfinance. The second volume
is a selection of academic and policy-relevant papers
on economic thought from an Islamic perspective with
discussions on topics such as fiscal and monetary policy,
among others. The third volume (this volume) presents
a wealth of empirical evidence on various issues related
to Islamic banks such as profitability, efficiency, product
development, and Islamic corporate financing. The
fourth volume touches upon ethical, governance, and
regulatory issues in Islamic finance. The fifth volume is
about financial stability and risk management in Islamic
financial institutions.
In the following sections, I briefly introduce the papers
in this volume. Part 1 includes a selection of papers on
Islamic capital markets and corporate finance. Part 2
contains empirical papers on Islamic banking – topics
such as efficiency and profitability comparisons between
Islamic and conventional banks. Part 3 comprises papers
on product development in Islamic banks using innovative
approaches.
1. Islamic capital markets and corporate
finance
The first part of this volume is a collection of papers on
Islamic capital markets and corporate finance. Islamic
capital markets refer to those capital market products that
comply with Islamic finance principles. These investment
products can be categorized into two broad asset classes:
Islamic equity markets and Islamic debt markets.
The Islamic equity markets are stocks and mutual funds that
pass certain financial and nonfinancial screening criteria
that render them to be permissible for investment by Islamic
institutional and individual investors. Financial screening
of stocks is generally based on Shariah principles that
discourage debt and prohibit interest-based transactions. In
identifying a stock as being Shariah-compliant, a common
financial screen would be, for example, the debt-to-total asset
ratio. If this ratio exceeds 33% for a company, its stock may
be deemed impermissible for investment. Due to the nature
Cite this chapter as: Tariq W (2015). Islamic banking and finance – Essays on corporate finance, efficiency and product
development: An introduction to the issues and papers. In H A El-Karanshawy et al. (Eds.), Islamic banking and finance
– Essays on corporate finance, efficiency and product development. Doha, Qatar: Bloomsbury Qatar Foundation
Developing Inclusive and Sustainable Economic and Financial Systems
Tariq
of modern enterprise, interest income may be tolerated up
to a certain percentage, e.g., as long as interest income is
less than 5% of all income. Nonfinancial screens are based
on the underlying business that the company is engaged in.
All industries are considered permissible as long as there is
no contravening of basic Shariah principles regarding trade.
If a company is engaged in the trading of goods and services
deemed impermissible by Shariah, then its stock would also
be impermissible. Examples of such goods and services that
Shariah scholars would prohibit include alcohol, gambling,
and profane/violent entertainment. Since the introduction
of Dow Jones Islamic Indices, a number of similar indices
have been developed in different countries. Each Islamic
stock index would have slightly different financial screening
criteria depending on which Shariah scholars are on the
respective Shariah boards.
The second asset class in the Islamic capital markets is the
Islamic debt markets, which largely constitute products
known as sukuk. In theory, sukuk are meant to be securitized
representations of undivided shares in an underlying
asset or service. In practice, however, the majority of
sukuk prospectuses are drafted by lawyers with the aim
of replicating bond structures. Effectively, the majority of
the over $600 billion sukuk that have been issued can be
considered as the Islamic equivalent of bonds.
Conventional corporate finance research is devoted
mainly to the study of valuation of securities, portfolio
management, capital structure of firms, dividend policy,
and market inefficiencies. A significant source of corporate
financing is lending by banks; a large theoretical and
empirical literature exists on this topic as well. In recent
years, research in the vein of corporate Islamic financing
has grown with the increasing availability of reliable data.
The first paper in this section is about the valuation of
sukuk. Ariff and Safari first explore whether sukuk are
equivalent to bonds. Using a unique dataset from the
Bondstream database, they compare sukuk and bonds
issued by the same issuers with the same rating in the same
market. They find evidence suggesting that bonds and
sukuk are priced differently and that sukuk yields are not
determined by bond yields. Their paper is also among the
very few papers that attempts to propose valuation models
for pricing sukuk.
The second paper by Fauzi, Locke, Basyith and Idris is
an empirical analysis that explores the impact of Islamic
debt (sukuk) on the value of the issuing company. They
find evidence in support of the tradeoff theory of capital
structure, which posits that companies actively decide
between equity or debt issuance by assessing costs and
benefits. In their sample, they found evidence that suggests
that the issuance of sukuk was positively associated with
an improvement in the financial performance of the
firm, perhaps due to the associated tax benefits. They
conduct further analysis on whether subsequent sukuk
issuances by the same firm also positively affect financial
performance, with mixed results. This paper is relevant to
readers interested in theories on capital structure and their
application in Islamic finance
The third paper by Zulkhibri of the Islamic Development
Bank focuses on bank financing as a source of capital. He
xii
explores the relationship between bank financing, financing
rate and bank-specific characteristics in Malaysia, a
country whose financial system operates on a dual-banking
system. Using Bankscope data, the result of his pooled
panel estimations is that Islamic banks financing behavior
is dependent on the characteristics of the banks, such as
its size, liquidity, and capital. This is consistent with the
behavior of conventional banks. The lending behaviors of
both types of banks do not differ significantly with respect
to interest rates. The author also echoes sentiments of
purist Islamic economists that Islamic banks should move
away from debt-based instruments and towards profitloss-sharing in order to differentiate themselves from
conventional banks.
The fourth paper by Bhatt and Sultan adds a leverage risk
factor to the Fama and French multifactor asset pricing
model in order to investigate whether Islamic stocks are less
sensitive to leverage than other stocks such as conventional
stocks and socially responsible stocks. In widely cited
papers, academics have reported anomalies in their
results on asset pricing models (Fama and French (1992);
Lakonishok, Shleifer and Vishny (1994); Kothari, Shanken
and Sloan (1995)). Despite the methodological challenges
of studying asset pricing models, such as concerns about
data-snooping, sample selection biases, irrational behavior
of market participants, and the empirical difficulties of
distinguishing between competing hypotheses, this area
of research remains quite popular. The application of such
factor models to Shariah-compliant stocks is a nascent area
of research. In Bhatt and Sultan’s paper, readers will find
a valuable reference on this topic. The authors find that
Shariah-compliant stocks exhibit lower risk premiums to
traditional risk factors, but that they are also sensitive to the
leverage risk factor. This last result may have implications
for asset management practices and will also be of interest
to researchers in this field.
The final paper in this section by Sadeghi, explores a
number of issues surrounding Shariah-compliant stocks.
He first compares Shariah-compliant sustainable stocks
and conventional sustainable stocks, and then, using
event study methodology, Sadeghi investigates the market
reaction to the addition and deletion of a stock from a
Shariah-compliant index. It is suggested that readers
should regard the results in this conference paper simply
as preliminary findings. Further research is needed before
drawing any conclusions.
2. Islamic banking – Efficiency, profitability
and international perspectives
The first paper in this section by Sufian and Zulkhibri
presents evidence on the relationship between economic
freedom and the performance of Islamic banks. Economic
freedom in this paper broadly refers to the extent of
private ownership of resources as compared to government
control. Measurements of economic freedom are taken
from the Heritage Foundation and include measures of
business freedom, trade freedom, investment freedom, etc.
The results of the study are mixed. The authors present
evidence of a positive association between financial/
business freedom and the profitability of Islamic banks in
the MENA region. However, monetary freedom is negatively
associated with the performance of Islamic banks, lending
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Islamic banking and finance – Essays on corporate finance, efficiency and product development
support to the benefits of government intervention. The
authors also offer some preliminary policy evidence based
on the results of their panel regressions.
Anouze’s paper is in application of two nonparametric
techniques to measure the efficiency and performance of
Islamic banks, namely: Data Envelopment Analysis (DEA)
and the Classification and Regression Tree (CART). The
use of DEA to analyze efficiency of Islamic banks is widely
applied. It essentially involves grouping data into inputs
and outputs, where input variables are the resources
to be minimized and output variables are the resources
and products/services to be maximized to ensure a high
relative efficiency score. On the other hand, there are only
a handful of papers that use the CART technique to analyze
Islamic banks. This data mining technique can be used to
descriptively represent and explore data in a decision-tree
type of format that makes it relatively simple to understand.
Combining these two techniques allows this paper to make
a useful contribution to the literature. Anouze analyzes the
period between 1997–2007 to explore the performance
of Islamic and conventional banks during the period of
geo-political crises as well as the onset of the financial
crises.2 Overall, Anouze identifies 15 factors that future
researchers can consider as being important in predicting
efficient banks.
The third paper by Srairi, Kouki, and Harrathi explores
the relationship between Islamic bank efficiency and
stock market performance for 25 Islamic banks in the GCC
region using Data Envelopment Analysis (DEA) during
the period 2003–2009. They find evidence that suggests
that pure technical efficiency is positively associated
with stock returns while changes in scale efficiency have
no association with stock performance. Pure technical
efficiency here refers to the ability of managers to efficiently
utilize the resources of the bank. Scale efficiency refers to
the proportional reduction in input usage if the bank can
operate at optimal scale in terms of outputs produced. So
in other words, the authors find evidence that suggests
that stock returns respond positively to improvements
in managerial efficiency. The authors also find evidence
suggestive of increasing technical efficiency over the years
in the sample of Islamic banks.
Aytug and Ozturk’s paper is an empirical study investigating
the differences between Islamic banks and conventional
banks in Turkey between 2003 and 2011. They specifically
investigate whether being Islamic has any unique effect
on profitability ratios. One of the distinguishing features
of this paper is the application of the propensity score
matching method to estimate the average treatment effect
of being an Islamic bank. This statistical matching method
can overcome the concerns of self-selection bias that may
occur when using more conventional techniques such
as the Ordinary Least Squares (OLS) and Generalized
Method of Moments (GMM) estimators. For the sake of
comparison, however, the authors do also use the OLS
method. According to the regression results, there is an
insignificant relationship between being Islamic and
profitability if profitability is measured using the Return
on Assets (ROA) and Return on Equity (ROE) ratios. When
using the Net Interest Margin (NIM) ratio as a measure for
profitability, the results suggest that Islamic banks may
have a negative relationship with NIM. The results of the
Eds. Hatem A. El-Karanshawy et al.
propensity score matching generally confirm the findings
of the regression results. More research in this area will be
required to uncover the distinctiveness of using NIM as a
profitability ratio for Islamic banks as opposed to ROA and
ROE, and whether such differences may be driving the
results in this study.
3. Product development in Islamic finance
The first paper in this section by Calkan is an application
of simple mathematical techniques to enhance the
competitiveness of Islamic banking products. The
mudharabah-based deposits of Islamic banks – Profit
Sharing Investment Accounts (PSIAs) – have an inherent
competitive disadvantage compared to conventional time
deposits because profit-rates are not known in advance,
which may deter some customers. Calkan proposes that
while Shariah scholars prohibit the fixing of profit rates
in mudharabah transactions in advance, there is no
prohibition on reliably estimating the expected profit rates
and disclosing this to the customer. Through simulation
techniques, the author provides results suggesting that
Islamic banks may be able to estimate the PSIA profit rates
within a 95% confidence interval. This may potentially
be beneficial to banks by increasing their competitive
advantage and improving their fund management;
hence, this academic research paper has direct practical
implications for the industry.
Nienhaus’s paper is an innovative proposal to structure a
sukuk product based on musharakah with a self-adjusting
profit sharing ratio. He promotes the use of profit-loss
sharing (PLS) products, which are viewed by some Islamic
economists as the ideal form of Islamic financing. However,
the uptake of these products has been met with resistance
from lenders due to the adverse selection and moral hazard
problems that PLS financing arrangements pose. Nienhaus’s
proposal of self-adjusting profit sharing ratios is an attempt
to overcome the information problems in standard
musharakah financing arrangements. Interestingly, the
concept is based upon existing Accounting and Auditing
Organization for Islamic Financial Institutions (AAOIFI)
standards that allow for parties in a musharakah
arrangement to adjust profit-sharing ratios, if mutually
agreeable. Nienhaus puts forward a formula (which
can be viewed as a customizable template) that would
automatically adjust the profit-sharing ratio as new
information emerges on the profitability of the venture.
This method, he argues, would be less costly than ex-post
renegotiations regarding the profit-sharing ratios, because
in ex-post renegotiations, each party would strive to protect
their gains/losses at the expense of the other party. On
the other hand, the ex-ante model of self-adjusting ratios
proposed by Nienhaus links the profit-sharing ratio to
the actual performance of the venture, so the adjustment
is non-discretionary. Prior research has been limited to
theoretical treatment of the topic (Ahmed, 2002); hence,
readers would welcome the numerical example that
Nienhaus provides.3 In short, PLS financing arrangements
do have a role to play in the financial system particularly
in the financing of Small and Medium Enterprises (SMEs),
which have been shown to be disproportionally affected by
economic downturns. Researchers would be keen to build
upon the model provided here to develop more feasible PLS
financing products for Islamic banks to use.
xiii
Tariq
The third paper by Rizvi and Arshad is a proposal
to introduce Gross Domestic Product (GDP)-linked
government securities. The authors also attempt to test
the performance/benefits of their proposal using simple
simulations on four countries, as well as using descriptive
statistics and correlation analysis. Their focus is on the
introduction of a GDP-linked sukuk for government
financing. The idea is that indebted countries may
face economic shocks that reduce their ability to repay
international debt. If the rate of repayment is linked to
the country’s GDP as opposed to an external benchmark
such as the London Interbank Offered Rate (LIBOR),
this would ensure greater risk-sharing between the
sovereign borrower and international lenders. In times
of economic downturns, for example, the repayments
on a GDP-linked sukuk would be lower. The authors
provide an overview of GDP-linked sukuk. They present
some Shariah issues that may arise in structuring such
products. They provide simple simulations to show the
potential benefits for countries that adopt GDP-linked
sukuk. Finally, the authors realize that investors/lenders
would be reluctant to partake in greater risk sharing;
hence, they discuss the viability of such products from
the perspective of lenders. More work would be needed
to ensure the viability of these products; in particular,
the Shariah issues that would arise from a musharakahbased GDP-linked sukuk merits a lengthier discussion
than the authors have provided. Nonetheless, the paper
makes a reasonable contribution to the growing literature
on GDP-linked securities.
The final paper in this section is by Aydin and Reyner.
The paper makes a contribution to the Islamic financial
engineering literature. Unlike previous studies, this
paper uses a rigorous mathematical framework of riskneutral valuation and arbitrage-free contract prices. The
authors argue that the reference to risk-free zero bonds
in conventional finance is unrealistic. Instead, the authors
propose the use of a commodity basket-linked currency,
which they describe in detail in the paper. The authors
compare valuation of a conventional forward contract
with a forward contract from an Islamic perspective and
discuss the challenges of using a commodity basket in
the latter case, particularly because their forward prices
may not be available. Stochastic processes may be useful
in this instance for estimation purposes. In short, readers
interested in a financial mathematics and probability
theory treatment of Islamic financial engineering will find
this paper a valuable starting point. The use of financial
mathematics in Islamic finance literature is scarce indeed;
this fact alone makes this paper noteworthy.
3. Conclusion
The papers presented in this volume represent significant
contributions to the body of research in Islamic banking,
and suggest fruitful areas of future research. Looking ahead,
the lack of standardized sukuk contracts and the absence
of liquidity in these markets calls for further research in
financial engineering and regulatory frameworks. There
are continuing calls for Islamic banks to differentiate
themselves from conventional banks in order to reduce their
dependence on markup-based debt financing; this has been
a contentious issue that has been ongoing for decades and
that will remain a topic for researchers and practitioners to
xiv
ponder in the spirit of academic and intellectual discourse.
It is expected that new large-scale evidence on the effects
of high levels of debt in the economy (including household
debt) will help to inform this debate. The topic of Shariah
equity screens will continue to evolve and there is expected
to be a convergence between Shariah-compliant and socially
responsible investing. In addition to prevailing negative
stock screening approaches, perhaps future research will
explore the viability of positive-screening approaches and
the role that ‘impact investing’ has to play in Islamic equity
markets. Questions on how policy-makers can promote
economic freedom and protect shareholders and creditors
rights in a way that makes it attractive for Islamic banks
in particular to do business, will continue to be explored.
Increasing the competitive advantage of Islamic banks can
be achieved perhaps through product development that
increases information transparency and caters to the needs
of customers. Finally, attempts at using alternative pricing
benchmarks for Islamic securities that tie more closely to
the real economy, such as GDP-linked securities, remains a
fruitful area for researchers.
Notes
1. For an excellent overview of the empirical literature, I
recommend readers to refer to the chapter by Ongena
and Zaheer (2013).
2. It can be argued, however, that limiting the period until
2007 may not sufficiently capture the effects of the
financial crisis.
3. Diaw et al (2011) also provide hypothetical examples
for their GDP-linked sukuk based on the ijarah.
References
Abedifar P, Molyneux P, Tarazi A. (2013) Risk in Islamic
Banking. Review of Finance. forthcoming.
Ahmed H. (2002) Incentive-Compatible Profit-Sharing
Contracts: A Theoretical Treatment. In: Munawar I,
Llewellyn DT (Eds.). Islamic Banking and Finance – New
Perspectives on Profit-Sharing and Risk. Edward Elgar.
Cheltenham. 40–54.
Ali SS. (2013) State of Liquidity Management in Islamic
Financial Institutions. Islamic Economic Studies.
21(1):63–98.
Baele L, Farooq M, Ongena S. (2014) Of Religion and
Redemption: Evidence from Default on Islamic Loans.
Journal of Banking & Finance. 44:141–159.
Beck T, Demirgüç-Kunt A, Merrouche O. (2013) Islamic vs.
Conventional Banking: Business Model, Efficiency and
Stability. Journal of Banking and Finance. 37:433–447.
Diaw A, Bacha O, Lahsasna A. (2011) Public Sector Funding
and Debt Management: A Case for GDP-Linked Sukuk.
Paper presented at the 8th International Conference on
Islamic Economics and Finance, 19–21 December 2011,
Doha, Qatar.
Fama E, French K. (1992) The Cross-Section of Expected
Stock Returns. Journal of Finance. 47(2):427–465.
Kothari S, Shanken J, Sloan R. (1995) Another Look at the
Cross-Section of Expected Stock Returns. Journal of
Finance. 50(1):185–224.
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Lakonishok J, Shleifer A, Vishny R. (1994) Contrarian
Investment, Extrapolation, and Risk. Journal of Finance.
49(5):1541–1578.
Ongena S, Şendeniz-Yüncü I. (2011) Which Firms Engage
Small, Foreign, or State Banks? And Who Goes Islamic?
Evidence from Turkey. Journal of Banking and Finance.
35:3213–3224.
Ongena S, Zaheer S. (2013) Chapter 4: Empirical evidence.
In: Islamic Finance in Europe, Occasional paper series.
No. 146, June. European Central Bank.
Rogoff K. (2011a). Global Imbalances Without Tears. Project
Syndicate. Available at: http://www.project-syndicate.org
/commentary/global-imbalances-without-tears.
Regulation in the Face of Global Imbalances. Banque
de France, 4 March 2011. Paris, France. Available at:
https://www.banque-france.fr/fileadmin/user_
upload/banque_de_france/Economie-et-Statistiques/
La_recherche/GB/Session1-Rogoff.pdf.
Roubini N. (2013) There is a Lot Islamic Finance Can
Teach Us. Blog. Available at: http://www.roubiniblog.
com/2013/12/roubini-there-is-lot-islamic-finance.html.
Zaheer S, Ongena S, Van Wijnbergen S. (2013) The
Transmission of Monetary Policy Through Conventional
and Islamic Banks. International Journal of Central
Banking. 8(5):175–224.
Rogoff K. (2011b) What Imbalances After the Crisis? Speech
at International Symposium of the Banque de France –
Eds. Hatem A. El-Karanshawy et al.
xv
Valuation of Islamic debt instruments, the Sukuk:
Lessons for market development
Mohamed Ariff1, Meysam Safari2
Faculty of Business, Bond University, Gold Coast, Australia, T: (617)-5595–2296, E: [email protected]
SEGi University, Selangor, Malaysia, T: (60)-17271–4676, E: [email protected]
1
2
Abstract - A key issue for the fast growing Islamic debt market is whether sukuk instruments
are equivalent to conventional bonds as is practiced today by the market operators. This major
research question is explored here using a large traded data set on sukuk matched with a sample
of conventional bonds issued by the same issuers with same risk-rating and traded in the same
market. Matched samples of sukuk and conventional bonds traded over seven years are used in
our analyses, using statistical and causality tests. The results suggest that sukuk instruments are
priced significantly differently and that their yields are not Granger-caused by conventional security
yields or vice versa. This empirical finding does not support the market’s current practices based
on the assumption that sukuk are like normal bonds. In addition, we describe the basic contract
specifications, core cash flow structures in order to develop and suggest valuation models for three
selected sukuk types. The major implication of these findings is that there is much more research
needed to first document the independent nature of sukuk market behavior and the need for a
re-examination of current market practices.
Keywords: Sukuk, bond, yield curve, yield to maturity, Islamic finance, fixed income securities,
securitization
1. Introduction
Sukuk is the plural form of sakk, which in Arabic means
legal instrument, deed, or cheque.1 It was used in preIslamic era as withdrawal certificate (a form of cheque) on
deposits in financial firm or authorized merchants. Later,
these certificates became instrument for trading. And then
it became debt instruments traded among willing holders
of already issued debt just as is a conventional bond.
Being a relatively new product, there is no consensus on the
exact definition of sukuk. Some of international regulatory
bodies such as AAOIF, attempted to define sukuk.
Accounting and Auditing Organization for Islamic Financial
Institutions (AAOIFI, 2004) defines sukuk as “Certificates of
equal value representing, after closing subscription, receipt of
the value of the certificates and putting it to use as planned,
common title to shares and rights in tangible assets, usufructs
and services, or equity of a given project or equity of a special
investment activity.” International Islamic Financial Market
(IIFM, 2010) defined sukuk as a commercial paper that
provides an investor with ownership in an underlying asset.
It is asset-backed trust certificate evidencing ownership of
an asset or its usufruct (earnings or fruits). It has a stable
income and complies with the principle of Shariah. Unlike
conventional bonds, sukuk needs to have an underlying
tangible asset transaction either in ownership or in a master
lease agreement.
Islamic Financial Services Board (IFSB, 2009) definition of
Sukuk is “Sukuk (plural of sakk), frequently referred to as
“Islamic bonds,” are certificates with each sakk representing
a proportional undivided ownership right in tangible assets,
or a pool of predominantly tangible assets, or a business
venture (such as a Mudarabah). These assets may be in a
specific project or investment activity in accordance with
Shariah rules and principles.”
2. Sukuk securities are not the same
as bonds
This section provides empirical evidences that sukuk
securities are different from conventional bonds. The
empirical evidence is first shown as yield curves; second,
paired sample mean test between pairs of sukuk and bonds
issued by same issuer and for the same maturity period;
and third, Granger causality test between abovementioned
security pairs. Tests were conducted using data on
aggregated monthly yield to maturity (YTM) data2 from
Malaysian securities over August 2005 to April 2013 (total
of 93 observations for each security).
Cite this chapter as: Ariff M, Safari M (2015). Valuation of Islamic debt instruments, the Sukuk: Lessons for market
development. In H A El-Karanshawy et al. (Eds.), Islamic banking and finance – Essays on corporate finance, efficiency
and product development. Doha, Qatar: Bloomsbury Qatar Foundation
Developing Inclusive and Sustainable Economic and Financial Systems
Ariff and Safari
Descriptive statistics
Descriptive statistics for various sukuk securities and
conventional bonds are presented in Table 1. The
statistics suggest that the mean yield of sukuk securities
for all types of issuers and for all forms of maturities is
3.92 percent. The yields vary within a minimum of 2.91
(3 months maturity Treasury sukuk securities) and the
maximum of 5.66 (sukuk securities issued by AAA rated
corporations with 20-year maturity). On the other hand,
the mean yield of conventional bonds of all types of issuers
and all maturities is 3.91 percent. These vary between a
minimum of 2.90 (3 months maturity Treasury bills) and
the maximum of 5.60 (conventional bonds issued by AAA
rated corporate with 20-year maturity). At the issuer level,
AAA rated corporate issue of sukuk securities yielded 4.29
percent while the mean of Treasury bill yields is 3.55
percent. On the other hand, the highest conventional
mean yield for AAA rated corporate issuers is 4.31 percent
while the lowest mean yield for conventional bills and
notes issued is by the Government of Malaysia with 3.51
percent.
Yield curves
The result data are presented in this section, as yield curves
in two plots. Yield curve is the relation between the cost of
borrowing and the time to maturity of a debt security for a
given issuer for a given rating quality. Yield curves for sukuk
securities and conventional bonds issued by government
and corporations are plotted as in Figure 1 A and Figure 1B.
The plots are presented as YTM of (i) conventional against
(ii) sukuk issues in two plots. The two issuer types are of
increasingly higher risk rating with sovereign being the
lowest risk – therefore with the lowest yields – on the one
end, and the AAA corporate issues with higher yield at the
other end.
As Figure 1(A) suggests, the yields of Government
Islamic Issues (GII) are higher than those of conventional
bonds issued by the same issuer (Malaysian Government
Securities, or MGS). The difference between sukuk yield
and conventional bond yield is larger as maturities increase
from 2 years to 15 years. The maximum difference between
the yields of sukuk securities and those of conventional
bonds for this category is with 3 years maturities. The
difference is 8.28 basis points. On average, there is a 4.04
basis points difference between yields of sukuk securities
and conventional bonds. The total outstanding value of
sukuk securities issued by Malaysian government as at
April 2013 was RM 155.9 billion (US$ 55 billion).
Multiplying yield difference and market size indicates
that the Malaysian government needs to pay an extra RM
6.3 billion per year to investors holding sukuk securities
compared to the amount conventional issues of same term
and quality. This means that the sukuk investors earn RM
6.3 billion higher return compared to the investors in
conventional bond market. Obviously sukuk yields are
systematically higher.
Figure 1(B) shows a plot of the yields of securities issued by
AAA rated corporate issuers. Yields of sukuk securities are
less than yields of conventional bonds for maturities of less
2
than 10 years; it is more for periods beyond 10 years. The
maximum difference between the yields of sukuk securities
and the conventional bonds issued by corporate issuers
with maturities less than 10 years is for those with 2-year
maturity with a -7.28 basis points. However, the maximum
amount for securities with maturities longer than 10year term is +6.38 basis points for securities with 20-year
maturity. Long-dated sukuk securities are perceived by
the market as being more risky, thereby attracting higher
yields. Long dated sukuk are perhaps more risky given the
risk of greater uncertainty beyond 10 years. It is puzzle
why the same firms issuing short-dated securities provide
a safer investment. Obviously again, the yield differences
are systematic.
On average, there is a -2.44 basis points difference between
yields of sukuk securities and conventional bonds. The total
outstanding value of sukuk securities issued by Malaysian
AAA Corporate issuers in April 2013 was RM 64.48 billion
(US$ 22 billion). Multiplying yield difference and the
market size indicates that the AAA corporate issuers would
save RM 1.57 billion (US$ 0.53 billion) per year on their
sukuk securities.
Comparison of yields of Sukuk securities and
conventional bonds
Results of the paired sample t-tests are summarized in
Table 2 (panels A and B) on the equality of means. Out of
the 20 tested pairs of mean yields of sukuk and conventional
bonds, 19 cases (95 percent) of all pairs showed significant
differences in their yields to maturities. In 15 cases, the null
hypotheses are rejected at 0.01 significance levels. Thus,
one can conclude that the yields of sukuk securities differ
from the case of conventional bonds, although the issuer
and the issue tenure are the same. Table 2 A is a summary
of the statistics pertaining to the mean yield of sukuk and
conventional bonds.
As the t-statistics suggests, the mean yield of sukuk
securities and conventional bonds are significantly
different for all issues by Government. The difference
between the means
of these are positive, indicating that sukuk securities
tend to yield more than conventional bonds issued by the
Government of Malaysia, ceteris paribus. Thus, the market
associates higher risks (for reasons, we still do not know) to
sukuk structures rather than conventional structures.
Table 2B is a summary of the statistics for sukuk securities
and conventional bonds issued by AAA rated corporations.
For these securities, the mean yields of sukuk securities and
conventional bonds are significantly different in all cases
except in the cases of 10-year maturity. The differences are
negative for securities with tenure 7 year or less, while, for
securities with 10 years maturity or more, the difference is
positive. This means the mean of yield of sukuk securities
issued by AAA rated corporations is lower than the yield of
conventional bonds for issues with 7 years or less maturity.
For the securities with long-term maturities (10 years and
more) the mean yield of sukuk securities is higher than the
conventional bonds.
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Eds. Hatem A. El-Karanshawy et al.
3.515
3.3394
3.4333
3.6178
3.8454
4.0665
4.3576
4.6380
4.9529
5.3006
5.6013
4.315
Mean
3-Month
6-Month
1-Year
2-Year
3-Year
5-Year
7-Year
10-Year
15-Year
20-Year
Mean
Corporate
2.903
2.942
3.015
3.17
3.315
3.564
3.759
3.943
4.186
4.357
3-Month
6-Month
1-Year
2-Year
3-Year
5-Year
7-Year
10-Year
15-Year
20-Year
Government
Mean
Tenure
Issuer
4.259
3.473
3.33
3.39
3.52
3.7
3.97
4.27
4.57
4.93
5.31
5.6
2.89
2.92
2.99
3.1
3.21
3.5
3.74
3.95
4.13
4.3
0.429
0.422
0.453
0.412
0.375
0.359
0.366
0.385
0.420
0.477
0.494
0.521
0.519
0.527
0.525
0.447
0.380
0.318
0.319
0.391
0.401
0.393
Std. Dev
Conventional
Median
Table 1. Descriptive statistics of Sukuk vs. Conventional bonds.
1.81
1.90
2.17
1.9
1.59
1.61
1.61
1.57
1.7
1.85
1.98
2.12
1.84
1.98
2.06
2.1
2.13
1.8
1.84
1.93
1.77
1.58
Range
3.60
2.58
2.28
2.6
3.1
3.32
3.62
3.87
4.03
4.21
4.41
4.61
1.82
1.85
1.92
2.2
2.37
2.78
2.91
3.09
3.35
3.6
Min
5.41
4.49
4.45
4.5
4.69
4.93
5.23
5.44
5.73
6.06
6.39
6.73
3.66
3.83
3.98
4.3
4.5
4.58
4.75
5.02
5.12
5.18
Max
4.291
3.556
3.3133
3.4009
3.5534
3.7726
4.0016
4.3082
4.5942
4.9433
5.3559
5.6651
2.9183
2.9544
3.0280
3.2039
3.3976
3.6359
3.8003
3.9967
4.2285
4.3946
Mean
4.237
3.509
3.29
3.35
3.47
3.67
3.93
4.24
4.53
4.93
5.3
5.66
2.89
2.92
2.99
3.11
3.31
3.63
3.75
3.99
4.18
4.32
Median
0.425
0.411
0.455
0.405
0.359
0.324
0.322
0.352
0.405
0.483
0.580
0.573
0.538
0.544
0.530
0.436
0.340
0.310
0.312
0.358
0.374
0.372
Sukuk
Std. Dev
1.80
1.85
2.17
1.9
1.61
1.56
1.5
1.56
1.69
1.84
2.15
2.11
1.96
2.04
2.11
2
1.84
1.8
1.79
1.81
1.67
1.5
Range
3.57
2.67
2.24
2.56
3.04
3.3
3.62
3.84
4
4.18
4.38
4.58
1.82
1.85
1.97
2.3
2.63
2.85
3
3.17
3.45
3.68
Min
5.38
4.45
4.41
4.46
4.65
4.86
5.12
5.4
5.69
6.02
6.53
6.69
3.78
3.89
4.08
4.3
4.47
4.65
4.79
4.98
5.12
5.18
Max
Valuation of Islamic debt instruments, the Sukuk: Lessons for market development
3
Ariff and Safari
Yield curve of government issued securities
A
Yield curve of corporate issued securities
B
4.4
5.7000
4.2
5.2000
4
3.8
4.7000
3.6
4.2000
3.4
3.2
3.7000
3
Government Conventional Issues
Government Sukuk Issues
2.8
0
5
10
15
Corporate Bonds
Corporate Sukuk
3.2000
20
0
5
10
15
20
Figure 1. Yield curve for Sukuk securities vs. Conventional bonds.
Table 2A. Paired samples T-test results: Government.
Table 2B. Paired samples T-test results: Corporate issues.
Tenure
Sukuk
Conv
Δ (Sukuk
- Conv)
t-Stat
Tenure
Sukuk
Conv
Δ (Sukuk
-Conv)
t-Stat
3M
6M
1Y
2Y
3Y
5Y
7Y
10Y
15Y
20Y
2.9183
2.9544
3.0280
3.2039
3.3976
3.6359
3.8003
3.9967
4.2285
4.3946
2.9031
2.9424
3.0147
3.1699
3.3148
3.5641
3.7588
3.9433
4.1865
4.3567
0.0152
0.0120
0.0132
0.0340
0.0828
0.0718
0.0415
0.0533
0.0420
0.0380
2.789***
2.461**
2.419**
4.789***
7.686***
8.465***
6.418***
7.037***
5.887***
5.043***
3M
6M
1Y
2Y
3Y
5Y
7Y
10Y
15Y
20Y
3.3133
3.4009
3.5534
3.7726
4.0016
4.3082
4.5942
4.9433
5.3559
5.6651
3.3394
3.4333
3.6178
3.8454
4.0665
4.3576
4.6380
4.9529
5.3006
5.6013
-0.0260
-0.0325
-0.0644
-0.0728
-0.0648
-0.0495
-0.0438
-0.0096
0.0553
0.0638
-6.173***
-7.483***
-7.360***
-6.789***
-6.047***
-5.685***
-7.318***
-0.953
2.414**
2.571**
**, ***: Significant at 0.05, and 0.01 significance levels,
respectively.
**, ***: Significant at 0.05, and 0.01 significance levels,
respectively.
Granger causality test of yields of Sukuk and
conventional bonds
change in yield of sukuk to verify if it causes a change in
yield of conventional bonds. Second test is on the change
in yield of conventional bonds to verify if it causes a
change in yield of sukuk. The latter test is on the yields
of conventional bonds Granger causing yields of sukuk.
Results of pair-wise Granger causality test on each pair is
presented in Table 3.
The previous section showed that the mean yield of sukuk
is statistically different from yield of conventional bonds.
Since each pair of securities issued by the same issuer for
the same period of time and for same rating, it is expected
that the correlation between yields of these securities
may be high. This may be a valid reason for a hypothetical
argument that this difference arises from an unidentified
causal relationship. One may wish to test if changes in yield
of one type of security may cause change in the other series?
In other words, one may want to test for Granger Causality
(Granger, 1969) between yields of sukuk securities and those
of conventional bonds to identify if some causal link exists.
In order to test this, two Granger causality tests were
conducted on each pair of securities. First test is on the
4
The first null hypothesis tested was “yield of sukuk security
does not Granger cause the yield of conventional bond
counterparts.” As the statistics in Table 3 suggest, out of 20
pairs of securities tested, the null hypothesis are rejected in
only 6 pairs at the 0.10 significance level. Yields of sukuk
securities Granger cause yield of conventional bonds in only
6 out of 20 (or 30 percent) pairs. This indicates that one
may not generally conclude that yield of sukuk securities
Granger cause the yields of conventional bonds. Results also
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Valuation of Islamic debt instruments, the Sukuk: Lessons for market development
Table 3. Pair-wise granger causality tests with lags = 2.
Issuer
Maturity
Government
3M
6M
1Y
2Y
3Y
5Y
7Y
10Y
15Y
20Y
3M
6M
1Y
2Y
3Y
5Y
7Y
10Y
15Y
20Y
Corporate
Sukuk security does not Granger
cause conventional bond
Conventional bond does not
Granger cause sukuk security
F-Statistic
Prob
F-Statistic
0.190
0.7611
2.5477**
3.7064**
2.4141*
3.6773**
0.6627
0.1797
0.1263
0.2575
0.0684
0.9525
3.1337**
3.0097*
0.8961
1.0998
1.7843
0.0696
0.6228
0.9677
0.8266
0.4702
0.0842
0.0286
0.0955
0.0293
0.5180
0.8358
0.8815
0.7736
0.9339
0.3898
0.486
0.0545
0.4119
0.3376
0.1724
0.9327
0.5389
0.3840
0.8972
1.2937
1.5399
3.7080**
5.1703***
0.8004
0.0314
0.2506
0.8322
1.1982
1.6904
3.3457**
5.6886***
1.6613
0.7585
0.3039
1.2941
0.7896
0.5011
0.9878
Prob
0.4115
0.2795
0.2202
0.0285
0.0076
0.4525
0.9690
0.7788
0.4325
0.3067
0.1905
0.0399
0.0048
0.1959
0.4714
0.7387
1.2794
0.4573
0.6078
0.3766
Note: *, **, ***: significant at 0.10, 0.05, and 0.01 significance levels, respectively.
show that yields of Government sukuk (1, 2, 3 and 5 years)
and those of AAA rated corporate (1 and 2 years) issues
Granger cause their conventional bond yields. Results do
not show a consistent pattern in terms of issuer or maturity
of the security for having a Granger causal effect.
The second test conducted was to check for the presence of
Granger causal relation between conventional bonds and
sukuk. The null hypothesis tested is “yield of conventional
bonds does not Granger cause the yield of sukuk security.”
Out of the 20 pairs of securities tested, the null hypothesis
is rejected in 4 pairs at the 0.05 significance levels. This
indicates that one may not generally conclude that the
conventional bond yields Granger cause sukuk security
yields. These results show that yield of conventional bonds
issued by Government (2 and 3 years) and AAA rated
corporate (6 months and 1 year) Granger cause their
sukuk counterparts. Results do not show consistent pattern
in terms of issuer or maturity of the security for having a
Granger causal effect.
Finally, as in Table 3, bi-directional Granger causality
(as expressed in: Enders, 1995, Hossain, 2005) between
yield of sukuk and yield of conventional bonds is
observable in 3 out of 20 (or 15 percent) pairs. In other
words, in 3 pairs of securities, both null hypotheses are
significantly rejected, or, yield of sukuk Granger cause
yield of conventional bonds and the other way around.
This may signal that both variables are Granger caused
by a third variable yet to be explored. Results show
that yield of sukuk and conventional bonds have bidirectional Granger causal relation in securities issued
by Government (2 and 3 years) and AAA rated corporate
Eds. Hatem A. El-Karanshawy et al.
(1 year). In summary, it is reasonable to conclude that,
with few exceptions, there is no causal relationship
between sukuk and conventional bonds. This is the
second statistically supported evidence to affirm an
argument that the two types of debt instruments are not
the same. This conclusion has important implication for
market operation, valuation practices, risk estimation
and regulatory rule setting. These are challenges to be
addressed in future research.
3. Structuring sukuk contracts
This section reviews the structure of three major types
of sukuk securities namely Ijarah, Musharakah, and
Mudarabah.
Ijarah sukuk
Ijarah, which means “to give something on rent” (Lewis and
Algaoud, 2001), is the reward or recompense that proceeds
from a rental contract between two parties, where the
lessor (the owner of the asset) leases capital asset to the
lessee (the user of the asset) (Gait and Worthington,
2007). There is a tendency toward lease financing (Ijarah)
in Islamic banking sector, since it promises higher yields
than in trade finance (Murabahah) and it also has longer
financing horizon, which is an important feature for
business investments (Daryanani, 2008). In order to be
permissible under the Shariah, the Ijarah contract should
satisfy some conditions. The primary requirement is that
the lessor must be the real owner and in possession of the
asset to be leased under contract. As a result, the lessor
should solely bear all risks and uncertainties associated
5
Ariff and Safari
to the asset and be responsible for all damage, repair,
insurance, and depreciation of the asset (Khan and Bhatti,
2008).
It could be inferred that charging rental payment is not
allowed until the lessee actually receives the possession
of the asset and shall pay the rental only as long as it is
in usable condition. Moreover, in case of manufacturing
defects, which are beyond the lessee’s control, the lessor
is responsible. However, the lessee is responsible for the
proper upkeep and maintenance of the leased asset. The
intention of posing such restriction in Ijarah contract by
Shariah is to protect both parties to the contract by reducing
the uncertainty and ambiguity from the agreement (Wilson,
2004). In addition to that, both lessor and lessee should be
clear on purpose of Ijarah and the usage of assets, moreover,
the Ijarah purpose must comply with Shariah (Al-Omar
and Abdel-Haq, 1996).
There are two forms of leasing contracts, or Ijarah, in
Islamic finance. Ijarah, or direct leasing contract, is the case
where the lessee uses the capital asset owned by the lessor,
with his/her permission, for a specific period of time for
a monthly or annual rent. The owner assumes ownership
title during the whole contract period, and the owner
should performs the ownership responsibilities such as
insurance (Zaher and Hassan, 2001). In this Ijarah contract
possession of asset should be transferred back to the owner
after the contract matures. In other words, in pure Ijarah
contracts, there is no option to transfer the ownership of
the asset at maturity.
Ijarah wa Iqtina, or hire purchase, is the case of contract
where the basic intention is transferring the ownership after
completing the leasing period. Ijarah wa Iqtina is popularly
practiced when Islamic bank purchases equipment or some
other capital asset based on the request of an individual
or institutional customer and then rents it to the customer
for a certain fixed rent. On the other hand, the customer
promises to purchase the equipment or asset within a
specified period to transfer the ownership from the Islamic
bank to the customer (Al-Jarhi and Iqbal, 2001). However,
it should be noted that the lease contract is completely
separate and independent from the contract of purchase
of residuals, which has to be valued on a market-basis and
cannot be fixed in advance.
Then the SPV leases back the assets to the issuer at a specific
predetermined rental fee and then the SPV securitize the
ownership in the assets by issuing sukuk certificates to the
public investors (Lewis, 2007). These sukuk certificates
represent an undividable share in the ownership of the
assets, which entitle the sukuk holders to distribution of
the rental payments on the underlying assets. However,
the rental payment could be fixed or floating for the whole
period, dependent on the leasing contract between the SPV
and originator. Since these sukuk certificates represent
ownership in real assets, they can be traded in a secondary
market.
The SPV manages the cash flows of the sukuk contract
by receiving periodic rentals and installments from the
originator and then disbursing the cash flows to the
sukuk holders (Aseambankers, 2005). SPV also manages
disbursement of lump sum maturity payments. At the
maturity of a sukuk contract, the SPV no longer has a role
and consequently will cease to exist. However, the Ijarah
sukuk is typically issued for periods longer than five years
and could be considered as long-term debt certificates.
This may raise the issue of SPV’s default risk, so, the
investors typically receive a direct guarantee from the
issuer’s guarantee of the SPV obligations (Wilson, 2008).
This guarantee also includes the obligation by the issuer to
repurchase the asset from the SPV at the end of the Ijarah
contract at the original sale price.
Wilson (2008) suggest that SPV does not have any of the
risks associated with banks due to SPV’s nature. In other
words, SPV is bankruptcy remote. If the issuer faces
the bankruptcy, the creditors to the issuer cannot claim
the assets held by the SPV or otherwise interfere with the
rights of the sukuk-holders with respect to the underlying
assets (Gurgey and Keki, 2008). As a result, SPV would be
attractive to both issuers and investors, and this may justify
the relatively high legal establishment costs. Figure 2 is a
contract specification.
Kamali (2007) claims that the fixed and predetermined
nature of rental cash flow introduces additional risk
because the Ijarah sukuk holders receive steady income,
The purchase contract should be an optional, non-binding
contract because the quality and the market price of the
asset at the end of the lease period are unclear (Chapra,
1998). One other approach is the case where the ownership
is gradually transferred to the customer. In this case, and
in addition to the regular rental payment, the customer
shall pay installments of the value of the asset in order to
reduce the ownership share of the lessor in the asset until
the ownership is fully transferred to the lessee (Metwally,
2006). Ijarah wa Iqtina, having strong agreement among
Shariah scholars, is widely used in the real estate, retail,
industry, and manufacturing sectors (Iqbal, 1998, Warde,
2000).
Ijarah sukuk is based on the Ijarah contract. In order to
issue Ijarah sukuk, the originator, who primarily owns the
assets, sells the assets to a Special Purpose Vehicle (SPV),
which is typically a company in an offshore tax-free site.
6
Figure 2. Ijarah sukuk structure – adapted from Wilson
(2008), Abdul Majid (2003), El-Gamal (2007),
Bose and McGee (2008).
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Valuation of Islamic debt instruments, the Sukuk: Lessons for market development
which is more risk averse than is the case of common
stocks. However, he mentioned general market conditions,
price movements of real assets, ability of the lessee to pay
the rental or installments, maintenance and insurance cost
are sources of risks to the Ijarah sukuk. He concluded that
because of these risk factors, the expected return on some
of Ijarah sukuk may not be precisely predetermined and
fixed. Thus, the fixed rental may only represent a maximum
that is subject to some possible deductions.
The major criticism of Ijarah sukuk is that the return is
variable or floating in most cases. Moreover, this variable
rate, sometimes for simplification reasons, is mostly
benchmarked or “pegged” to an interest-based index such
as the London Interbank Offered Rate (LIBOR) for US$
based sukuk and in local currencies the local rates. Usmani
(2002) criticized this practice by associating riba to this
form of Ijarah sukuk practice. Shariah scholars suggest the
usage of other non-interest benchmarks for pricing and
evaluation purposes. In order to overcome the riba issue,
government sukuk could be assessed by macroeconomic
indicators and corporate sukuk could be assessed based on
the company performance indicators.
Musharakah sukuk
Iqbal and Molyneux (2005) defined Musharakah as “an
arrangement where two or more parties establish a joint
commercial enterprise and all contribute capital as well
as labor and management as a general rule.” In contrast to
Mudarabah contract, Musharakah investors have the right
to participate in management of the business partnership,
however, this right is entrusted to each investor (Shinsuke,
2007). It could be argued that Musharakah contract may
require establishment of a partnership or company, where
Musharakah contract parties are the participants and
owners (Wilson, 2004).
Musharakah sukuk securities could be issued based on
such financing concept. Musharakah type of equity finance
demands that both a profit-sharing ratio and length of
the joint venture agreement is decided in advance. Similar
to Mudarabah, loss is shared in proportion to the capital
contribution unless the loss is proven to be due to negligence
of one party (Daryanani, 2008). Therefore, all profits and
losses generated from the Musharakah are shared among
the parties on the basis of the pre-agreed ratio. As a result,
Musharakah is basically suitable for financing private or
public companies as well as projects as also practiced by
Islamic banks, where it is typically performed through joint
ventures between banks and business firm for a certain
operation (Gait and Worthington, 2007).
Musharakah, due to its nature and advantages in providing
equal (but proportionate) benefits for all parties, has the
support of all scholars and is valid under Shariah principles.
However, El-Gamal (2000) suggests that parties to
Musharakah usually need the help of legal expert to ensure
that any potential Riba or Gharar is carefully avoided. On
the discussion about Musharakah contracts, Chapra (1998)
concluded that “The only requirement of the Shariah would
be justice, which would imply that the proportional shares
of partners in profit must reflect the contribution made to the
business by their capital, skill, time, management ability,
goodwill and contacts. Anything otherwise would not only
Eds. Hatem A. El-Karanshawy et al.
shatter one of the most important pillars of the Islamic value
system, but also lead to dissatisfaction and conflict among the
partners and destabilize the partnership. The losses must,
however, be shared in proportion to capital contribution and
the stipulation of any other proportion would be ultra vires
and unenforceable.”
Lewis and Algaoud (2001) suggest two ways to structure a
Musharakah contact. However, both types are based on the
same general concept of Musharakah, where parties (capital
owner and entrepreneur) are ensured an equitable share in
the profit or loss on pre-agreed terms. The difference lies
in the pre-agreed sharing ratio. In the first method, this
pre-agreed ratio is pre-fixed and remains constant for the
whole period of the contract while in the second type, the
ratio is declining. The diminishing Musharakah contract
is preferred by some financiers since it allows release of
their capital from the investment by reducing its equity
share each year and receiving periodic profits based on the
remaining balance. On the other hand, the equity share of
the entrepreneur increases over time to the extent that he/
she becomes the sole owner of the firm.
Sukuk based on diminishing Musharakah are gaining
momentum since they enable Islamic banks or Shariahcompliant investment companies to provide up-front
investment funding to the issuer. In this regard, both parties
establish a Special Purpose Vehicle (SPV) to administer the
sukuk. In order to issue a diminishing Musharakah sukuk,
the issuer transfers the ownership of an asset to the SPV to
enter the partnership agreement. The investors enter the
agreement by paying cash. Therefore, both the investors
and the issuer are equity partners in the SPV. However,
the investors share in the SPV diminishes over time as the
issuer pays installments to investors to repurchase their
respective shares in the asset. These installment payments
plus the issuer’s rental payments for use of asset (asset’s
generated income) so the contract becomes a Musharakah
sukuk with cash flow stream for sukuk-holders. In fixedratio Musharakah sukuk, the cash flow stream for the sukukholder is only from the income generated from the asset
and not the installment part. The structure of diminishing
Musharakah sukuk is depicted in the Figure 3.
Figure 3. Diminishing musharakah sukuk structure
(Wilson, 2008).
7
Ariff and Safari
Flexibility in payment schedule and amounts has made
diminishing Musharakah sukuk more convenient to use.
However, it should be highlighted that all arrangements
should be agreed upon ex-ante by all parties to the SPV. The
payments are usually monthly or quarterly, but not necessary
in equal amounts (Wilson, 2008). Smaller installments
could be made during the initial period of the sukuk, with
most of the asset value or SPV capital remaining with the
investors, but the amount of the installment payments
could increase in a linear fashion, or according to some
predetermined formula. As the issuer’s share in the asset
increases through the buy-back process, the periodic rental
might be expected to decrease due to a decline in remaining
share. However, this does not necessarily have to be the case,
especially if there is capital appreciation in the value of the
asset. In other words, when installment and rental payments
are aggregated, they might be constant, diminishing or
increasing over time, provided both parties agree to the
formula used and the documentation is transparent.
4. Sukuk cash flows
This section reviews cash flow patterns of two major types
of sukuk securities namely Ijarah sukuk and Musharakah
sukuk.
Figure 4. Cash flows pattern for sukuk with fixed promised
regular payments and promised maturity
payment.
of predetermined maturity payment, r is the discount rate,
and N is the number of time periods remaining to maturity.
Another variation of cash flows of Ijarah sukuk happens
when the periodic payments follow a growth pattern. In
Ijarah sukuk, maturity payment (M) should be based on
market value (hence, a priori undetermined), while the
Ri, the amount of promised regular payment (rental fees)
in period i is following a predetermined constant growth
model with growth rate of g. However, in practice, maturity
payment of Ijarah sukuk is predetermined and mentioned
in contract documents. Therefore, its cash flow is similar to
the one in Figure 5.
Ijarah sukuk
Ijarah sukuk can have various types of payback structures.
In the simplest form, Ijarah sukuk payback could be fixed
promised regular payments and not a predetermined
promised maturity payment. The formal Ijarah contract
does not have the option for parties to transfer the
ownership of the asset at the end of the period. Thus, at
the end of an Ijarah contract, the asset should be returned
to the owner (capital owner or the SPV). In order to
transfer the ownership back to the issuer at the maturity,
one should use Ijarah wa Iqtina (lease and purchase)
contract. Ijarah wa Iqtina sukuk is form of Ijarah contract
where the ownership of the asset will be transferred to
lessee (issuer) at the maturity of the sukuk. However, the
maturity payment is not determined at the issuance time
of sukuk. The valuation of the asset in this case should be
conducted at the maturity time, when the market value
of the asset is revealed and maturity payment is set to be
equal to that.
Ijarah sukuk has fixed and predetermined rental payment
(rewards) and a market valued maturity payment.
However, in practice, the maturity value of the property is,
sometimes, fixed and predetermined to both sides of the
contract. Therefore, the cash flow pattern of Ijarah sukuk
would be similar to that of a coupon-bearing conventional
bond such securities are depicted in Figure 4. The valuation
of this form of sukuk is similar to a straight bond because
of the similarity of cash flows pattern and the tradability.
Thus, a sukuk could be priced by using the Equation 1.
N
P=∑
t =1
R
+
M
(1 + r ) (1 + r )
t
N
=
R
M
1 
1 +
(1)
r  (1 + r ) N  (1 + r ) N
In Equation 1, P is the price of sukuk security, R is the
amount of periodical promised payment, M is the amount
8
Figure 5. Growing promised regular payments pattern
with predetermined promised maturity
payment.
The cash flow pattern of sukuk consists of a growing
annuity of promised regular payments and a promised
maturity payment. Thus, using the formula for calculating
the present value of an annuity, one can formulate the price
of a sukuk security as Equation 2.
N
P=∑
t =1
R
+
M
(1 + r ) (1 + r )
t
N
=
R1
(r - g )
 1+ g
. 1 – 

  1 + r 
N

M
+
N
 (1 + r )
(2)
In Equation 2, P is the price of sukuk, r is the discount rate,
N is the number of periods to maturity, g is the growth
rate of promised regular payments, and R1 is the amount
of first promised regular payment. It is assumed that the
promised payments are growing at a constant rate of g,
thus, R2 = R1(1 + g).
Musharakah Sukuk
The simplest form of cash flow generated by a Musharakah
sukuk security is obtained from those securities that
only pay a lump sum amount of cash in a certain and
predetermined point of time in future at maturity.
There is no cash payment to investors prior to maturity
as shown in Figure 6. The amount of maturity payment,
however, is not determined and should be based on the
performance of the underlying investment. However, as
in most practice cases, the maturity payment is fixed and
predetermined.
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Valuation of Islamic debt instruments, the Sukuk: Lessons for market development
Figure 6. Zero-promised regular payment cash flow
pattern of Musharakah Sukuk securities.
Figure 8. Declining promised payments cash flows
pattern.
This form of cash flow is the same as conventional discount
bond cash flow or a bond with balloon payment. Thus the
same valuation process is applicable. The valuation of these
forms of sukuk could be performed using the conventional
pricing approach as follow. The current price of sukuk is the
maturity payment (face value) discounted to the present
time, similar to discount bonds.
P=
M
T -t
(1 + r )
(3)
In Equation 3, P is the price of sukuk, M is maturity
payment (face value), T is maturity date, t is time, and r
is the discount rate. It should be noted that r should not be
based on any interest bearing benchmark.
Diminishing Musharakah sukuk may possess payoff
structure in a manner that it only pays some promised
regular payments at certain periods of time with zero
maturity payment. Amount of promised regular payments
are fixed and predetermined. Cash flow pattern of such
security is depicted in Figure 7.
Figure 7. Fixed promise regular payment cash flows
pattern of Sukuk without promised maturity
payment.
Cash flow represented in Figure 7 is identical to a constant
annuity. Thus, in order to evaluate the price of a diminishing
Musharakah sukuk, present value of annuity is applicable.
This will result into Equation 4.
N
P=∑
t =1
R
(1 + r )
t
=
R
1 
1 
r  (1 + r ) N 
(4)
In Equation 4, P is the price of diminishing Musharakah
sukuk security, R is the amount of periodical promised
payment, r is the discount rate, and N is the number of time
periods remaining to maturity.
In diminishing Musharakah sukuk, amount of promised
regular payment in period i is following a predetermined
constant negative growth (declining) model with growth
rate of g, which is a negative number. In other words, the
cash flow stream of regular payments declines to zero at
maturity (M is also equal to zero). The cash flow diagram is
presented in Figure 8.
Eds. Hatem A. El-Karanshawy et al.
Thus, price of a diminishing Musharakah sukuk can be
formulated as Equation 5.
N
P=∑
t =1
R
(1 + r )t
=
N


R1
. 1 –  1 + g   (r - g )   1 + r  
(5)
In Equation 5, P is the price of diminishing Musharakah
sukuk, r is the discount rate, N is the number of periods to
maturity, g is the negative growth rate of promised regular
payments, and R1 is the amount of first promised regular
payment. It is assumed that the promised regular payments
are declining at a constant rate of g, thus, R2 = R1 (1 + g) < R1.
5. Sukuk world markets
The sukuk market has grown rapidly in recent years.
Emergence of more than 250 Takaful (Islamic Insurance)
companies, 350 Islamic equity funds, and 370 Islamic banks
worldwide has made a great demand for sukuk. Khan and
Bhatti (2008) highlighted that sukuk constitute about 85
percent of the Middle Eastern capital market, US$13bn of
them have been issued there with an average growth rate
of over 45 percent during 2002–2007. The Middle East and
Asian regions will primarily rely on sukuk to meet their US$1.5
trillion infrastructure needs over the next ten years (John,
2007) is the kind of statements one reads in commentaries.
McKenzie (2008) investigated the sukuk market statistics
and reported that, while sovereign sukuk issues by Bahrain
and Malaysia played an initial role in establishing the
market, about 80 percent of issues between 2001 and
2006 have been corporate issues. However as reported in
McKenzie (2010), most sukuk issuers are government or
quasi-government organizations in 2009. In addition to
that, he found that the most important market for corporate
sukuk issues totaling US$44 billion over the period of 2001–
2006 has been for infrastructure finance, with issuance
of US$17billion, 39 percent of the total. The next largest
markets were for financial services (18 percent) and energy
(6 percent). McKenzie (2009) claims that there is a growing
appetite and demand for investment in sukuk that goes well
beyond Islamic investors amongst those investors wishing
to gain exposure to diverse but high quality assets. The IFSL
Islamic Finance (2012) reported that the sukuk issuance
has increased significantly in past three years to the level of
US$84.4billion in 2011. The amount of new sukuk issues in
last ten years is shown in plots in Figure 9.
According to International Islamic Financial Market
(IIFM) 2010 report, the sukuk market size grew to over
US$136billion in mid-2009. As at the end of 2010, the total
asset value of publicly-listed sukuk was more than US$
197 billion in 13 markets (Alvi, et al., 2011). Mass media
reports suggest a much higher asset value by including
9
Ariff and Safari
90.0
80.0
70.0
60.0
50.0
40.0
30.0
20.0
10.0
0.0
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011
Figure 9. Amount of worldwide Sukuk issues (Source:
IFSL, Zawya Sukuk monitor).
privately-issued sukuk in a number major financial centers
such as Zurich, London, Frankfurt, Singapore and others.
Hence, a figure of US$ 840 billion is suggested in Ariff
et al. (2012). Currently, sukuk are offered in specialized
exchanges such as the Labuan Exchange in Malaysia, the
Third market in Vienna, the Dubai International Finance
Exchange, and the London Stock Exchange (Asad, 2009b).
Governments and regulators in a variety of countries have
recognized the important role that sukuk can play in capital
markets and have been giving priority to developing their
countries as sukuk centers (Abd Razak and Abdul Karim,
2008).
3%
3%
2%
3%
Malaysia
4%
Qatar
UAE
5%
Indonesia
11%
S. Arabia
Bahrain
69%
Pakistan
Others
Figure 10. Sukuk issuance by country, 2011 (Source:
IFSL, Zawya Sukuk monitor).
6. Sukuk regulations
All Islamic financial and banking interactions, similar to
their conventional counterparts, are subject to regulations
of professional authorities. Moreover, in order to be
recognized as an “Islamic” transaction, sukuk products
must follow some extra procedures required by Islamic
authorities (Iqbal, 1999). Shariah regulations governing
sukuk, similar to other Islamic finance and banking
practices in general, are dictated from three sources;
international organizations, local authorities, or in-house
Shariah boards.
Some major Islamic financial institutions have their
own in-house Shariah Supervisory Boards (SSB). This is
10
mostly practiced among Islamic investment companies
and banks. In the process of issuing sukuk, according to
Liquidity Management Center of Bahrain (2008), Shariah
advisors should study proposed sukuk structures and
suggest a Shariah structure which fulfills the set economic
aims sought in the issue. In addition, they should work
closely with legal counsel of the issuer and the arranger
(investment banker) to ensure that the legal documents
are in line with Shariah requirements. Finally, they should
Issue a fatwa, which is an opinion, as to the compliance of
the sukuk with Islamic principles—similar to the opinion
tax authorities provide on a tax question—on the whole
sukuk deal before it can be put into circulation.
Islamic financial institutions are subject to rules and
regulations of the local Shariah authority. Some countries
such as Malaysia have set up their own Shariah Advisory
Council (SAC) at the national level, which oversees the
consistent application across similar situations in financial
interactions. These councils are part of the Securities
Commission or also part of the Central Bank of the country
at national level. At the international level (the Organization
of Islamic Countries) there is a Shariah Council in Saudi
Arabia.
In a global perspective, there are few international
organizations that attempt to regulate and screen the
conduct of sukuk issuance and trade. Among these
international organizations, AAOIFI, IFSB, and IIFM are
the most influential ones (DIFC, 2009). Although these
organizations try to base their rulings on Shariah principles,
there are occasions that sukuk based on their guidelines have
variations in formation. Not being national bodies, these
rules cannot be imposed, so remain voluntary. Therefore,
Siddiqui (2008) highlighted that more communication
between these organizations will bridge the differences
existing between sukuk contracts.
The decisions regarding permissibility of each sukuk
contract, as mentioned above, is made by expert Muslim
scholars who are appointed to the Shariah Board. Number
of these experts are estimated to fall between 100 and
200, worldwide (Asad, 2009a). This indicates the urgent
need for training of expert Muslim scholars to sit on
Shariah boards in Islamic financial institutions. In order
to address this shortfall, some Islamic institutes such as
ISRA3 and INCEIF4 are planning to design special programs
for training certified Shariah scholars. However, there are
some other institutions that are currently providing short
time courses on this topic.5
AAOIFI
The Accounting and Auditing Organization for Islamic
Financial Institutions (AAOIFI) in its website6 introduces
itself as “an Islamic international autonomous non-profit
corporate body that prepares accounting, auditing,
governance, ethical and Shariah standards for Islamic
financial institutions and the industry.” AAOIFI was
established in accordance with the Agreement of
Association, which was signed by Islamic financial
institutions in 1990 in Algiers. Then, it was registered
in 1991 in Bahrain. As an independent international
organization, AAOIFI is supported by institutional
members (more than 200 members from 45 countries).
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Valuation of Islamic debt instruments, the Sukuk: Lessons for market development
Members include central banks, Islamic financial
institutions, and other participants from the international
Islamic banking and finance industry, worldwide.
Esta­
blishment of AAOIFI represented a shift from the
authority of independent Shariah supervisory boards set
up in individual Islamic banking and finance operations,
to a centralized model for the dissemination of standards,
procedures and best practices (Gambling et al., 1993;
Maurer, 2002, Pomeranz, 1997)
In May 2003, AAOIFI issued Shariah standard FAS 17
titled “Investment sukuk.” In this standard, AAOIFI issued
standard for 14 different types of sukuk, where some of these
sukuk are classified as tradable and others are classified
as non-tradable based on the type and characteristics
of the issued sukuk (AlBuolayan, 2006). AAOIFI and
its Shariah Board chairman, Shaikh Muhammad Taqi
Usmani, have special attention to sukuk because it is one
of the most favored Islamic financing instruments and
there is a possibility of variation in contract formation by
practitioner. Shaikh Usmani has given some comments
on the practice of sukuk over time. For instance, he has
highlighted that “… since Ijarah sukuk represent the pro
rata ownership of their holders in the tangible assets of the
fund, and not the liquid amounts or debts, they are fully
negotiable and can be sold and purchased in the secondary
market. Anyone who purchases these sukuk replaces the
sellers in the pro rata ownership of the relevant assets and
all the rights and obligations of the original subscriber are
passed on to him. The price of these sukuk will be determined
on the basis of market forces, and are normally based on their
profitability” (Usmani, 2001).
However, his most cited and debated comment regarding
sukuk is his statement in November 2007 which declared
that some 85 percent of outstanding sukuk had failed the
Shariah-compliance test on the basis that they were ‘assetbased’ rather than ‘asset-backed’ with the guaranteed
return of the face value of the sukuk on maturity and in the
absence of a transfer in asset ownership to sukuk holders
(Usmani, 2007). Since then, the juridical validity of sukuk
became suspect (Hasan, 2010, Alsayyed and Malik, 2010).
His reasons for such declaration were in brief as under.
• There have been cases where the assets in the sukuk
were the shares of companies that do not confer true
ownership but which merely offer to sukuk holders a
right to returns.
•Most sukuk issued are identical to conventional
bonds with regard to the distribution of profits from
their enterprises at fixed percentage bench-marked
on interest rates. The legal presumption regarding
sukuk is that no fixed rate of profit or the refund of
capital can be guaranteed.
•Virtually, all sukuk issues guarantee the return
of the principal to holders at maturity (just as
in conventional bonds) through a binding promise
from either the issuer or the manager to repurchase
the assets at the stated price regardless of their true
or market value at maturity.
Later on, in February 2008, AAOIFI issued a guidance
statement on accounting for investments and amendment
in FAS 17 (AAOIFI, 2008). Summary of important issues
raised in this guideline are:
Eds. Hatem A. El-Karanshawy et al.
a) Sukuk issuances have to be backed by real assets, the
ownership of which has to be legally transferred to
sukuk holders in order to be tradable
b) Sukuk must not represent receivables or debts, except
in the case of a trading or financial entity selling
all its assets or a portfolio with a standing financial
obligation, in which, some debts owing by third
parties, incidental to physical assets or usufruct, are
unintentionally included
c) The manager of the sukuk is prohibited from extending
“loans” to make up for the shortfall in the return on
the assets, whether acting as a mudarib (investment
manager), or sharik (partner) or wakil (agent)
d) Guarantees to repurchase the assets at nominal value
upon maturity with the exception of Ijarah sukuk
structures are also prohibited
e)Closer scrutiny of documentation and subsequent
execution of the transaction is required by Shariah
Supervisory Boards
Maurer (2010) investigated this controversial issue
and concluded that for some, Usmani’s salvo was a long
overdue and much needed corrective to what they saw as
the excesses of sukuk issuances and structured financing
vehicles that came very close to mimicking conventional
bonds. To others, it was an overreaction, born of impatience
with the pace of development of Islamic financial
institutions and markets, and an unrealistic appraisal of
what Islamic finance can actually accomplish in a globally
interconnected and interdependent world.
IFSB
The Islamic Financial Services Board (IFSB), which is based
in Kuala Lumpur, was established in 2002 and started
operations in early 2003. It serves as an international
standard-setting body of regulatory and supervisory
agencies that have vested interest in ensuring the soundness
and stability of the Islamic financial services industry, which
is defined broadly to include banking, capital market and
insurance. In advancing this mission, the IFSB promotes
the development of a prudent and transparent Islamic
financial services industry through introducing new or
adapting existing international standards consistent with
Shariah principles and recommend them for adoption.
To this end, the work of the IFSB complements that of the
Basel Committee on Banking Supervision, International
Organization of Securities Commissions and the
International Association of Insurance Supervisors.
Drafting of standards in IFSB is done on a task force working
group method. The IFSB council appoints members of
technical committee which are responsible for advising the
council on technical issues within its terms of reference.
Islamic Development Bank (IDB) Shariah Supervisory
Board is responsible for Shariah supervision of IFSB’s
standards. IFSB has issued three standards that affect the
issuance, trading, or investing in sukuk:
• IFSB-1: Guiding principles of risk management
for institutions (other than insurance institutions)
offering only Islamic financial services, issued in
December 2005. This guideline also includes the
various risk elements affecting institutions offering
investment certificates such as sukuk and operational
11
Ariff and Safari
consideration regarding them. This guideline offers
a general perspective toward risk sources and risk
management and is not specific for sukuk.
• IFSB-2: Capital adequacy standard for institutions
(other than insurance institutions) offering only
Islamic financial services, issued in December 2005.
This standard overviews various Islamic contracts
(some of which are underlying contracts of sukuk)
and provide capital requirement for each.
• IFSB-7: Capital adequacy requirements for sukuk,
securitizations and real estate investment, issued
in January 2009. The first part of this guideline
investigates the sukuk and securitization of it, sukuk
structures, operational requirements pertaining to
sukuk, treatment for regulatory capital purposes of
sukuk and securitization exposures, and treatment of
credit risk exposures of sukuk.
IIFM
The International Islamic Financial Market (IIFM), located
in Bahrain, is a global standardization body for the
Islamic capital and money market segment of the Islamic
financial services industry. Its primary focus lies in the
standardization of Islamic products, documentation and
related processes. IIFM was founded with the collective
efforts of Central Bank of Bahrain, Bank Indonesia, Central
Bank of Sudan, Labuan Financial Services Authority
(Malaysia), Ministry of Finance (Brunei Darussalam) and
Islamic Development Bank (a multilateral institution based
in Saudi Arabia). Besides the founding members, IIFM is
supported by its permanent members, namely State Bank
of Pakistan and Dubai International Financial Centre
Authority (UAE). IIFM is further supported by a number
of regional and international financial institutions as well
as other market participants as its members. IIFM activities
are under supervision of its Shariah Advisory Panel,
which currently has ten members. Focus of IIFM’s work
is on Islamic capital and money markets. Presently, IIFM
has no specific standard or guideline pertaining to sukuk.
However, it has released two reports on sukuk in 2010
and 2011. In these reports, IIFM investigated the current
sukuk market from an international as well as domestic
perspective. It also investigated various sukuk structures
international issues. Moreover, it has studied some of sukuk
issues as case study.
Sukuk rating
Rating is an evaluation of a corporate or municipal
bond’s relative safety from an investment standpoint.
In conventional sense, it scrutinizes the issuer’s ability
to repay principal and make interest payments. Then, a
grade (the most controversial part of rating process) is
given to the bond that indicates its credit quality. Private
independent international rating companies such as
Standard & Poor’s, Moody’s, and Fitch, or domestic rating
agencies like RAM and MARC of Malaysia, provide these
evaluations of issuer’s financial strength, or the ability
to pay a bond’s principal and interest in a timely fashion.
As a result, bonds are rated in a range from AAA or Aaa
(the highest), to C or D, which represents a company that
has already defaulted. Each rating company has its own
definition and methodology for rating and own set of
rating ranges.
12
The introduction of sukuk rating in the 90 s represents
another critical milestone in the development of sukuk
market. Bond ratings are principally designed to arrive at
a reasoned judgment on credit risk via a careful analysis of
the critical issues surrounding a specific debt on the issuer
(Mohd Asri, 2004).
From the global capital markets point of view, sukuk can be
rated just like any conventional bond, and can be traded as
such, as well (Maurer, 2010). In general, rating agencies
have the same criteria for corporate bonds rating. The
criteria incorporate issue structure (repayment schedule
and debt types), business risk analysis, financial risk
analysis, management, ownership and other qualitative
factors (Mohd Asri, 2004). However, realizing the
uniqueness and types of sukuk, the rating methodology
should be different to that of conventional bonds rating
(Jalil, 2005). Rosly (2007) argued that sukuk structures
falls in 2 categories:
•Asset-Backed sukuk, for which ratings are dependent
on a risk analysis of the asset. However, investors
hold rights to underlying assets through SPV and
not directly; hence, sukuk performance is driven by
assets and not linked to the originator.
•Unsecured sukuk, for which ratings are primarily
dependent on the riskiness of the sponsor, originator,
or the borrower.
Therefore, similar to conventional bonds, risk elements
affecting sukuk should be thoroughly investigated by rating
agencies. Among risk elements, Rosly (2007) mentioned
that credit risk is the most critical one. Other factors he
highlighted are currency risk (for international issues),
tax risk, and reserve funds. In contrast to highly sensitive
conventional bonds, sukuk are less sensitive to interest
rate. Zurich based investment bank Credit Suisse believes
investment in Islamic Finance and Banking products
are not expose to interest rates since Islam prohibits
charging interest and sukuk securities are unaffected to
the credit crisis in the international finance and banking
industry (Farook, 2009). Tariq (2004) summarized risk
characteristics of each type of sukuk structures which is
depicted in Table 4.
Security commission
Security Commissions in each country is a statutory body
that investigates the conduct of financial markets and has
enforcement power. Among all of their responsibilities,
some are related to origination and exchange of sukuk
securities. These responsibilities are supervising
exchanges, clearing houses, and central depositors;
approving authority for corporate bond issues (including
sukuk securities); and regulating all matters relating to
securities and futures contracts. Hence, one may assume
that the all sukuk securities issued require permission
and approval by Security Commission of that particular
jurisdiction. Beyond the regulatory obligations, they also
act as a strategic policy making body on financial markets.
Among these incentives,7 some directly affect the issuance
of sukuk securities in Malaysia. Issuers of sukuk securities
in Malaysia are offered tax deduction on expenses incurred
in due course of issuance of Wakalah, Murabahah, Bai
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Eds. Hatem A. El-Karanshawy et al.
Salam Sukuk
Musharakah
Term Finance
Sukuk (MTFS)
Fixed rate Hybrid/
Pooled Sukuk
Floating
Rate Ijarah Sukuk
Very high due to
fixed rate,
remains for the
entire maturity
of the issue
Similar to floating
rate, but, unique
because the rate
is not indexed
with a benchmark
like LIBOR,
hence least
exposed to
this risk
Very high due
to fixed rate
Default on rent
payment, floating
rate makes default
risk lesser serious –
see previous case
Credit risk of debt
part of pool,
default on rents,
fixed rate makes
credit risk serious
Musharakah has
high default risk
(Khan and Ahmed,
2001), however,
MTFS could be
based on the
strength of the
entire balance sheet
Unique credit risk
(Khan and Ahmed,
2001)
Securitized salam,
fixed-rate and
non-tradable
Default on rent
payment, fixed
rate makes credit
risk more serious
Securitized Ijarah,
certificate holder
owns part of asset
or usufructs and earns
fixed rent – tradable
Certificate holder
owns part of asset
or usufructs and
earns floating rent
indexed to market
benchmark such as
LIBOR – tradable
Securitized pool of
assets; debts must
not be more than
49%, floating rate
possibility exists –
tradable
Medium term
redeemable
Musharakah
certificate based
on diminishing
Musharakah –
tradable as well
as redeemable
Fixed Rate Ijarah
Sukuk
Unique basis of
credit risks exist,
(Khan and
Ahmed, 2001)
Rate of return
(Interest rate risk)
Very high due to
fixed rate,
remains for the
entire maturity
of the issue
Very high due to
fixed rate,
remains for the
entire maturity
of the issue
Exists only within
the time of the
floating period
normally
6 months
Istisna, Murabahah
debt certificates –
non-tradable
Description of Sukuk
structure
Credit Risk
Zero coupon Sukuk
Type of Sukuk
Table 4. Summary of risk characteristics of sukuk structures – source: (Tariq, 2004).
If all other conditions
are similar, FX risk
will be the same for
all cases of Sukuk.
However, those
Sukuk which are
liquid or which are
relatively short term
in nature will be
less exposed.
The composition
of assets in the pool
will also contribute
to the FX risk in
different ways.
Hence this can be
very useful tool to
overcome the FX
risk by diversifying
the pool in different
currencies.
FX risk
Other risks
Liquidity Risk is serious
as far as the non-tradable
Sukuk are concerned.
Business risk of the
issuer is a risk underlying
Sukuk as compared to
traditional fixed incomes.
Shariah compliance risk
is another one unique in
case of Sukuk.
Infrastructure rigidities,
lack of efficient
institutional support
increases the risk of
Sukuk as compared
to traditional fixed
incomes, (Sundararajan
and Errico, 2002)
Price risk
Related to the
underlying
commodities prices
and assets in relation
to the market prices.
Ijarah Sukuk is most
exposed to this as
the values of the
underlying assets
may depreciate
faster as compared
to market prices.
Maintenance of the
assets will play an
important part in this
process. Liquidity
of the Sukuk will also
play an important part
in the risk. Salam is
also exposed to price
risks. However,
through parallel
contracts these risks
can be overcome
Valuation of Islamic debt instruments, the Sukuk: Lessons for market development
13
Ariff and Safari
Bithaman Ajil, Musharakah, Mudarabah, Ijarah, and Istisna
sukuk securities until 2015. Moreover, the Malaysian
Security Commission (SC) has recognized SPV as a
channel to transfer funds and hence, exempted them from
income taxes. In addition to that, the issuer is permitted to
tax deduct the cost of issuance of sukuk incurred by SPV
Company. On the other hand, sukuk investors are receiving
some incentives such as tax exemption on profits earned
from investing in sukuk securities (except for the profits
due to convertible loan stocks).
To add to the sukuk market flavor, SC offers intermediaries
income tax exemption for qualified institutions “in respect
of statutory income derived from regulated activity of dealing
in securities and advising on corporate finance relating to
arranging, underwriting and distributing non-ringgit sukuk
originating from Malaysia which are issued or guaranteed
by the Government or approved by the SC until the year of
assessment 2014.”
7. Institutional developments
Similar to conventional practices of accounting, there is
no unique standard for Islamic accounting and finance.
However, in a similar fashion to conventional counterparts,
there are attempts to develop one. AAOIFI, which gained
respect of many Islamic countries and institutions, has
developed standards for accounting statements. In the
section 6.2 of the “Statement of Financial Accounting,”
AAOIFI states that the following is the main focus of
financial accounting in Islam: “on the fair reporting of the
entity’s financial position and results of its operations, in a
manner that would reveal what is halal (permissible) and
haram (forbidden). Moreover, they highlighted one of the
key objectives of financial reports: Information about the
Islamic bank’s compliance with the Islamic Shariah and its
objectives and to establish such compliance; and information
establishing the separation of prohibited earnings and
expenditures, if any, which occurred, and of the manner in
which these were disposed of.”
Such an objective is not merely the same as the objectives
set by the IASB (International Accounting Standards
Board) in IFRS (International Financial Reporting
Standards). In the 2001 “Framework for the Preparation
and Presentation of Financial Statements,” IASB
mentioned that the objective of financial statements
is to “provide information about the financial position,
performance and changes in financial position of an entity
that is useful to a wide range of users in making economic
decisions.” Hence, one may conclude that these two
objectives are not necessarily the same.
Amin (2011) examined the implication of such difference
in objectives and showed that there would be an actual
difference in outcomes. He claimed that due to the
difference in standards, particularly with respect to the
presence of SPV in the structuring of sukuk securities,
there would be a different outcome in financial reports.
He argues that the root cause of such difference lies in the
different interpretation of the role of the SPV, especially the
fact that AAOIFI requires the ownership transfer of assets
to it, while the IFRS considers sukuk transaction purely as a
financing transaction and does not require the transferring
of title of assets in the balance sheet statement.
14
Amin (2011) suggests that as there are more attempts to
develop an internationally accepted accounting standards
by IASB, Islamic accounting should follow their lead
and focus more on requiring such standards to adhere to
the demands of the Muslim world. He suggests that IFRS
may include Islamic requirements in the standards by
mandating practitioners to provide such information as
footnote disclosures or in other formats.
Former issues indicate that there are potential conflicts
between the outcome of following AAOIFI standards
and IFRS standards in practice. Hence, firms may face a
dilemma in adopting the proper accounting standards.
Since issuance of sukuk securities is one of the many
accounting and finance practices of firms, they may
generally follow the IFRS standards, which are more
comprehensive. In addition to that, audit firms may prefer
conducting audit service for firms who follow a more
acceptable standard (IFRS) rather than a less common
one (AAOIFI). This might result in higher fees for auditing
firms that adopt AAOIFI standards and make them less
favored. Moreover, majority of accountant practitioners
are educated with IAS methods and standards. The
number of well-versed accountants in Islamic regulations
is still lower than the demand; hence, many firms ought to
hire conventional accountants.
These issues have hindered the acceptance and adoption
of AAOIFI standards by many firms. Therefore, similar
to Amin’s (2011) recommendation, we suggest that
the AAOIFI should focus more on collaboration with
international standard setting bodies and negotiate with
them to incorporate and embed the demands of Muslim
practitioners as well as investors in the internationally
recognized standards such as IFRS.
8. Future developments
Pro-market regulatory framework is a key feature for longterm growth and sustainability of any financial market
development. Presently, the contracts are arranged and
organized individually and hence, the cost of issuance
of sukuk securities is higher compared to conventional
bonds. Lack of standardized sukuk contracts results in
similar costly structuring of contracts. Standardization of
contracts strengthens the convenience of securitization
process, and will boost the cooperation among regulatory
bodies. This will lead to one-stop servicing of primary
market making, thus lessening the burdens of issuance
process.
Market deepening activities should be undertaken to
improve the lack of liquidity in the market. Majority of
sukuk markets suffers from illiquidity to the extent that 70
percent or more are not traded as at 2011 has never been
publicly traded since the issuance until their maturity.
Introduction of discount houses as the intermediary
between the exchange and the custormers would improve
liquidity.
Market broadening activities are also required to further
develop sukuk markets. Introducing more securities
targeting at specific needs of customers would help
introduce a broader list of securities. For example, the
sukuk market has 6 traded ones although conceptually
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Valuation of Islamic debt instruments, the Sukuk: Lessons for market development
there are 14 different types. These securities, though
combination of two or more types of existing contracts into
one structure, would result in new a security to address the
unique needs of the issuer. More efforts should be put into
financial engineering of sukuk securities.
Finally, education is another critical area that needs more
attention in order to ensure the continuity and sustainability
of sukuk markets. Currently, education and training efforts
across the world is highly un-coordinated. Standardization
of education and training on Islamic finance in general
and in particularly sukuk securities is absent. There is no
accreditation for different programs offered to the public.
Although an international accreditation body for education
programs is ideal, national accrediting bodies may serve
the need for the time being. Existence of accreditation for
Islamic finance programs may not only ensure the quality
of the programs, but also is vehicle for standardization of
these programs.
Although some limited number institutions (like INCEIF in
Malaysia) offer specialized programs, majority of universities
do not offer Islamic finance programs. Underlying reasons
for shortage should be further investigated. Some potential
reasons might be lack of enough syllabuses, or even lack
of demand for such programs. However, universities may
collaborate and share their resources to develop “market
intelligence” database. Such database can be used to
identify the potential programs, courses and syllabus. For
instance, especial programs may be developed for training
of Shariah board members.
Notes
1 During the 3rd century AD, financial firms in Persia
(currently known as Iran) and other territories in
the Persian Sassanid Dynasty issued letters of credit
known as “chak.” KHARAZMI, A. A.-A. M. B. A. (1895)
Maf t h al-Ool m. IN VLOTEN, G. V. (Ed. Leiden. In postIslamic Arabic documents this word has transformed
into “sakk” FLOOR, W. (1990) AK (legal document,
testament, money draft, check). IN YARSHATER, E.
(Ed. Encyclopedia Iranica Online Edition. New York.,
Columbia University.
2 Data is collected from Bondstream database, a product
of Bond Pricing Agency Malaysia. Sample is limited
to Malaysia due to unavailability of parallel markets
for sukuk and bonds in other countries. Moreover,
Malaysia accommodate almost two third of world
sukuk issues.
3International Shariah Research Academy for Islamic
Finance, based in Kuala Lumpur, Malaysia.
4 International Centre for Education in Islamic Finance,
based in Kuala Lumpur, Malaysia.
5 Besides the ISRA and INCEIF which offer long term
programs, we can mention, among all, Dr. Kahf
Institute, REDmoney Islamic Finance training institute,
and UK Islamic Finance Council.
6http://www.aaoifi.com/aaoifi/TheOrganization/Over­
view/tabid/62/language/en-US/Default.aspx accessed
on 9/4/2013.
7 For complete list of such incentives refer to the website
of Security Commission of Malaysia http://www.
sc.com.my/main.asp?pageid=593&menuid=322&ne
wsid=&linkid=&type=(Accessed on 11/4/2013)
Eds. Hatem A. El-Karanshawy et al.
References
AAOIFI (2004) Shariah Standards for Financial Institutions.
Accounting and Auditing Organization for Islamic
Financial Institutions (AAOIFI).
AAOIFI (2008) Guidance Statement on Accounting for
Investments and Amendment in FAS 17. Accounting and
Auditing Organization for Islamic Financial Institutions
(AAOIFI).
Abd Razak, D. & Abdul Karim, M.A. (2008) Development
of Islamic Finance in Malaysia: A Conceptual Paper. 8th
Global Conference on Business & Economics. Florence, Italy.
Abdul Majid, A.R. (2003) Development of Liquidity
Management
Instruments:
Challenges
and
Opportunities. International Conference on Islamic
Banking: Risk Management, Regulation and Supervision.
Jakarta –Indonesia.
Al-Jarhi, M.A. & Iqbal, M. (2001) Islamic Banking: Answers
to Some Frequently Asked Questions Jeddah, Saudi
Arabia, Islamic Research and Training Institute, Islamic
Development Bank.
Al-Omar, F. & Abdel-Haq, M. (1996) Islamic Banking: Theory,
Practice & Challenges, Karachi, Oxford University Press.
Albuolayan, A. (2006) Rapid Surge in Sukuk Market.
ArabNews, Jeddah, Saudi Arabia. Jeddah, Saudi Arabia,
Saudi Research & Publishing Co.
Alsayyed, N.A. & Malik, F. (2010) Sukukization: Islamic
Economic Risk Factors in Shariah View. Working Paper
of International Shari’ah Research Academy for Islamic
Finance (ISRA). Kuala Lumpur, Malaysia, International
Shari’ah Research Academy for Islamic Finance (ISRA).
Alvi, I.A., Mohammed, A.R., Khan, G.Z., Nasser, U.M., Naseer,
B. & Khan, M.S. (2010) Sukuk Report, A Comprehensive
Study of The International Sukuk Market. In Alvi, I.A. (Ed.
International Islamic financial market. 1 ed. Manama,
Bahrain, International Islamic financial market.
Alvi, I.A., Mohammed, A.R. & Naseer, U.M. (2011) Sukuk
Report, A Comprehensive Study of The International
Sukuk Market. In Alvi, I. A. (Ed. International Islamic
Financial Market. 2 ed. Manama, Bahrain, IIFM.
Amin, M. (2011) Accounting for sukuk under IFRS and
AAOIFI accounting standards. In Jaffer, S. (Ed.) Global
Growth, Opportunities and Challenges in the Sukuk
Market. Euromoney Trading Inc.
Asad, S.I.-U.-D. (2009a) The business of shariah advice The
DAWN newspaper, Karachi, Pakistan. Karachi, Pakistan,
Pakistan Herald Publication (Pvt) Limited.
Asad, S.I.-U.-D. (2009b) An overview of the sukuk market.
Jang Newspaper. Karachi, Pakistan, Jang Group of
Newspapers.
Aseambankers (2005) Capitalising on Opportunities in the
Sukuk Industry. Aseambankers Malaysia Berhad. Kuala
Lumpur, Aseambankers.
Bose, S., SR. & McGee, R.W. (2008) Islamic Investment
Funds: An Analysis of Risks and Returns. SSRN eLibrary.
Chapra, M.U. (1998) The Major Modes of Islamic Finance. 6th
Intensive Orientation Course on “Islamic Economics, Banking
and Finance.” the Islamic Foundation, Leicester, U.K.
15
Ariff and Safari
Daryanani, N. (2008) A Deeper Understanding on
the Prohibition of Riba. Management. Nottingham,
University of Nottingham, UK.
DIFC (2009) Dubai International Financial Centre Sukuk
Guidebook, Dubai, UAE.
El-Gamal, M.A. (2000) A Basic Guide to Contemporary
Islamic Banking and Finance. Plainfield, Indiana, Islamic
Society of North America.
El-Gamal, M.A. (2007) Mutuality as an antidote to rentseeking Shariah arbitrage in Islamic finance. Thunderbird
International Business Review, 49, 187–202.
Enders, W. (1995) Applied Econometrics Time Series,
New York, John. Wiley & Sons.
Farook, R. (2009) Global Financial Crisis Unthinkable
Under Islamic Banking Principles. Sunday Observer. Sri
Lanka, The Associated Newspapers of Ceylon Ltd.
Floor, W. (1990) AK (legal document, testament, money
draft, check). In Yarshater, E. (Ed. Encyclopedia Iranica
Online Edition. New York., Columbia University.
Gait, A.H. & Worthington, A.C. (2007) A Primer on Islamic
Finance: Definitions, Sources, Principles and Methods.
Working Papers Series of University of Wollongong.
Wollongong, Australia, University of Wollongong.
Gambling, T., Jones, R. & Karim, R.A.A. (1993) Credible
Organizations: Self-Regulation V. External StandardSetting in Islamic Banks and British Charities. Financial
Accountability & Management, 9, 195–207.
Granger, C.W.J. (1969) Investigating Causal Relations
by Econometric Models and Cross-spectral Methods.
Econometrica, 37, 424–438.
Gurgey, U. & Keki, E. (2008) Sukuk in Turkey. International
Financial Law Review, 27, 111–111.
Hasan, Z. (2010) Islamic Finance: What Does It Change,
What It Does Not – The Structure – Objectives Mismatch
and Its Consequences. Working Paper. Kuala Lumpur,
Malaysia, International Centre for Education in Islamic
Finance (INCEIF).
Hossain, A. (2005) Granger-Causality Between Inflation,
Money Growth, Currency Devaluation and Economic
Growth in Indonesia, 1951–2002. International Journal
of Applied Econometrics and Quantitative Studies, 2, 23.
IFSB (2009) Capital Adequacy Requirements for Sukuk,
Securitizations and Real Estate Investment. In Board,
I.F.S. (Ed. IFSB. Kuala Lumpur, Malaysia, Islamic
Financial Services Board.
IFSL (2012) Islamic Finance. IN Mckenzie, D. (Ed. Finaicial
Market Series. London, International Financial Services
London.
Iqbal, M. (1998) Islamic banking. In Kahf, M. (Ed.) Lessons
in Islamic Economics. Jeddah, Islamic Research and
Training Institute.
Iqbal, M. & Molyneux, P. (2005) Thirty years of Islamic
banking: History, Performance and Prospects,
Houndmills, Basingstoke, Hampshire; New York,
Palgrave Macmillan.
16
Iqbal, Z. (1999) Financial engineering in Islamic finance.
Thunderbird International Business Review, 41, 541–559.
Jalil, A. (2005) Islamic Bonds Issues: The Malaysian
Experience. In T. Ramanayanah, Mohamed Sulaiman,
Hasnah Harun, Ruhaini Ali, Aizzat Mohd. Nasurdin,
Nabsiah Abdul Wahid & Intan Osman (Eds.) The
6th Asian Academy of Management Conference, 9–11
December 2005. Casuarina Ipoh, Perak, Malaysia, Asian
Academy of Management.
John, I. (2007) Sukuks key to meeting $1.5tr infrastructure
needs of ME, Asia. Khaleej Times. Dubai, Galadari
Printing and Publishing.
Kamali, M.H. (2007) A Shari’ah Analysis of Issues in Islamic
Leasing. J.KAU: Islamic Econ, 20, 3–22.
Khan, M.M. & Bhatti, M. I. (2008) Development in Islamic
banking: a financial risk-allocation approach. The
Journal of Risk Finance, 9, 40–51.
Khan, T. & Ahmed, H. (2001) Risk Management: An Analysis
of Issues in the Islamic Financial Industry, Jeddah, Saudi
Arabia, Islamic Development Bank – Islamic Research
and Training Institute.
Kharazmi, A.A.-A.M.B.A. (1895) Maf t h al-Ool m. In
Vloten, G.V. (Ed.) Leiden.
Lewis, M.K. (2007) Islamic Banking in Theory and Practice.
Monash Business Review, 3, 1–8.
Lewis, M.K. & Algaoud, L.M. (2001) Islamic banking,
Cheltenham, UK, Edward Elgar.
Liquidity Management Center (2008) The Guide to Sukuk
Market. Bahrain, November 2008.
Maurer, B. (2002) Anthropological and Accounting
Knowledge in Islamic Banking and Finance:
Rethinking Critical Accounts. The Journal of the Royal
Anthropological Institute, 8, 645–667.
Maurer, B. (2010) Form versus substance: AAOIFI projects
and Islamic fundamentals in the case of sukuk. Journal
of Islamic Accounting and Business Research, 1, 32–41.
McKenzie, D. (2008) Islamic Finance 2008. International
Financial Services London, Islamic Finance Working
Group, 8.
McKenzie, D. (2009) Islamic Finance 2009. International
Financial Services London (IFSL), 8.
McKenzie, D. (2010) Islamic Finance 2010. International
Financial Services London (IFSL), 8.
Metwally, M.M. (2006) Economic Consequences of
Applying Islamic Principles in Muslim Societies. Journal
of Islamic Banking and Finance, 23, 11–33.
Mohd Asri, N. (2004) The Effect of Islamic Private Debt
Securities Rating Changes on Firm’s Common Stock
Returns. The Journal of Muamalat and Islamic Finance
Research, 1, 25–38.
Pomeranz, F. (1997) The Accounting and Auditing
Organization for Islamic Financial Institutions: An
Important Regulatory Debut. International Journal of
Accounting, Auditing and Taxation, 6, 123–30.
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Valuation of Islamic debt instruments, the Sukuk: Lessons for market development
Rosly, S.A. (2007) Islamic Capital Market: Shariah Stocks,
Sukuk, i-Reits, Islamic unit trust funds. LOFSA-INCEIF
Workshop on Islamic Finance, 14th-15thNovember 2007,
Labuan, Malaysia. Labuan, Malaysia, International
Center For Education in Islamic Finance (INCEIF).
Shinsuke, N. (2007) Beyond the Theoretical Dichotomy in
Islamic Finance: Analytical Reflections on Mur ba ah
Contracts and Islamic Debt Securities. Kyoto Bulletin of
Islamic Area Studies,, 1, 72–91.
Siddiqui, R. (2008) Contributing to the Development of the
Islamic Capital Market: The Dow Jones Citigroup® Sukuk
Index. Dow Jones Islamic Market Indexes NewsLetter,
December, 1–2.
Sundararajan, V. & Errico, L. (2002) Islamic financial
institutions and products in the global financial system:
Key issues in risk management and challenges ahead. Risk
Management in an Islamic Financial System, September 1,
2002, Tehran, Iran. Tehran, Iran Iran Banking Institute,
Central Bank of the Islamic Republic of Iran.
Tariq, A.A. (2004) Managing Financial Risks of Sukuk
Structures. School of Business and Economics.
Leicestershire, UK, Loughborough University.
Usmani, M.T. (2001) Principles Of Shari’ah Governing
Islamic Investment Funds.
Usmani, M.T. (2002) An Introduction to Islamic Finance,
The Hague, Kluwer Law International.
Eds. Hatem A. El-Karanshawy et al.
Usmani, M.T. (2007) Sukuk and their Contemporary
Applications. Accounting and Auditing Organization for
Islamic Financial Institutions (AAOIFI), 14.
Warde, I. (2000) Islamic Finance in the Global Economy,
Edinburgh Edinburgh University Press.
Wilson, R. (2004) Overview of the sukuk market. In Adam,
N.J. & Thomas, A. (Eds.) Islamic Bonds: Your Guide to
Issuing, Structuring and Investing in Sukuk. University
of Durham, United Kingdom, Euromoney Institutional
Investor PLC.
Wilson, R. (2008) Innovation in the structuring of Islamic
sukuk securities. Humanomics, 24, 170–181.
Zaher, T.S. & Hassan, M.K. (2001) A Comparative Literature
Survey of Islamic Finance and Banking. Financial
Markets, Institutions & Instruments, 10, 155–199.
Acknowledgment
An earlier version of this paper was presented at the 23rd
Global Finance Conference, Chicago, USA in May 2012
at De Paul University. Authors would like to appreciate
the feedbacks received from reviewers and participants of
that conference. This research was funded by Khazanah
Nasional Bhd., Malaysia, and the Maybank Endowed
chair professorship at the University Putra Malaysia. The
remaining errors are solely those of the authors.
17
The impact of Islamic debt on company value
Fitriya Fauzi1, Stuart Locke2, Abdul Basyith, Muhammad Idris
Faculty of Economics, University of Muhammadiyah, Palembang, Indonesia, Email: [email protected]
Waikato Management School, The University Waikato
1
2
Abstract - This study uses micro-econometric analysis to examine the impact of Islamic debt on
firm value and firm financial performance by observing Malaysian firms. A number of significant
contributions to corporate finance arise from this research in relation to Islamic debt instruments
and firm financial performance. First, it provides evidence of the Islamic debt impact on firm value
and firm financial performance. Second, and very importantly it provides new insights, adding
substantially to the very few studies that have been conducted on these types of instruments.
The choice of model employed is specified according to its diagnostic testing results for nonnormality, heteroskedasticity, multicollinearity, endogeneity and linearity in. A test is conducted
to confirm that there are no outliers in the data set prior to the diagnostic testing. Poolability and
co-integration testing are also included. Based on the diagnostic results, data are analysed using
the dynamic panel generalised method of moment (GMM using a quarterly balanced panel of 80
Malaysian firms issuing Islamic debt which spans from 2000 to 2009. This method is employed to
investigate the impact of Islamic debt issues on firm value and/or firm financial performance.
The result reveals that Islamic debt has a significant positive impact on company value and firm
financial performance. It also confirms that trade-off theory holds well in the Malaysian context for
Islamic debt financing. Furthermore, the coefficient for Islamic debt is higher than the coefficient
for non-Islamic debt, suggesting that the Islamic debt provides a higher contribution to firm value
and to the improvement of firms’ financial performance compared to non-Islamic debt.
Keywords: Islamic debt, firm value
1. Introduction
Modes of financing are a part of capital structure, and the
relationship between capital structure and firm value is
important and continues to be debated in the literature.
An unleveraged firm can be seen as an all equity firm,
whereas a leveraged firm is made up of ownership equity
and debt. A firm’s debt to equity ratio provides a measure
of the leverage or gearing. The influence of debt to equity
on the value of a firm is the subject of multiple articles,
including the seminal work of Modigliani and Miller
(M&M). Modigliani and Miller achieved notoriety with
their proof that capital structure made no different to firm
value. Subsequent relaxation of the M&M assumptions
suggested that debt to equity choice does impact on firm
value. More recent research considering bankruptcy costs
has contributed to empirical estimates of optimal leverage.
The M&M argument is that, in a perfect market, how a firm
is financed is irrelevant to its value. However, even though
it is widely accepted that the world is not made up of perfect
markets, the issue of optimal D/E ratio remains contentious
and unresolved.
Moreover, there are considerable studies that have
investigated the impact of conventional debt on firm
value, and those studies have generated few theories
which are implemented up to now. However, so far there
is no investigation of the financial performance and value
consequences of using Islamic debt as opposed to nonIslamic debt (conventional debt). The lack of prior research
into the financial aspect of Islamic debt does not reflect a
lack of concern for such matters, but rather is a reflection
of the newness of the topic. Therefore, the extent to which
prior research is applicable to Islamic debt, particularly in
emerging market, requires analysis, and this study is useful
to supplement existing studies in this field and serves as a
reference for studies in the future.
Cite this chapter as: Fauzi F, Locke S, Basyith A, Idris M (2015). The impact of Islamic debt on company value. In
H A El-Karanshawy et al. (Eds.), Islamic banking and finance – Essays on corporate finance, efficiency and product
development. Doha, Qatar: Bloomsbury Qatar Foundation
Developing Inclusive and Sustainable Economic and Financial Systems
Fauzi et al.
The literature relating to Islamic debts has predominantly
focused on the legal aspects of Islamic law, concept,
basic requirements and the validity of how the debts are
conducted in Islamic finance as general Islamic debt (Cakir,
2007; Mirakhor, 1996; Ashhari, 2009; Somolo, 2009;
Tariq, 2007; Wilson, 2008). So far, researchers have been
unable to find research that looks at the effects of Islamic
debts on the value of the company in international contexts.
Haneef (2009) discusses the history of Sukuk, explaining
how it has evolved from an asset backed structure, where
Sukuk holders have ownership rights over the underlying
asset, to an asset based structure, where Sukuk holders
rank paripassu with unsecured creditors. Other scholars
(Abd.Sukor, 2008; Al-Amine, 2001; Juan, 2008; Kamali,
2007; Mohd Yatim, 2009; Mokhtar, 2009; Al-Amine, n.d.;
Al Amine, 2008; Usmani, 1999, n.d.; Vishwanath, 2009;
Wilson, 2008, n.d.; Yean, n.d.) also discuss the structure
and the regulation of the Sukuk market in relation to
Shariah perspective and Shariah compliancy. Therefore,
this study attempts to examine the impact of Islamic debt
on company value.
The rest of this paper is organised as follows. Section two
presents the significance emergence of Islamic debt in the
fast growing form of financing in emerging and mature
markets. Section three provides literature review followed
by section four, five and six which present the methodology,
analysis and conclusion.
2. Significant emergence of Islamic debt
The Islamic financial and economic system has existed since
the time of the prophet Muhammad SAW. During that time,
buying and selling, and savings and loans activities were
not as extensive as they are now. However, the principle
remains the same; no interest charged and no non halal
products and activities permitted. The interest system is
not used at all because it is forbidden by Allah SWT. The
banning was declared in the Quran and the Hadith.
Islamic debt, known as Sukuk, has evolved to become a
significant part of corporate capital trading in the secondary
market. The Accounting and Auditing Organization of
Islamic Financial Institutions (AAOIFI) has also defined
Islamic debt as certificates of equal value representing
undivided shares in the ownership of tangible assets,
usufruct and services or (in the ownership of) assets of
the particular projects or any specified investment activity.
Investment of Sukuk should be distinguished from common
shares and bonds. While shares represent the ownership
of a company as a whole and are for an indefinite period,
Sukuk represent specified assets and are for a given period
of time. Sukuk, unlike bonds, carry returns based on cash
flow originating from the assets on the basis of which they
are issued (Ayub, 2007; p. 392).
Islamic debt includes no periodic interest payments and
provides a different cash flow profile when compared with
non-Islamic debt instruments for borrowing companies
and lenders. There is a socio-religious dimension relating
to major principles that underlie all business transactions
under Islamic law. All business transactions must adhere
the teaching of the Islamic foundation, which is the Quran
and Sunnah. There are at least four major prohibitions in
Islamic business transactions. The first is the prohibition of
20
riba, known as adding any interest payments to a loan or
other financing contract. The second is the prohibition from
gharar and maisir, known as uncertainty and gambling; so
transactions embodying these attributes will be considered
invalid. The third is the prohibition of non-halal business
transactions, such as alcohol, gambling and any other
things that are prohibited and considered as non-halal. The
fourth is the general prohibition of contracts that fail to
meet the highest Shariah standards (Ayub, 2007).
The development of Sukuk is supported by many factors,
including the development of Islamic banking (takaful)
and an increasing demand for Islamic products in the debt
market. The development of Sukuk with its associated
types of structure has given rise to much discussion and
debate among scholars of Islamic law. The uniqueness of
Islamic debt compared to non-Islamic debt is that Islamic
debt offers a secure investment based on the principle of
rent and profit sharing without legalised interest system.
It is constituted by pure motive of cooperation based on
Islamic law. How the market prices this security in term
of the yield curve and how risk pricing is embedded in the
value and performance of the firms.
Recent innovations in Islamic finance have changed the
dynamics of the Islamic finance industry, especially in
the debt markets. Sukuk became increasingly popular
as companies sought to raise funds by offering corporate
Sukuk. It has become significant for raising funds in the
international capital markets through Islamic Shariah.
Increases in this market have been strong all over the
world, especially in Malaysia, UAE and Saudi Arabia. In
1996 total Sukuk issued was USD 0.05b rising to USD
15.5b by the end of 2008. The most significant increase
occurred in 2007 with more than 130 issues valued at
USD 34.3b. The trend is apparent in Table 1 which shows
a rapid expansion by value through to the financial
crisis in 2008. Malaysia accounts for 43.7% of Sukuk
issues followed by UAE with 30.1% and Saudi Arabia
representing 10.4%. The size of offering by country for
2009 is shown in Table 2.
In the early years of Sukuk’s emergence as a financial
instrument, murabahah and istisna were the most
significant forms of issuance, accounting for 62.5% and
19.5% respectively. This changed between 2002 and 2007
when musyarakah and ijarah become the largest type of
issue, accounting for 36.3% and 28.3% of the total market.
In 2008 to 2009 the ranking reversed with ijarah and
musyarakah accounting for 43.4% and 20.8% respectively
as reflected in Table 3.
The increase in the issue size from year to year indicates that
this market was gradually developing. It became lucrative
for both the Sukuk issuer and the Sukuk holder, receiving
increased support in the form of market surveillance and
regulation, and from market participants.
The evolution of Sukuk structure is presented in Figure 1.
The evolving of the original structure is due to the
needs of this market for its product development. At the
beginning of the emergence of Sukuk is debt based Sukuk.
Murabahah Sukuk is one of the debt-based forms. The
second stage of the evolution is asset-based Sukuk. One of
the forms of this structure is Ijarah Sukuk. The last stage
Islamic banking and finance – Essays on corporate finance, efficiency and product development
The impact of Islamic debt on company value
Table 1. Global Sukuk issuance by year.
200
Year
180
160
1
1
–
4
16
23
32
50
96
99
130
174
140
120
100
80
60
40
20
08
07
20
06
20
05
20
04
20
03
20
02
20
01
20
99
20
98
19
19
97
0
96
0.05
0.9
–
0.2
1.6
2.9
4.2
3.5
7.8
19.5
34.3
15.5
Value USD billion
Number
19
1996
1997
1998
1999
2001
2002
2003
2004
2005
2006
2007
2008
Number
19
Value USD billion
Source: ZawyaSukuk Monitor, 2009
31.5
5.2
0.3
21.7
1.5
0.7
1.8
7.5
1.3
0.2
0.13
0.16
0.14
72.13
Value in%
43.67
7.21
0.42
30.08
2.08
0.97
2.50
10.40
1.80
0.28
0.18
0.22
0.19
100
Source: ZawyaSukuk Monitor, 2009
of the evolution is equity-based Sukuk or partnershipbased Sukuk. The forms of this structure are musyarakah
and istisna’.
3. Literature review
Capital structure is widely discussed in the finance literature.
The mixture of debt to equity in the financial structure
of companies and whether it will impact upon financial
performance risk and valuation is the subject of theoretical
and empirical studies. In general, capital structure theories
are classified into three categories; first, the zero impact
hypotheses or the Modigliani and Miller theory; second,
the positive impact hypotheses or trade-off theory; third,
the negative impact hypotheses or pecking order theory.
Modigliani and Miller theory
Modigliani and Miller (1958) argue that capital structure
is irrelevant, thus the total cash flows a company makes for
all investors (debt holders and shareholders) are the same
Eds. Hatem A. El-Karanshawy et al.
45
40
Value USD Billion
Value in %
35
30
25
20
15
10
5
0
ay
s
Ba ia
hr
In ain
do
ne
sia
Br
UA
un
ei Pak E
D is
ar ta
us n
sa
la
m
K
Sa u
ud wa
i A it
ra
bi
a
Q
at
ar
U
K
Su
da
n
U
G SA
er
m
an
y
Malaysia
Bahrain
Indonesia
UAE
Pakistan
Brunei Darussalam
Kuwait
Saudi Arabia
Qatar
UK
Sudan
USA
Germany
Total
Value USD Billion
al
Country
50
M
Table 2. Global Sukuk issuance by country in 2009.
regardless of capital structure. On the other hand Jensen
and Meckling (1976) states that the amount of leverage in a
firm’s capital structure is associated with its performance.
Furthermore, several researchers have conducted numerous
studies which aim to examine the relationship between
capital structure and firm performance. However, until now
the evidence regarding this study is contradictory and mixed.
(Ebaid, 2009; Ni & Yu, 2008; Phillips, 2004) find consistent
results with Modigliani and Miller (M&M) theorem. On the
other hand, (Abor, 2007; Bhabra, Liu & Tirtiroglu., 2008)
find inconsistent results with M&M theorem. The focus on
country effect, for example, developed and emerging markets
has been done in some of the studies (Bhabra et al., 2008;
Ebaid, 2009). Other studies document a focus on firm size,
such as large, medium or small companies (Abor, 2007).
The Modigliani and Miller theory (1958) assumes that
a capital market is perfect (no transaction or bankruptcy
costs, perfect information), individuals and corporations
can borrow at the same rate, and no taxes. It does not
21
Fauzi et al.
Table 3. Global Sukuk issuance by structure type.
Year
Type of
structures
Value USD
billion
Value
in%
Phase I
(1996–2001)
Murabahah
Al Salaam
Istisna
Ijarah
Mudarabah
Musyarakah
Al Istithmar
Hybrid
Other
1.6
0.16
0.5
0.25
0.05
–
–
–
–
62.5
6.3
19.5
9.8
2.0
–
–
–
–
Total
2.56
100.0
Phase II
(2002–2007)
Murabahah
Al Salaam
Istisnaa
Ijarah
Mudarabah
Musyarakah
Al Istithmar
Hybrid
Other
4.9
1.9
4.1
20.5
8
26.3
2.9
2.8
1
6.8
2.6
5.7
28.3
11.0
36.3
4.0
3.9
1.4
Total
72.4
100.0
Murabahah
Al Salaam
Istisnaa
Ijarah
Mudarabah
Musyarakah
Al Istithmar
Hybrid
Al Wakalah
4
0.05
0.08
13.6
2.5
6.5
3.5
0.075
1
12.8
0.2
0.3
43.4
8.0
20.8
11.2
0.2
3.2
Total
31.305
100.0
Phase III
(2008–2009)
Source: ZawyaSukuk Monitor, 2009.
matter if the firm’s capital is raised by issuing stock or debt,
or what the firm’s dividend policy is. Therefore, the M&M
theory concludes that capital structure is irrelevant.
Modigliani and Miller made two propositions under these
conditions. Their first proposition was that the value of a
leveraged firm is the same as the value of an unleveraged firm.
Their second proposition was that the expected return on
equity is positively related to leverage because the risk to equity
holders increases with leverage. These two propositions stand
on the assumption that taxes are ignored, and bankruptcy cost
and other agency costs were not considered.
When taxes were taken into account, they had two
propositions as well. First, they state that the value of the
firm is positively related to leverage. It means that corporate
leverage lowers tax payments because corporations can
deduct interest payments but not dividend payments.
Secondly, they state that the cost of equity rises with
leverage because the risk to equity rises with leverage.
These propositions assume that firms have a capital
22
structure almost entirely composed of debt. But in the real
world, firms cannot stand only with debt or a hundred
percent leverage because an increase in debt will increase
bankruptcy cost and agency cost. Consequently, it means
that no optimal capital structure exists.
Trade-off theory
After the seminal work of M&M, little research has
been done to explore capital structure in which some
assumptions were proposed; trade off theory and pecking
order theory. The trade-off theory derived from the models
based on taxes and agency cost. Modigliani and Miller
(1963), DeAngelo and Masulis (1980) and Jensen and
Meckling (1976) suggest the firm has an optimal capital
structure by offsetting the advantages of debt and the cost
of debt. Therefore, trade off theory refers to the idea that
a company chooses how much debt finance and how much
equity finance to use by balancing the costs and benefits. It
states that there is an advantage to financing with debt, the
tax benefits of debt, and tax benefits to be had, but there
is also a cost to financing with debt, the costs of financial
distress including bankruptcy costs, and agency costs.
This theory suggests that there is a positive relationship
between debt level and firm performance. Moreover, the
implication of this trade off theory is that firms have target
leverage and they adjust their leverage toward the target
over time. In addition, Harris and Raviv (1990) imply
that higher leverage can be expected to be associated with
larger firm value, higher debt level relative to expected
income, and lower probability of reorganization following
default.
The empirical relevance of the trade-off theory has often
been questioned. Some research has been conducted
to investigate this theory and the results from various
contexts are mixed and inconclusive. The evidence does
indicate there are likely to be differences attributable
to firm size, country and the maturity of the respective
capital market.
Pecking order theory
Pecking order theory was developed by Myers and
Majluf (1984). Myers and Majluf (1984) consider firms
must issue common stock to raise cash to undertake a
valuable investment opportunity. Management is assumed
to know more about the firm’s value than potential
investors. Investors interpret the firm’s actions rationally.
The evolution of Sukuk
1. Debt based Sukuk
2. Asset based Sukuk
3. Equity based Sukuk
4. Hybrid/mixed Sukuk
Figure 1. The evolution of Sukuk structure.
Islamic banking and finance – Essays on corporate finance, efficiency and product development
The impact of Islamic debt on company value
An equilibrium mode1 of the issue-investment decision is
developed under these assumptions. The model shows that
firms may refuse to issue stock, and therefore may pass up
valuable investment opportunities. The model suggests
explanations for several aspects of corporate financing
behaviour, including the tendency to rely on internal
sources of funds, and to prefer debt to equity if external
financing is required.
In addition, Frank and Goyal (2003) state that capital
structure is acquired in accordance with the priority of the
firm in which internal funding is preferable and external
funding is less preferable. If it is needed, firms could use
external funding from the lowest risk debt. Therefore,
pecking order theory refers to the idea that companies prefer
to use their sources of financing from internal financing
to equity. If external financing is required, firms issue the
safest security first. That is, they start with secure debt, then
perhaps equity as a last choice. In addition, issue costs are
least for internal funds, low for debt and highest for equity.
There is also the negative signaling to the stock market
associated with issuing equity, positive signaling associated
with debt, and asymmetric information between managers
and investors. This theory suggests that there is a negative
relationship between debt level and firm performance.
Therefore, the implication of this pecking order theory is
that firms prefer to depend on internal sources of funds and
prefer debt to equity if external financing is required. Thus,
a firm’s leverage is not driven by the trade-off theory, but
rather by results of the firm’s attempts to mitigate signalling
effect and information asymmetry.
The majority of studies have been conducted in mature
markets with some based on developing markets including
Asia, Africa and the Middle East. Previous studies also
consider firm size, investigating whether large corporations
and small and medium sized corporations behave differently.
The results are mixed and inconclusive. In addition, previous
studies examined different institutional structure (Booth,
Aivazian, Demirguc-Kunt & Maksimovic., 2001); different
governance mechanisms (Wiwattanakantang, 1999);
different market power and firms investment (Eriotis,
Frangouli & Neokosmides., 2002); different regional
risk (Zeitun & Tian, 2007); different firm characteristics,
ownership structure and industry membership (Bhabra
et al., 2008). In comparison with the abundance of studies
on the relationship between capital structure and firm
performance and the determinant factors of capital structure
in conventional debt, only a few studies focus on Islamic
debt. This study will provide evidence about how Sukuk
impacts upon financial performance and corporate value in
markets.
Furthermore, the study of leverage impact on firm value
has become an important aspect of capital structure theory.
Myers (1984) claims that the firm value depends on the debt
ratio, similarly, many studies have focused on the impact
of the debt level and the debt type on a firm’s financial
performance (Ebaid, 2009; Ghosh & Cai, 1999; Hatfield,
Cheng & Davidson, 1994; Coleman, 2007; Talberg, Winge,
Frydenberg & Westgaard, 2008). However, few recent
studies investigate the impact of the Islamic debt type on a
firm’s value and financial performance. This study attempts
to investigate the impact of Islamic debt on company value
and a firm’s financial performance.
Eds. Hatem A. El-Karanshawy et al.
4. Methodology
Data
The data for this study were obtained from the Islamic
Finance Information Service (IFIS) database. The sampling
period is 2000 to 2009, which is ten years and, this study
used quarterly data. This quarterly data is important since
the issuance of Islamic debt for every firm is in different
quarters. Initially, this study proposed to investigate the
debt choice impact on a company value and firm’s financial
performance using Malaysian firms as a sample. Further,
this study notes that 227 companies from Malaysia issued
Islamic debt from 2000 to 2009. From those 227 Malaysian
companies, 106 companies are public companies, and 121
companies are limited companies. From the 106 public
companies, 31 companies have been excluded from the
data list because of the unavailability of their financial
statement data. In addition, the sample of Islamic debt
offering must have data availability on the size of the
offering, the maturity length, the history of the issuance,
and other accounting data information.
For panel data analysis, the availability of ten years’ worth
of data is required, particularly quarterly data. To mitigate
the problem of missing values, this study uses multiple
imputations by including the weighted value to compensate
the missing value excluded in the model (Raghunathan,
2004).
Variables
Firm value and firm financial performance are dependent
variables. Each of the performance indicators measures a
different aspect of performance. Tobin’s Q is used as a firm
value indicator,and it is considered as a market reflection
of the firm’s activities and performances. Return on Asset
(ROA) and Return on Equity (ROE) metrics are used as a
firm financial performance indicator. ROE measures the
performance from the perspective of the equity holders;
meanwhile ROA measures the asset productivity and
operating profit margin. It is important to note that none
of these measures truly reflect the complete picture by
themselves but have to be seen in conjunction with other
metrics.
The proportion of the Islamic debt to non-Islamic debt,
the proportion of Islamic debt, the frequency of Islamic
debt issuance and the type of Islamic debt issued are used
as independent variables. The proportion of Islamic debt
is calculated as the total Islamic debt divided by the total
assets/or total Islamic debt divided by the total Islamic
debt plus total equity. The proportion of non-Islamic debt
is calculated as the total of non-Islamic debt divided by the
total assets/or total non-Islamic debt divided by the total
non-Islamic debt plus total equity.
For the proportion of Islamic debt, the frequency of Islamic
debt issued and the type of Islamic debt, dummy variables
are employed. To avoid too many parameters and to find
the unique least square estimates for the model, this study
uses only n-1 dummy; therefore, the baseline is chosen
for every set of the specifications. In addition, n-1 dummy
may mitigate the problem of multicollienarity among the
regressors (Baltagi, 2005). The choice of a baseline category
23
Fauzi et al.
is essentially arbitrary, for this study fits precisely with all
regression models regardless of which category is selected
for this role. The value and meaning of the individual
dummy-variable coefficients δ1, δ2, ζ1, ζ1, η1 and η2 depend,
however, on which category is chosen as the baseline.
The first group dummy is aimed at examining the effect of
the Islamic debt proportion on each company, and three
categories are set: first, a company having an Islamic
debt proportion below the average of the Islamic debt
proportion; second, a company having an average Islamic
debt proportion; and third, a company having an Islamic
debt proportion higher than the average of the Islamic debt
proportion. The first category will be set as “1” if companies
have a below average Islamic debt proportion; otherwise
it is set equal to “0”. The second category will be set as “1”
if companies have an average Islamic debt proportion;
otherwise it is set equal to “0”. The third category will be set
as “0” if companies have a higher than average Islamic debt
proportion. The baseline category is used for this dummy
if the company has a higher than average proportion
Islamic debt. The average of the Islamic debt proportion is
8.06%, which is calculated by the total of the Islamic debt
proportion over the total number of companies.
The second group dummy is aimed at examining the effect
of the Islamic debt issuance frequency on each company,
and three categories are set: first, a company issuing an
Islamic debt only once; second, a company issuing an
Islamic debt for the second time; and third, a company
issuing an Islamic debt more than twice. The first category
will be set as “1” if companies issue an Islamic debt only
once; otherwise it is set equal to “0”. The second category
will be set as “1” if companies issue an Islamic debt for
the second time; otherwise it is set equal to “0”. The third
category will be set as “0” if companies issue an Islamic
debt more than twice. The baseline category is used for this
dummy if the company has more than twice of the Islamic
debt issuance.
The third group dummy is aimed at examining the effect of
the Islamic debt type on each company, and three categories
are set: first, a company issuing a debt-based type of Islamic
debt; second, a company issuing an asset-based type of
Islamic debt; and third, a company issuing an equity-based
type of Islamic debt. The first category will be set as “1” if
companies issue a debt-based type; otherwise it is set equal
to “0”. The second category will be set as “1” if companies
issue an asset-based type; otherwise it is set equal to “0”.
The third category will be set as “0” if companies issue an
equity-based type. The baseline category is used for this
dummy if the company issues an equity-based type of the
Islamic debt.
Prior research suggests that the performance of each firm
may differ according to their size, because larger firms
have greater economies of scale in the transaction costs
associated with long term debt, which may influence the
results and inferences (Ramaswamy, 2001; Frank & Goyal,
2003; Coleman, 2007; Jermias, 2008; Ebaid, 2009). In
addition, larger firms have less potential of bankruptcy
cost; therefore, firm size should be positively related to the
borrowing capacity (Krishnan & Moyer, 1997). This study
uses a natural logarithm of the total assets as a proxy for
firm size as the control variable (Naceur & Goaied, 2002;
24
Akhtar, 2005; Zeitun & Tian, 2007; Talberg et al., 2008).
The natural logarithm is applied for the firm size variable
owing to the skewness and kurtosis problem. Further,
natural logarithm ensures that the actual regressor has less
statistical noise in the regression model, and moderates the
effects of the large size of the firm.
Model specification
This study uses the panel data method which allows
the unobservable heterogeneity for each observation
in the sample to be eliminated and multicollinearity
among variables to be alleviated. Unobservable
heterogeneity might result in spurious correlations with
the dependent variables, which would bias the coefficient
obtained (Baltagi, 2005). Before proceeding to the
model specification, diagnostic testing of normality,
heteroskedasticity, multicollinearity, and autocorrelation,
was conducted to determine the appropriate method used
in this study. The specification testing results are provided
in Tables 4 and 5.
The heteroskedasticity result is 462.99 with p-value 0.0000,
suggesting that there is a heteroskedaticity problem.
Therefore, this problem needs to be catered to obtain
efficient and unbiased results. The skewness and kurtosis
results are 45.72 with p-value 0.000 and 4.25 with p-value
0.0392, suggesting that non-normal distribution, thus
this non-normal distribution has to be treated. Therefore,
outliers’ checking is conducted prior to data analysing.
The multicollinearity result is 10.9100 with p-value 0.000,
suggesting no multicollinearity problem among the
explanatory variables. Before proceeding to the endogeneity
test and linearity test results, a brief conclusion made is that
heteroskedasticity and non-normality problems have to be
Table 4. Summary of the specification testing results.
Tests
p-value
Heteroskedasticity
Skewness
Kurtosis
Multicollinearity
Linearity
Endogeneity
462.99*
0.0000
45.72*
0.0000
4.25*
0.0392
10.91*
0.0000
2.9055*
0.0000
Endogeneity exist
*Sig. at 1% significance level.
Table 5. The DWH test for endogeneity of regressors.
Variables
Islamic Debt
Proportion
Non-Islamic
Debt Proportion
Tobin’s Q
0.0820
(0.7745)
37.4723*
(0.0000)
ROA
6.9848*
(0.0083)
0.0036
(0.9523)
ROE
0.0035
(0.9526)
14.2038*
(0.0002)
*Sig. at 1% significance level.
Islamic banking and finance – Essays on corporate finance, efficiency and product development
The impact of Islamic debt on company value
treated. The linearity test result for group 1 is 2.905506
with p-value 0.0000, which rejects the null hypothesis of
nonlinearity. Similar to the result for group 1, the linearity
test result for group 2 is -4.659 with p-value 0.000, which
supports the linear model. The endogeneity test result
reveals that the regressors in the model present endogeneity.
Supported by numerous previous studies, by the
assumptions above, by the endogeneity tests and by
the linearity tests, the Generalised Method of Moments
is appropriate as it corrects for heteroskedasticity, the
endogeneity problems and reduces multicollinearity, hence
improving the efficiency of the estimates.In conclusion,
according to specification testing results, a linear dynamic
panel GMM is employed.
Before constructing the dynamic panel GMM model, the
equation below is a starting point for this study to establish
if the debt choice has an impact on a firm’s value and
firm’s financial performance. A model for the regression of
Islamic debt, non-Islamic debt, the proportion of Islamic
debt, the frequency of Islamic debt issuance, and the type
of the Islamic debt issued is then:
y it = a + b i1 X i1 + b i 2 X i 2 + δ 1 K i1 + δ 2 K i 2 + ζ 1 N i1
+ζ 2 N i 2 + η1 Z i1 + η 2 Z i 2 + uit
uit = µ i + λ t + v it
002.e
(1)
ps(2)
i = 1, …, N; t = 1, …, T,
where yi is firm’s value and/or firm’s financial performance.
Xi1 is Islamic debt, Xi2 is non-Islamic debt and, Xi3 is firm
size. K is the dummy proportion for Islamic debt, N is the
dummy frequency for Islamic debt and Z is the dummy
Islamic debt type. mi denotes the unobservable individual
effect, lt denotes the unobservable time effect, and vit is
the remainder stochastic disturbance term. This model
describes three parallel regression planes, which can differ
in their intercepts. Hereafter, the X, K, N, Z will be referred
as Xit (set of regressors):
3.eps(3)
where xit is a K × 1 vector of regressors, β is a K × 1 vector
of parameters to be estimated, and ai represents timeinvariant individual nuisance parameters. Under the
null hypothesis, uit is assumed to be independent and
identically distributed (i.i.d.) over periods and across
cross-sectional units.
The GMM equation model (Blundell & Bond, 1998) is
specified as follows (Cameron & Trivedi, 2010):
(4)
where yit is the Tobin’s Q at time t for firm i, Xit is a set of
regressors, hi is an unobserved firm-specific effect and eit is
a stochastic error. Moving to the right yit−1, it is to obtain:
y it = a y it-1 + b ′ X it + η i + e it . (5)
Eds. Hatem A. El-Karanshawy et al.
y it - y it-1 = a ( y it-1 - y it-2 ) + b ′( X it - X it-1 ) + (e it - e it-1 ) (6)
where X includes lag performance for explanatory
varia­bles, yit−1 as well as dependent variables. The firstdifferencing eliminates potential bias that arises from
unobservable heterogeneity. After first-differencing, GMM
estimation uses lagged values as instruments for Xit − Xit−1:
Supposing that the regressors are predetermined, it is
possible to obtain consistent estimates of coefficients
performing a GMM estimator that exploits the following
orthogonality conditions;
E  y it- s ( e it - e it-1 )  = 0 for s ≥ 2 and t = 3, …, T (7)
E  X it- s (e it - e it-1 )  = 0 for s ≥ 2 and t = 3, …, T (8)
Then, the instrumental variable estimation in the first
difference model is:
Δy it = γ 1 Δy it-1 +…+ γ p Δy it- p + ΔX ′it b + Δe it ,
t = p + 1, …, T
(9)
where ∆ is the first difference operator. The first difference
in equations and levels using their past levels/first
differences are used for the instrumented variables.
Further, a test of overidentifiying restrictions is necessary
to test the validity of overidentifying instruments in an
overidentified model to identify that the parameters of
the model are estimated using optimal GMM. This test
is called Hansen’s test, and the null hypothesis is that all
instruments are valid. At last, weak instruments testing is
done to identify whether the instrument is weak, and the
overidentified model is used because the model has only
one endogenous regressor that is overidentified (Cameron
& Trivedi, 2010; p. 191–199).
5. Analysis
y it = a i + β ′ x it + uit , i = 1, …, N and t = 1, … T ,
y it - y it-1 = (a - 1) y it-1 + b ′ X it + η i + e it Taking into account the first difference, one can elide the
unobserved firm-specific effect:
The sample used consists of 80 listed firms issuing
Islamic debt for the period of 2000 to 2009. Therefore,
there are approximately 3,200 observations used. Table 6
provides the descriptive statistics used in this study.
The table depicts the number of observations, mean,
standard deviation, minimum and maximum value of
each variable. The dependent variables are Tobin’s Q,
ROA and ROE,and each of these dependent variables is
regressed toward its explanatory variables.This study
divides all explanatory variables into four categories. The
first category is the debt structure used by the firm. The
second category is the frequency of Islamic debt issuance.
The third category is the Islamic debt proportion issued.
The fourth category is the Islamic debt type issued. Firm
size and year of Islamic debt issued are used as control
variables.
The mean value for Tobin’s Q is 0.1679 with a range of
-1.6600 to 1.9938, suggesting that most of the firms
experienced low firm performance based on the market
measure. A low Tobin’s Q may indicate that the stock is
25
Fauzi et al.
Table 6. Descriptive statistics.
Variables
Obs.
Mean
Dependent variables
Tobin’s Q
80
0.1679
ROA
80
0.0925
ROE
80
0.0156
Explanatory variables the debt structure of the firm
Islamic Debt Proportion
80
0.0806
Non-Islamic Debt Proportion
80
0.2174
The frequency of Islamic debt issuance
First Issuance
80
0.0000
Second Issuance
80
0.1316
More Than two Issuance
80
0.4211
The proportion of Islamic debt issued
Islamic Debt Below Average
80
0.8813
Islamic Debt Average
80
0.0000
Islamic Debt Above Average
80
0.1164
The type of Islamic debt issued
80
Debt Type of Islamic Debt
80
0.0000
Asset Type of Islamic Debt
80
0.1053
Equity Type of Islamic Debt
80
0.1316
Control Variables Size effect
Firm Size
80
6.0388
Year effect
Year 2001
80
0.0119
Year 2003
80
0.0625
Year 2004
80
0.1563
Year 2005
80
0.3438
Year 2006
80
0.1719
Year 2007
80
0.1406
Year 2008
80
0.0938
Year 2009
80
0.0313
undervalued. Theoretically, stock being undervalued is
likely to happen in a firm which has a stable earning history,
a historically consistent return on equity and a higher
earnings growth rate compared to the market average.
Apparently, this seems to be consistent with the sample
used for this group, in which the majority of firms are large
firms (see the mean value of firm size, which suggests that
most of the firms are big firms).
The mean value for ROA is 0.0925 with a range of 0.0100
to 0.1526. Though the mean value of ROA is considerably
small, this positive value indicates that the firms in the
sample create shareholder value over the sampling period.
This positive value also indicates an effective utilisation
of firm assets in generating an operating surplus in the
business. This lower value of ROA may indicate that the
firms are asset-intensive firms. If so, they thus require
more money to be invested into the business to continue
generating earnings. According to a common rule, ROA
below 5% indicates asset-heavy firms (for example;
manufacturing, railroads, telecommunication providers,
car manufacturers, etc); meanwhile ROA above 20%
26
Std. Dev.
Min
Max
0.2129
0.0004
0.0352
-1.6600
0.0100
0.0021
1.9938
0.1526
0.2292
0.0847
0.1725
0.0102
0.0598
0.4576
0.8732
0.0000
0.3381
0.4938
0.0000
0.0000
0.0000
0.0000
1.0000
1.0000
0.3235
0.0000
0.3208
1.0000
0.0000
1.0000
0.0000
0.3069
0.3381
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
1.0000
1.0000
0.7254
4.6032
8.4924
0.1081
0.2440
0.3660
0.4787
0.3803
0.3504
0.2938
0.1754
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
1.0000
1.0000
1.0000
1.0000
1.0000
1.0000
1.0000
1.0000
indicates asset-light firms (for example, agency firms,
software firms, advertising firms, etc). The ROA is
approximately 9% which may indicate that the majority
of the firms used in the sample are asset-heavy firms, and
represent a variety of sectors. These are a few examples of
the firms used in the sample: Esso Malaysia is one of the
biggest fuel providers in Malaysia, Hubline is one of the
biggest shipping service providers, Kinsteel is one of the
largest steel millers, Kuala Kepong is the largest rubber
plantation and manufacturer, and Zecon is a construction,
infrastructure, toll concession and property development
company.
The mean value for ROE is 0.0156 with a range of 0.0021
to 0.2292, suggesting that most of the firms experienced
low firm performance based on accounting measures.
However, the positive value indicates that the firms in the
sample create shareholder value and operating efficiency
is positively translated into benefits to the owners.
Furthermore, the lower value of ROE may indicate that
the majority of the firms require more capital invested as
discussed in point two, where it is noted that the majority
Islamic banking and finance – Essays on corporate finance, efficiency and product development
The impact of Islamic debt on company value
of the firms are asset-heavy. Therefore, the lower value of
ROE does not mean that they have lower performance.
Moreover, those asset-heavy firms have less competition as
the entry barrier is high. This can be said to be one of the
competitive advantages of these firms.
The mean value for Islamic debt proportion is 0.0806 with a
range of 0.0102 to 0.4576, indicating that most of the firms
issued small amounts of Islamic debt. This may be due to
the fact that this Islamic debt is traded in the thin trading,
moreover, some of the Islamic debt type certificates cannot
be traded in the stock exchange due to its Islamic law issue.
The mean value for non-Islamic debt proportion is 0.2174
with a range of 0.0598 to 0.8732, indicating that most
of the firms are not highly leveraged. This also suggests
that the majority of the firms are less risky since excessive
debt can lead to greater interest payments and principal
repayment burden.
First issuance is used as a baseline category for the
frequency of Islamic debt issuance, and it takes the
value of zero. The mean value for the second issuance of
Islamic debt is 0.1316 with a range of 0.0000 to 1.0000,
suggesting that only 13.16% of the firms issued Islamic
debt for the second time. The mean value for more than
two issuance is 0.4211 with a range of 0.0000 to 1.0000,
suggesting that most of the firms issued Islamic debt more
than twice.
The average of the Islamic debt proportion is 8.06%, which
is calculated by the total of the Islamic debt proportion over
the total firms in the sample, and thus, this 8.06% average
value is used as the average category. The mean value for
Islamic debt below average is 0.8831 with a range of 0.0000
to 1.000, suggesting that most of the firms issued Islamic
debt no greater than 10% (below the average). Islamic debt
average is used as a baseline category for the proportion
of Islamic debt issued, and it takes the value of zero. The
mean value of Islamic debt above average is 0.1164 with a
range of 0.0000 to 1.0000, suggesting that only a few firms
issued Islamic debt greater than the average. This may be
due to the fact that excessive debt issued might increase
the probability of default. Therefore, the issuers have to
assess the trade-off between the Islamic debt and any other
potential risks arising as a result of this debt.
Debt type is used as a baseline category for the Islamic
debt type and it takes the value of zero. The mean value for
asset type of Islamic debt is 0.1053 with a range of 0.0000
to 1.0000, suggesting that only 10.53% of the firms in the
sample issued this type of Islamic debt. The mean value for
the equity type of Islamic debt is 0.1316 with a range of
0.0000 to 1.0000, suggesting that only 13.16% of the firms
in the sample issued this type of Islamic debt.
The mean value for firm size is 6.0388 with a range of
4.6032 to 8.4924, suggesting that most of the firms are big
firms (see explanation on point two). During the sampling
period 2000 to 2009, Islamic debt is only issued during
these eight years – 2001, 2003 to 2009. Islamic debt is
mostly issued in 2005 which accounted for 34.38%. The
mean value for 2001, 2003, 2004, 2005, 2006, 2007, 2008
and 2009 are 1.19%, 6.25%, 15.63%, 34.38%, 17.19%,
14.06%, 9.38% and 3.13% respectively from the total
sample.
Eds. Hatem A. El-Karanshawy et al.
Table 7 provides a pairwise correlation matrix of the
explanatory variables. The highest correlation is between
the Islamic debt proportion and Tobin’s Q, which counts
for 0.4379 (p-value 0.0000) and this value is significant.
The second highest correlation is between the proportion
of Islamic debt and the Islamic debt above average, which
counts for 0.5979 (p-value 0.0000). The third highest
correlation is between the proportion of Islamic debt and
the Islamic debt below average, which counts for −0.5965
(p-value 0.0000). The rest of the correlation coefficient
is less than 0.5, and it is considered as a low correlation
between the explanatory variables, thus, giving less cause
for concern about the multicollinearity problem.
Table 8 presents the dynamic GMM panel regression
results. There are three regression equations, and there are
four explanatory variable categories.
The debt structure of the firm and Tobin’s
Q, ROA and ROE
The coefficient of Islamic debt and non-Islamic debt are a
positive and significant, indicating that these two variables
have a positive effect on a firm’s financial performance.
Both variables are statistically significant at a 1% level.
This finding can be better explained by trade-off theory.
According to prior literature, a firm has an optimal capital
structure by offsetting the advantages of debt and the cost
of debt (Modigliani & Miller, 1963; DeAngelo & Masulis,
1980; Jensen & Meckling, 1976; Haris & Raviv, 1990;
Frank & Goyal, 2003), and this theory apparently can
also be applied to Islamic debt. Trade-off theory refers to
the idea that a company chooses how much debt finance
and how much equity finance to use by balancing the
costs and benefits. It states that there is an advantage
to financing with debt and the tax benefits of debt, and
fortunately Islamic debt is exempted from the taxes.
Moreover, the use of leverage is one way to improve firm
performance (Champion, 1999), and firms prefer debt
financing because they anticipate a higher return (Hadlock
& James, 2002). Furthermore, this finding is in line with
Krishnan and Moyer (1997) and Abor (2005) who find a
positive relationship between capital structure choice and
firm financial performance in developing countries. In
particular, Krishnan and Moyer (1997) include Malaysia as
one of the sample in their study.
The positive result for Islamic debt coefficient obtained
supports the trade-off theory, which was derived from the
models based on taxes and agency cost. From the point of
view of internal management, having Islamic debt in their
debt structure brings more pressure to the management as
Islamic debt is more expensive compared to non-Islamic
debt, hence, improving the firm’s efficiency is important
to maximise asset utilisation due to the Islamic debt
obtained. At the end, this action leads to improvement in
the firm’s performance. Moreover, debt may reduce agency
costs by reducing cash flows available for expropriation
and investments in negative net present value projects
(Harris & Raviv, 1990; Jensen, 1986), as does Islamic
debt. Furthermore, compared to equity issues, the issue
of debt will not dilute the managers’ equity holdings
as a proportion of total equity, but further enhance the
alignment of interests (Fleming, Heaney & McCosker,
2005). In addition, though conventional debt and Islamic
27
28
1.0000
0.1525***
(0.0000)
0.0382**
(0.0353)
0.4379***
(0.0000)
0.0997***
(0.0000)
0.0000
(0.0000)
-0.0681***
(0.0002)
-0.0699***
(0.0001)
-0.3119***
(0.0000)
0.0000
(0.0000)
0.3119***
(0.0000)
0.0000
(0.0000)
-0.0556***
(0.0021)
0.0045
(0.8056)
-0.2079***
(0.0000)
-0.0056
(0.7587)
0.0573***
(0.0016)
0.0348**
(0.0547)
0.2012***
(0.0000)
0.0689***
(0.0001)
0.0078
(0.6681)
-0.0361**
(0.0466)
-0.0688***
(0.0001)
Tobin’s Q
ROA
Year 2009
Year 2008
Year 2007
Year 2006
Year 2005
Year 2004
Year 2003
Year 2001
Firm Size
Equity Type of Islamic Debt
Asset Type of Islamic Debt
Debt Type of Islamic Debt
Islamic Debt Above Average
Islamic Debt Average
Islamic Debt Below Average
More than 2 Issuance
Second Issuance
First Issuance
Non-Islamic Debt Proportion
Islamic Debt Proportion
ROE
Tobin’s Q
Variables
ROA
0.1829***
(0.0000)
0.0393**
(0.0304)
0.0528***
(0.0036)
0.0000
(0.0000)
0.0285
(0.1162)
0.0346**
(0.0565)
0.0780***
(0.0000)
0.0000
(0.0000)
0.0813***
(0.0000)
0.0000
(0.0000)
0.0228
(0.2080)
0.0073
(0.6892)
0.0082
(0.6878)
0.0000
(1.0000)
0.0000
(1.0000)
-0.1001***
(0.0000)
0.0097
(0.5919)
0.0000
(1.0000)
0.0355**
(0.0501)
0.0000
(1.0000)
0.0000
(1.0000)
1.0000
Table 7. Pairwise correlation matrix for explanatory variables.
0.1006***
(0.0000)
0.0466***
(0.0101)
0.0000
(0.0000)
0.003
(0.8666)
0.0920***
(0.0000)
0.1048***
(0.0000)
0.0000
(0.0000)
0.1059***
(0.0000)
0.0000
(0.0000)
0.0087
(0.6304)
0.1006***
(0.0000)
0.1971***
(0.0000)
-0.0483***
(0.0078)
-0.0292
(0.1074)
-0.0884***
(0.0000)
-0.0105
(0.5624)
0.0037
(0.8368)
0.0542***
(0.0028)
0.0732***
(0.0001)
0.0010
(0.9572)
1.0000
ROE
-0.2773***
(0.0000)
0.0000
(0.0000)
-0.0847***
(0.0000)
-0.0334*
(0.0659)
-0.5965***
(0.0000)
0.0000
(0.0000)
0.5979***
(0.0000)
0.0000
(0.0000)
-0.0400**
(0.0274)
0.1094***
(0.0000)
-0.1369***
(0.0000)
0.1882***
(0.0000)
0.0611***
(0.0007)
0.0271
(0.1349)
0.3000***
(0.0000)
0.0496***
(0.0062)
0.0237
(0.1917)
0.0153
(0.398)
-0.0177
(0.3304)
1.0000
Islamic Debt
Proportion
0.0000
(0.0000)
-0.0919***
(0.0000)
-0.0780***
(0.0000)
0.0788***
(0.0000)
0.0000
(0.0000)
-0.0817***
(0.0000)
0.0000
(0.0000)
-0.0587***
(0.0012)
0.2026***
(0.0000)
0.4492***
(0.0000)
0.1362***
(0.0000)
0.0971***
(0.0000)
-0.1106***
(0.0000)
-0.0901***
(0.0000)
-0.0866***
(0.0000)
0.0624***
(0.0006)
0.0330***
(0.0690)
-0.0148
(0.4135)
1.0000
Non-Islamic
Debt Proportion
0.0000
(0.0000)
0.0000
(0.0000)
0.0000
(0.0000)
0.0000
(0.0000)
0.0000
(0.0000)
0.0000
(0.0000)
0.0000
(0.0000)
0.0000
(0.0000)
0.0000
(0.0000)
0.0000
(0.0000)
0.0000
(0.0000)
0.0000
(0.0000)
0.0000
(0.0000)
0.0000
(0.0000)
0.0000
(0.0000)
0.0000
(0.0000)
0.0000
(0.0000)
0.0000
(0.0000)
First
Issuance
Fauzi et al.
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Eds. Hatem A. El-Karanshawy et al.
Year 2009
Year 2008
Year 2007
Year 2006
Year 2005
Year 2004
Year 2003
Year 2001
Firm Size
Equity Type of Islamic Debt
Asset Type of Islamic Debt
Debt Type of Islamic Debt
Islamic Debt Above Average
Islamic Debt Average
Islamic Debt Below Average
Second Issuance
More than 2 Issuance
Table 7. (Continued)
1.0000
-0.3320***
(0.0000)
0.1188***
(0.0000)
0.0000
(0.0000)
-0.1170***
(0.0000)
0.0000
(0.0000)
-0.1335***
(0.0000)
-0.0364**
(0.0450)
0.1010***
(0.0000)
-0.0426***
(0.0188)
0.0088
(0.6293)
0.0458***
(0.0115)
0.0449***
(0.0132)
-0.1011***
(0.0000)
0.0231
(0.2020)
0.0046
(0.8011)
-0.0200
(0.2704)
Second
Issuance
-0.0309*
(0.0886)
0.0000
(0.0000)
0.0290
(0.1103)
0.0000
(0.0000)
0.0548***
(0.0025)
-0.0166
(0.3603)
-0.0076
(0.7086)
-0.0934***
(0.0000)
0.0000
(0.0000)
-0.0926***
(0.0000)
-0.0038
(0.8321)
0.1730***
(0.0000)
0.0733*
(0.0001)
-0.0772***
(0.0000)
0.0082
(0.6508)
1.0000
More Than
2 Issuance
0.0000
(0.0000)
-0.9890***
(0.0000)
0.0000
(0.0000)
0.0199
(0.2732)
0.0075
(0.6785)
0.2083***
(0.0000)
-0.1008***
(0.0000)
-0.1189***
(0.0000)
-0.0096
(0.5979)
-0.3430***
(0.0000)
-0.0761***
(0.0000)
-0.0665***
(0.0002)
-0.0102
(0.5737)
0.0189
(0.2987)
1.0000
Islamic Debt
Below Average
0.0000
(0.0000)
-0.0209
(0.2487)
-0.0109
(0.5496)
-0.2060***
(0.0000)
0.1025***
(0.0000)
0.1218***
(0.0000)
0.0119
(0.5132)
0.3406***
(0.0000)
0.0744***
(0.0000)
0.0591***
(0.0011)
0.0037
(0.8405)
-0.0186
(0.3040)
1.0000
Islamic Debt
Above Average
0.0000
(0.0000)
0.0000
(0.0000)
0.0000
(0.0000)
0.0000
(0.0000)
0.0000
(0.0000)
0.0000
(0.0000)
0.0000
(0.0000)
0.0000
(0.0000)
0.0000
(0.0000)
0.0000
(0.0000)
0.0000
(0.0000)
1.0000
Debt Type of
Islamic Debt
-0.1335***
(0.0000)
-0.0363*
(0.0740)
-0.0375**
(0.0384)
-0.0900***
(0.0000)
-0.0144
(0.4266)
-0.0897***
(0.0000)
-0.0186
(0.3065)
0.1052***
(0.0000)
0.0168
(0.3550)
0.0661***
(0.0003)
1.0000
Asset Type of
Islamic Debt
The impact of Islamic debt on company value
29
30
-0.0070
(0.6981)
1.0000
Year 2009
-0.0287
(0.1131)
-0.0338*
(0.0627)
-0.0474***
(0.0089)
-0.0284
(0.1171)
-0.0244
(0.1784)
-0.0150
(0.4085)
-0.0056
(0.7566)
1.0000
Year 2001
***Sig. at 1% significance level, **Sig. at 5% significance level and *Sig. at 10% significance level.
Year 2009
1.0000
Year 2008
Year 2007
Year 2007
Year 2008
Year 2009
Year 2008
Year 2007
Year 2006
Year 2005
Year 2004
Year 2003
1.0000
-0.0305*
(0.0922)
-0.0115
(0.5278)
0.0085
(0.6746)
0.0167
(0.4107)
-0.1020***
(0.0000)
-0.0172
(0.3959)
-0.0618***
(0.0023)
0.0917***
(0.0000)
0.1215***
(0.0000)
0.0244
(0.2299)
Year 2001
1.0000
0.2151***
(0.0000)
-0.0426***
(0.0188)
-0.1022***
(0.0000)
-0.1200***
(0.0000)
-0.0618***
(0.0007)
-0.1011***
(0.0000)
0.0781***
(0.0000)
0.2361***
(0.0000)
-0.0200
(0.2704)
Firm Size
Firm Size
1.0000
Equity Type of
Islamic Debt
Equity Type of Islamic Debt
Table 7. (Continued)
-0.1335***
(0.0000)
-0.0801***
(0.0000)
-0.0688***
(0.0001)
-0.0422***
(0.0198)
-0.0158
(0.3826)
1.0000
Year 2004
-0.1124***
(0.0000)
-0.0966***
(0.0000)
-0.0593***
(0.0011)
-0.0222
(0.2202)
1.0000
Year 2005
-0.0579***
(0.0014)
-0.0356**
(0.0499)
-0.0133
(0.4623)
1.0000
Year 2006
Fauzi et al.
Islamic banking and finance – Essays on corporate finance, efficiency and product development
The impact of Islamic debt on company value
Table 8. Summary of the regression result.
Dynamic GMM
Variables
L1
Constant
The debt structure of the firm
Islamic Debt Proportion
Non-Islamic Debt Proportion
The frequency of Islamic debt issuance
D1_First Issuance D2_Second Issuance
D3_More Than 2 Issuance
The proportion of Islamic debt issued
D1_ID Below Average
D2_ID Average
D3_ID Above Average
The type of Islamic debt issued
D1_Debt Type of ID
D2_Asset Type of ID
D3_Equity Type of ID
Control Variables Size effect
Firm Size
Year effect
Year 2001
Year 2003 Year 2004
Year 2005
Year 2006
Year 2007
Year 2008
Year 2009
J-Statistics
Chi2
Eds. Hatem A. El-Karanshawy et al.
Tobin’s Q
ROA
ROE
0.5742***
(0.0174)
0.3180***
(0.0988)
0.1220***
(0.0041)
8.7500***
(0.0410)
0.1547***
(0.0015)
0.8463**
(0.4264)
0.7441***
(0.0723)
0.5372***
(0.0159)
0.0159***
(0.0025)
0.0177***
(0.0015)
0.0276**
(0.2540)
0.3357***
(0.0544)
(Omitted)
-0.0961*
(0.0951)
0.2090***
(0.0562)
(Omitted)
-0.0774***
(0.0140)
0.0323***
(0.0026)
(Omitted)
-0.7018***
(0.2332)
1.2480***
(0.1135)
0.0006*
(0.0129)
(Omitted) -0.0248**
(0.0111)
0.0181***
(0.0010)
(Omitted) -0.0036***
(0.0010)
0.1319*
(0.1146)
(Omitted)
-0.0135**
(0.1266)
(Omitted)
0.0423
(0.1076)
0.1764*
(0.1049)
(Omitted)
0.0250***
(0.0074)
0.0346***
(0.0027)
(Omitted)
0.3337
(0.2695)
1.0458***
(0.0532)
-0.0844***
(0.0142)
-0.0040***
(0.0005)
-0.2354***
(0.0654)
-1.5848*
(0.9510)
0.0458***
(0.0098)
0.0130
(0.0109)
0.0171***
(0.0048)
-0.0346***
(0.0072)
0.1160***
(0.0177)
0.0397*
(0.0236)
0.0587**
(0.0294)
21514.85
0.0000
-0.0169*
(0.0097)
0.0092***
(0.0024)
0.0173***
(0.0019)
0.0058***
(0.0014)
-0.0018***
(0.0003)
0.0080***
(0.0020)
0.0017***
(0.0006)
0.0048***
(0.0019)
78177.74
0.0000
-0.5066***
(0.1019)
-0.0786*
(0.0433)
0.3969***
(0.0657)
0.0981***
(0.0159)
-0.0392***
(0.0078)
-0.0889***
(0.0298)
0.8061***
(0.1146)
0.7980**
(2.8289)
1.27e+06
0.0000
31
Fauzi et al.
debt are fundamentally different, they perform similarly in
a competitive market as these two instruments are affected
by the same factors (Kraciska & Nowak, 2012).
Although both debt types have a positive impact, the
coefficient for Islamic debt is higher than the coefficient
for non-Islamic debt (only for Tobins’ Q), suggesting that
the Islamic debt provides a higher contribution to the
improvement of firms’ financial performance compared to
non-Islamic debt. Furthermore, it can be concluded that
when Islamic debt is chosen as a tool of firm financing,
(1) the markets react positively to firm performance, thus
this positive reaction might lead to the stock becoming
overvalued;(2) Islamic debt not only improves the
effectiveness of the firm’s management s in managing their
assets to generate profits, but it also improves the operating
efficiency of the total business; (3) firms are effective in
managing their operation efficiency which in the end
contributes to the owners’ wealth because ROE measures
the performance from the perspective of the equityholders. There are a few reasons for this significant positive
contribution of Islamic debt issuance. First, Islamic debt is
claimed and advertised as a secure investment due to its
structure. Second, Islamic debt is given a special privilege
such as stamp duty and exempted tax for both issuers and
investors. Third, Islamic debt is guaranteed by the special
purpose vehicle (SPV); in case of default the Islamic debt
holders may recourse the assets underlying the Islamic
debt. Fourth, though there were a few cases of default in
Middle East, those cases have no impact on the investors’
perspective, as some investors investing in Islamic debt only
do so only to comply with the religious matter. Fifth, the
majority of investors are non-Muslim, with an increasing
presence of foreign investors (PricewaterhouseCoopers
Malaysia, 2008). Sixth, the Islamic debt issuance contributes
to an increase in the issuer’s stock returns (Nagano, nd.).
Moreover, from the issuers’ perspective, there are benefits
issuing Islamic securities, in particular, Islamic debt. The
key benefits are tax incentives, value proposition and
regulatory process. First, for tax incentives, the issuers are
exempted from stamp duty, tax deductible of issuance cost,
and the special purpose vehicle (SPV) is exempted from tax,
and tax neutrality. Second, for value proposition, there is a
wider investors’ base, Islamic debt is attractively priced due
to the strong demand, there is strong structuring expertise
in the Islamic finance industry, and Islamic debt enhances
the issuers’ profile. Third, in terms of regulatory process,
the process facilitates the issuance process, the rating
of Islamic debt is automatically approved for AAA-rated
for Islamic debt issued in domestic (Malaysian) currency
and A-rated Islamic debt issued in foreign currency, any
amendment to terms of approved Islamic debt need only
to inform the Securities Commission, and exchangeable
Islamic debt is exempted from rating. From the point of
view of shareholders, the usage of debt increases their
wealth, and because of this, markets believe that Islamic
debt positively contributes to the firm performance.
Moreover, Islamic debt issuance contributes to an increase
in the issuers’ total factor productivity (Nagano, n.d.).
Furthermore, the positive result may be due to the stabilised
nature of the Malaysian financial system which has evolved
in line with the changing structure of the economy. The
changes in the economic structure and financial system
32
in turn have had an important influence in shaping the
increasing complexity and sophisticated nature of its capital
market along with the implementation of regulations, and
these changes support firms to operate more effectively and
efficiently, increasing the confidence of markets. Moreover,
a more diversified financial system, in particular, the rapid
growth of the Malaysian Islamic Capital Market and the
Malaysian debt market, has increased the alternative
sources of financing available to corporations.
These key benefits supports the theory that the choice of
capital structure may help mitigate agency costs (Jensen
& Meckling, 1976). According to the agency costs theory,
high leverage or a low equity/asset ratio reduces the
agency costs of outside equity and increases firm value by
constraining or encouraging managers to act more in the
interests of shareholders. Moreover, corporate debt has a
disciplining effect on management, since it serves to reduce
the free cash flow and therefore minimises management’s
discretionary spending.
Overall, the finding for ROA and ROE are similar to Tobin’s
Q which also supports the trade-off theory (Modigliani &
Miller, 1963; DeAngelo & Masulis, 1980; Jensen & Meckling,
1976; Haris & Raviv, 1990; Frank & Goyal, 2003), and this
theory apparently can also be applied to Islamic debt.
The frequency of Islamic debt issuance and Tobin’s
Q, ROA and ROE
The coefficient for first issuance of Islamic debt is a positive
and significant at 1% level of significance (but for Tobins’ Q
which significant at 10% level of significance), suggesting
that the first issuance of Islamic debt affects higher firm
performance. This also indicates that (1) the markets
react positively to the issuance of Islamic debt when it is
first introduced to the market; (2) the firm effectively
utilises its assets to generate profits for the shareholders,
and additional debt, in particular Islamic debt, pushes the
management to perform better. There are several factors
that might contribute to this positive finding. First, the
managers of the firms are compelled to put more effort into
generating more profits. Because some of Islamic debt is in
the form of partnership (profit and loss sharing agreement),
Islamic debt tends to place greater pressure on the managers
to manage the firms effectively. Second, there is a broadbased coordination of government policies which resulted
in a comprehensive public policy that supports growth and
innovation in the Islamic financial market, in particular,
Islamic debt. Third, the importance of government
intervention, such as tax incentives and required ratings
improves issuers’ and investors’ confidence. Fourth, the
rapid growth of Islamic finance signifies that Islamic debt
has moved from the pioneering stage to being an established
financing instrument that serves as a commercially viable
and effective tool for mobilising investment assets to finance
productive economic activities. Fifth, in the beginning of
the Islamic finance initiation, Islamic debt offered those
competitiveness features, particularly cost effectiveness,
secureness and efficiency. As such, the market had high
expectations of this new instrument, the upshot was
that Islamic debt brought more pressure on managers
to manage their firms effectively in order to meet market
expectations. Sixth, apart from being well-regulated by
various standards and guidelines, Malaysia is also the only
Islamic banking and finance – Essays on corporate finance, efficiency and product development
The impact of Islamic debt on company value
country that makes it compulsory for all tradable corporate
debt securities to be rated to enhance investors’ confidence
and to assist in the investment decision-making process.
Another distinguishing factor for the Malaysian Islamic
debt market is the establishment of a centralised, national
level Shariah supervisory board, which ensures that every
Islamic debt issued in Malaysia, is in full compliance with
the Shariah. All these factors provide sufficient protection
to investors in the Islamic debt and conventional debt
markets.
However, the coefficient for the second issuance of Islamic
debt is a negative and significant at 5% level of significance,
suggesting that the issuance of Islamic debt for a second
time lowers firm performance. This negative finding is
similar to the study by Godlewski et al. (2010), which
suggest that Islamic debt expansion has a detrimental
effect on firm value. This negative finding may indicate
that (1) either the management of the firms have loosened
their control because of overconfidence from the first
successful issuance of Islamic debt or that the management
have expropriated the firms’ previous profits; (2) the
markets have experienced, observed and learnt from the
first Islamic debt issuance, leading underconfidence in
the markets over this second issuance, which in turn may
affect the share price of those firms issuing Islamic debt;
(3) low credit rating of firms issuing Islamic debt as this
is associated with high risk. The gap between the first and
the second Islamic debt issuance ranges between two to six
years. Presumably, in that time period, investors observed
the firm’s performance, their Islamic debt rating, the
market conditions such as the frequency of default cases of
Islamic debt. In Malaysia, cases of Islamic debt default were
few and it is something that raises concern on the investors’
protection because a default occurs due to the breach of
any binding obligations under the original terms of the
agreement between the issuer and the Sukuk holders. Thus
this factor may contribute to the negative result.
Furthermore, debt is also a source of information which
indicates the firm’s current condition that investors can
use to monitor and evaluate major operating decisions of
the firm in two ways. Firstly, the mere ability of the firm
to make its contractual payments to debt-holders provides
information. Secondly, in the event that the organisation
fails to make the payments, their ways to resolve the
matter either through informal negotiation or formal
bankruptcy proceedings will disseminate considerable
information to the investors (Harris & Raviv, 1990). In
sum, the negative relationship of the second issuance of
Islamic debt and its firm’s performance is probably either
a result of the previous firm performance in meeting their
obligation of payment or a result of inefficient utilisation
of their firms’s assets.
Fortunately, the coefficient for more than two issuance
of Islamic debt is a positive and significant at 1% level of
significance, suggesting the issuance of Islamic debt for
more than two improves a firm’s financial performance.
This may indicate that after having a few experiences
in issuing Islamic debt, the issuance of Islamic debt later
on impacts positively on firm performance. This may be
caused by the fact that (1) the debt-holders of Islamic debt
closely monitor the management of the firm to ensure
that the firm can generate profits and distribute a periodic
Eds. Hatem A. El-Karanshawy et al.
stream of cash flow over time. Thus, Islamic debt also
reduces the agency problem within the company and hence
increases firm value. (2) That as the industry grows, it is
more apparent that there is more demand by non-Muslim
investors and issuers to play a role in the industry. Here in
Malaysia, for instance, there is just as strong a demand for
Shariah compliant products among non-Muslims as there
is among Muslims (PricewaterhouseCoopers Malaysia,
2008).From the view point of markets, this may indicate
that the markets have learnt through several issuances of
Islamic debt and therefore they have greater confidence in
subsequent issuances compared to the second issuance of
Islamic debt. However, investors are irrational according
to the behavioural finance theory. Their decision may be
influenced by the magnitude issue, their bias selection and
the lucky event issue.
The proportion of Islamic debt issued
and Tobin’s Q
The coefficients for the proportion of Islamic debt below
the average and at the average are a positive and significant
varies at 10% and 1% level of significance. These positive
and significant results may be caused by internal and
external factors. In terms of internal factors, the proportion
of Islamic debt issued at a certain level stimulates the
management to work effectively. For external factors,
there are two views; first from the markets’ view, second
from the view of government support. From the markets’
view, the proportion of a certain level of Islamic debt may
be considered as tax exempted stimulation as the profits
derived from Islamic debt are exempted from the taxes.
Furthermore, the markets have confidence over the assets/
projects underlying the Islamic debt contract which may
bring profits in future; therefore, this market confidence
affects their stock price. With regards to government
support, the Malaysian government has provided an
interesting model to promote the co-existence of an ethical
and societal-based finance through issuing a few regulations
that appeal to Muslim and non-Muslim investors; hence
these regulations issued can assure the credibility of this
instrument. Furthermore, the regulating body has taken
vital steps to develop a facilitative regulatory framework, to
create a large pool of players, to introduce a comprehensive
range of innovative and competitive Islamic financial
product and services, and to ensure sufficient depth to
facilitate liquidity management, hence creating market
confidence.
Though debt reduces the agency costs of free cash flow
by reducing the cash flow available for spending at the
discretion of managers (Jensen, 1986), an increased
leverage also has costs; as leverage increases the risk of
default also increases. This theory supports the result for
Islamic debt above the average which is a negative and
significant at 1% level of significance. This finding suggests
that the greater the proportion of Islamic debt issued, the
lower the firm performance. This result is similar to the
empirical result for non-Islamic debt, in that the proportion
of debt at a certain level may hamper firm performance as
an additional incurrence of debt gives no guarantee that
firm performance will be higher. This is mainly because as
the leverage increases, so does the risk of default, which
provides a greater incentive for lenders to monitor the firm.
Though it is claimed that Islamic debt is more secure than
33
Fauzi et al.
the conventional debt, this result finds no support for that
claim. On the contrary, this finding supports the notion
that as the leverage increases, the probability of default
also increases, and Islamic debt is no exception to this rule.
Overall, the result for Islamic debt proportion Tobins’ Q,
ROA and ROE has similarity.
The type of Islamic debt and Tobin’s Q
The coefficients for the debt-type and equity-type are a
positive and significant at 1% and 10% level of significance
for Tobins’ Q and ROE. While all types of Islamic debt are
a positive and significant at 1% level of significance for
ROA. Though the finding for ROA is slightly different than
for Tobin’s Q, this result does not impair on the Tobin’s Q
result, as it is common for different methods of calculation
to give different results. The finding suggests that debt-types
and equity-types affect higher firm performance. The result
supports the notion that certain types of debts have a different
impact on shareholders’ wealth (Mikkelson & Partch, 1986);
hence, this finding can also be applied to Islamic debt.
Furthermore, the finding can be explained by the different
Islamic debt structure. This is important since the structure
determines the obligation of the originator/issuers. There is
typically a requirement that on maturity of the Islamic debt
or upon an event of default, the originator has a purchase
obligation to repurchase the assets which enables the
Special Purpose Vehicle (SPV) to redeem the outstanding
certificates and repay the Sukuk holders. In this regard,
the rights of Sukuk holders in the event of default will vary
depending on whether the Sukuk structure is an assetbased or an asset-backed structure. The positive result
for debt-based and equity-based Sukuk may be caused by
their structure. The assumptions that may be raised is that
debt-based and equity-based are in the structure of assetbacked Sukuk, and asset-based is in the structure of assetbased Sukuk. Thus, the rights of the Sukuk-holders depend
on the structure of Islamic debt. For example, in the case
of Sukuk ijarah, if the Sukuk is asset-backed, this allows
the holders to liquidate the underlying asset in the event
of default to recover most of their investments. On the
other hand, if the Sukuk is asset-based, this only represents
beneficial ownership on the underlying asset and it restricts
the holders’ rights in the event of a default.
The coefficient for firm size is a negative and significant for
all four regression equations, suggesting that bigger firms
which having Islamic debt in their debt structure have a
lower firm performance. The negative result may be due to
the fact that bigger firms are already well-stabilised in terms
of cash flows and profits because of their well-stabilised
capital structure; hence changing its capital structure
with a new unproven instrument may endanger the firm’s
credibility and ability to maintain their stable cash flows
and profits. This notion leads to the markets’ perspective
on the firms’ capability in the future; the markets may have
lower confidence and in turn, this affects the stock price of
the firms.
Apart from year 2004, all the years reveal a significant
result. All the years (2003, 2004, 2005, 2007, 2008, and
2009) have a positive coefficient except year 2001 and year
2006. Malaysia, with its economic strength, supportive
government policies, educated workforce, developed
34
infrastructure, vibrant business environment and quality
of life, has always been an attractive market for foreign
investors. Therefore, the coefficients for year 2003, 2004
and 2005 are supported.
Despite the challenging global economy, Malaysia
has continued to pursue liberalisation, enhancing the
entrepreneurial and investment environments. The
economy scores above the world average in many of the
ten economic freedoms (World Bank, 2011). The trade
regime is relatively open despite lingering non-tariff
barriers. However, corruption and a judicial system that
remains vulnerable to political influence pose significant
challenges to economic freedom. 2001 and 2006 were two
years which yielded a negative and significant impact. The
first, 2001, may be due to the global economic slowdown
overall. Significantly, though, a general election was held in
2003 and again in 2008, revealing a pattern in which there
is a two year gap between this political event and a year
yielding a negative and significant impact. This may indicate
that before the general election, the political situation in
Malaysia heats up, which affects the market players.
The Malaysian economy has been surprisingly resilient in
spite of the global slowdown in 2007.Malaysia has only
felt a minor impact from the slowing US economy, but
emerging challenges in the form of soaring food prices and
the persistent rise in global oil prices are weighing down
heavily on economic prospects. Furthermore, to avoid the
fiscal deficit, the government announced a revamp in oil
subsidies, pushing up the price of petrol diesel, which has
adverse implications for inflation and economic growth.
However, in 2008 and 2009, the business confidence index
increased as it indicates by the rise of sales and production,
higher export sales, higher capacity utilisation, higher
domestic demands and higher capital investment. The
gross domestic product growth was sustained at a certain
targeted level. This growth was driven by high commodity
prices, strong private consumption and steady investment,
and supported by fiscal spending. The business condition
index would be a better indicator of current economic
activity as it relies on firm-level information. Therefore,
the positive and significant coefficients for year 2007, 2008
and 2009 are supported.
6. Conclusion
In sum, the findings for all three categories of explanatory
variables, along with their control variables for all metrics
(Tobins Q, ROA and ROE), are only slightly different in
their coefficient value. Almost all the coefficient signs and
significance values reveal the same direction and a similar
significance value. The coefficients for Islamic debt is higher
than the coefficient for non-Islamic debt and, overall, the
findings for Tobin’s Q, ROA, ROE and EVA support the
trade-off theory (Modigliani & Miller, 1963; DeAngelo &
Masulis, 1980; Jensen & Meckling, 1976; Haris & Raviv,
1990; Frank & Goyal, 2003) and this theory apparently can
also be applied to Islamic debt.
References
Abd. Sukor, M.E., Muhamad, R, Gunawa, A.Y. (2008)
Malaysian Sukuk: Issues in accounting standard.
Shariah Journal, 16(1), 63–74.
Islamic banking and finance – Essays on corporate finance, efficiency and product development
The impact of Islamic debt on company value
Abor, J. (2005) The effect of capital structure on
profitability: An empirical analysis of listed firms in
Ghana. Journal of Risk Finance, 6, 438–447.
Ghosh, A., Cai, F. (1999) Capital structure: New evidence of
optimality and pecking order theory. American Business
Review, 17(1): 32–38.
Abor, J. (2007) Debt policy and performance of SMEs:
Evidence from Ghanaian and South African firms. The
Journal of Risk Finance, 8(4), 364–379.
Hadlock, C.J., James, C.M. (2002) Do banks provide
financial slack? Journal of Finance, 57, 1383–1420.
Al Amine, M.A.B. (2008) Sukuk market: Innovations and
challenges. Islamic Economic Studies, 15(2), 1–22.
Haneef, R. (2009) From asset-backed to asset-light
structures: The intricate history of Sukuk. ISRA
International Journal of Islamic Finance, 1(1).
Al Amine, M.A.B. (n.d.). The Islamic bonds market:
Possibilities and challenges. International Journal of
Islamic Financial Services, 3(1).
Harris, M., Raviv, A. (1990) Capital structure and the
informational role of debt. The Journal of Finance,
45(2), 321–349.
Al-Amine, M.A.B.M. (2001) Istisna’ and Its application in
Islamic banking. Arab Law Quarterly, 16(1), 22–48.
Hatfield, G.B., Cheng, L.T.W., Davidson, W.N. (1994) The
determination of optimal capital structure: The effect of
firm and industry debt ratios on market value. Journal of
Financial and Strategic Decisions, 7(3):1–14.
Ashhari, M.Z., Chun, L.S., Nassir, A. Md. (2009)
Conventional vs Islamic bond announcements: The
effects on shareholders’ wealth. International Journal of
Business and Management, 4(6).
Ayub, M. (2007) Understanding Islamic Finance. England:
John Wiley & Sons, Ltd.
Baltagi, B.H. (2005) Econometric Analysis of Panel Data (3
ed.). West Sussex, England: John Wiley & Son Ltd.
rd
Bhabra, H., Liu, T., Tirtiroglu, D. (2008) Capital Structure
Choice in a Nascent Market: Evidence from Listed Firms
in China. Financial Management, 37(2), 341.
Blundell, R., Bond, S. (1998) Initial conditions and moment
restrictions in dynamic panel data models. Journal of
Econometrics, 87, 115–143.
Booth, L., Aivazian, V., Demirguc-Kunt, A., Maksimovic, V.
(2001) Capital Structures in Developing Countries. Journal
of Finance, 56(1), 87–130.
Cakir, S., Raei, F. (2007) Sukuk vs. Eurobonds: Is there
a difference in value-at-risk, IMF Working Paper
(Vol. WP/07/237): International Monetary Fund.
Cameron, A.C., Trivedi, P.K. (2010) Microeconometrics using
stata (Rev. Ed.).College Station, Texas: Stata Press.
Jensen, M.C. (1986) Agency costs of free cash flow,
corporate finance and takeovers. American Economic
Review, 76(2), 323–329.
Jensen, M.C., Meckling, W.H. (1976) Theory of the firm:
Managerial behavior, agency costs and ownership
structure. Journal of Financial Economics, 3(4),
305–360.
Jermias, J. (2008) The relative influence of competitive
intensity and business strategy on the relationship
between financial leverage and performance. The British
Accounting Review, 40(1), 71–88.
Juan, S. (2008) Prospects and challenges for developing
corporate Sukuk and bond markets: International
Monetary Fund. International Journal of Islamic and
Middle Eastern Finance and Management, 1(1), 20–30.
Kamali, M.H. (2007) A shari‘ah analysis of issues in Islamic
leasing. Islamic Economic, 20(1), 3–22.
Kraciska, O., Nowak, S. (2012) What’s in it for me? A
primer differences between Islamic and conventional
finance in Malaysia. IMF Working Paper, 151.
Champion, D. (1999) Finance: The joy of leverage. Harvard
Business Review, 77(4), 19–22.
Krishnan, V.S., Moyer, R.C. (1997) Performance, capital
structure and home country: An analysis of Asian
corporations. Global Finance Journal, 8(1), 129–143.
Coleman, A.K. (2007) The impact of capital structure
on the performance of microfinance institutions. The
Journal of Risk Finance, 8(1), 56–71.
Mikkelson, W.H., Partch, M.M. (1986) Valuation effects of
security offerings and the issuance process. The Journal
of Financial Economics, 15, 31–60.
DeAngelo, H. & Masulis, R.W. (1980) Optimal capital
structure under corporate and personal taxation.
Journal of Financial Economics, 8(1), 3–29.
Mirakhor, A. (1996) Cost of capital and investment in a
non-interest economy. Islamic Economic Studies, 4(1),
35–46.
Ebaid, I.E.S. (2009) The impact of capital-structure choice
on firm performance: empirical evidence from Egypt.
The Journal of Risk Finance, 10(5), 477–487.
Modigliani, F., Miller, M. (1963) The cost of capital,
corporation finance and the theory of investment.
American Economic Review, 48, p.261–297.
Eriotis, N.P., Frangouli, Z., Neokosmides, Z.V. (2002) Profit
margin and capital structure: An empirical relationship.
The Journal of Applied Business Research, 18(2), 85–88.
Modigliani, F., Miller, M.H. (1958) The cost of capital,
corporation finance and the theory of investment.
American Economic Review, 48(3), 261.
Fleming, G., Heaney, R., McCosker, R. (2005) Agency
costs and ownership structure in Australia. Pacific-Basin
Finance Journal, 13, 29–52.
Mohd Yatim, M.N., Shah, M. (2009) Sukuk (Islamic bond): A
crucial financial instrument for securitisation of debt for
the debt-holders in shari’ah-compliant capital market.
International Journal of Business and Management.
Frank, M.Z., Goyal, V.K. (2003) Testing the pecking
order theory of capital structure.Journal of Financial
Economics, 67(2), 217–248.
Eds. Hatem A. El-Karanshawy et al.
Mokhtar, S. (2009) A synthesis of Shariah issues and
market challenges in the application of wa’d in equity-
35
Fauzi et al.
based Sukuk. ISRA International Journal of Islamic
Finance, 1(139–145).
empirical study of the Indian Manufacturing sectors.
Strategic Management Journal, 22, 989–998.
Myers, S., C., Majluf, N., S. (1984) Corporate financing and
investment decisions when firms have information that
investors do not have. Journal of Financial Economics,
13, 187–221.
Somolo, E. (2009) Islamic Pricing Benchmark. ISRA Inter­
national Journal of Islamic Finance, 1(1).
Myers, S.C. (1984) The capital structure puzzle. Journal of
Finance, 39, 575–592.
Naceur, S.B., Goaied, M. (2002) The relationship between
dividend policy, financial structure, profitability and
firm value. Applied Financial Economics, 12; 843–849
Nagano, M. (n.d.). Islamic finance and the theory of capital
structure. Discussion Papers in Economics, Society of
Economics Nagoya City University, 51.
Ni, J., Yu, M. (2008) Testing the Pecking-Order Theory:
Evidence from Chinese Listed Companies. The Chinese
Economy, 41(1), 97–113.
Phillips, P.A., Sipahioglub, M.A. (2004) Performance
implications of capital structure: evidence from quoted
UK organisations with hotel interests. Service Industries
Journal, 24(5), 31–51.
PricewaterhouseCoopers Malaysia.(2008) Malaysia, Asia’s
Islamic Finance hub (Kuala Lumpur). Retrieved from:
http://www.pwc.com/my/en/publications/islamicfinance-hub.jhtml
Raghunathan, T.E. (2004) What do we do with missing
data? Some options for analysis of incomplete data.
Annual Reviews Public Health, 25, 99–117.
Ramaswamy, K. (2001) Organisational ownership,
competitive intensity and firm performance: An
36
Talberg, M., Winge, C., Frydenberg, S. & Westgaard, S.
(2008) Capital structure across industries. International
Journal of the Economics of Business, 15(12), 181–200.
Tariq, A.A., Dar, H. (2007) Risks of Sukuk structures:
implications for resource mobilization. Thunderbird
International Business Review, 49(2), 203–233.
Usmani, M.T. (1999) The Concept of Musharakah and Its
Application as an Islamic Method of Financing. Arab
Law Quarterly, 14(3), 203–220.
Vishwanath, S.R., Azmi, S. (2009) An overview of Islamic
Sukuk bonds. The Journal of Structured Finance, 14(4),
58–67.
Wilson, R. (2008) Innovation in the structuring of Islamic
Sukuk securities. Humanomics, 24(3), 170–181.
Wilson, R. (n.d.). Islamic Bonds: Your Guide to Issuing,
Structuring and investing in Sukuk.
Wiwattanakantang, Y. (1999) An empirical study on the
determinants of the capital structure of Thai firms.
Pacific-Basin Finance Journal, 7(3–4), 371–403.
World Bank. (2011).
Yean, T.W. (n.d.). Sukuk: Issues and the way forward.
Zeitun, R., Tian, G.G. (2007) Capital structure and
corporate performance: Evidence from Jordan.
Australasian Accounting Business and Finance Journal,
1(4): 40–61.
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Islamic financing and bank characteristics in a
dual banking system: Evidence from Malaysia
Muhamed Zulkhibri
Economic Research and Policy Department, Islamic Development Bank, P.O Box 5925, Jeddah 21432, Saudi Arabia,
E-mail: [email protected]
Abstract - An understanding of financing behaviour explains the performance of Islamic banks
as an alternative to the conventional finance model as suppliers of capital for businesses and
entrepreneurs. In the Islamic banking system, banks are suppliers of capital and not lenders unlike
that stipulated in a traditional banking system. To date, Islamic banks have become the major
source of capital in Malaysia and lending behaviour is an important policy variable. In this context,
the paper examines the relationship between bank financing, bank financing rate and bank-specific
characteristics in a dual banking system. The evidence suggests that the bank-specific characteristics
are important for Islamic banking financing behaviour. The Islamic banks’ financing behaviour is
consistent with behaviour of conventional banks in that the bank lending operates through banks
depending on the size, and level of liquidity and capital. The findings also suggest that that there is
no significant difference between Islamic bank financing and conventional bank lending behaviour
with respect to interest rates.
Keywords: Islamic banks, base financing rate, bank financing, panel regression analysis
1. Introduction
An Islamic bank is a deposit-taking institution, which
includes all functions currently known as banking activities.
The bank mobilizes funds on the basis of mudaraba (profitsharing) or wakalah (as an agent charging a fixed fee for
managing funds), which form part of its liabilities, while
financing on a profit-and-loss sharing (PLS) basis or through
the purchase of goods (on cash) and sale (on credit) or
other trading, leasing and manufacturing activities, form
part of the assets. Apart from demand deposits, which are
treated as interest-free loans from the clients to the bank
and are guaranteed to be repaid in full, it plays the role of
an investment manager for the owners of deposits, akin to
a universal bank.
In contrast, conventional banks are understood in a
generic sense as financial intermediaries. Their main task
is to provide indirect finance, in contrast to direct finance
through financial markets. Banks are understood to
channel surplus funds from the household sector into the
corporate sector facing a deficit, as they invest more than
internal or direct finance. Banks play a vital role in the
economy enabling more productive investment than would
be possible merely on the basis of profits and financial
market funds. Hence, an asymmetric information and
understanding of the functioning of banks must lead to the
conclusion that Islamic banks are not viable.
The distinguishing features of Islamic banks are
the prohibition of charging or paying interest, the
impermissibility of demanding collateral and, to a small
extent, compulsory charitable spending (Khan and
Mirakhor 1992). Profit has to be generated merely by
primary and secondary modes of Islamic finance (Chapra
2000). Primary modes include profit-sharing arrangements
such as mudaraba (partnership) and musharakah (equity
participation), while secondary modes are essentially
mark-up pricing or leasing arrangements.
Many scholars also argue that Islamic banking resembles
conventional banking schemes. This includes the claim that
the majority of Islamic lending has a debt-like character
(Aggarwal and Youssef 1996). El-Gamal (2005) has
concluded that Islamic finance is primarily a form of rentseeking legal arbitrage and simply seeks to replicate the
operations of conventional financial instruments. However,
some researchers have argued that Islamic financial
institutions have huge potential over the conventional
banking model as an alternative finance to absorb macrofinancial shocks and promote economic growth (Dridi and
Hasan 2010; Mills and Presley 1999).
Cite this chapter as: Zulkhibri M (2015). Islamic financing and bank characteristics in a dual banking system: Evidence
from Malaysia. In H A El-Karanshawy et al. (Eds.), Islamic banking and finance – Essays on corporate finance, efficiency
and product development. Doha, Qatar: Bloomsbury Qatar Foundation
Developing Inclusive and Sustainable Economic and Financial Systems
Zulkhibri
In term of deposit, Islamic banks use mainly the risksharing PLS instruments, while in financing, most Islamic
banks rely on debt-like instruments (mark-up financing
and a guaranteed profit margin) that are based on deferred
obligation contracts. Moreover, conventional interest
rates (the London Interbank Offered Rate (LIBOR) or
a domestic equivalent) will always be a benchmark for
Islamic banks’ mark-up. As a result, in the case of such debtlike instruments, the pricing of Islamic financing is not a
function of real economic activity but is based on a predetermined interest rate plus a credit risk premium.
The objective of the paper is to investigate the determinants
of Islamic financing while taking into consideration bankspecific characteristics. Understanding this behaviour
indicates how efficiently Islamic banks perform their roles
as suppliers of capital for businesses and entrepreneurs.
However, little is known about the determinants of bank
financing which operate alongside conventional banks in
the dual banking system. Moreover, due to the fact that the
rate of return on retail PLS accounts closely follows interest
rates offered by conventional banks in Malaysia (Chong
and Liu 2009; Cervik and Charap 2011), the paper employs
a panel-pooled regression methodology by investigating
the cross-sectional differences in the way that Islamic
banks respond to base financing rates across bank-specific
characteristics.
2. Overview of Malaysian Islamic financial
industry
Malaysia’s Islamic finance industry has been in existence
for over 30 years. The enactment of the Islamic Banking
Act 1983 enabled the country’s first Islamic Bank to be
established. Malaysia’s overall strategy in the development
of Islamic banking can be summarized under four pillars:
100%
90%
1. A full-fledged Islamic banking system operating on a
parallel basis with a full-fledged conventional system
(dual banking system).
2. A step-by-step approach, in the context of an overall
long term strategy.
3. A comprehensive set of Islamic banking legislation
and a common Shariah Supervising Council for all
Islamic banking institutions.
4. A practical and open-minded approach in developing
Islamic financial interests.
An important feature in the implementation of Islamic
banking in Malaysia and creating a viable Islamic banking
system is that three basic elements have been adopted:
1. A large number of instruments and range of different
types of financial instruments must be available to
meet the different needs of different investors and
borrowers.
2.A large number of institutions with an adequate
number of different types of institutions participating
in the Islamic banking system to provide depth to the
Islamic banking system.
3. An Islamic interbank market to support an efficient
and effective system linking the system to the
institutions and the instruments.
As it can be observed in Figure 1, the share of Islamic assets
in the overall banking system is growing significantly, from
around 7% in 2006 to 20% in 2012. As at the end of 2012, the
country’s Islamic banking system had accumulated a total of
RM119 billion in assets, or about 20% of the total assets of
the banking sector, which is RM0.6 trillion. To date, Malaysia
has 16 Islamic Banks, which comprises nine local Islamic
Banks and seven foreign Islamic Banks. Figure 2 shows the
composition of the Islamic financing modes. It shows that
6%
7%
6%
4%
4%
4%
4%
3%
8%
14%
16%
17%
18%
20%
87%
86%
82%
80%
79%
78%
77%
2006
2007
2008
2009
2010
2011
2012
80%
70%
60%
50%
40%
30%
20%
10%
0%
Commercial Bank
Islamic Bank
Investment Bank
Source: Bank Negara Malaysia (2013)
Figure 1. Total assets: Islamic and conventional banks.
38
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Islamic financing and bank characteristics in a dual banking system: Evidence from Malaysia
100%
90%
19%
21%
16%
15%
18%
30%
29%
27%
20%
20%
26%
23%
80%
70%
60%
29%
30%
50%
40%
30%
20%
41%
37%
33%
32%
34%
32%
32%
2006
2007
2008
2009
2010
2011
2012
10%
0%
Bai Bithaman Ajil
Ijarah
Ijarah Thumma Al-Bai
Murabahah
Musyarakah
Mudharabah
Istisna'
Others
Source: Bank Negara Malaysia (2013)
Figure 2. Composition of Islamic financing modes.
the Bai Bithaman Ajil and Ijarah Thumma Al Bai dominate
the composition with 32% and 23%, respectively, over the
period 2006–2012. This dominant trend of using mark-up
and debt-like instruments in Islamic financing practices
support some of the arguments that Islamic banking is akin
to conventional banking in practical terms.
As shown in Figure 3, the data reveal a high degree
of correlation between the base financing rate on retail
7.00
financing and conventional lending rates on loans in
Malaysia. Between 2009 and 2012, the correlation of
base lending rate of conventional banks and the Islamic
base financing rate was about 76 percent. Accordingly,
despite the fact that conventional and Shariah-compliant
Islamic banks operate in different banking environments,
it is surprising that the Islamic base financing rate closely
tracks interest rates offered by conventional banks in
Malaysia.
Correlation=0.76
6.50
Islamic
6.00
5.50
5.00
Conventional
4.50
2009
2010
2011
2012
Source: Bank Negara Malaysia (2013)
Figure 3. Islamic rate of return and conventional average lending rate.
Eds. Hatem A. El-Karanshawy et al.
39
Zulkhibri
3. Literature review
Islamic banking is different from conventional banking
from a theoretical perspective because interest (riba) is
prohibited in Islam (rate of return on deposits cannot be fixed
by the bank and interest cannot be charged on loans). The
prohibition of interest is combined with the common belief
that banks channel funds towards productive investment,
which makes Islamic banking and Western economic
theory inconsistent with each other. A unique feature of
Islamic banking is the PLS paradigm, which is largely based
on the mudaraba (profit-sharing) and musharakah (equity
participation) concepts of Islamic contracting.
More recent literature on Islamic finance tries to establish the
difference between Islamic rates of return and conventional
banks interest rates based on empirical assessments. Cervik
and Charap (2011) compare the empirical behaviour of
conventional bank deposit rates and the rate of returns
on retail Islamic PLS investment accounts in Malaysia and
Turkey. The findings show that conventional bank deposit
rates and PLS rate of returns exhibit co-integration in the
long-run, and that conventional bank deposit rates cause
returns on PLS accounts. Moreover, the time-varying
volatility of conventional bank deposit rates and PLS
returns is correlated and is statistically significant.
The concepts of Islamic finance in using the rate of returns as
a replacement for interest can be divided into two strands of
argument. The idealist literature attempts to look at the key
concepts of Islamic finance such as PLS, money, interest and
profit from an ideal perspective. A pre-determined return to the
lender, dependent on the borrowing period and independent
of the borrower’s uncertainty, is not permissible under Islamic
banking. This means that the ideal and most ‘Islamic’ form
of each concept should be accepted as valid. Much of the
literature on Islamic banking and finance in the 1960s and the
theoretical studies on Islamic banking fall under this category.
Such correlations have been observed in other studies. In
the case of Malaysia, Chong and Liu (2009), for example,
find that retail Islamic deposit rates mimic the behaviour
of conventional interest rates. The study shows that only
a small portion of Islamic bank financing is strictly PLSbased, and that Islamic deposits are not interest-free,
but are very much pegged to conventional deposits. The
findings also suggest that the Islamic resurgence worldwide
drives the rapid growth in Islamic banking rather than the
advantages of the PLS paradigm, implying that regulations
similar to those of conventional banks should be applied for
the Islamic bank.
Another line of argument based on maslaha-oriented
literature would be at the extreme end of the continuum.
According to this view, riba should not be interpreted in a
simplistic fashion as modem bank interest. Any interestbased bank could theoretically be an Islamic bank provided
that the Islamic ideals of justice, equity, fairness, nonexploitation were its guiding principles; humane terms of
providing finance to those ‘needing’ them were practised;
and, it provided one way of helping the economically
disadvantaged classes of society to raise their standard of
living. Nonetheless, it can be seen from this that there has
been a gradual shift from the idealist position to a more
pragmatic, mark-up based and less risky version.
Similarly, Kasri and Kassim (2009) examined the
relationship between investment deposits and rate
of return, including interest rate for Islamic banks in
Indonesia over the period 2000 to 2007. Using a vector
autoregressive model (VAR) model, the study reveals that
the mudaraba investment deposit in the Islamic banks are
co-integrated with return of the Islamic deposit, interest
rate of the conventional banks’ deposit, number of Islamic
banks’ branches, and national income in the long-run. The
finding also suggests that rate of return and interest rate
move in tandem, indicating that Islamic banks in Indonesia
are exposed to benchmark risk and rate of return risk.
In the conventional literature, the interest rate has long
been recognized not only by classical and neo-classical
economists, but also by contemporary economists as
one of the factors that determine the level of savings in
the economy, and that the interest rate has a positive
relationship with savings. However, Haron (2001) found
similar positive relationship behaviour for the profit rate
declared by Islamic banks. In other words, Islamic bank
customers are guided by the profit maximization theory
since there is no pre-determined rate of return involved
in the Islamic banking system. Since depositors at Islamic
banks possess similar attitudes to those at the conventional
banks, the interest rate will continue to have an influence
on the operations of Islamic banks.
On a similar note, while funding activities are carried out
mainly through the participatory PLS model, it is wellestablished in the literature that Islamic banks follow their
conventional counterparts in creating assets through nonPLS, debt-like instruments with a pre-determined, fixed
rate of return; in line with the findings of Beck et al. (2010),
there are “few significant differences in business orientation,
efficiency, asset quality or stability” between conventional
and retail Islamic banks. As a result, given the implicit link
to interest rates on the asset side of the balance sheet, PLS
rates of return follow conventional bank deposit rates.
40
In practice, the main explanation of the similarity between
Islamic bank profit rate and conventional banks can be
attributed to the differences in perceptions of riskiness
(theoretically and practically) at the institutional and
systemic level, particularly on the asset side. In addition,
Islamic banks lose on the grounds of liquidity, assets
and liabilities concentrations and operational efficiency,
whereas they tend to win in the field of profitability.
Nevertheless, Islamic banking could provide a further
guarantee, albeit still marginal, against systemic risks in
certain emerging financial markets.
4. Data and estimation methodology
The study employs a panel of annual bank level data of
all Islamic banks operating in the Malaysia covering the
period 2006–2012. The financial statements of Islamic
banks operating in Malaysian Islamic banking sectors are
collected from the Bankscope database of Bureau van Dijk’s
company. The macroeconomic variables: consumer price
index, real gross domestic product, Islamic base financing
rate, and monetary policy rate are taken from various issues
of Quarterly Statistical Bulletin published by Central Bank
of Malaysia.
Table 1 reports the basic descriptive statistics for the
sample:
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Islamic financing and bank characteristics in a dual banking system: Evidence from Malaysia
Table 1. Descriptive statistics.
Variable
Financing
Assets
Liquidity
Capital
GDP
Base Financing
Rate
Overnight Rate
Prices
# of
Obs. Mean
Std.
Dev.
Min
Max
69
71
70
66
119
119
1.28
0.95
0.04
0.04
0.08
0.05
7.41
11.83
-0.02
-0.02
20.17
4.51
16.57
16.96
0.20
0.24
20.42
4.66
119
119
14.15
14.91
0.08
0.09
20.32
4.60
2.92
3.29
0.46
0.24
Using a panel of information for 17 individual banks,
we initially estimate the benchmark model for Islamic
bank financing to allow for asymmetric response to bank
characteristics and monetary policy. This has usually been
done by introducing interaction terms between Islamic base
financing rate and bank discriminatory variables. Beside
these variables, we control for economic activities and
consumer prices, which allow us to control for demandside effects on Islamic bank financing. By combining time
series of cross-sectional observation, panel data give more
informative data, more variability, less co-linearity among
the variables, more degrees of freedom and more efficiency
(Gujarati and Sangeetha 2007). The panel-data estimation
method of both pooled-regression and fixed-effect model
is preferred. Fixed-effects specification is mainly used to
account for time-invariant unobservable heterogeneity that
is potentially correlated with the dependent variable. To test
for estimation robustness of the models, we employ randomeffect estimations and use all diagnostic tests to validate the
models. Our baseline model specification is as follows:
l
ΔFIN it = m i + ∑ β j ΔBFRt- j + γ 0 SIZE it-1 + ω 0 LIQUIDITYit-1
l
+ϕ 0CAPITALit-1 + ∑ γ j ΔBFRt- j * SIZE it-1
j =1
l
+ ∑ ω j ΔBFRt- j * LIQUIDITYit-1
j =1
l
+ ∑ ϕ j ΔBFRt- j * CAPITALit-1
j =1
l
l
j =1
j =1
+ ∑κ j ΔGDPt-1 + ∑ l j ΔPRICESt-1 + υ i + e i ,t
(1)
where ∆ is the first-difference operator, FIN is the Islamic
banks financing, GDP is the logarithm of real GDP, PRICES
is the logarithm of consumer price index, BFR is the Islamic
financing rate, SIZE, LIQUIDITY and CAPITAL are the bank
size, liquidity and capitalisation respectively. The subscript
i denotes banks where i = 1, …, N; t denotes time where
t = 2006–2012; ni denotes individual bank effects and ei,t
denotes error-term.
Eds. Hatem A. El-Karanshawy et al.
Size:
SIZE it = ln A it
2.00 3.50
2.93 3.70
Note: Financing, Assets, GDP and Prices are in logarithmic
forms; Liquidity is defined as a ratio of liquid assets (cash
and short-term funds) to total assets; Capital is defined as
ratio of capital and reserve to total assets.
Source: Author’s own computation from Bankscope.
j =1
The choice of bank-specific characteristics is based on
the theoretical assumption that a certain type of bank
is expected to be more responsive to financing shocks
since it operates in a dual banking system, and these
characteristics are widely used in the empirical literature.
Following Gambarcota (2005), the three measures for
bank characteristics of size (SIZE), liquidity (LIQUIDITY)
and capitalisation (CAPITAL) are defined as follows:
∑
-
N
i=1
ln A it
NT
s
Liquidity:
LIQUIDITYit =
N
LAit  T ∑ i=1 LAit / Ait 
 /T
-∑
Ait  t=1
NT

ps
Capitalisation:
CAPITALit =
s
N
K it  T ∑ i=1 K it / Ait 
 /T
-∑
Ait  t=1
NT

Bank size (SIZE) is measured by the logarithm of total
assets (A). Relatively, banks with a smaller size may face
higher constraints in raising external funds, thus forcing
them to reduce their lending (Kasyhap and Stein 1995,
2000). Liquidity (LIQUIDITY) is measured by the ratio of
liquid assets (cash and short-term funds) to total assets
(LA). More liquid banks can draw down on their liquid
assets to shield their financing portfolios and are less likely
to cut back on financing in the face of rising cost or rate
of return. Capitalisation (CAPITAL) is measured by the
ratio of capital and reserve to total assets (K). Since raising
bank capital is costly, the bank tends to adjust the lending
behaviour to meet the required level of capital. In the face
of a rising rate of return, a bank’s cost of financing rises
while the remuneration of bank assets remains the same.
Hence, the financing of highly-leveraged banks is expected
to be more responsive to changes in the rate of return than
the financing of well-capitalised banks (Kishan and Opiela
2006).
All three criteria are normalised with respect to their average
(NT) across all the banks in the respective sample in order to
get indicators that sum to zero over all observations. For the
Eqation (1), the average of the interaction term (∆BFR*SIZE,
∆BFR*LIQUIDITY and ∆BFR*CAPITAL) is, therefore, zero
and the parameters are directly interpretable as the overall
Islamic rate of return effect on Islamic bank financing. To
remove the upward trend in the case of size (reflecting the
fact that size is measured in nominal terms), or the overall
mean in the case of liquidity and capitalisation, the bank
characteristic variables are defined as deviations from their
cross-sectional means at each time period.
The assumption is that small, less liquid and poorly
capitalised banks react more strongly to changes in base
financing rate. This would correspond to a significant
41
Zulkhibri
positive coefficient for the interaction terms ∆BFR*SIZE,
∆ BFR*LIQUIDITY, and ∆BFR*CAPITAL, and means that
banks with these characteristics reduce their financing
growth rate more strongly in response to a restrictive shock
of base financing rate than do larger, more liquid and wellcapitalised banks.
Since Islamic banks operate in the dual banking system,
conventional interest rates may influence the Islamic
bank financing behaviour. Equation (1) may represent the
overall effect of Islamic bank financing without monetary
policy. Huang (2003) argues that, under the conventional
system, changes in interest rates have a larger effect on
bank loans supplies because banks’ ability to insulate their
financing supplies from changes in monetary policy will be
restricted, in particular, during periods of tight monetary
conditions. We try to test this hypothesis for Islamic
financing behaviour by including monetary policy rate,
where MP is monetary policy shock proxy by overnight
policy rate in Equation (2) and estimate the following
model:
ΔFIN it
l
= m i + ∑ β j ΔBFRt- j + γ 0 SIZE it-1 + δ 0 LIQUIDITYit-1
j =1
l
+ ϕ 0CAPITALit-1 + ∑ γ j ΔBFRt- j * SIZE it-1
j =1
l
l
+ ∑ ω j ΔBFRt- j * LIQUIDITYit-1 + ∑ ϕ j ΔBFRt- j * CAPITALit-1
j =1
l
l
l
j =1
j =1
j =1
+ ∑ ζ MPt- j + ∑κ j ΔGDPt-1 + ∑ l j ΔPRICESt-1 + υ i + e i ,t
j =1
(2)
5. Empirical results
Table 2 reports the results for our benchmark model of
Islamic bank financing, while Table 3 to Table 4 report
the results from fixed-effect and random-effect. The direct
impact of changes in the base financing rate on bank
financing is negative and significant. The coefficients for
base financing rate range from 1.78 to 5.47, which means
that an increase of base financing rate by one percentage
point leads to a decrease in the bank financing in the range
of between 1.7% to 5.5%. The result of our benchmark
models in line with the basic theoretical prediction is
similar to the bank lending channel of conventional banks
(Ehrmann et al. 2003). Since the Islamic rate of return
implicitly tracks interest rates offered by conventional
banks (Chong and Liu 2009), the results also explain
that the reduction in Islamic bank deposits may not
be completely substituted by other forms of financing
in order to continue to meet financing demand, thus
leading to a reduction in Islamic bank financing. The
results for fixed-effect and random-effect provide similar
observations, albeit with a lower impact of base financing
rate on bank financing between -0.21 to -1.76. The
estimated regression equations for all models explain
the behaviour of financing in the range of 30% - 97%. All
diagnostic tests confirm the good fit of the models.
The results from Table 2 to Table 4 also show the
importance of bank-specific characteristics with respect to
42
the bank lending behaviour. The variable of SIZE is positive
and highly significant for all models. In the fixed-effect
and random-effect model, SIZE is positive and significant,
ranging from 0.51 to 1.39. Larger banks might be more
efficient due to scale economies, while the theoretical and
empirical literature on the relationship between size and
stability is ambiguous (Beck et al. 2013). This suggests
that size is an important factor characterising the banks’
financing reaction with large banks being expected to
minimise cost. This finding is also consistent with Fadzlan
and Zulkhibri (2009), who suggest that larger financial
institutions in Malaysia attain a higher level of technical
efficiency in their operations and exhibit an inverted
U-shape behaviour.
In the case of the liquidity characteristic, the results show
that the coefficient of LIQUIDITY is positively associated
and highly significant with bank financing, and is between
1.16 and 9.47. Only banks that have a larger share of liquid
assets, or that are bigger, are able to shield their lending
relationships. This evidence points to the fact that Islamic
banks are able to protect their financing portfolios by
drawing down on their liquid assets and are, therefore, less
likely to cut on financing, whereas the latter have better
access to external finance due to their size in order to retain
their preferred liquidity ratio. This finding also implies that,
in periods of rising base financing rate, a borrower from a
less liquid bank, on average, tends to suffer from a sharp
decline in financing more than does a customer of a more
liquid bank. The result is in line with the findings by Brooks
(2007) that liquidity is the main determinant explaining
credit supply in Turkey.
Looking at the coefficient of capitalisation, CAPITAL, it
appears that bank financing is positively associated with
bank capitalisation, or the bank capital structure. The
results suggest that market participants may perceive
highly capitalised banks as being less risky (Kishan and
Opiela 2000). Consequently, it should be more expensive
for poorly-capitalised banks to finance externally. Such
poorly-capitalised banks try to avoid the cost of falling
below the regulatory minimum capital requirements or
the increased risk of violating the capital requirement by
holding capital buffers and asset buffers (short-term riskweighted assets rather than customer financing) that
can be liquidated if the bank runs into problems with the
capital requirement. The more short-term risk-weighted
assets (other than customer loans) the bank holds on its
balance sheet (i.e., the higher the bank’s asset buffer),
the lower the risk of violating the capital requirements
will be. The short-term risk-weighted assets will soon be
liquid, thereby reducing the capital requirement in the
near future. Also, the higher the bank’s capital buffer,
the lower the risk of violating the capital requirement
will be.
The macroeconomic variables included in the bank
financing models control for the demand-side effect, and
only the real GDP growth variable is significant in the
equation, where it has a positive coefficient. The response
of credit to economic activity is consistent with expectation.
The facts that the coefficient of real GDP is significant may
imply that the economic activities are taken into account in
financing decisions in an important way. On the other hand,
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Islamic financing and bank characteristics in a dual banking system: Evidence from Malaysia
Table 2. OLS estimation: Islamic bank financing and characteristics models.
Bank-specific characteristics
ΔFINANCEit
ΔBFRt
Bank-specific char.
SIZE
LIQUIDITY
CAPITAL
Impact of BFR
ΔBFRxSIZE
ΔBFRxLIQUIDITY
SIZE
SIZE
-1.841***
(0.165)
-1.787***
(0.259)
1.110***
(0.050)
1.113***
(0.052)
–
–
-0.145***
(0.021)
–
ΔBFRxCAPITAL
–
Impact of MP
ΔONt
–
Control Variables
ΔGDPt
ΔPRICESt
R-square
1.495*
(0.635)
-1.311
(4.105)
0.848
–
–
-0.146***
(0.021)
–
–
0.053***
(0.194)
1.142*
(0.710)
-1.828
(4.561)
0.848
LIQUIDITY
LIQUIDITY
CAPITAL
CAPITAL
-3.156***
0.272
-3.509***
0.516
-4.363***
(0.155)
-5.478***
(0.681)
–
2.282***
(1.827)
–
–
-4.591***
(0.657)
–
–
6.227*
(4.329)
-6.205
(9.318)
0.332
–
1.811***
(1.924)
–
–
-4.489***
(0.671)
–
0.052***
(0.194)
8.363*
(4.969)
-9.292
10.106
0.341
–
–
–
–
3.065***
(0.884)
2.887***
(0.873)
–
–
–
–
-1.806***
(0.550)
–
8.636*v
(2.115)
-2.899
(16.869)
0.771
-1.032***
(0.542)
1.237***
(0.737)
9.778**
(7.256)
-2.277
(18.852)
0.664
Notes: Standard errors are reported in parenthesis. Standard errors and covariances are White-heteroskedasticityconsistent. The subscript i denotes banks and the subscript t denotes time, where t = 2006–2012. Bank-specific
characteristics: SIZE is defined as logarithm of total asset; LIQUIDITY is defined as ratio of liquid asset (interbank
deposits and securities); CAPITAL is defined as capital and reserves to total assets; BFRt is the base financing rate;
MPt is the overnight interest rate; GDPt is the logarithm of real GDP; PRICESt is the logarithm of consumer price index;
*significant at 10%, **significant at 5%, and ***significant at 1%.
the price variable is negatively related to bank financing,
but is insignificant. The rise in inflation may be associated
with the variability of the inflation rate and will generate
uncertainty about the future return on investments. This,
in turn, discourages firms from undertaking investments
which, consequently, reduces their financing demand.
However, the price variable is insignificant for the results
of all regressions.
Due to the potential interrelations between bank
financing and conventional interest rate, all bank
financing models are run with overnight policy rate
(ON). The coefficients in all regressions are negatively
related to bank financing and vary within a reasonable
magnitude (0.05 to 0.52), but are broadly lower than
the base financing rate. The results of these regressions
suggest that the reaction of banks to changes in interest
rates remains the same as the change base financing rate
and is robust in terms of a different type of econometric
Eds. Hatem A. El-Karanshawy et al.
specifications. This finding broadly supports the findings
by Chong and Liu (2009), Cervik and Charap (2011),
and Beck et al. (2013) that there is no significant
difference in bank financing behaviour with respect to
interest rates. Furthermore, Kasri and Kassim (2009)
confirm that the conventional interest rate is one of
the determinants for saving deposits in Indonesia. This
evidence explains why the bulk of Islamic bank financing
is based on the mark-up principle and is very debt-like
in nature (i.e. murabaha and ijarah), rather than using
the principle of PLS. Despite their operations being
different from those of conventional banks, Islamic
banks seem to face asymmetric information, severe
adverse selection and moral hazard problems similar to
those of their counterparts in their attempts to provide
funds to entrepreneurs. However, the use of debt-like
instruments is a rational response on the part of Islamic
banks to informational asymmetries in the environments
in which they operate.
43
Zulkhibri
Table 3. Fixed effect estimation: Islamic bank financing and characteristics models.
Bank-specific characteristics
ΔFINANCEit
ΔBFRt
Bank-specific char.
SIZE
LIQUIDITY
CAPITAL
Impact of BFR
ΔBFRxSIZE
ΔBFRxLIQUIDITY
SIZE
-0.208***
(0.128)
-0.294***
(0.194)
0.595***
(0.036)
0.508**
(0.554)
–
–
-0.222**
(0.010)
–
ΔBFRxCAPITAL
–
Impact of MP
ΔONt
–
Control Variables
ΔGDPt
ΔPRICESt
Constant
Firm Dummy
R-square
F-statistic
SIZE
–
–
-0.021**
(0.159)
–
–
-0.043***
(0.017)
LIQUIDITY
LIQUIDITY
CAPITAL
CAPITAL
-0.470***
(0.415)
-1.764***
(0.516)
-0.820***
(0.171)
-0.339***
(0.186)
–
1.792***
(1.894)
–
–
-1.605**
(0.511)
–
–
1.193*
(0.746)
-0.844
(2.197)
7.588***
1.538*
(1.207)
-0.357
(2.385)
8.784
5.170*
(3.116)
-0.361
(7.153)
10.902***
(0.357)
(7.917)
(0.937)
Yes
0.972
72.29
Yes
0.972
66.77
Yes
0.738
5.989
–
1.161**
(0.815)
–
–
-1.458*
(1.108)
–
-0.528*
(0.317)
5.021*
(3.410)
-0.305
(7.446)
19.164***
–
–
–
–
6.067**
(2.354)
8.815**
(2.526)
–
–
–
–
-3.011**
(1.599)
–
-3.671**
(1.335)
-0.368**
(0.169)
5.386**
(2.494)
-0.980
(7.967)
18.058***
8.222**
(4.319)
-0.229
(9.035)
17.516***
(16.003)
(2.494)
(2.553)
Yes
0.751
5.856
Yes
0.735
5.231
Yes
0.743
4.995
Notes: Standard errors are reported in parenthesis. Standard errors and covariances are White-heteroskedasticityconsistent. The subscript i denotes banks and the subscript t denotes time, where t = 2006–2012. Bank-specific
characteristics: SIZE is defined as logarithm of total asset; LIQUIDITY is defined as ratio of liquid asset (interbank
deposits and securities); CAPITAL is defined as capital and reserves to total assets; BFRt is the base financing rate;
MPt is the overnight interest rate; GDPt is the logarithm of real GDP; PRICESt is the logarithm of consumer price
index; *significant at 10%, **significant at 5%, and ***significant at 1%.
To analyze further the impact of banks reducing their
bank financing in response to a change in base financing
rate with respect to specific characteristics, we have
interacted the base financing rate variable with bank
size (∆BFR*SIZE), liquidity (∆BFR*LIQUIDITY) and
capitalization (∆BFR*CAPITAL); this is to investigate the
economic arguments that there is a unique role for Islamic
banks in the dual banking system, and the importance of
heterogeneity among banks. Tables 2, 3 and 4 also report
the results of the base financing rate with respect to bankspecific characteristics for bank financing using a fixed-effect
and random-effect model. The estimates of bank-specific
characteristics coefficients provide interesting results. The
estimate coefficients of BFR*LIQUIDITY, BFR*CAPITAL and
BFR*SIZE consistently show a positive sign, and are highly
44
significant at the conventional level. These results suggest
that banks’ lending and their ability to obtain other sources
of funding are factors that are affected indirectly through
bank-specific characteristics.
In summary, we find strong evidence of the asymmetric
adjustment of bank financing through bank-specific
characteristics, as reported in the literature for conventional
banks and in line with the arguments of Kashyap and Stein
(1995, 2000) and Kishan and Opiela (2000). Moreover,
banks react differently to the base financing rate depending
on their own specific characteristics; a bank with higher
capitalisation, in particular, is expected to increase
financing more than a bank with greater size and liquidity.
Furthermore, since an Islamic bank is operating in a dual
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Islamic financing and bank characteristics in a dual banking system: Evidence from Malaysia
Table 4. Random effect estimation: Islamic bank financing and characteristics models.
Bank-specific characteristics
ΔFINANCEit
ΔBFRt
Bank-specific char.
SIZE
LIQUIDITY
CAPITAL
Impact of BFR
ΔBFRxSIZE
ΔBFRxLIQUIDITY
SIZE
-0.368***
(0.028)
-0.324***
(0.025)
0.618***
(0.035)
1.399***
(0.496)
–
–
-0.221**
(0.010)
–
ΔBFRxCAPITAL
–
Impact of MP
ΔONt
–
Control Variables
ΔGDPt
ΔPRICESt
Constant
Firm Dummy
R-square
F-statistic
SIZE
–
–
-0.244**
(0.143)
–
–
-0.041**
(0.021)
1.222*
(0.842)
-0.657
(2.191)
7.101***
(0.346)
1.423*
(1.112)
-1.394
(0.502)
4.052
(7.067)
Yes
0.908
93.58
Yes
0.904
72.83
LIQUIDITY
LIQUIDITY
CAPITAL
CAPITAL
-0.452***
(0.214)
-0.334***
(0.215)
-0.377***
(0.142)
-0.371***
(0.105)
–
9.471***
(1.805)
–
–
-1.539***
(0.102)
–
–
5.365*
(4.104)
-0.698
(7.112)
10.873***
(0.861)
Yes
0.394
6.256
–
4.639**
(2.858)
–
–
-0.189***
0.075
–
-0.472**
(0.231)
9.664*
(4.380)
-0.808
(7.406)
13.830
(14.852)
Yes
0.413
5.515
–
–
–
–
7.307***
(2.284)
8.521***
(2.331)
–
–
–
–
-3.406***
(1.129)
–
6.131*
(3.517)
-3.444
(2.066)
17.810***
(2.297)
Yes
0.400
5.455
-3.667**
(1.777)
-0.323***
(0.036)
8.530*
(5.198)
-5.773
(8.947)
17.412
(2.340)
Yes
0.409
4.968
Notes: Standard errors are reported in parenthesis. Standard errors and covariances are White-heteroskedasticityconsistent. The subscript i denotes banks and the subscript t denotes time, where t = 2006–2012. Bank-specific
characteristics: SIZE is defined as logarithm of total asset; LIQUIDITY is defined as ratio of liquid asset (interbank
deposits and securities); CAPITAL is defined as capital and reserves to total assets; BFRt is the base financing rate;
MPt is the overnight interest rate; GDPt is the logarithm of real GDP; PRICESt is the logarithm of consumer price
index. *significant at 10%, **significant at 5%, and ***significant at 1%.
banking system, the asymmetric information problems
faced by Islamic banks are expected to affect the ability to
protect the financing lines from policy-induced reductions
in deposits, and result in Islamic bank financing behaviour.
6. Conclusion
A significant number of empirical studies have explored the
bank lending behaviour of conventional banks over the past
decades, while studies on Islamic bank financing behaviour
remain scarce due to lack of bank-level data. Under the dual
banking system, to the extent that financial constraints vary
with banks’ ability to access other sources of financing, the
implication is that Islamic bank financing responses to the
bank financing rate and the conventional interest rate are
Eds. Hatem A. El-Karanshawy et al.
contingent on observable bank-specific characteristics.
Understanding this mechanism is crucially important, in
the context that Islamic banks have increasingly played a
dominant role in the Malaysian financial system.
This paper analyses the importance of bank-specific
characteristics with respect to Islamic bank financing
in Malaysia. The results obtained from pooled panel
estimation allow us to draw several significant conclusions
about Islamic bank financing behaviour in Malaysia within
a dual banking system. The evidence gathered in this study
suggests that bank-specific characteristics are important
for Islamic banking financing behaviour. The financing
behaviour of Islamic banks is consistent with the behaviour
of conventional banks in that the bank lending operates
45
Zulkhibri
according to banks’ size, and level of liquidity and capital
(Goldniuk 2006). The results of these regressions also
suggest that the reaction of Islamic banks financing to
changes in interest rates is the same as for conventional
banks, and are robust to different types of econometric
specifications.
Many problems and challenges relating to Islamic
instruments, financial markets, and regulations must
be addressed and resolved. A complete Islamic financial
system with its identifiable instruments and markets is
still at a relatively early stage of evolution. The functioning
of Islamic banks should rapidly differentiate itself from
conventional banking. Due to the existence of moral
hazards and adverse selection in the industry, an Islamic
bank is not able to provide a full-fledged alternative
finance to conventional finance. Moreover, an Islamic bank
does not develop in the path that was envisioned by the
Islamic scholars (Saeed 1996). One of the drawbacks is
the low level of participation in PLS arrangements, which
seems contradictory to the essential concept of Islamic
banking. In practice, it would beneficial for Islamic banks
to stop replicating the conventional banking models that
concentrate mainly on debt-based instruments and markup models, but instead to move over to the PLS model.
References
Aggarwal R, Yousef T. “Islamic Banks and Investment
Financing.” Journal of Money, Credit, and Banking,
32(1), 2000. 93–120.
Bank Negara Malaysia, Quarterly Bulletins, Kuala Lumpur.
Beck T, Demirguc-Kunt A, Merrouche O. “Islamic vs.
Conventional Banking: Business Model, Efficiency and
Stability.” Journal Banking and Finance, 37(2), 2013.
433–447.
Brooks PK. “Does the Bank Lending Channel of Monetary
Transmission Work in Turkey?” IMF Working Paper
07/272, Washington DC: International Monetary Fund.
2007.
Cervik S, Charap J. “The Behavior of Conventional and
Islamic Bank Deposit Returns in Malaysia and Turkey.”
IMF Working Paper, No. WP/11/156, Washington DC:
International Monetary Fund, 2011.
El-Gamal M. “Mutuality as an Antidote to Rent-Seeking
Sharia-Arbitrage in Islamic Finance”. http://www.ruf.
rice.edu/~elgamal/. 2005.
Ehrmann M, Gambacorta L, Martinez-Pages J, Sevestre P,
Worms A. “Financial Systems and the Role of Bank in
Monetary Transmission in the Euro Area”. In I. Angeloni,
A. Kashyap, and B. Mojon, (eds.), Monetary Transmission
in the Euro Area: A Study by the Eurosystem Monetary
Transmission Network, Cambridge University Press,
2003. 235–269.
Fadzlan S, Zulkhibri M. “Post-Crisis Productivity Change
in Non-Bank Financial Institutions: Efficiency Increase
or Technological Progress?”Journal of Transnational
Management, 14(2), 2009. 124–154.
Haron S. “Islamic banking and finance.” Leading issues in
Islamic Banking and Finance, 20 (1), 2001. 17–32.
Huang Z. “Evidence of Bank Lending Channel in the U.K.”
Journal of Banking and Finance, 27, 2003. 491–510.
Gambacorta L. “Inside the Bank Lending Channel.”
European Economic Review, 49, 2005. 1737–1759.
Gujarati DN, Sangeetha S. “Basic Econometric.” 4th Edition,
McGraw-Hill Education Books Ltd., India, 2007.
Golodniuk I. “Evidence on the Bank-Lending Channel
in Ukraine.” Research in International Business and
Finance, 20, 2006. 180–99.
Kasri RA, Kassim, S. “Empirical Determinants of Saving of
the Islamic Banks in Indonesia.” Journal of King Abdul
Aziz University, Islamic Economics, 22(2), 2009.
Khan M, Mirakhor A. “Islamic Banking”. The New Palgrave
Dictionary of Money and Finance, London: Macmillan
Press, 1992.
Kashyap A, Stein J. “The Impact of Monetary Policy on Bank
Balance Sheets.” Carnegie Rochester Conference Series
on Public Policy, 1995. 51–195.
Kashyap A, Stein J. “What Do a Million Observations Say
About the Transmission Mechanism of Monetary Policy.”
American Economic Review, 90(3), 2000. 407–428.
Kishan P, Opelia T. “Bank Capital and Loan Asymmetry
in the Transmission of Monetary Policy.” Journal of
Banking and Finance, 30, 2006. 259–285.
Chong B, M-H Liu. “Islamic Banking: Interest-Free or
Interest-Based?” Pacific- Basin Finance Journal, 17,
2009.124–144.
Khan M. “Time Value of Money and Discounting in Islamic
Perspective.” Review of Islamic Economics, Vol. 1, No.
2, 1991. 35–45.
Chapra MU. “The Future of Economics: An Islamic
Perspective.” The Islamic Foundation, Leicester, UK,
2000.
Mills P, Presley J. “Islamic Finance: Theory and Practice.”
London, UK: Macmillan, 1999.
Dridi J, Hasan M. “Have Islamic Banks Been Impacted
Differently than Conventional Banks During the Recent
Global Crisis?” IMF Working Paper, No. WP/10/201,
Washington DC: International Monetary Fund, 2010.
46
Saeed A. “Islamic Banking and Interest: A Study of
the Prohibition of Riba’ and its Contemporary
Interpretation.” Leiden: EJ Brill, 1996.
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Leverage risk, financial crisis, and stock returns:
A comparison among Islamic, conventional, and
socially responsible stocks
Vaishnavi Bhatt1, Jahangir Sultan2
Ramaiah Institute of Management Studies, Banglaore, India, Email: [email protected]
The Hughey Center for Financial Services, Bentley University, Email: [email protected]
1
2
We thank Marcia Cornet, Vishwanathan Iyer, Kartik Raman, and seminar participants at Bentley University for their
useful suggestions. We thank Dow Jones Indexes for providing us with information on proprietary indexes. We remain
responsible for all remaining errors. Please do not quote without permission.
Abstract - According to the financial press, firms with low leverage have lower distress risk due
to their reduced exposure to the credit market, especially during credit crises. Compared to their
conventional and socially responsible (SRI) counterparts, Shariah compliant (SC) stocks are
low-leverage stocks. Our hypothesis is that SC firms would be less sensitive to leverage risk and
thus would be ideal for wealth preservation during declining market environment. We find that
the leverage risk factor performs consistently across various categories of firms and its impact is
more pronounced during the recent financial crisis. However, we also find that compared to the
conventional stocks, SC stocks are also quite sensitive to the leverage factor. In contrast, the SRI
class of stocks has the least sensitivity to leverage risk factor, suggesting they can be attractive for
wealth preservation during credit crises.
Keywords: asset pricing, leverage, returns, financial crisis
JEL Codes: G00, G01, G10, G11, G12, G32
1. Introduction
As the divine code of law, the Shariah is a code of conduct
that guides business transactions for the Muslims and are
based on the Quran and the edicts of Prophet Muhammad
(pbuh). Hence, the guidelines set forth in the Shariah
become imperative to every Muslim and govern all aspects
of life, whether they may be of personal, social, political,
economic or financial nature. Shariah compliant (SC)
stocks1 are low-leverage stocks with high asset backing,
compared to their conventional and socially responsible
(SRI) counterparts. It is a widely held belief that firms
with low leverage have lower distress risk due to their
reduced exposure to the credit market. Naturally, these
firms are capable of promoting flight to safety, especially in
a declining market environment.
In this paper, we examine if SC stocks have lower sensitivity
to economy wide leverage risk. To this extent, we create a
new leverage risk factor (LEV) on the basis of firm-specific
financial leverage (total debt over assets)2. The risk factor
LEV (defined as the return on high leverage stocks minus
the return on low leverage stocks) is a non-diversifiable risk
premium and therefore should be included in any multifactor
asset pricing model. The evidence that high leverage requires
higher risk premium can be indicative of the notion that high
leverage can be value destructive, especially when equity
prices are falling in a persistent fashion. The fact that Islamic
stocks may have lower credit market exposure is important
for wealth preservation during both good and bad times.
Milly and Sultan (2009) report that Islamic stocks listed
globally have outperformed conventional stocks and SRI
stocks during the 2007–2009 economic crisis3. It would
be interesting to examine how these stocks respond to the
traditional risk factors (such as market risk premium, size,
and value) as well as the leverage risk factor. If indeed
Islamic stocks have lower sensitivity to the leverage risk
factor, it would be indicative of their attractiveness for wealth
preservation when investors are looking for safer assets.
In this paper, using a sample of 3704 globally traded stocks
for the period January 2000- April 2009, we construct a risk
factor based on firm-specific leverage find that the inclusion
Cite this chapter as: Bhatt V, Sultan J (2015). Leverage risk, financial crisis, and stock returns: A comparison among
Islamic, conventional, and socially responsible stocks. In H A El-Karanshawy et al. (Eds.), Islamic banking and finance –
Essays on corporate finance, efficiency and product development. Doha, Qatar: Bloomsbury Qatar Foundation
Developing Inclusive and Sustainable Economic and Financial Systems
Bhatt and Sultan
of the leverage risk factor leads to a weakening of the
significance of the traditional FF variables. Furthermore,
we show that, in comparison to the traditional FF factors,
the economic and statistical significance of the leverage
risk factor is high, especially during the financial crisis. We
also demonstrate that the leverage risk factor contributes to
the systematic risk of a firm and represents the underlying
macroeconomic fundamentals. Finally, we show that
compared to the conventional stocks, SC stocks display
substantially lower risk premium to traditional risk factors.
We also find that similar to the conventional stocks, Islamic
stocks are also sensitive to the leverage factor, thus leading
us to suggest that a leveraged based screening of Islamic
stocks may not be ideal for wealth preservation especially
during a credit crises. An investor must search for other
redeemable characteristics in Islamic stocks that can help
preserve equity value during falling equity prices.
The remainder of the paper is as follows. In Section II, we
review the link between leverage and stock returns. In
Section III, we discuss the recent financial crisis to motivate
the empirical model. In Section IV we offer empirical
results, and the final section concludes the paper.
2. Review of literature
A detailed analysis of the sensitivity of SC stocks to
the leverage risk is tricky. In the first place, one must
demonstrate that, in the context of a multifactor asset
pricing model, the previous risk factors are incapable of
capturing economy wide leverage risk. Once a reliable risk
factor is constructed, a researcher can proceed to the next
stage to investigate whether such risk factor is significant
in an asset pricing model. Finally, the analysis can proceed
to examine if there are differences in the way different
categories of firms respond to this newly created risk factor.
Consider the following multifactor asset pricing model
(Fama-French (1992))
rt - r ft = b 0 + b 1(rmt - r ft ) + b 2 Rt ,SMB + b 3 Rt ,HML + e t (1)
shows that excess return on a portfolio (rt – rft) is explained
by the sensitivity of its return to three factors: the excess
return on a broad market portfolio (rmt – rft); the difference
between the return on a portfolio of small stocks and the
return on a portfolio of large stocks (SMB, small minus
big); and the difference between the return on a portfolio
of high-book-to-market stocks and the return on a portfolio
of low-book-to-market stocks (HML, high minus low).
Our analysis thus leads us to first address an important
question which has largely been ignored in the literature.
Fama and French (1992) note that SMB (return on a
portfolio of small firms minus the return on a portfolio of
large firms) and HML (return on a portfolio of high book
to market firms minus return on a portfolio of low book
to market firms) are statistically important in explaining
the cross-section of equity returns. Subsequent work by
academics and practitioners has sought to verify the effects
of these factors (FF factors, from hereafter) on cross-section
of equity returns (for example, see Fama and French (1993,
1995, and 1998), Liew and Vassalou (2000), Davis, Fama
and French (2000), Sivaprasad and Muradoglu (2009),
48
and Vassalou and Xing (2004)). A common finding in the
literature is that value stocks earn a premium over growth
stocks. Similarly there is evidence that small sized stocks
earn a premium over big stocks.
These so-called empirical anomalies continue to generate
controversies in the literature. For instance, are value and
size premiums caused by the underlying risk factors of
firms falling within these categories? Similarly, the notion
of whether value and size premiums reflect incorrect
extrapolation of past earnings growth by the market and
subsequent correction of the mispricing errors, continues
to receive attention in the literature (see Eom and Park
(2008) for a recent survey).
How well do FF risk factors capture financial distress risk?
Fama and French (1992) note that the combination of book
to market and size describes the cross-section of average
stock returns and absorb the apparent roles of other variables
like leverage and E/P. The authors note that the SMB and
HML factors are correlated with leverage and, therefore,
adequately represent financial distress. The ability of the
traditional FF factors to directly capture leverage risk is
critical for asset management, especially when leverage
risk becomes a source of systemic risk in the economy. The
implication for an investor facing such catastrophic shocks
is simple. If size and value based strategies do not perform
consistently well across good and bad times, the rationale
behind such investing strategy is at risk.
However, Fama and French (1992 and 1993) deal with the
market leverage (assets over market value of equity) and
the book leverage (assets over book value of equity), which
may not directly capture the sensitivity of the firms to
economy wide leverage risk4. In particular, the debt market
exposure of a firm is a major determinant of the distress
risk that may not be directly captured by the FF factors.
Furthermore, to the extent that excessive leveraging and
major credit events can lead to correlated defaults, we
may find that the debt market exposure is monotonically
increasing in financial leverage. In essence, the resulting
credit crisis produces contagion-like effects with leverage
risk as being the primary catalyst. According to Fama and
French (1996), if default risk becomes correlated across
firms, market participants, especially workers in distressed
firms, tend to avoid all distressed firms in general. We
believe that this presents an ideal opportunity for volatility
spillover among firms in the economy, with the extent of
spillover monotonically rising in leverage.
Surprisingly, very few studies have empirically examined
the role of leverage risk factor in asset pricing. Chan and
Chen (1991) examine the effects of financial leverage
(book value of debt and preferred stock over market value
of equity) on stock returns and find a positive relationship.
Unfortunately, their analysis does not investigate if factor
loadings on the financial leverage can subsume the effects
of HML and SMB. As Fama and French (1992) write, “It
would be interesting to check whether loadings on their
distress factors absorb the size and book-to-market equity
effects in average returns documented here.” Ferguson and
Shockley (20003) write, “... a three-factor empirical model
that includes factors based on relative leverage and relative
distress should outperform the Fama and French (1993)
three-factor model in the cross section”.
Islamic banking and finance – Essays on corporate finance, efficiency and product development
A comparison among Islamic, conventional, and socially responsible stocks
An investigation into this topic is timely given the recent
financial crisis when economy-wide leverage played a key
role in exacerbating the risk exposure especially for the
leveraged5 financial and non-financial firms. As the subprime
crisis deepened, coupled with escalating liquidity crisis, the
credit market virtually dried up, limiting access to funds.
The TED spread (difference between the interest rates on
Eurodollar loans and short-term U.S. T-bill) rose in July 2007,
then spiked even higher in September 2008, reaching as high
as 4.65% on October 10, 2008. While the impact was felt
mostly by the hedge funds, insurance agencies, banks, and
firms directly involved in construction business and mortgage
lending, the effects of the liquidity crisis also had affected the
non-financial firms as well. Thus, the financial crisis in 2007–
2008 had a devastating contagion-like effect on credit risk,
with leverage risk acting as the centrepiece. An analysis of the
Islamic stocks and their conventional counterparts is critical
from the point of view of academic as well as the practitioner
community. If Islamic stocks have lower sensitivity to the
leverage risk factor, then these stocks would be ideal for
wealth management, especially during financial crises.
There are several studies on the relationship between
leverage and stock returns. See Chou, Ko, and Lin (2010)
for a recent survey. In one strand of the literature, leverage
is positively related to stock returns, especially for weak
firms with poor investment opportunities. Accordingly, as
debt increases the risk exposure of such firms, investors
demand a premium. Sivaprasad and Muradoglu (2009)
find that leverage has a significant positive relation with
stock returns. Gomes and Schmid (2009) show that equity
returns are increasing in market leverage. Ho, Strange
and Piesse (2008) conduct a similar study for the Hong
Kong stock exchange and conclude that market leverage
(Assets/Market value of equity) exhibit a significant
conditional relationship with the stock returns. Bhandari
(1988) performs cross sectional regressions between
monthly average returns and the leverage ratios for the
period 1948–1979 and finds that the debt equity ratio has
a positive effect on stock returns. Ferguson and Shockley
(2003) include relative leverage (D/E) and relative
distress risk, based on Altman’s Z score. They find that
their model performs better than the three factor FF model
in explaining stock returns. On similar lines, Chou et al
(2010) propose an augmented five factor model which
incorporates both FF factors as well as Ferguson and
Shockley factors and demonstrate that this augmented five
factor model explains most of the asset pricing anomalies.
In contrast, there are several studies that offer rationales
for supporting a negative relationship between financial
leverage and stock returns. The debt-overhang theory
(Meyers, 1977) provides a convenient framework to suggest
why leverage reduces equity return. Accordingly, as leverage
increases, the distress risk increases, and shareholders pass
up positive NPV projects. As a result, the stock price decreases,
reflecting underinvestment in successful projects and a
decline in firm value (Meyers (1977)). Other explanations
include firms substituting debt for equity especially during
economic crisis when the cost of equity financing is higher
than the cost of debt financing (Dimitrov and Jain (2006));
managerial preference for equity over debt because high
debt payments can reduce equity returns, especially when
firms do not take advantage of growth opportunities (Lang,
et al (1995)); the benefit of external disciplining mechanism
Eds. Hatem A. El-Karanshawy et al.
of debt financing (Jensen and Meckling (1976) and Fama
and Jensen (1983)); and a reduction of the manager’s ability
to waste free cash (Jensen (1986)). Overall, these studies
imply that debt reduces agency costs and managerial waste,
improves disclosure, and thus reduces equity risk premium.
As a result, leverage is decreasing in stock returns.
The previous discussion suggests that the leverage risk
factor is important for asset pricing models. Our focus in
this paper is to examine the extent to which the well-known
anomalies (size and book to market effects) are resolved
by directly adding leverage as a systematic risk factor.
Leverage risk becomes fundamental risk especially when
firms’ exposure to the debt market becomes pervasive and
correlated across the economy. Fama and French (1996)
recognize that investors avoid financially distressed firms
because distress risk is correlated across the economy.
We suggest that when leverage risk becomes correlated
across the economy, it has a contagion-like effect on firms
in general, especially those with high exposure to the
debt market. To this extent, while size and book to market
factors are correlated with the leverage of the firm, they
may not adequately capture the firm’s direct exposure to
the economy wide systemic risk due to excessive leverage.
Finally, to the extent that Islamic stocks tend to have low
leverage and are involved only in permissible economic
activities under the guidelines of the Quran and Sunnah,
may have reduced exposure to interest rate volatility.
This simple and powerful proposition has not been fully
addressed in the literature. If Islamic stocks continue to act
like their conventional counterparts, it only goes to reaffirm
the harmful effect of riba as firms take on more debt.
Our suggestion is consistent with the anecdotal evidence
from the recent financial crisis when leverage risk became
one of the primary drivers of the global economic crisis.
There was plenty of evidence of such systemic risk in the
recent financial crisis: debt markets such as the commercial
paper market, the repo market, and short-term bank
borrowing virtually dried up. Altogether, increased leverage
of firms, especially of hedge funds, insurance agencies,
banks, and mortgage companies, coupled with a liquidity
crisis, took a heavy toll on the global economy.
In the next section, we discuss the link between leverage
risk factor and selected macroeconomic variables such as
the industrial production, unemployment, inflation, credit
spread and term spread. Our intent is to draw inferences on
the effects of the leverage risk factor on stock returns across
various time periods.
3. Leverage risk and the financial
crisis—contemporary evidence
In 2004, the US Securities and Exchange commission
granted a waiver of the international standards of
maximum accounting leverage ratio6 (which was about
12) for five major securities firms – Goldman Sachs, Merrill
Lynch, Morgan Stanley, Lehman Brothers and Bear Sterns.7
Subsequently, many of the investment banks boosted their
leverage ratios to as high as 30. Mortgage giants Freddie
Mac and Fannie Mae had leverage levels close to 60 to 1
(2008 data), which can be very lucrative if the asset prices
rise, but is disastrous when asset prices fall. A recent
report8 cites excessively high leverage ratios prevailing in
49
Bhatt and Sultan
the housing market and the underlying mortgage backed
securities as the culprit behind the credit crisis. Towards
the end of the year 2009, the global economy was afflicted
with excessive indebtedness which adversely affected the
worldwide economy. For example, average household
sector debt increased 141 per cent of disposable income in
the United States and 177 percent in the United Kingdom.
Furthermore, the best known banks in the US and Europe
had their leverage (assets/equity) rising to forty, sixty or
even hundred times the size of their equity capital.9
There is a broad consensus that increased leverage affects
stock returns during the financial crisis. According to the
popular press10, under normal circumstances where stock
prices deviate from their underlying fundamentals, prices
tend to bounce back to their intrinsic values, thereby
restoring the efficiency of the equity markets. However,
during a prolonged crisis, price discovery process takes
longer, and stocks move away from their intrinsic values
for a longer period of time. In addition, when investors
are pessimistic about the financial markets, they may miss
out on profitable arbitrage opportunities as prices move.
In fact, due to the significant mispricing in the market, the
US subprime crisis caused share prices of various US and
European banks to fall and exerted immense pressure on
these banks in the form of deteriorating profit margins.11
From a balance sheet perspective, companies reduce their
leverage ratios either by selling off their assets (thereby
restructuring their balance sheets) or by issuing new
shares. Both of these strategies have different implications
on the expected returns from the investor perspective.
According to James Lee, Vice Chairman of JP Morgan,12 in
spite of the efforts by the financial sector to augment their
capital levels to as high as $300 billion firms have not been
able to bring down the leverage to pre-crisis levels.
While many financial institutions and asset managers
have been deleveraging since 2008, the process might
eventually diminish the ability of these institutions to
produce attractive returns, especially when they are unable
to grow their balance sheets. In such circumstances, as
financing gets costlier, firms focus on augmenting their
capital level rather than investing it. In this process—“The
big get bigger and the rest get smaller”13—has a direct
impact on the stock returns of these firms. In other words,
higher leverage levels increase the risk exposure of the
firms and present higher growth opportunities, which
should lead to higher stock returns. In contrast, lower
leverage levels shrink the balance sheet of the firm and also
reduce their competitiveness, having a negative impact on
the shareholder value and stock returns.
Leverage risk during the financial crisis has macroeconomic
implications. Notwithstanding the de-leveraging efforts of
banks and other financial institutions, as of November 6,
2009, banks in particular exhibited 40 to 1 leverage (assets
over equity capital). Similarly, the deleveraging efforts
undertaken by many governments have also led to adopting
restrictive monetary policy, resulting in higher interest
rates. However, analysts argue that increasing interest
rates and withdrawing funds from the financial system may
cause the economy to exacerbate the effects of the credit
risk. It has also been forecasted14 that deliberate attempts
by the governments to deleverage will lead to lower wages
in developed countries and a permanent unemployment
50
of 15% to 25%. Such macroeconomic instability has the
potential to push investors away from the stock and bond
markets. Furthermore, an increase in the perceived risk in
the financial markets would prompt investors requiring a
higher risk premium, which directly affects the expected
returns on these stocks. So, deleveraging could have
negative effects and is expected to reduce productivity.
Overall, leverage affects expected returns not as a firm
specific variable but as a systematic risk factor.
4. Empirical results
Our initial sample includes weekly data for approximately
4000 stocks from 55 countries from January 2000 to April
2009. Our sample includes both financials (banks, S&Ls,
credit unions, mortgage financing companies, real estate
firms, and insurance companies) and non-financial firms.
Since financial firms, especially banks and insurance firms,
operate with high leverage, we will also separate financials
from the aggregate sample to examine if financials stocks
have different sensitivity to the risk factors.
We eliminate stocks having negative book to market equity
from the sample in the construction of the risk factors.15 Also,
the number of stocks each year used in the construction of
factors varies depending on the availability of data for
the corresponding year. This eliminates the problem of
survivorship bias in the sample. The data for the weekly
stock returns are extracted from Datastream, while the data
related to economic fundamentals like size, book to market
equity and leverage are extracted from FactSet. Stock returns
are in US dollar terms and are based upon log relatives of
weekly stock prices. The Dow Jones Global Index is used as
the market benchmark, and the US risk-free rate is used as
a proxy for global risk free rate16. We use previous year-end
fundamentals to form portfolios for each successive year;
the rationale behind this is that investors use information
contained in the balance sheets and financial statements to
predict future returns. Investors are assumed to follow a buy
and hold policy with annual portfolio rebalancing.
Construction of risk factors
We sort all stocks in the sample by size, book to market and
leverage and categorize them in 3 groups (top 30%, middle
40%, and bottom 30%). Using the independent sorting
procedure we construct value weighted portfolios formed
by the intersection of three portfolios based on size, three
portfolios based on book to market equity and three portfolios
based on leverage (Debt/Assets). In all, we have 3*3*3 = 27
portfolios. The returns on these annually rebalanced
portfolios create the dependent variable. In addition to
the XMKT (market risk premium), the FF factors are: SMB
(size mimicking portfolio constructed each week by taking
the simple average of the returns on small sized portfolios
minus returns on big sized portfolios), HML (book to market
mimicking portfolios constructed each week by taking the
simple average of the returns on high book to market portfolios
minus the returns on low book to market portfolios) and
LEV (leverage mimicking portfolios constructed each week
by taking simple average of the returns on high leveraged
portfolios minus the returns on low leverage portfolios).
Table 1 reports the number of stocks used for the construction
of factors and portfolios each year which varies depending
on the availability of data and meeting specific requirements
Islamic banking and finance – Essays on corporate finance, efficiency and product development
A comparison among Islamic, conventional, and socially responsible stocks
Table 1 A. Number of stocks.
Year
No. of stocks
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
3745
4344
4378
4379
4396
4403
4399
4399
4406
4391
Table 2A. Correlation matrix of the explanatory factors
for—all stocks (Jan 2000–April 2009).
XMKT
SMB
LEV
HML
SMB
1.000
-0.087
0.031
-0.057
-0.087
1.000
-0.030
0.086
LEV
HML
0.031
-0.030
1.000
0.451
-0.057
0.086
0.451
1.000
Table 2B. Correlation matrix of the explanatory factors—
all stocks (Jan 2000–June 2007).
XMKT
SMB
LEV
HML
XMKT
SMB
LEV
HML
1.000
-0.151
-0.145
-0.213
-0.151
1.000
0.205
0.283
-0.145
0.205
1.000
0.407
-0.213
0.283
0.407
1.000
Table 2C. Correlation matrix of the explanatory factors—
all stocks (July 2007–April 2009).
XMKT
SMB
LEV
HML
XMKT
SMB
1.000
-0.007
0.222
0.143
-0.007
1.000
-0.448
-0.301
Eds. Hatem A. El-Karanshawy et al.
LEV
0.222 -0.448
1.000
0.523
XMKT
Mean
Median
Maximum
Minimum
Std. Dev.
Skewness
Kurtosis
(for e.g. positive book to market equity). The correlation
matrices for the sample across the three periods are reported
in Tables 2A-2C. For the aggregate period, there is a positive
correlation of .45 between HML and LEV, which is expected
since both these factors closely represent the distress
risk of the firm. In Table 2B, for the non-crisis period, the
correlation coefficients are as follows: 0.407 (LEV, HML),
-.1451 (LEV, XMKT), and .205 (LEV, SMB). In Table 2C, for
the crisis period, there are some interesting changes. For
example, the correlation between LEV and HML increases
further, and the correlation between LEV and XMKT actually
turns positive. Finally, the correlation between LEV and SMB
actually turns negative. Table 3 reports summary statistics
for the risk factors, XMKT, SMB, HML and LEV.
XMKT
Table 3. Descriptive statistics of the returns on Market
factor, SMB, HML, and LEV factors.
HML
0.143
-0.301
0.523
1.000
-0.002
0.001
0.115
-0.221
0.027
-1.398
13.623
SMB
LEV
0.002
0.002
0.054
-0.063
0.017
-0.272
4.883
0.000
0.000
0.045
-0.035
0.008
-0.077
6.630
HML
0.001
0.001
0.053
-0.049
0.013
0.127
6.155
XMKT is defined as rm-rf where rf is the return on the risk
free asset and rm is the return on the market portfolio. SMB
is the return on the size mimicking portfolio constructed
by taking the simple average of the returns each week
of all “small” portfolios minus “big” portfolios. HML
is the return on book to market mimicking portfolio
constructed by taking the simple average of the returns
each week of all “high BE/ME” portfolios minus “low BE/
ME” portfolios. LEV is the return on leverage mimicking
portfolios constructed by taking the simple average of the
returns each week of all “high leverage” portfolios minus
“low leverage portfolios”.
Macroeconomic variables and factor loadings
In this section we demonstrate that the FF and leverage
risk factors have macroeconomic implications. Several
studies have shown that macroeconomic variables predict
expected returns on stocks and bonds. See for example,
Abel (1999), Fama (1981), Elton, et. al. (2001), Vassolou
(2003) and Petkova (2006) and references therein. These
studies show a significant positive relationship between the
excess market returns and indicators of economic growth.
We extend this analysis and test for the relationship
between selected macroeconomic variables and returns on
SMB, HML, and LEV factors.
We choose the following variables to represent the
world economic environment for these globally traded
stocks: growth rate in industrial production (world),
unemployment rate (world), inflation (U.S.), credit
spread (U.S.) and term spread (U.S.) during our sample
period. Credit spread is defined as the difference in the
weekly yield on Moody’s AAA corporate bonds and 1
year maturity government Treasury notes. Term spread
is calculated as the difference between the weekly yield
on 1 year treasury notes and 3 month treasury bills. The
source of the data is the FRED database at the St. Louis
Federal Reserve. Monthly data for industrial production
for the world and unemployment rates have been obtained
from the database of IHS Global Insights (http://www.
ihsglobalinsight.com/EconomicFinancialData). Monthly
inflation rates for the U.S. are obtained from the website
www.Inflationdata.com.
Table 4 represents the results for multivariate regressions
of each of the macroeconomic variables on lagged excess
market returns and returns on SMB, HML and LEV. We use
three lags to extract the maximum information content of
these factors. The regressions are estimated at following
periodicity: monthly for inflation, industrial production
51
Bhatt and Sultan
growth rate, unemployment rate and weekly for credit
spread and term spread. The regression model is:
Ykt = b 0 + b 1(rm - r ft )t-i + b 2 Rt-i ,SMB
+ b 3 Rt-i ,HML + b 4 Rt-i ,LEV + e t (2)
where, Ykt represents each of the following macroeconomic
variables: monthly percentage change in industrial
production growth rate, inflation and unemployment rate,
weekly credit spread and term spread, and i represents the
number of lagged terms 1 to 3 to reduce serial correlation.
All macroeconomic variables have been tested for unit root
and those with unit root have been differenced once to
induce stationarity.
Panel A represents the coefficients and t-statistics for each
of the above macroeconomic variables regressed against
one lag of the independent variables. The results indicate
that the leverage risk factor affects the unemployment rate
and inflation (at 5% level of significance) while remains
insignificant for term spread, industrial production and
credit spread. In Panel B, unemployment exhibits significant
sensitivity to LEV lagged one period and inflation shows
significant sensitivity to LEV lagged two periods. Panel C
also shows significant factor loadings on LEV for all the
macroeconomic variables at different lag lengths. With
respect to the other factors, SMB shows significant factor
loadings for industrial production and term spread (at the
first and second lags) while the impact of excess market
returns on these macroeconomic variables seems to be
weak. Note that HML seems to have limited ability to
predict these economic variables at the first and the second
lags but tends to exhibit a significant impact on these
variables at the third lag (significant for unemployment and
inflation). Overall, the results emphasize that the “leverage
risk factor” is a systematic risk factor, though its effects on
macroeconomic variables are not uniform.
pool of assets like mortgages or credit card receivables
(commonly known as collateralized debt obligations).
Investors relied on the major credit rating agencies like
Moody’s and Standard & Poor’s for the acceptance of these
products. The collapse of the subprime lending sector
and the resulting credit crisis in 2007 and 2008 exposed
a colossal failure of the credit rating agencies; which also
paved the way for a near-complete closure of markets for
these products. In a nutshell, the credit spreads did not
reflect the true economic risk underlying the corporate
debt, hence it is difficult to establish a true empirical
relationship between the leverage risk factor and the credit
spread variables.
Notwithstanding the previous discussion, we find a positive
relationship between LEV and the credit spread (see
Table 4, Panel C). This is consistent with the evidence that
the firms hit hard by the credit crisis were those that relied
heavily on debt to finance growth like Home Depot, Toyota
Motor and FedEx.18 Stock prices of these firms, including
investment banks like Citigroup and UBS AG, plummeted
during the recent market meltdown. Bear Stearns and
American Home Mortgage are notable examples of firms
which were coerced to sell their holdings at far below their
book values. In general, there was a continuous re-pricing
of risk in the stock market and stock prices plummeted.
In contrast, the US treasury yields were falling due to
flight to safety, while the rates on mortgage debts failed
to decline at the same pace. This resulted in higher credit
spreads because mortgage debts were most risky and
demanded a premium over Treasury bonds. Thus, the
positive relationship between leverage risk factor and
the credit spread as seen in Table 4 is plausible since both
the variables are representative of the increased exposure
of the firm to distress risk caused by over leveraging.
Finally, when inflation is uncertain, investors demand
inflation risk premium. Inflation induces volatility in
the returns on debt and hence there is a leverage risk
premium. Whether the relationship between inflation and
leverage risk premium is positive or negative depends on
the interaction between inflation, taxes (corporate tax and
personal tax), expected return on assets, and the amount of
debt used in the project. According to Armitrage (2005), as
inflation increases, the real tax adjusted weighted average
cost of capital decreases because higher inflation alleviates
the corporate taxes on the firms’ real profits and increases
the tax advantage on debt. However in the presence of
personal taxes, higher inflation causes an increase in the tax
rates on real returns to debt. This increases the leverage risk
of the firms which are heavily dependent on debt and thus
demand a premium over firms which rely less on external
debt. For our sample, we find mixed evidence (positive and
negative) of the relationship between inflation and the risk
factors (See Table 4).
These results have powerful implications for the US
economy bouncing back from a severe financial crisis. First,
researchers argue that “deeper the decline in GDP, peak
to trough, the more rapid the post recession rebound.” A
recent report17 suggests that this is the case only if there
is a significant increase in the private sector liabilities.
According to the report, a 0.3% drop in employment rate
requires the real GDP growth higher than 3%, which in
turn requires a 5% rise in the private sector liabilities,
and subsequently, has a significant impact on the level
of industrial production. In fact, we have witnessed slow
moving recoveries following the 1980, 1991 and 2001
recessions, with the slowness being attributed to low levels
of private liabilities during these periods. This supports
the positive relationship between the leverage risk factor
and industrial production and a consistently negative
relationship between unemployment rate and the leverage
risk factor which has been documented in the earlier
section.
Explaining cross-section of returns
Next, the credit spread is a representative of firm’s default
risk. A high credit spread indicates stringent credit markets
and higher risk levels. However, the last quarter century
witnessed some of the major developments in finance,
for e.g. “securitisation” and introduction of “structured
products” which generate cash flows from underlying
In this section, we present our regression results by
including leverage factor as a systematic risk factor. First, we
test for the significance of the FF risk factors. Next we add
LEV to the regression model to compare results across three
periods: January 2000 –April 2009 (aggregate), January
2000 – June 2007 (non-crisis), and July 2007 – April 2009
(crisis)19. To check on the robustness of these results, we will
52
Islamic banking and finance – Essays on corporate finance, efficiency and product development
A comparison among Islamic, conventional, and socially responsible stocks
Table 4. Multivariate regressions of macroeconomic variables conditional on factor returns during the aggregate period.
The following regression is estimated to demonstrate the link between Fama-French factors and economic variables:
Ykt = b 0 + b 1 (rm - rft )t -i + b 2 R t -i ,SMB + b 3 R t -i ,HML + b 4 R t -i ,LEV + e t
where Ykt represents each of these macroeconomic variables (monthly Industrial production growth rates, monthly unemployment
rate, monthly data for percentage change in inflation rates and weekly data for credit spread and term spread) for the combined period
(January 2000 to April 2009). i represents the number of lagged terms 1 to 3. rf is the return on the risk free asset and rm is the return
on the market portfolio. rf is the return on the risk free asset and rm is the return on the market portfolio. RSMB is the return on the size
mimicking portfolio constructed by taking the simple average of the returns each week of all “small” portfolios minus “big” portfolios.
RHML is the return on book to market mimicking portfolio constructed by taking the simple average of the returns each week of all “high
BE/ME” portfolios minus “low BE/ME” portfolios. RLEV is the return on leverage mimicking portfolios constructed by taking the simple
average of the returns each week of all “high leverage” portfolios minus “low leverage portfolios”.
Panel A
Industrial
Coeff.
t-stats
Unemployment rate
Coeff.
XMKT t-1 −3.241 −0.307
0.007
SMB t-1
−41.724 −2.249** 0.338
0.729
HML t-1 −29.364 −1.090
62.315
1.540
LEV t-1
−1.667
R-square
0.105
0.114**
Credit Spread
t-stats
Coeff.
0.044
1.110
1.721*
−2.302**
−0.0001
0.0001
0.0001
−0.0003
0.030**
t-stats
Term Spread
Coeff.
t-stats
% change in
inflation rate
Coeff.
t-stats
1.847 −0.172
−2.081** 0.0000 −1.660*
1.326
1.365
−0.0001 −2.965** 1.532
0.681
0.959
−0.0001 −1.393
−6.337
0.0001
1.348
5.998 −2.207**
−1.404
0.028**
0.152**
Panel B
Industrial
Coeff.
XMKT t-1
SMB t-1
HML t-1
LEV t-1
XMKTt-2
SMB t-2
HML t-2
LEV t-2
R-square
−2.582
−47.422
−20.096
24.896
13.845
−23.054
−21.410
61.648
0.224
t-stats
Unemployment rate
Coeff.
0.016
−0.205
−3.096** 0.438
0.578
−0.794
0.903
−0.910
2.104** −0.350
0.087
−1.497
0.447
−0.897
1.553
−1.043
0.219
t-stats
0.098
1.786*
1.588
−1.840*
−2.413**
0.305
1.018
−1.412
Credit Spread
Coeff.
−0.0001
0.0001
0.0001
−0.0002
−0.0001
0.0001
0.0000
−0.0001
0.054
t-stats
Term Spread
Coeff.
0.0000
−1.619
1.928** −0.0001
0.507
−0.0001
0.0001
−0.960
−3.493** 0.0000
0.835
0.0000
−0.183
−0.0002
0.0001
−0.687
0.042
t-stats
% change in
inflation rate
Coeff.
2.308
−1.501
−2.992** 1.172
−1.802* −5.263
1.362
3.367
0.361
−0.388
0.187
−0.519
−2.547** 0.306
0.951
5.379
0.186
t-stats
1.707
0.792
−1.897*
0.814
0.454
0.123
0.165
2.274**
Panel C
Industrial
Coeff.
t-stats
XMKT t-1 −5.246 −0.410
SMB t-1
−48.352 −3.397
HML t-1
−7.562 −0.297
LEV t-1
44.872
1.699
2.842
XMKTt-2 18.343
SMB t-2
−25.884 −2.343
HML t-2
−6.985 −0.384
1.522
0.045
LEV t-2
2.282
XMKTt-3 16.418
SMB t-3
−25.738 −1.619
2.303
0.149
HML t-3
2.740**
LEV t-3
R-square
0.430
Unemployment rate
Coeff.
0.090
0.480
0.230
−1.224
−0.458
0.196
0.270
−0.043
−0.476
0.087
−0.631
−1.109
0.412
t-stats
0.535
2.004**
0.579
−2.597**
−3.220**
0.941
0.946
−0.084
−3.596**
0.316
−2.107**
−2.042**
Credit Spread
Coeff.
−0.0001
0.0001
0.0001
−0.0002
−0.0001
0.0000
0.0000
−0.0001
0.0000
0.0000
−0.0001
0.0003
0.066
t-stats
Term Spread
Coeff.
0.0000
−1.651*
1.835* −0.0001
0.540
−0.0001
0.0001
−1.038
−4.013** 0.0000
0.603
0.0000
−0.320
−0.0001
0.0001
−0.787
0.0000
−0.529
0.354
0.0000
0.0000
−0.923
1.953*
0.0000
0.044
t-stats
−1.402
−2.961
−1.733
1.267
−0.505
−0.585
−2.321
0.812
0.473
0.225
0.738
0.153
% change in
inflation rate
Coeff.
2.403
0.803
−4.378
3.673
0.739
0.067
0.110
6.470
0.308
0.393
5.296
−2.493
0.215
t-stats
1.678*
0.537
−1.353
0.859
0.991
0.049
0.053
2.284**
0.338
0.214
2.655**
−0.861
(*indicates significant at 10% level, ** indicates significance at 5% level)
Eds. Hatem A. El-Karanshawy et al.
53
Bhatt and Sultan
further classify firms into two groups: financial and nonfinancial. Financial firms include all financial institution as
well as real estate and mortgage firms. The popular adage is
that leverage is a two-way sword. It magnifies returns in an
up market and magnifies losses in a down market. Finally,
we test our main hypothesis that Islamic stocks would be
less sensitive to the leverage risk factor than conventional
and socially responsible stocks. Our primary rationale is
that low leverage of Islamic stocks would lessen the interest
rate exposure of these firms.
We use the following firm-specific GARCH model:
rt - r ft = β 0 + β 1(rmt - r ft ) + β 2 Rt ,SMB + β 3 Rt ,HML
(3)
+ β 4 Rt ,LEV + e t
e t|ψ t-1 ~ N(0, σ t2 ), (4)
q
p
i=1
j =1
σ t2 = W + ∑ a ie 2 t-i + ∑ δ iσ t- j (5)
where rt – rft in the mean equation is the weekly excess
return on asset i, rft is the weekly risk free rate (US T-bill),
rmt – rft is the market risk premium (XMKT), and SMB, HML
and LEV are Fama-French factors and the leverage risk
factor, defined earlier. The variance equation (5) models
the conditional variance as a GARCH(p,q) process where
p and q denote the lag length. W is the intercept term, a
is the ARCH term and d is the GARCH term. a and d terms
are expected to be positive and significant determinants
of the conditional variance of changes in the excess
return. The primary reason for using the GARCH model
is that preliminary diagnostics suggest that the weekly
excess returns have time varying variance with volatility
clustering and fat tails. The GARCH models are estimated
using the Bollerslev-Wooldridge (1992) corrections to deal
with excess kurtosis. As noted earlier, standard t-statistics
based inferences in the presence of excess kurtosis in the
residuals are asymptotically invalid because standard errors
are biased downward, leading to false acceptances.
Factor loadings at the firm level
We test the above model at both the firm and portfolio level
for all 3,707 financial and non-financial firms. Each week
from January 2000 to April 2009 we run cross sectional
regressions of weekly excess stock returns on XMKT,
SMB, and HML factors. Next, we add LEV to test for its
significance in addition to the market factor and the FamaFrench factors. For robustness check, we test for the partial
F-statistics of LEV to see whether this additional factor
contributes significantly in explaining the cross section of
expected returns (in addition to the market factor and the
traditional FF factors).
Tables 5 exhibit the summary of the impact of XMKT, SMB,
HML and LEV factors on the returns of firm and portfolios.
Model 1 is the traditional FF case and Model 2 includes
the LEV factor in addition to the FF factors. As shown, we
have 3707stocks in the sample. Note that in Table 5 and
subsequent tables, we only include regression results that
are significant at least at the 5% level. In Panel A, the
results show that for the aggregate period (2000–2009),
in 3,304 instances XMKT is positive. A positive sign for the
54
XMKT is consistent with the single factor CAPM model.
The distribution of the SMB is about half positive and half
negative. The HML is positive in 1,539 and negative in
216 cases. When LEV is added to the model (Model 2), we
find that, there is a .24% increase in the number of cases
((3312/3304)-1) where XMKT is significant. With the
addition of LEV, there is a 2.54% increase in the number
of cases where SMB is significant. Surprisingly, the number
of cases HML is positive and significant drops by 16.58%.
Finally, in 2,208 instances, LEV is positive, though in
125 instances it is negative.
The results (Panel B) for the non-crisis period (2000-June
2007) are similar. The number of cases where the factors
is significant changed as follows: .39% (XMKT), 1.83%
(SMB), and -2.05% (HML). With regard to positive and
negative impact of the factors on stock returns, there are
some changes compared to the aggregate period (Model
2). For example, SMB, the number of negative cases is now
680, representing a 40% decline from the previous model.
In contrast, HML, now has 539 instances for which the
coefficients are negative, indicating a 17.43% increase from
the previous value. Finally, we have 862 instances of positive
and 149 cases of negative coefficients for LEV. It appears
that, compared to Model 2 (aggregate period), there is a
large number of instances the regression coefficients are
insignificant. Altogether, the number of significant cases
drops by 56%, suggesting that the LEV factor is able to
capture systemic risk in the economy across good and bad
times quite well.
However, the contribution of LEV in capturing leverage risk
is evident when we estimate firm-specific regressions for
the crisis period (Panel C). During July 2007-April 2009,
compared to the non-crisis period, there is a 201.07%
increase in the number of the cases where LEV is significant.
This increase is indicative of several stylized facts during
the escalating financial crisis afflicting the global economy.
It appears that the credit crisis had a contagion-like effect,
impacting firms across all spectrums of leverage. In essence,
firms were hard hit especially when access to the debt market
was severely limited because of reluctance among financial
institutions to lend. The results suggest that for 3,038 firms,
the sensitivity to LEV is positive and significant. Only in
66 cases the variable has negative coefficients. Compared
to the non-crisis period, the addition of LEV during the
credit crisis leads to a change in the number of significant
cases for the remaining factors: XMKT (-84.91%), SMB
(-14.97%), and HML (-27.54%). In particular, the number
of negative coefficients for HML is higher than the positive
ones, indicating that during the recent credit crisis, a value
based investment strategy would have earned investors
negative risk premium. Again, it supports the notion that
the HML may not have been a good proxy for the distress
risk during this period.
Factor loadings at the portfolio level
Table 5 also the highlights portfolio-specific regressions
(Panels D-F) for the three periods. Based on the intersection
of these three factors, we have 27 portfolios with annual
rebalancing. The results reconfirm our earlier finding
that the addition of LEV weakens the significance of the
traditional FF factors. For the aggregate period (Panel D),
we find that LEV is significant and positive for 19 out of 27
portfolios. The number drops to 17 when we estimate the
Islamic banking and finance – Essays on corporate finance, efficiency and product development
i =1
j =1
σ t2 = W + ∑ a ie 2 t -i + ∑ δ iσ t - j
q
2
t
p
i =1
q
σ t2 = W + ∑ a ie 2 t -i +
p
j =1
i
∑δ σ
e t |ψ t -1 ~ N(0, σ t2 ),
t- j
rit - r ft = β 0 + β 1(rmt - r ft ) + β 2 R t ,SMB + β 3 R t ,HML + β 4 R t ,LEV + e t
rit - r ft = β 0 + β 1(rmt - r ft ) + β 2 R t ,SMB + β 3 R t ,HML + e t
e t |ψ t -1 ~ N(0, σ ),
Model 2
Model 1
Eds. Hatem A. El-Karanshawy et al.
3584
0.39%
3584
0
3570
Total
Positive
Negative
3570
0
Positive
Negative
Total
%change in
significance (by model)
Model 2
Model 1
XMKT
Panel B: Non-crisis Period
3312
0.24%
3312
0
3304
Total
Positive
Negative
3304
0
Positive
Negative
Total
% change in
significance (by model)
Model 2
Model 1
2111
1.83%
1431
680
2073
1461
612
SMB
2381
2.54%
1234
1147
2322
1208
1114
SMB
1627
-2.05%
1088
539
1661
1244
417
HML
1464
-16.58%
1005
459
1755
1539
216
HML
1031
882
149
0
LEV
2333
2208
125
LEV
Positive
Negative
Total
Positive
Negative
27
0.00%
27
0
27
27
0
Positive
Negative
Total
Positive
Negative
Total
%change in
significance (by model)
Model 2
Model 1
27
0.00%
27
0
27
27
0
XMKT
Panel E: Non-crisis Period
Total
% change in
significance (by model)
Model 2
Model 1
XMKT
Panel D: Aggregate Period
Panel A: Aggregate Period
XMKT
All Portfolios: 27
All stocks: 3707
24
0.00%
15
9
24
15
9
SMB
21
-4.55%
12
9
22
13
9
SMB
23
9.52%
18
5
21
18
3
HML
23
9.52%
17
6
21
18
3
HML
(Continued)
20
17
3
LEV
20
19
1
LEV
where, ri is the return on portfolio i; rf is the return on the risk free asset and rm is the return on the market portfolio. RSMB is the return on the size mimicking portfolio constructed by taking
the simple average of the returns each week of all “small” portfolios minus “big” portfolios. RHML is the return on book to market mimicking portfolio constructed by taking the simple
average of the returns each week of all “high BE/ME” portfolios minus “low BE/ME” portfolios. RLEV is the return on leverage mimicking portfolios constructed by taking the simple average
of the returns each week of all “high leverage” portfolios minus “low leverage portfolios”. All indicated coefficients with (*) are significant at least at the 5% level of significance.
Table 5. Summary of results showing the number of stocks and portfolios which showed significant sensitivities to XMKT, SMB, HML, and LEV factors.
A comparison among Islamic, conventional, and socially responsible stocks
55
56
35.00%
-43.48%
-41.67%
-100.00%
%change in
significance (by period)
-14.97%
-84.91%
% change in
significance (by period)
-27.54%
201.07%
27
13
-18.75%
14
-17.65%
0
0.00%
Total
%change in
significance (by model)
3104
1179
-17.32%
1795
-8.65%
541
-2.35%
Model 2
3038
66
Model 2
854
941
157
384
Positive
Negative
439
740
1426
1965
554
Total
Total
%change in
significance (by model)
27
0
1
12
7
7
0
0
Positive
Negative
16
17
4
Total
16
0
0
17
4
0
0
1330
96
401
1564
439
115
Positive
Negative
Model 1
Panel C: Crisis period
Table 5. (Continued)
XMKT
SMB
HML
LEV
Model 1
Positive
Negative
Panel F: Crisis period
XMKT
SMB
HML
LEV
Bhatt and Sultan
model for the non-crisis period (Panel E). In contrast, we
find that in all cases, LEV is positive and significant during
the crisis period (Panel F). We also note that, in comparison
to the aggregate period, there is a 100% reduction in the
number of cases XMKT is significant during the crisis
period. For the remaining variables, percentage change
in significance is as follows: SMB (-41.67%) and for HML
(-43.48%), indicating an across the board weakening of
the FF factors during the financial distress. In contrast,
there is a 35% increase in the number of instances where
LEV is positive and significant.
Overall, the FF factors seem to lose their significance when
LEV as a systemic risk is included in the model. In particular,
during a financial crisis period, sensitivity to LEV at the firm
and portfolio level suggests that the traditional FF factors
may not be adequately capturing the effects of economywide distress arising from excess leverage. Therefore,
sensitivity to this systemic risk translates into additional
risk premium that is not adequately captured by the FF risk
factors.
Portfolio-specific regression results across aggregate, noncrisis and crisis periods are provided next in Tables 6–8 to
highlight the magnitude of the coefficients and to check
for robustness of adding LEV. In Table 6, there are several
stylized facts for the aggregate period. First, there is a
noticeable increase in the adjusted R2 when LEV is added as
an explanatory variable. Second, as reported earlier, with
the addition of LEV in Model 2, the traditional FF factors
tend to lose their statistical significance. Finally, we find
that in many instances, the coefficient of HML actually
turns negative.
In Table 7, we report the results for the non-crisis period
and the results indicate that while the variable LEV is an
important explanatory power, its addition makes only
marginal impact on Model 2. There is an increase in the
adjusted R2 but by a small margin. In contrast, during
the crisis period (Table 8), the addition of the LEV makes
a substantial contribution to the overall forecast ability
of Model 2. The adjusted R2 increases by a considerable
margin. In addition, the size of the coefficients across the
27 portfolios is large, ranging from 1.56% (portfolio #7)
to 4% (portfolio #18). The magnitude of the coefficients
indicates the heightened sensitivity of firms to the economic
distress during the period. As indicated earlier, the results
indicate that our leverage risk factor performs quite well in
representing systemic risk in the global economy. Also note
that the number of negative significant coefficients for HML
increases considerably (from 5 to 12)20 which suggests that
the value based investment strategy may not with falling
equity prices. In contrast, LEV has a positive relationship
with stock returns for all 27 portfolios, suggesting that
investors demand a premium for investing in high leverage
portfolios during the credit crisis.
The negative effects of leverage risk factor on stock
returns
In a number of cases (see Tables 5–8), leverage risk has
a negative effect on stock returns, which is consistent
with several existing studies. For example, Penman et al.
(2007) decompose the book to price ratio of a firm into two
components. The first component is the enterprise book
to price (which represents the operating risk of the firm),
Islamic banking and finance – Essays on corporate finance, efficiency and product development
A comparison among Islamic, conventional, and socially responsible stocks
Table 6. Factor loadings of all firms for the aggregate period (January 2000 to April 2009).
Model 1
Model 2
rit - rft = b 0 + b 1 (rmt - rft ) + b 2 R t ,SMB + b 3 R t ,HML + e t
rit - rft = b 0 + b 1 (rmt - rft ) + b 2 R t ,SMB + b 3 R t ,HML + b 4 R t ,LEV + e t
e t |ψ t-1 ~ N(0,σ t2 ),
e t |ψ t -1 ~ N(0,σ t2 ),
q
σ t2 = W + ∑ a i e t2-i +
i=1
p
∑δ σ
i
q
t- j
σ t2 = W + ∑ a i e t2-i +
j =1
i=1
p
∑δ σ
i
t- j
j =1
where, ri is the return on portfolio i; rf is the return on the risk free asset and rm is the return on the market portfolio. RSMB is the return
on the size mimicking portfolio constructed by taking the simple average of the returns each week of all “small” portfolios minus “big”
portfolios. RHML is the return on book to market mimicking portfolio constructed by taking the simple average of the returns each week of
all “high BE/ME” portfolios minus “low BE/ME” portfolios. RLEV is the return on leverage mimicking portfolios constructed by taking the
simple average of the returns each week of all “high leverage” portfolios minus “low leverage portfolios”. All indicated coefficients are
significant at 5% level of significance. GARCH models are estimated using the Bollerslev-Wooldridge corrections to the standard errors.
Model 1 excludes LEV. Model 2 includes LEV. Coefficients of the GARCH variance equations are not reported to conserve space. They are
available upon request.
Aggregate Period
Model 1
Portfolio
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
Intercept
0.002*
0.002*
0.001
0.002
0.000
0.000
0.000
0.000
0.000
0.000
0.001
0.000
0.000
0.001
0.001
0.000
0.001
0.001
0.000
0.000
0.000
0.001
0.001
0.001
0.002*
0.000
0.002*
MKT
Model 2
SMB
HML
0.794* 0.388*
0.663* 0.344*
0.742* 0.598*
0.557* 0.390*
0.661* 0.546*
0.723* 0.565*
0.622* 0.640*
0.765* 0.731*
0.840* 0.842*
0.896* 0.053
0.690* 0.062
0.718* 0.009
0.630* 0.083
0.631* 0.052
0.632* 0.138*
0.630* 0.226*
0.749* 0.250*
0.810* 0.188*
0.772* -0.470*
0.656* -0.396*
0.613* -0.379*
0.698* -0.387*
0.721* -0.283*
0.671* -0.253*
0.804* -0.839*
0.773* -0.321*
0.846* -0.486*
-0.235*
-0.047
0.053
0.261*
0.431*
0.565*
0.744*
0.812*
0.923*
-0.264*
-0.072
0.016
0.372*
0.340*
0.462*
0.892*
0.684*
0.872*
-0.288*
0.035
0.082
0.255*
0.352*
0.505*
0.856*
0.901*
0.880*
Adj.
R-square Intercept
-0.015
0.000
-0.004
0.014
-0.001
0.024
0.055
0.022
0.080
-0.041
0.009
-0.021
-0.014
0.042
0.032
0.108
0.051
0.086
0.198
0.094
0.122
0.064
0.073
0.103
0.297
0.179
0.255
measured as the ratio of book value of operating assets to
their market value. The second component is the financial
leverage component (which represents the financing risk
of the firm), measured as the ratio of market value of debt
to market value of equity. The authors find that enterprise
book to price ratio has a significant positive relationship
with the expected stock returns while the “leverage”
component of book to price ratio has negative relationship
Eds. Hatem A. El-Karanshawy et al.
0.002*
0.002*
0.001
0.001
0.000
0.000
-0.001
0.000
0.000
0.000
0.001
0.000
0.000
0.001
0.001
0.000
0.001
0.001
0.000
0.000
0.000
0.001
0.001
0.001
0.002*
0.000
0.002*
MKT
SMB
HML
0.794*
0.642*
0.727*
0.547*
0.665*
0.701*
0.639*
0.767*
0.862*
0.900*
0.700*
0.733*
0.628*
0.648*
0.634*
0.628*
0.752*
0.814*
0.775*
0.651*
0.622*
0.671*
0.718*
0.667*
0.806*
0.769*
0.857*
0.389*
0.301*
0.560*
0.364*
0.521*
0.510*
0.671*
0.720*
0.890*
0.044
0.023
-0.050
0.050
-0.003
0.089
0.222*
0.203*
0.130*
-0.467*
-0.406*
-0.358*
-0.414*
-0.338*
-0.262*
-0.848*
-0.290*
-0.562
-0.231*
-0.193*
-0.254*
0.248*
0.353*
0.444*
0.753*
0.740*
0.725*
-0.289*
-0.169*
-0.155
0.306*
0.214*
0.306*
0.874*
0.590*
0.716*
-0.274*
-0.022*
-0.075
0.180*
0.261*
0.357*
0.937*
0.670*
0.688*
Adj.
LEV R-square
-0.023
0.722*
1.303*
0.145
0.396*
0.457*
-0.118
0.306*
0.741*
0.141
0.453*
0.807*
0.337*
0.548*
0.659*
0.053
0.515*
0.939*
-0.100
0.409*
0.650*
0.312
0.599*
0.543*
-0.430*
1.035*
1.271*
-0.021
0.103
0.169
0.042
0.064
0.103
0.021
0.059
0.141
-0.028
0.079
0.101
0.034
0.129
0.151
0.112
0.132
0.215
0.189
0.148
0.202
0.104
0.158
0.184
0.277
0.255
0.401
with the expected stock returns. Johnson (2004) documents
a negative relationship between leverage and cross section
of expected returns after controlling for firm specific
characteristics like volatility. See Arditti (1967), Dimitrov
and Jain (2006) for similar results. In particular, Dimitrov
and Jain (2006) note that during economic distress, raising
equity is costlier than debt (e.g., bank financing or line
of credit), so firms would prefer to increase leverage. So,
57
Bhatt and Sultan
Table 7. Factor loadings of all firms for non-crisis period (January 2000 to June 2007).
Model 1
Model 2
rit - rft = b 0 + b 1 (rmt - rft ) + b 2 R t ,SMB + b 3 R t ,HML + e t
rit - rft = b 0 + b 1 (rmt - rft ) + b 2 R t ,SMB + b 3 R t ,HML + b 4 R t ,LEV + e t
e t |ψ t -1~N( 0 ,σ t2 ),
e t |ψ t -1 ~ N(0, σ t2 ),
q
σ t2 = W + ∑ a i e t2-i +
i=1
p
∑δ σ
i
q
t- j
σ t2 = W + ∑ a i e 2t -i +
j =1
i=1
p
∑δ σ
i
t- j
j =1
where, ri is the return on portfolio i; rf is the return on the risk free asset and rm is the return on the market portfolio. RSMB is the return
on the size mimicking portfolio constructed by taking the simple average of the returns each week of all “small” portfolios minus “big”
portfolios. RHML is the return on book to market mimicking portfolio constructed by taking the simple average of the returns each week of
all “high BE/ME” portfolios minus “low BE/ME” portfolios. RLEV is the return on leverage mimicking portfolios constructed by taking the
simple average of the returns each week of all “high leverage” portfolios minus “low leverage portfolios”. All indicated coefficients with
(*) are significant at 5% level of significance. GARCH models are estimated using the Bollerslev-Wooldridge corrections to the standard
errors. Model 1 excludes LEV. Model 2 includes LEV. Coefficients of the GARCH variance equations are not reported to conserve space.
They are available upon request.
Non-crisis Period
Model 1
Portfolio
Intercept XMKT
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
0.002*
0.002*
0.001
0.001
0.000
0.000
-0.001
0.000
0.000
-0.001
0.000
0.000
0.000
0.001
0.001
0.000
0.001
0.001
0.000
0.000
0.000
0.001
0.001
0.001
0.002*
0.000
0.002*
0.856
0.778
0.822
0.673
0.709
0.781
0.785
0.807
0.878
0.973
0.743
0.791
0.699
0.687
0.662
0.701
0.781
0.852
0.865
0.693
0.664
0.823
0.769
0.715
0.883
0.951
0.891
Model 2
SMB
HML
0.449*
0.457*
0.678*
0.397*
0.565*
0.618*
0.686*
0.764*
0.882*
0.142*
0.069
0.081
0.127*
0.077
0.176*
0.294*
0.298*
0.233*
-0.360*
-0.367*
-0.308*
-0.335*
-0.226*
-0.205*
-0.747*
-0.496*
-0.416*
-0.259*
-0.139
0.005
0.278*
0.435*
0.562*
0.712*
0.817*
0.925*
-0.367*
-0.073
-0.070
0.346*
0.346*
0.436*
0.846*
0.668*
0.834*
-0.354*
0.040
0.037
0.255*
0.324*
0.498*
0.812*
0.940*
0.846*
Adj.
R-square Intercept XMKT
0.473
0.320
0.378
0.394
0.440
0.433
0.532
0.501
0.471
0.466
0.446
0.324
0.383
0.405
0.400
0.482
0.469
0.446
0.542
0.502
0.437
0.420
0.478
0.453
0.401
0.363
0.395
falling equity returns during economic distress and rising
leverage support the empirical finding that leverage and
return on equity may be negatively correlated.
Managerial preference for debt over equity financing is also
related to the value of the firm and its future prospects.
Lang et al. (1995) find a negative relationship between
financial leverage and future growth of a firm. The authors
58
0.002*
0.002*
0.001
0.001
0.000
0.000
-0.001
0.000
0.000
-0.001
0.000
0.000
0.000
0.001
0.001
0.000
0.001
0.001
0.000
0.000
0.000
0.001
0.001
0.001
0.002*
0.000
0.002*
SMB
0.851* 0.454*
0.783* 0.447*
0.846* 0.646*
0.674* 0.396*
0.722* 0.540*
0.777* 0.572*
0.777* 0.699*
0.810* 0.760*
0.899* 0.916*
0.974* 0.142*
0.764* 0.045
0.812* 0.027
0.703* 0.106*
0.715* 0.049
0.682* 0.119*
0.700* 0.300*
0.797* 0.271*
0.886* 0.171*
0.929* -0.349*
0.694* -0.374*
0.682* -0.304*
0.814* -0.345*
0.779* -0.280*
0.721* -0.215*
0.867* -0.753
1.014* -0.478
0.921* -0.450
HML
-0.224*
-0.204*
-0.253*
0.277*
0.378*
0.479*
0.772*
0.766*
0.775*
-0.369*
-0.143
-0.178
0.310*
0.256*
0.338*
0.870*
0.599*
0.729*
-0.415*
-0.003
-0.085
0.221*
0.266*
0.392*
0.930*
0.707*
0.677*
Adj.
LEV R-square
-0.195
0.343
1.119*
0.005
0.313*
0.311*
-0.221*
0.212*
0.637
0.008
0.322*
0.661*
0.213*
0.429*
0.559*
-0.084
0.392*
0.832*
-0.430*
0.314*
0.539*
0.170
0.523*
0.447*
-0.699*
0.666*
0.992*
0.473
0.328
0.419
0.392
0.449
0.450
0.531
0.502
0.475
0.464
0.456
0.361
0.386
0.422
0.437
0.481
0.484
0.473
0.552
0.511
0.461
0.422
0.504
0.480
0.441
0.363
0.456
emphasize that the negative relationship between leverage
and growth is more visible for firms with a low Tobin’s q
since these firms are characterised by negligible growth
opportunities not recognised by the capital markets. The
study further rationalises that managers of firms with
considerably lucrative growth opportunities generally do
not opt for a high leverage21 because high interest payments
on debt tend to erode the profitability of the firm which
Islamic banking and finance – Essays on corporate finance, efficiency and product development
A comparison among Islamic, conventional, and socially responsible stocks
Table 8. Factor loadings of all firms during crisis period (July 2007 to April 2009).
Model 1
Model 2
rit - r ft = β 0 + β 1(rmt - r ft ) + β 2 R t ,SMB + β 3 R t ,HML + e t
rit - r ft = β 0 + β 1(rmt - r ft ) + β 2 R t ,SMB + β 3 R t ,HML + β 4 R t ,LEV + e t
e t |ψ t -1 ~ N(0, σ ),
e t |ψ t -1~N( 0 , σ t2 ),
2
t
q
p
i =1
j =1
σ t2 = W + ∑ a ie 2 t -i + ∑ δ iσ t - j
q
p
i =1
j =1
σ t2 = W + ∑ a ie 2 t -i + ∑ δ iσ t - j
where, ri is the return on portfolio i; rf is the return on the risk free asset and rm is the return on the market portfolio. RSMB is the return
on the size mimicking portfolio constructed by taking the simple average of the returns each week of all “small” portfolios minus “big”
portfolios. RHML is the return on book to market mimicking portfolio constructed by taking the simple average of the returns each week of
all “high BE/ME” portfolios minus “low BE/ME” portfolios. RLEV is the return on leverage mimicking portfolios constructed by taking the
simple average of the returns each week of all “high leverage” portfolios minus “low leverage portfolios”. All indicated coefficients with
(*) are significant at 5% level of significance. GARCH models are estimated using the Bollerslev-Wooldridge corrections to the standard
errors. Model 1 excludes LEV. Model 2 includes LEV. Coefficients of the GARCH variance equations are not reported to conserve space.
They are available upon request.
Crisis period
Model 1
Portfolio Intercept
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
-0.003
-0.005
-0.003
-0.003
-0.003
-0.004
-0.004
-0.002
-0.002
-0.003
-0.004
-0.004
-0.003
-0.003
-0.004
-0.003
-0.003
-0.005
-0.002
-0.001
-0.003
-0.002
-0.002
-0.004
-0.004
-0.004
-0.005
XMKT
0.075
0.084
0.077
0.087
0.145
0.167*
0.078
0.196*
0.112
0.086
0.037
0.109
0.100
0.127
0.095
0.098
0.031
0.051
0.173*
0.119
0.194*
0.146
0.160
0.157
0.196
0.043
-0.008
SMB
Model 2
HML
0.161
-0.320
0.221
-0.164
0.498
-0.326
0.608*
-0.079
0.711*
-0.151
0.060
0.757*
0.129
0.782*
0.855*
-0.081
0.077
1.347*
-0.520* 0.361
-0.619* 0.522
-0.791* 0.463
-0.370* 0.702*
-0.728* 0.724*
-0.586* 0.908*
0.871*
-0.275
-0.556* 1.278*
-0.863* 1.541*
-1.069* -0.026
-0.964* -0.011
-1.017* 0.092
-0.903* 0.278
-1.064* 0.445
-0.897* 0.749*
-1.176* 0.887*
-1.124* 1.435*
-1.782* 1.765*
Adj.
R-square Intercept XMKT
-0.034
-0.040
-0.005
-0.010
0.012
0.004
0.053
0.053
0.136
0.026
0.053
0.105
0.059
0.132
0.130
0.134
0.175
0.264
0.182
0.113
0.117
0.089
0.178
0.199
0.374
0.386
0.454
prevents the firm from utilizing the benefits of these growth
opportunities. Hence, a negative relationship between
leverage and growth seems rational, which implies a
negative relationship between leverage and stock returns.22
The negative effect of leverage on return on equity is also
consistent from a corporate governance perspective. Jensen
and Meckling (1976) suggest that increased debt levels
Eds. Hatem A. El-Karanshawy et al.
-0.001
-0.002
0.000
-0.003
-0.003
-0.004
-0.003
-0.001
0.000
-0.001
-0.003
-0.003
-0.003
-0.002
-0.003
-0.003
-0.002
-0.003
-0.002
-0.002
-0.002
-0.001
-0.002
-0.003
-0.002
-0.003
-0.002
-0.034
0.020
-0.028
0.038
0.029
0.064
-0.033
0.019
0.018
0.063
-0.007
0.063
0.057
0.039
0.034
0.028
0.012
0.045
0.028
0.000
0.044
-0.042
0.009
0.024
0.129
0.048
-0.048
SMB
HML
0.264
0.461*
0.393*
0.238
0.385*
0.598*
0.436*
0.544*
0.748*
0.022
-0.128
-0.247
0.031
-0.119
-0.016
0.114
0.020
-0.134
-0.554*
-0.349
-0.475*
-0.480*
-0.485*
-0.386
-0.735*
-0.587*
-1.153*
-0.942*
-0.974*
-0.831*
-0.180
-0.364
-0.196
0.177
-0.087
-0.080
-0.927*
-0.910*
-0.859*
-0.274
-0.442*
-0.490*
0.096
0.039
-0.112
-0.877*
-1.069*
-0.853*
-0.416
-0.600*
-0.413
-0.004
0.333
0.486*
Adj.
LEV R-square
2.931*
3.193*
3.633*
1.784*
2.610*
2.643*
1.563*
2.736*
3.851*
3.042*
3.149*
3.078*
2.140*
3.061*
3.328*
1.895*
3.287*
4.033*
2.388*
2.744*
2.596*
2.083*
2.883*
2.875*
2.702*
2.972*
3.693*
0.410
0.444
0.549
0.312
0.450
0.427
0.311
0.473
0.587
0.458
0.506
0.538
0.385
0.552
0.561
0.373
0.609
0.657
0.522
0.479
0.531
0.356
0.534
0.561
0.581
0.680
0.771
have direct implications on the cash flow of the company by
enforcing regular interest payments on debt which controls
managerial expropriation. Fama and Jensen (1983)
explain that increased debt levels adds to the default risk
of the firm and affects the manager’s reputations adversely
in case the firm defaults on its interest payments or debt.
This imposes a constraint on manager expropriation and
leads to better corporate disclosures. In addition, Jensen
59
Bhatt and Sultan
(1986) suggests that leverage increasing transactions such
as LBOs, new debt issues (bonds), and stock repurchase
reduce the manager’s access to free cash, thus reducing
their waste. He further suggests that debt reduces the
agency cost. This implies that as leveraging increases,
external monitoring increases, and managerial efficiency
is expected to rise. Furthermore, this may be imply that
as firms become efficient, shareholders demand less risk
premium for leverage, and as a result, stock prices fall with
higher leverage.
Consistent with the above discussion, there are also more
instances of significant negative coefficients for LEV during
the non-crisis period (which was a period with profitable
investment opportunities in the market). For example
as reported in Table 5, there are 149 cases of negative
coefficients on LEV (Model 2, Panel B) during the noncrisis period, but the number reduces to 66 during the crisis
period (Panel C). At the portfolio level (Panel F), compared
to the non-crisis period (Panel E), the number of negative
coefficients for LEV reduces from 3 to 0.
Leverage and investment strategy
These results have investment implications that suggest
investing in highly leveraged firms. However, an investor
needs to decide between excessively high leverage level
and the negative effects of leverage on financial distress
(Luoma and Spiller (2002)). See Bris and Koskinen (2002)
for further evidences. A recent report23 elaborates that the
regular interest payments on debt for those companies
which fund their investments through debt tend to erode the
cash flow levels of the company by adding to the operating
expenses of the firm. The flip side of the argument is that a
firm with highly profitable growth opportunities and with
a strong cash flow position would still earn a higher return
on equity since they yield high profit margins. The report
claims that a period of economic recovery is characterized
by a strong economic momentum which bolsters earnings
potentials of levered firms. The rationale behind this is that
debt is cheaper for firms with promising growth prospects,
and such they perform at the peak levels when debt is
easily available24. The economic recovery in 2003 provides
strong evidence to this fact when the federal funds rate was
approximately 1.25%, which in turn stimulated economic
growth to jump from 1% to 7%. During this period, levered
companies, high yield bonds and bank loans yielded
attractive returns.
These results do not suggest that as efficiency increases,
stock price decreases. Rather, as firms become more
efficient, debt becomes cheaper and such companies can
afford to have high debt levels in their capital structure
(thereby decreasing the overall cost of capital) without
increasing their credit risk. Due to lower risk levels,
investors do not need additional compensation for excessive
leverage as in the case of firms which are not efficient. Also,
in efficient markets, due to strong corporate governance
principles and better disclosures, the probability of insider
information is reduced and information of the company is
quickly reflected in the stock prices. Hence there is no scope
for mispricing or arbitrage opportunities; so returns fall.
Leverage risk of financial and real estate firms
We perform additional robustness tests by separating
the financial stocks in the sample from the non-financial
60
stocks25. Such an examination is critical because it removes
industry-specific effects of the credit crisis since the effects
may not have been uniformly distributed among financial
and non-financial firms. Financial firms included in the
sample include banks, S&Ls, credit unions, mortgage
financing companies, real estate firms, and insurance
companies. Clearly, these firms bore the brunt of the credit
crisis due to over speculation, deregulation, and over
leveraging. We re-construct FF and LEV factors and estimate
firm26 and portfolio-specific regressions using two separate
samples of firms: the first sample with 645 financial stocks
and the second sample with 2,975 non-financial stocks.
A summary of the regression results are reported in Panels
A-F, Table 9. In Panels A-C, there is evidence that the
leverage risk factor performs well across the three periods,
especially during the crisis period. The results support
the hypothesis that the addition of LEV weakens the
significance of the traditional FF factors. For the aggregate
period (Panel A), LEV is significant and positive for 17 out
of 27 portfolios. During the non-crisis period (Panel B), the
significance of LEV drops, we now have 14 positive and 4
negative instances. During the crisis period (Panel C), in 23
out of 27 cases, LEV is positive and significant. Note that,
in comparison to the non-crisis period, there is a -85.19%
change in the number of cases XMKT is significant during
the crisis period. For the remaining risk factors, the change
in significance is as follows: SMB (−5.26%) and HML
(−59%), suggesting a weakening of the FF factors during
the financial distress. In contrast, there is an increase of
27.78% in the number of instances where LEV is positive
and significant during the crisis period.
In Panels D-F, we report a summary of statistically
significant results for the non-financial firms in the sample.
We confirm our previous findings that the addition of
LEV weakens the significance of the traditional FF factors
considerably. For the aggregate period (Panel D), LEV is
positive in 16 out of 27 portfolios. During the non-crisis
period (Panel E), in 14 instances LEV has positive and
significant coefficients, and in 4 instances the coefficients
are negative and significant. Similar to our earlier findings,
in 27 out of 27 cases, LEV is positive and significant during
the crisis period (Panel F). In comparison to the non-crisis
period, there is a -100% change in the number of cases
where XMKT is significant during the crisis period. For the
SMB, the number of significant cases changes by -39.13%
and for HML, the number of significant cases changes by
-57%, confirming the fact that the power of the FF factors
weakens during the financial distress. In contrast, there is
a 50% increase in the number of instances where LEV is
positive during the crisis period.
Details of these portfolio-specific regressions across
aggregate, non-crisis and crisis periods are provided in
Tables 10–15 to demonstrate the contribution of the LEV
on a case by case basis. Tables 10–12 report the results for
the financial stocks while Tables 13–15 report the results
for the non-financial stocks. In these tables we also report
the adjusted R2 for Model 1 (without LEV) and Model 2
(with LEV) and the results confirm our earlier results.
In Table 10, first, there is a noticeable increase in the
adjusted R2 when LEV is added as an explanatory variable,
indicating increased forecasting power of Model 2 during
the aggregate period. Second, as reported earlier, with the
Islamic banking and finance – Essays on corporate finance, efficiency and product development
i =1
j =1
σ t2 = W + ∑ a ie 2 t -i + ∑ δ iσ t - j
p
p
i =1
j =1
σ t2 = W + ∑ a ie 2 t -i + ∑ δ iσ t - j
q
rit - r ft = β 0 + β 1(rmt - r ft ) + β 2 R t ,SMB + β 3 R t ,HML + e t
e t |ψ t -1~N( 0 , σ t2 ),
rit - r ft = β 0 + β 1(rmt - r ft ) + β 2 R t ,SMB + β 3 R t ,HML + e t
e t |ψ t -1~N( 0 , σ t2 ),
q
Model 2
Model 1
Eds. Hatem A. El-Karanshawy et al.
Positive
Negative
Total
Positive
Negative
Positive
Negative
Total
Positive
Negative
Total
% change in
significance (by model)
Model 2
Model 1
Panel B: Non-crisis Period
Total
% change in
significance (by model)
Model 2
Model 1
0.00%
27
0
27
27
0
27
XMKT
27
0.00%
27
0
27
27
0
15
7
22
15
7
22
0.00%
-5.00%
HML
22
0.00%
15
7
22
15
7
HML
8
12
20
8
11
19
SMB
23
4.55%
7
15
22
7
16
SMB
14
4
18
0
LEV
17
17
0
LEV
Positive
Negative
Total
Positive
Negative
Positive
Negative
Total
Positive
Negative
Total
% change in significance
(by model)
Model 2
Model 1
Panel E: Non-crisis Period
Total
% change in
significance (by model)
Model 2
Model 1
0.00%
27
0
27
27
0
27
XMKT
27
0.00%
27
0
27
27
0
XMKT
Panel D: Aggregate Period
Pane A: Aggregate Period
XMKT
Non financial stock portfolios
Financial stock portfolios
0.00%
14
9
23
14
9
23
SMB
22
4.76%
13
8
21
13
9
SMB
0.00%
18
3
21
18
3
21
HML
24
14.29%
18
3
21
18
6
HML
14
4
18
LEV
17
16
1
LEV
where, ri is the return on portfolio i; rf is the return on the risk free asset and rm is the return on the market portfolio. RSMB is the return on the size mimicking portfolio constructed by taking
the simple average of the returns each week of all “small” portfolios minus “big” portfolios. RHML is the return on book to market mimicking portfolio constructed by taking the simple average of the returns each week of all “high BE/ME” portfolios minus “low BE/ME” portfolios. RLEV is the return on leverage mimicking portfolios constructed by taking the simple average of
the returns each week of all “high leverage” portfolios minus “low leverage portfolios”.
Table 9. Summary of factor loadings for financial and non financial stock portfolios.
A comparison among Islamic, conventional, and socially responsible stocks
61
50.00%
-57.14%
-39.13%
-100.00%
-59.09%
-5.26%
-85.19%
% change in
significance (by period)
23
0
23
27.78%
% change in significance
(by period)
6
4
10
0
9
9
-10.00%
0
6
6
10
4
14
133.33%
0
0
0
0
0
0
0.00%
0
3
4
7
6
3
9
28.57%
0
25
25
0
18
18
-28.00%
15
0
15
4
0
4
-73.33%
Positive
Negative
Total
Model 2
Positive
Negative
Total
% change in significance
(by model)
Model 1
HML
SMB
XMKT
LEV
Model 1
Positive
Negative
Total
Model 2
Positive
Negative
Total
% change in significance
(by model)
HML
SMB
XMKT
Panel F: Crisis period
Panel C: Crisis period
Table 9. (Continued)
62
27
0
27
LEV
Bhatt and Sultan
addition of LEV in Model 2, the statistical significance of
the traditional FF factors tend to weaken. Finally, we find
that in many instances the coefficient of HML actually
turns negative. During the non-crisis period (Table 11), the
addition of LEV to the model makes only marginal impact
on the forecast power of the Model 2. The adjusted the R2
changes by a small margin. In contrast, we find that during
the crisis period (Table 12), the addition of the LEV makes
a substantial contribution to the overall forecast ability of
Model 2. The adjusted R2 increases by a substantial margin.
Furthermore, the size of the coefficient for LEV across
portfolios is large, similar to the results reported earlier. The
magnitude of the coefficient clearly indicates an increased
sensitivity of firms to the economic distress. Similar results
are obtained for non-financial stocks in Tables 13–15.
Overall, our analysis shows that financial and nonfinancial categories of stocks have similar exposure to the
debt market, despite the fact that the concept of leverage
and its use varies across these two categories of firms. It
again reinforces the notion that the financial crisis had a
contagion-like effect on all types of firms. It also establishes
the fact that our leverage risk factor is able to capture
economy-wide risk from over leveraging during the
financial crisis period.
Test for robustness
Earlier, we reported that the correlation between LEV and
HML as follows: .45 (aggregate period), .40 (non-crisis
period), and .51 (crisis period). These correlations may be
viewed as high, raising a criticism that LEV factor may be
collinear with HML, and as such, rendering the effects of
HML insignificant in majority of the cases. We argue that
while HML and LEV are both balance sheet variables and
are therefore should be correlated, they do not represent
similar risk factor in the economy. That is to say that HML
does not represent LEV and LEV does not represent HML.
Both are capturing economy wide risk. However, while
HML is supposed to be capturing distress risk (Fama and
French 1993)), it does not adequately capture systemic risk
of firms when their exposure to the debt market rises due
to economy-wide problems with over leveraging. To this
extent, LEV adds unique information to the model and does
a good job in capturing systemic risk related to the over
exposure of firms to the debt market.
While a correlation of .51 may not be indicative of
multicollinearity, it is important to examine if these results
are robust to such statistical artefact. In the present context,
we do not find a glaring evidence of multicollinearity
because it would have been reflected in high F-statistics
with insignificant t-statistics for the estimated coefficients.
Despite the fact that multicollinearity is not an issue, we
decided to estimate the partial F-statistics to check the
robustness of these results to multicollinearity. The partial
F-statistic determines the incremental explanatory power
of adding additional variables to the basic model. In the
present context, a significant partial F statistic (critical
value is 3.32 at the 1% significance level) provides
justification for adding LEV to the model containing the
traditional FF factors.
Table 16 reports the partial F-statistics (across all the three
groups - all stock portfolios, financial stocks only portfolios,
Islamic banking and finance – Essays on corporate finance, efficiency and product development
A comparison among Islamic, conventional, and socially responsible stocks
Table 10. Factor loadings for financial and non-financial firms for the aggregate period (January 2000 to April 2009).
Model 1
Model 2
rit - rft = b 0 + b 1 (rmt - rft ) + b 2 R t ,SMB + b 3 R t ,HML + e t
rit - rft = b 0 + b 1 (rmt - rft ) + b 2 R t ,SMB + b 3 R t ,HML + b 4 R t ,LEV + e t
e t |ψ t -1~N( 0 , σ t2 ),
e t |ψ t -1~N( 0 , σ t2 ),
q
p
q
σ t2 = W + ∑ a i e t2-i + ∑ δ iσ t - j
p
σ t2 = W + ∑ a i e t2-i + ∑ δ iσ t - j
i=1
j =1
i=1
j =1
where, ri is the return on portfolio i; rf is the return on the risk free asset and rm is the return on the market portfolio. RSMB is the return
on the size mimicking portfolio constructed by taking the simple average of the returns each week of all “small” portfolios minus “big”
portfolios. RHML is the return on book to market mimicking portfolio constructed by taking the simple average of the returns each week of
all “high BE/ME” portfolios minus “low BE/ME” portfolios. RLEV is the return on leverage mimicking portfolios constructed by taking the
simple average of the returns each week of all “high leverage” portfolios minus “low leverage portfolios”. All indicated coefficients are
significant at 5% level of significance. GARCH models are estimated using the Bollerslev-Wooldridge corrections to the standard errors.
Model 1 excludes LEV. Model 2 includes LEV. Coefficients of the GARCH variance equations are not reported to conserve space. They are
available upon request.
Aggregate Period
Model 1
Portfolio
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
Intercept
0.002*
0.001
0.003*
0.001
0.000
0.003*
0.000
0.002
0.002*
0.002
0.001
0.002*
0.001
0.001
0.002*
0.000
0.003*
0.002*
0.001
0.001*
0.001
0.001
0.002*
0.002*
0.001
0.001
0.001
MKT
0.709*
0.379*
0.783*
0.485*
0.473*
0.463*
0.383*
0.796*
0.756*
0.546*
0.552*
0.520*
0.503*
0.683*
0.369*
0.413*
0.713*
0.414*
0.668*
0.593*
0.743*
0.670*
0.693*
0.740*
0.609*
0.627*
0.679*
SMB
-0.026
0.514*
0.138*
0.151*
0.069
0.082*
0.173*
0.211*
0.399*
-0.315*
-0.190*
-0.154*
-0.120*
-0.117*
-0.085
-0.064
-0.074
-0.112*
-0.591*
-0.582*
-0.598*
-0.911*
-0.458*
-0.676*
-0.617*
-0.639*
-1.120*
Model 2
HML
-0.189*
-0.395*
-0.253*
0.337*
0.203*
0.188*
0.822*
0.687*
0.783*
-0.245*
-0.125*
-0.302*
0.312*
0.164*
0.085
0.718*
0.627*
0.482*
-0.067
-0.171*
0.007
0.035
0.051
0.207*
0.711*
0.611*
0.785*
Adj.
R-square Intercept
-0.060
-0.098
0.060
0.023
-0.015
0.038
0.121
-0.052
-0.002
0.178
0.088
0.109
0.133
-0.022
0.071
0.187
0.072
0.142
0.316
0.341
0.298
0.498
0.270
0.420
0.373
0.357
0.478
and non-financial stock only portfolios) for the aggregate,
non-crisis, and crisis periods, respectively. For the combined
stock portfolios, the partial F statistic is significant in 22
out of 27 portfolios during the aggregate period. During
the non-crisis period, the number of cases of significant
partial F statistics is reduced to 18. Similar results can be
seen for financial and non-financial stock portfolios during
the aggregate and the non-crisis period. However, the effect
Eds. Hatem A. El-Karanshawy et al.
0.002*
0.001
0.002
0.001
0.000
0.003*
0.000
0.001
0.001
0.002
0.001
0.001
0.001
0.001
0.002*
0.000
0.002*
0.002*
0.001
0.001
0.001
0.001
0.001
0.002
0.001
0.001
0.000
MKT
SMB
0.709*
0.263*
0.716*
0.478*
0.441*
0.423*
0.409*
0.702*
0.673*
0.535*
0.505*
0.491*
0.501*
0.633*
0.356*
0.423*
0.643*
0.371*
0.653*
0.564*
0.706*
0.661*
0.664*
0.682*
0.608*
0.579*
0.505*
-0.026
0.431*
0.163*
0.159*
0.066
0.085*
0.175*
0.193*
0.368*
-0.317*
-0.220*
-0.139*
-0.122*
-0.127*
-0.092*
-0.068
-0.108
-0.177*
-0.598*
-0.605*
-0.609*
-0.905*
-0.475*
-0.771*
-0.617*
-0.646*
-1.087*
HML
Adj.
LEV R-square
-0.189* -0.001 -0.063
-0.277* 0.857* 0.113
-0.189* 1.112* 0.348
0.333* 0.149
0.043
0.256* 0.419* 0.149
0.192* 0.462* 0.204
0.827* -0.105
0.094
0.753* 0.544* 0.121
0.797* 0.802* 0.267
0.209
-0.230* 0.112
-0.126* 0.332* 0.214
-0.247* 0.809* 0.359
0.312* 0.030
0.140
0.202* 0.384* 0.123
0.092
0.489* 0.262
0.713* -0.071
0.175
0.707* 0.602* 0.241
0.584* 0.737* 0.408
0.119
0.340
-0.070
0.391
-0.158* 0.169
0.010
0.374* 0.392
0.037
0.108* 0.510
0.118
0.324
0.368
0.190* 0.577* 0.566
0.711* 0.028* 0.375
0.689* 0.452* 0.439
1.113* 1.562* 0.699
of LEV is predominantly high during the crisis period with
significant partial F statistics in 27 cases for combined and
non-financial stock portfolios, and in 26 cases for financial
stocks only portfolios. This supports the evidence presented
earlier suggesting that compared to HML, LEV incorporates
additional and unique information concerning distress
risk exposure of the firms. In particular, the effect of LEV is
particularly dominant during the crisis period.
63
Bhatt and Sultan
Table 11. Factor loadings for financial and non-financial firms for non-crisis period (January 2000 to June 2007).
Model 1
Model 2
rit - r ft = β 0 + β 1(rmt - r ft ) + β 2 R t ,SMB + β 3 R t ,HML + e t
rit - r ft = β 0 + β 1(rmt - r ft ) + β 2 R t ,SMB + β 3 R t ,HML + β 4 R t ,LEV + e t
e t |ψ t -1~N(0 , σ ),
e t |ψ t -1~N(0 , σ t2 ),
2
t
q
q
p
p
σ t2 = W + ∑ a ie 2 t -i + ∑ δ iσ t - j
σ t2 = W + ∑ a ie 2 t -i + ∑ δ iσ t - j
i =1
j =1
i =1
j =1
where, ri is the return on portfolio i; rf is the return on the risk free asset and rm is the return on the market portfolio. RSMB is the return
on the size mimicking portfolio constructed by taking the simple average of the returns each week of all “small” portfolios minus “big”
portfolios. RHML is the return on book to market mimicking portfolio constructed by taking the simple average of the returns each week of
all “high BE/ME” portfolios minus “low BE/ME” portfolios. RLEV is the return on leverage mimicking portfolios constructed by taking the
simple average of the returns each week of all “high leverage” portfolios minus “low leverage portfolios”. All indicated coefficients with
(*) are significant at 5% level of significance. GARCH models are estimated using the Bollerslev-Wooldridge corrections to the standard
errors. Model 1 excludes LEV. Model 2 includes LEV. Coefficients of the GARCH variance equations are not reported to conserve space.
They are available upon request.
Non-crisis Period
Model 1
Portfolio
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
Intercept
MKT
0.002*
0.001
0.003*
0.002
0.000
0.003*
0.000
0.001
0.002*
0.002*
0.001
0.002*
0.001
0.002*
0.003*
0.001
0.003*
0.002*
0.001
0.001*
0.001
0.001
0.002*
0.002
0.001
0.002
0.002
0.757*
0.547*
0.873*
0.586*
0.508*
0.497*
0.643*
0.842*
0.792*
0.592*
0.600*
0.590*
0.551*
0.444*
0.362*
0.524*
0.778*
0.417*
0.727*
0.663*
0.776*
0.745*
0.747*
0.789*
0.655*
0.728*
0.780*
SMB
0.110
0.922*
0.365*
0.311*
0.167*
0.145*
0.362*
0.354*
0.547*
-0.252*
-0.104*
-0.056
-0.054
-0.095*
-0.033
0.063
0.043
-0.017
-0.493*
-0.454*
-0.512*
-0.795*
-0.367*
-0.523*
-0.380*
-0.472*
-0.850*
Factor loadings by types of firms
Model 2
HML
-0.172*
-0.542*
-0.325*
0.296*
0.171*
0.208*
0.800*
0.678*
0.759*
-0.282*
-0.143*
-0.311*
0.295*
0.137*
0.030
0.702*
0.588*
0.419*
-0.040
-0.224*
0.003
0.061
-0.008
0.235*
0.651*
0.497*
0.655*
Adj.
R-square Intercept
0.168
0.143
0.265
0.262
0.182
0.236
0.342
0.265
0.307
0.301
0.224
0.168
0.299
0.280
0.158
0.316
0.233
0.242
0.452
0.504
0.522
0.518
0.376
0.442
0.236
0.272
0.347
Previously, we noted that LEV is a good proxy for distress
risk across financial and non-financial stocks. We find that
stocks have similar sensitivities to the leverage risk factor.
In this section, we further conduct an additional robustness
check to examine whether there are differences in the
way various categories of firms respond to the economy
wide risk factors because they are classified by stock
64
0.002*
0.001
0.002
0.002
0.000
0.003*
-0.001
0.001
0.002
0.002*
0.001
0.002
0.001
0.001
0.002*
0.001
0.003*
0.002*
0.001
0.001*
0.001
0.002
0.002*
0.002
0.001
0.002
0.001
MKT
0.784*
0.521*
0.867*
0.589*
0.499*
0.466*
0.683*
0.834*
0.753*
0.594*
0.589*
0.562*
0.557*
0.732*
0.370*
0.553*
0.768*
0.405*
0.731*
0.663*
0.765*
0.755*
0.749*
0.773*
0.668*
0.726*
0.712*
SMB
0.110
0.897*
0.411*
0.316*
0.160*
0.136*
0.443*
0.312*
0.513*
-0.251*
-0.121*
-0.055
-0.038
-0.045
-0.035
0.065
0.020
-0.020
-0.483*
-0.454*
-0.535*
-0.781*
-0.366*
-0.544*
-0.360*
-0.472*
-0.862*
HML
-0.223*
-0.496*
-0.283*
0.294*
0.193*
0.208*
0.826*
0.693*
0.786*
-0.286*
-0.143*
-0.276*
0.293*
0.149*
0.052
0.666*
0.624*
0.452*
-0.044
-0.224*
0.017
0.056
0.009
0.234*
0.602*
0.525*
0.956*
Adj.
LEV R-square
-0.269*
0.375*
0.673*
-0.035
0.186*
0.296*
-0.408*
0.359*
0.477*
-0.038
0.126
0.411*
-0.153
0.277*
0.335*
-0.480*
0.377*
0.417*
-0.096
0.001
0.209*
-0.110
0.088
0.212*
-0.436*
0.184
1.241*
0.170
0.155
0.317
0.260
0.185
0.257
0.333
0.274
0.347
0.300
0.232
0.215
0.301
0.198
0.192
0.351
0.249
0.283
0.449
0.503
0.537
0.516
0.378
0.461
0.253
0.268
0.497
exchanges as meeting desired criteria for various style of
investing. For instance, the Dow Jones classifies investing
in certain stocks (popular household names) under broad
categories such as socially responsible investing because
these firms promote social, environmental, and corporate
responsibility. To this extent, we consider conventional,
Islamic and Socially Responsible Investing (SRI) stocks,
where each group exhibits distinct characteristics27. There
Islamic banking and finance – Essays on corporate finance, efficiency and product development
A comparison among Islamic, conventional, and socially responsible stocks
Table 12. Factor loadings for financial and non-financial firms for the crisis period (July 2007 to April 2009).
Model 1
Model 2
rit - r ft = β 0 + β 1(rmt - r ft ) + β 2 R t ,SMB + β 3 R t ,HML + e t
rit - r ft = β 0 + β 1(rmt - r ft ) + β 2 R t ,SMB + β 3 R t ,HML + β 4 R t ,LEV + e t
e t |ψ t -1~N(0 , σ ),
e t |ψ t -1~N(0 , σ t2 ),
2
t
q
q
p
p
σ t2 = W + ∑ a ie 2 t -i + ∑ δ iσ t - j
σ t2 = W + ∑ a ie 2 t -i + ∑ δ iσ t - j
i =1
j =1
i =1
j =1
where, ri is the return on portfolio i; rf is the return on the risk free asset and rm is the return on the market portfolio. RSMB is the return
on the size mimicking portfolio constructed by taking the simple average of the returns each week of all “small” portfolios minus “big”
portfolios. RHML is the return on book to market mimicking portfolio constructed by taking the simple average of the returns each week of
all “high BE/ME” portfolios minus “low BE/ME” portfolios. RLEV is the return on leverage mimicking portfolios constructed by taking the
simple average of the returns each week of all “high leverage” portfolios minus “low leverage portfolios”. All indicated coefficients with
(*) are significant at 5% level of significance. GARCH models are estimated using the Bollerslev-Wooldridge corrections to the standard
errors. Model 1 excludes LEV. Model 2 includes LEV. Coefficients of the GARCH variance equations are not reported to conserve space.
They are available upon request.
Crisis period
Model 1
Model 2
Portfolio
Intercept
MKT
SMB
HML
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
-0.004
-0.003
-0.005
-0.003
-0.005
-0.005
-0.005
-0.006*
-0.003
-0.005
-0.004
-0.006
-0.004
-0.003
-0.009*
-0.002
-0.004
-0.007*
-0.001
-0.001
-0.006
-0.003
-0.004
-0.005
-0.006
-0.001
-0.008*
-0.126
-0.051
0.596*
0.101
0.332*
0.390*
0.092
-0.034
0.252
0.344*
0.261*
0.257*
0.239*
0.266*
0.326*
0.143
0.203*
0.304*
0.308*
0.207
0.264
0.292*
0.197*
0.154
0.235
0.034
0.393*
-0.577*
-0.644*
-1.159*
-0.197
-0.794*
-0.636*
-0.115
-0.553*
-0.724*
-0.998*
-1.037*
-1.394*
-0.568*
-1.062*
-1.256*
-0.338*
-1.145*
-1.314*
-1.277*
-1.383*
-1.458*
-1.459*
-1.537*
-1.729*
-1.570*
-1.683*
-2.280*
-0.168
-0.708*
-1.536*
0.265
-0.222
-0.394
0.554*
0.103
0.190
-0.588
-0.531
-1.299*
-0.145
-0.332
-0.687
0.393*
-0.110
0.109
-0.542
-0.437
-1.107*
-0.528
-0.452
0.039
0.070
0.908*
0.121
Adj.
R-square Intercept
0.130
0.137
0.348
0.069
0.217
0.136
0.059
0.144
0.292
0.361
0.398
0.499
0.238
0.350
0.383
0.231
0.423
0.488
0.469
0.552
0.473
0.614
0.625
0.704
0.647
0.695
0.720
are distinct differences among these groups with respect to
the fundamentals such as size, ROA, ROE, leverage, return
on capital, PE ratio, and EPS. (See Milly and Sultan (2009)
for further evidences.)
We use stocks included in the Dow Jones Islamic Index
(DJIM) which is a proprietary index of stocks classified as
Islamic stocks by the Dow Jones Shariah Board. Because of
proprietary nature of such classifications, the names of the
Eds. Hatem A. El-Karanshawy et al.
-0.003
0.001
-0.002
-0.003
-0.003
-0.003
-0.004
-0.004
0.001
-0.003
-0.003
-0.004
-0.004
-0.002
-0.005
-0.002
-0.002
-0.006
0.000
0.000
-0.003
-0.002
-0.002
-0.001
-0.005
0.002
-0.005
MKT
SMB
-0.154 -0.283
-0.104 -0.002
0.266* -0.263*
0.048 -0.065
0.074 -0.140
0.201 0.227
0.069 -0.027
0.284* -0.184
0.178 0.094
0.238* -0.695*
0.140 -0.664*
0.133 -0.611*
0.193* -0.434*
0.152 -0.544*
0.103 -0.422*
0.054 -0.161
0.140 -0.600*
0.087 -0.596*
0.227 -0.869*
0.107 -0.913*
0.118 -0.837*
0.178* -1.127*
0.107 -0.881*
0.106 -1.108*
0.130 -1.146*
-0.052 -0.970*
0.161 -1.374*
HML
0.051
-0.187
-0.952*
0.336
0.184
0.131
0.619*
0.241
0.978*
-0.404
-0.276
-0.809*
-0.083
0.050
-0.132
0.442*
0.421
0.519*
-0.225
-0.011
-0.497*
-0.292
0.056
0.268
0.462
1.247*
0.854*
Adj.
LEV R-square
0.529
1.207*
1.642*
0.257
1.169*
1.551*
0.167
0.813*
1.691*
0.637*
0.733*
1.446*
0.228
0.843*
1.616*
0.268*
1.005*
1.275*
0.758*
0.917*
1.275*
0.615*
1.182*
1.275*
0.725*
1.255*
1.686*
0.203
0.357
0.653
0.102
0.483
0.408
0.085
0.251
0.542
0.449
0.506
0.696
0.277
0.466
0.603
0.265
0.539
0.671
0.543
0.644
0.626
0.661
0.736
0.809
0.671
0.770
0.826
stocks are withheld though some of the common household
names in the US may be classified as Islamic stocks because
they meet the requirements set by the Dow Jones Shariah
Board. On October 29, 2010, the market capitalization of
the Dow Jones Islamic World Index was $20 billion with
2,369 stocks. The weights (%) for some of the major
countries in the index are as follows: US (50.54), UK
(6.71), Japan (5.42), Canada (5.27), Switzerland (3.45),
Australia (3.26), France (2.97), India (2.5), Taiwan (2.2),
65
Bhatt and Sultan
Table 13. Factor loadings for the non-financial stock portfolios for aggregate period (January 2000 to April 2009).
Model 1
Model 2
rit - r ft = β 0 + β 1(rmt - r ft ) + β 2 R t ,SMB + β 3 R t ,HML + e t
rit - r ft = β 0 + β 1(rmt - r ft ) + β 2 R t ,SMB + β 3 R t ,HML + β 4 R t ,LEV + e t
e t |ψ t -1~N(0 , σ ),
e t |ψ t -1~N(0 , σ t2 ),
2
t
q
q
p
p
σ t2 = W + ∑ a ie 2 t -i + ∑ δ iσ t - j
σ t2 = W + ∑ a ie 2 t -i + ∑ δ iσ t - j
i =1
j =1
i =1
j =1
where, ri is the return on portfolio i; rf is the return on the risk free asset and rm is the return on the market portfolio. RSMB is the return
on the size mimicking portfolio constructed by taking the simple average of the returns each week of all “small” portfolios minus “big”
portfolios. RHML is the return on book to market mimicking portfolio constructed by taking the simple average of the returns each week of
all “high BE/ME” portfolios minus “low BE/ME” portfolios. RLEV is the return on leverage mimicking portfolios constructed by taking the
simple average of the returns each week of all “high leverage” portfolios minus “low leverage portfolios”. All indicated coefficients are
significant at 5% level of significance. GARCH models are estimated using the Bollerslev-Wooldridge corrections to the standard errors.
Model 1 excludes LEV. Model 2 includes LEV. Coefficients of the GARCH variance equations are not reported to conserve space. They are
available upon request.
Aggregate Period
Model 1
Portfolio Intercept
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
0.002*
0.002*
0.000
0.001
0.000
0.000
0.000
0.000
0.000
-0.001
0.001
0.000
0.000
0.001
0.000
0.000
0.000
0.001
0.001
0.000
0.000
0.001
0.001
0.001
0.001
0.001
0.002
MKT
SMB
0.776*
0.646*
0.698*
0.584*
0.668*
0.775*
0.621*
0.741*
0.838*
0.958*
0.709*
0.704*
0.682*
0.629*
0.665*
0.686*
0.723*
0.848*
0.745*
0.638*
0.589*
0.685*
0.672*
0.647*
0.666*
0.772*
0.791*
0.453*
0.380*
0.614*
0.345*
0.533*
0.570*
0.594*
0.729*
0.812*
0.098
0.094
0.021
0.092
0.026
0.168*
0.341*
0.270*
0.253*
-0.459*
-0.372*
-0.353*
-0.394*
-0.243*
-0.165*
-1.074*
-0.187*
-0.408*
Model 2
HML
-0.425*
-0.097
0.048
0.215*
0.376*
0.586*
0.690*
0.787*
0.893*
-0.309*
-0.144
-0.094
0.281*
0.287*
0.474*
0.790*
0.641*
0.796*
-0.372*
-0.008
0.000
0.224*
0.257*
0.438*
1.002*
0.701*
0.796*
Adj.
R-square Intercept
0.076
0.011
0.041
0.023
0.048
0.014
0.109
0.075
0.103
-0.054
0.025
0.004
-0.016
0.036
0.032
0.056
0.051
0.033
0.194
0.054
0.102
0.043
0.031
-0.003
0.236
0.054
0.125
Germany (1.73), South Korea (1.56), Brazil (1.5), Russia
(1.47), China (1.39), Hong Kong (1.25), and Sweden
(1.09). Among some of the traditionally Muslim majority
countries, the weights are: Malaysia (.35), Kuwait (.22),
Qatar (.08), UAE (.03), and Bahrain (.01).
Our selection of SC stocks is in line with the recent
interest in the performance of faith based investing, with
its overarching goal to promote the betterment of society,
66
0.002
0.002
0.000
0.001
0.000
0.000
0.000
0.000
0.000
-0.001
0.001
0.000
0.000
0.001
0.000
0.000
0.000
0.000
0.001
0.000
0.000
0.001
0.001
0.000
0.001
0.000
0.001
MKT
0.774*
0.630*
0.686*
0.587*
0.677*
0.779*
0.626*
0.748*
0.871*
0.943*
0.727*
0.740*
0.696*
0.673*
0.692*
0.687*
0.753*
0.856*
0.743*
0.651*
0.604*
0.685*
0.682*
0.659*
0.672*
0.767*
0.794*
SMB
0.468*
0.391*
0.548*
0.342*
0.521*
0.554*
0.613*
0.723*
0.818*
0.112
0.084
-0.007
0.081
-0.019
0.125*
0.331*
0.210*
0.182*
-0.454*
-0.392*
-0.364*
-0.396*
-0.301*
-0.195*
-1.089*
-0.233*
-0.443*
HML
-0.395*
-0.246*
-0.185*
0.206*
0.336*
0.551*
0.708*
0.752*
0.767*
-0.276*
-0.197*
-0.224*
0.246*
0.185*
0.358*
0.757*
0.528*
0.699*
-0.347*
-0.053
-0.131
0.210*
0.178*
0.312*
1.105*
0.585*
0.589*
Adj.
LEV R-square
-0.196 0.057
0.679* 0.095
1.123* 0.166
0.043 0.023
0.211* 0.071
0.181 0.031
-0.101 0.094
0.136 0.083
0.454* 0.130
-0.161 -0.060
0.238* 0.047
0.558* 0.064
0.189 -0.007
0.463* 0.088
0.531* 0.100
0.137 0.065
0.498* 0.110
0.809* 0.140
-0.166 0.187
0.369* 0.098
0.628* 0.185
0.049 0.043
0.493* 0.094
0.504* 0.067
-0.505* 0.236
0.814* 0.140
1.068* 0.241
relative to conventional investment strategies, which lack
such ethical ambition. SC stocks are popular among a
new class of investors that, in addition to profit motives,
is also driven by their desire to live ethically and invest
morally. Compared to the conventional Western financial
system, Islamic finance is a newcomer to the global
financial world, encompassing somewhere between
$750 billion to $1 trillion of investments in firms and
projects that are classified as SC. Yet, over the past few
Islamic banking and finance – Essays on corporate finance, efficiency and product development
A comparison among Islamic, conventional, and socially responsible stocks
Table 14. Factor loadings for non-financial stocks portfolio for non-crisis period (January 2000 to June 2007)
Model 1
Model 2
rit - r ft = β 0 + β 1(rmt - r ft ) + β 2 R t ,SMB + β 3 R t ,HML + e t
rit - r ft = β 0 + β 1(rmt - r ft ) + β 2 R t ,SMB + β 3 R t ,HML + β 4 R t ,LEV + e t
e t |ψ t -1~N(0 , σ ),
e t |ψ t -1~N(0 , σ t2 ),
2
t
q
q
p
p
σ t2 = W + ∑ a ie 2 t -i + ∑ δ iσ t - j
σ t2 = W + ∑ a ie 2 t -i + ∑ δ iσ t - j
i =1
j =1
i =1
j =1
where, ri is the return on portfolio i; rf is the return on the risk free asset and rm is the return on the market portfolio. RSMB is the return
on the size mimicking portfolio constructed by taking the simple average of the returns each week of all “small” portfolios minus “big”
portfolios. RHML is the return on book to market mimicking portfolio constructed by taking the simple average of the returns each week of
all “high BE/ME” portfolios minus “low BE/ME” portfolios. RLEV is the return on leverage mimicking portfolios constructed by taking the
simple average of the returns each week of all “high leverage” portfolios minus “low leverage portfolios”. All indicated coefficients with
(*) are significant at 5% level of significance. GARCH models are estimated using the Bollerslev-Wooldridge corrections to the standard
errors. Model 1 excludes LEV. Model 2 includes LEV. Coefficients of the GARCH variance equations are not reported to conserve space.
They are available upon request.
Non-crisis Period
Model 1
Portfolio
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
Intercept
MKT
0.002
0.002*
0.000
0.001
0.000
0.000
-0.001
0.000
0.000
-0.001
0.001
0.000
0.000
0.001
0.000
0.000
0.000
0.001
0.000
0.000
0.000
0.001
0.001
0.000
0.001
0.001
0.001
0.846*
0.732*
0.782*
0.673*
0.718*
0.825*
0.758*
0.789*
0.903*
1.023*
0.775*
0.764*
0.756*
0.689*
0.711*
0.759*
0.759*
0.898*
0.849*
0.664*
0.639*
0.801*
0.727*
0.699*
0.820*
0.927*
0.865*
SMB
0.462*
0.415*
0.627*
0.363*
0.518*
0.573*
0.638*
0.729*
0.807*
0.121
0.057
0.044
0.140*
0.026
0.160*
0.368*
0.280*
0.249*
-0.432*
-0.367*
-0.323*
-0.370*
-0.242*
-0.173*
-1.110*
-0.342*
-0.388*
Model 2
HML
-0.390*
-0.098
0.115
0.268*
0.415*
0.613*
0.673*
0.815*
0.953*
-0.323*
-0.095
-0.038
0.277*
0.328*
0.499*
0.798*
0.657*
0.801*
-0.375*
0.013
-0.002
0.278*
0.308*
0.472*
1.026*
0.879*
0.849*
Adj.
R-square Intercept
0.483
0.238
0.347
0.406
0.451
0.453
0.551
0.527
0.471
0.465
0.458
0.357
0.401
0.411
0.444
0.475
0.495
0.447
0.549
0.475
0.414
0.403
0.460
0.416
0.436
0.358
0.410
years Islamic investments have become more competitive
and consequently attractive not only to Muslim but also
non-Muslim investors seeking alternative investments
opportunities, which live up to high ethical as well as
nominal performance standards. As a result, the number
of Islamic mutual funds and exchange traded funds
world-wide has increased considerably from merely 8
before 1992 to more than 300 in 2008, with an estimated
Eds. Hatem A. El-Karanshawy et al.
0.002
0.002*
0.000
0.001
0.000
0.000
-0.001
0.000
0.000
-0.001
0.001
0.000
0.000
0.001
0.000
0.000
0.000
0.001
0.000
0.000
-0.001
0.001
0.001
0.000
0.001
0.001
0.001
MKT
SMB
HML
0.830*
0.741*
0.821*
0.666*
0.726*
0.835*
0.750*
0.795*
0.939*
0.982*
0.790*
0.812*
0.766*
0.731*
0.751*
0.761*
0.794*
0.932*
0.912*
0.676*
0.769*
0.799*
0.769*
0.731*
0.764*
0.977*
0.890*
0.482*
0.407*
0.599*
0.366*
0.513*
0.561*
0.647*
0.726*
0.800*
0.125
0.051
0.026
0.136*
0.000
0.132*
0.366*
0.240*
0.192*
-0.380*
-0.385*
-0.400*
-0.369*
-0.268*
-0.208*
-1.118*
-0.360*
-0.412*
-0.321*
-0.164
-0.126
0.278*
0.394*
0.590*
0.719*
0.791*
0.862*
-0.243*
-0.127
-0.125
0.255*
0.257*
0.408*
0.791*
0.570*
0.726*
-0.428*
-0.026
-0.109
0.290*
0.242*
0.381*
1.202*
0.763*
0.656*
Adj.
LEV R-square
-0.364*
0.360*
0.923*
-0.057
0.109
0.130
-0.147
0.092
0.385*
-0.292*
0.150
0.401*
0.111
0.366*
0.440*
0.031
0.394*
0.710*
-0.402*
0.312*
0.649*
-0.048
0.396*
0.422*
-0.918*
0.552*
0.885*
0.496
0.251
0.371
0.408
0.450
0.451
0.554
0.525
0.469
0.480
0.457
0.365
0.396
0.419
0.458
0.473
0.502
0.461
0.565
0.486
0.418
0.404
0.470
0.435
0.492
0.358
0.444
market capitalization of $300 billion and numerous
traditional US financial institutions joining to partake in
this development.
Similarly, the SRI class of stocks is a relative newcomer,
which has gained popularity in recent years. In the
early 2000s, we have seen a dramatic interest in socially
responsible investing that poured billions of dollars
67
Bhatt and Sultan
Table 15. Factor loadings for non-financial stock portfolios during crisis period (July 2007 to April 2009).
Model 1
Model 2
rit - r ft = β 0 + β 1(rmt - r ft ) + β 2 R t ,SMB + β 3 R t ,HML + β 4 R t ,LEV + e t
rit - r ft = β 0 + β 1(rmt - r ft ) + β 2 R t ,SMB + β 3 R t ,HML + e t
e t |ψ t -1~N(0 , σ t2 ),
e t |ψ t -1~N(0 , σ t2 ),
σ t2 = W + ∑ a ie 2 t -i + ∑ δ iσ t - j
σ t2 = W + ∑ a ie 2 t -i + ∑ δ iσ t - j
q
p
i =1
j =1
q
p
i =1
j =1
where, ri is the return on portfolio i; rf is the return on the risk free asset and rm is the return on the market portfolio. RSMB is the return
on the size mimicking portfolio constructed by taking the simple average of the returns each week of all “small” portfolios minus “big”
portfolios. RHML is the return on book to market mimicking portfolio constructed by taking the simple average of the returns each week of
all “high BE/ME” portfolios minus “low BE/ME” portfolios. RLEV is the return on leverage mimicking portfolios constructed by taking the
simple average of the returns each week of all “high leverage” portfolios minus “low leverage portfolios”. All indicated coefficients with
(*) are significant at 5% level of significance. GARCH models are estimated using the Bollerslev-Wooldridge corrections to the standard
errors. Model 1 excludes LEV. Model 2 includes LEV. Coefficients of the GARCH variance equations are not reported to conserve space.
They are available upon request.
Crisis Period
Model 1
Model 2
Portfolio
Intercept
MKT
SMB
HML
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
-0.005
-0.005
-0.004
-0.004
-0.005
-0.006
-0.004
-0.004
-0.004
-0.005
-0.006*
-0.005
-0.004
-0.005
-0.006*
-0.004
-0.005
-0.006
-0.004
-0.004
-0.005*
-0.003
-0.004
-0.005
-0.003
-0.006*
-0.005
0.156
0.124
0.043
0.062
0.086
0.143
0.079
0.110
0.151
-0.004
-0.004
0.047
0.025
0.041
0.064
0.033
0.015
-0.004
0.177
0.034
0.114
0.118
0.098
0.101
-0.033
0.171
-0.018
0.586
0.500
0.392
-0.008
0.047
0.427
0.233
0.393
0.646
-0.268
-0.549*
-0.535
-0.410
-0.471
-0.234
-0.027
-0.527
-0.415
-0.643*
-0.667*
-0.818*
-0.640*
-0.543
-0.468
-0.337
-1.072*
-0.622
-1.208*
-0.967*
-0.652
0.459
0.359
0.191
0.579*
0.305
0.032
0.055
0.112
-0.161
0.674*
0.241
0.221
0.539
0.772*
0.834*
-0.673*
-0.626*
-0.241
-0.076
-0.309
0.036
0.756*
0.708*
-0.079
Adj.
R-square Intercept
-0.092
-0.087
-0.092
-0.073
-0.043
-0.030
0.033
0.001
-0.040
-0.066
-0.066
-0.043
-0.073
-0.042
-0.045
-0.010
-0.022
0.004
-0.020
-0.020
-0.009
-0.070
-0.040
-0.044
0.039
0.010
-0.024
into companies known for their efforts to offer ethical
investments and projects that promoted environmental
sustainability. In terms of the portfolio allocation and
structure, Islamic and socially responsible investing (SRI)
stocks exhibit strong similarities, whereas conventional
stocks are not subject to any other qualitative or
quantitative constraints. Although SRI funds were initially
68
-0.004*
-0.003*
-0.002
-0.004
-0.005
-0.005
-0.003
-0.003
-0.002
-0.002
-0.004
-0.003
-0.003
-0.003
-0.003
-0.002
-0.002
-0.003
-0.002
-0.003
-0.003
-0.003
-0.003
-0.003
-0.003
-0.004
-0.003
MKT
SMB
HML
LEV
Adj.
R-square
0.051
0.044
0.017
0.047
0.049
0.053
-0.019
-0.008
0.018
0.078
0.019
0.070
0.092
0.034
0.032
-0.008
-0.014
0.090
0.054
-0.012
0.058
-0.021
0.000
0.017
-0.013
0.141
0.002
0.718*
0.832*
0.756*
0.377*
0.537*
0.733*
0.540*
0.716*
0.840*
0.212
0.055
-0.087
0.032
-0.056
0.147
0.445*
-0.015
0.104
-0.451*
-0.255
-0.427*
-0.302
-0.276
-0.119
-0.085
-0.379*
-0.546*
-1.277*
-1.075*
-0.717*
-0.081
-0.216
-0.074
0.307
0.141
0.150
-0.604*
-0.769*
-0.736*
-0.116
-0.287
-0.227
0.150
0.161
0.035
-0.773*
-0.933*
-0.616*
-0.430
-0.420
-0.201
0.351
0.093
0.192
2.808*
2.866*
3.192*
1.710*
2.292*
2.438*
1.427*
2.412*
3.242*
2.539*
2.855*
2.957*
1.839*
2.786*
2.963*
1.938*
3.016*
4.151*
2.140*
2.589*
2.331*
1.827*
2.608*
2.566*
1.865*
2.762*
3.637*
0.391
0.442
0.535
0.292
0.439
0.417
0.315
0.471
0.575
0.417
0.479
0.504
0.332
0.487
0.515
0.359
0.533
0.612
0.444
0.445
0.487
0.239
0.448
0.453
0.331
0.488
0.609
conceived in a religious context as well, socially responsible
investing has expanded to take in consideration “the socalled ‘triple bottom line’, commonly known as the ‘three
P’s rule: people, planet and profit’” (Forte & Miglietta,
2007, p. 3). Most recently, assets under SRI management
were estimated to have increased “from $639 billion in
1995 … to $2.71 trillion in 2007”, while “assets in all types
Islamic banking and finance – Essays on corporate finance, efficiency and product development
A comparison among Islamic, conventional, and socially responsible stocks
Table 16. Partial f-statistics testing for the significance of contribution made by the LEV factor.
Restricted Model
Unrestricted Model
rit - r ft = β 0 + β 1(rmt - r ft ) + β 2 R t ,SMB + β 3 R t ,HML + e t
rit - r ft = β 0 + β 1(rmt - r ft ) + β 2 R t ,SMB + β 3 R t ,HML + β 4 R t ,LEV + e t
e t |ψ t -1~N(0 , σ t2 ),
e t |ψ t -1~N(0 , σ t2 ),
q
p
i =1
j =1
σ t2 = W + ∑ a ie 2 t -i + ∑ δ iσ t - j
q
p
i =1
j =1
σ t2 = W + ∑ a ie 2 t -i + ∑ δ iσ t - j
where ri is the return on portfolio i; rf is the return on the risk free asset and rm is the return on the market portfolio. RSMB is the return
on the size mimicking portfolio constructed by taking the simple average of the returns each week of all “small” portfolios minus “big”
portfolios. RHML is the return on book to market mimicking portfolio constructed by taking the simple average of the returns each week
of all “high BE/ME” portfolios minus “low BE/ME” portfolios. RLEV is the return on leverage mimicking portfolios constructed by taking
the simple average of the returns each week of all “high leverage” portfolios minus “low leverage portfolios”. Partial f-statistics and the
p-values test for the significance in the contribution of R-square made by the new model (which includes the LEV factor). The factors
SMB, HML and LEV have been rebalanced for financial stock portfolios and non financial stock portfolios. (*) indicates significance at 5%
level of significance. GARCH models are estimated using the Bollerslev-Wooldridge corrections to the standard errors. Model 1 excludes
LEV. Model 2 includes LEV. Coefficients of the GARCH variance equations are not reported to conserve space. They are available upon
request.
All stock portfolios
Aggregate Non-crisis
period
period
Portfolio
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
Partial
f-statistic
-1.62
56.73*
101.36*
14.96*
34.46*
43.57*
-15.57
19.58*
35.62*
6.96*
37.50*
66.64*
25.01*
49.24*
68.41*
3.29
46.24*
80.45*
-4.56*
31.43*
49.63*
22.87*
49.86*
48.31*
-12.15*
50.41*
118.71*
Crisis
period
Financial stock portfolios
Aggregate Non-crisis
period
period
Partial
Partial
Partial
f-statistic f-statistic f-statistic
1.04
5.39*
28.68*
-0.01
7.13*
13.12*
0.34
2.22
4.19*
-0.11
8.58*
23.42*
2.55
12.04*
26.13*
0.71
12.02*
21.10*
9.18*
8.18*
18.60*
2.70
21.54*
21.56*
28.19*
1.07
44.78*
71.90*
82.82*
116.58*
45.12*
75.91*
70.42*
36.19*
75.98*
103.75*
76.02*
87.33*
89.13*
50.96*
89.40*
93.21*
36.76*
105.33*
108.63*
68.14*
67.16*
83.87*
39.97*
72.94*
78.73*
47.55*
87.66*
131.34*
-0.17
115.83*
214.21*
11.11*
94.36*
101.84*
-13.06
95.79*
178.07*
19.93*
78.43*
189.67*
4.82*
80.92*
126.42*
-6.22
108.33*
217.60*
18.88*
41.14*
75.18*
12.65*
75.48*
162.88*
2.89*
72.14*
355.94*
of socially and environmentally screened funds [… in the
US] rose to $201.8 billion.” (2007 Report on Socially
Responsible Investing Trends in the United States, 2008,
p. ii) The premise of the “three P’s rule” is reflected in a
definition of socially responsible investing, which can be
found in the 2005 Report on SRI Trends in the United States
released by the Social Investment Forum:
Eds. Hatem A. El-Karanshawy et al.
Crisis
period
Partial
Partial
f-statistic f-statistic
2.04
6.41*
30.94*
-0.07
2.16
11.67*
-3.84
5.93*
24.36*
0.39
4.73*
24.14*
2.34
-38.57
17.04*
22.03*
9.49*
23.24*
-0.99
0.03
13.06*
-0.83
2.52
15.23*
9.85*
-0.97
116.64*
9.64*
33.22*
83.68*
4.60*
49.55*
44.31*
3.66
14.53*
52.41*
16.06*
21.61*
62.10*
6.09*
21.53*
53.15*
5.42*
24.76*
53.18*
16.25*
25.47*
39.69*
13.91*
40.51*
52.25*
7.90*
31.74*
58.15*
Non-financial stock portfolios
Aggregate Non-crisis
period
period
Crisis
period
Partial
f-statistic
Partial
Partial
f-statistic f-statistic
-8.36
45.46*
72.96*
1.26
12.75*
9.16*
-6.98
5.07*
15.69*
-1.71
12.12*
31.73*
5.04*
28.36*
37.70*
5.44*
33.00*
61.30*
-3.45
24.38*
50.31*
1.05
34.80*
37.14*
0.62
49.04*
74.25*
10.65*
7.58*
16.11*
1.73
0.13
-0.18
2.99
-0.88
-0.56
12.03*
0.92
5.81*
-1.87
6.12*
10.87*
-0.52
6.07*
11.21*
15.24*
9.57*
3.95*
1.35
8.93*
13.95*
43.70*
1.20
25.09*
75.57*
90.27*
127.92*
49.54*
81.80*
73.26*
39.88*
84.74*
137.36*
78.90*
99.36*
104.67*
58.03*
97.85*
109.78*
55.13*
112.73*
148.07*
79.61*
79.64*
91.88*
39.27*
84.35*
86.59*
42.22*
88.80*
153.14*
Socially responsible investing (SRI) is an investment
process that considers the social and environmental
consequences of investments, both positive and negative,
within the context of rigorous financial analysis… It is a
process of identifying and investing in companies that meet
certain standards of Corporate Social Responsibility (CSR)
69
Bhatt and Sultan
(2004 Report on Socially Responsible Investing Trends in
the United States. 10 Year Review, 2005, p. 2).
The congruence of Islamic and SRI stocks stems from the
fact that both do not have profit maximization as their sole
objective, but rather strive to achieve a paramount, ethical
obligation and a social-utilitarian function. In the case of
Islamic funds, the religious responsibilities and regulations
outlined in the Shariah, take precedence over profit in
order to further the establishment of a just and moral
Islamic economic system and ultimately society.
In contrast, profit maximization is the dominant objective
in traditional fund management. Conventional equity
portfolio strategies include neither positive nor negative
screens, whose purpose it is to align the portfolio with certain
ethical, qualitative standards. As such, conventional funds
are not subject to the qualitative screening procedures that
are so imperative to Islamic and SRI funds. Additionally,
Islamic funds differ from SRI and conventional ones, since
their provisions incorporate quantitative screens that are
directly based on ethical paradigms found in the Shariah.
Furthermore, Islamic funds have to comply with certain
income purification requirements, which are derived from
the teachings of the Holy Quran and Sunnah.
The hypothesis tested is that high leverage increases
exposure to the credit market and subsequently translates
into shareholders demanding higher risk premium. Recall
that Islamic stocks have low leverage, they are significantly
more asset-backed than conventional firms, and are
not involved in the business of speculation, production
of weapons, alcohol, pork, and entertainment. More
specifically, Islamic funds typically screen out companies
with excessive reliance on debt, where the typical
maximum level of total debt to market capitalization is set
at 33 percent28.
The first step towards applying our leverage risk factor to
these index classifications is to recreate the FF and LEV
specific to each category of stocks. This is followed by
estimating GARCH regressions at the firm and portfolio
level.
Factor loadings of conventional stocks
In Table 17, we report a summary of firm and portfolio
specific regressions by groups. The first panel reports the
results for the firms belonging to the conventional stock
category. We find that at the firm level, the inclusion of
LEV produced some interesting results. Compared to the
aggregate period, the number of instances where the
XMKT is significant drops by 79.57% at the firm and by
100% at the portfolio level. The change in significance
for SMB is as follows: 8.89% at the firm and -17.39% at
the portfolio level. The results for the HML are consistent
across both the firm and the portfolio level. The number of
instances where HML is significant at the firm level drops
by 56.13% and by 44.44% at the portfolio level. Finally, the
number of instances LEV is significant increases by 231%
at the firm and by 58.82% at the portfolio level. Overall,
these results are qualitatively similar to the ones reported
earlier and confirm our earlier finding that the inclusion of
LEV subsumes the effects of the traditional FF factors to a
great extent.
70
Factor loadings of Islamic stocks
Panel B reports the results for the Islamic group of stocks.
Compared to the aggregate period, there is a remarkable
change in the number of cases of where XMKT is significant
(-87.6% at the firm and by -92.59% at the portfolio level).
The change in significance for SMB is as follows: 9.89%
at the firm and -44.44% at the portfolio level. The results
for the HML are again consistent across both the firm and
the portfolio level. The change in statistical significance
for HML is as follows: -56.44% at the firm and by -60%
at the portfolio level. Finally, the number of instances
LEV is significant increases by 98.9% at the firm and by
73.33% at the portfolio level. Again, our results are quite
consistent with the previous results reported without
the index classifications. Islamic stocks behave similar to
the conventional stocks when it comes to sensitivities to
economic risk factors.
Factor loadings of SRI stocks
In Panel C, we report the results for 238 stocks classified
as SRI group of stocks. Compared to the previous groups,
we have some unusual results. We find that, compared to
the aggregate period, the number of instances where the
XMKT is significant drops by 89% at the firm and by 100%
at the portfolio level. For SMB, the changes in the number
of significant cases are: -66% (firm-level) and -80.95%
(portfolio-level). In contrast to our previous results, the
number of instances where HML is significant at the firm
level increases by 56.43% and by 35% at the portfolio level.
Finally, the number of instances where LEV is significant
drops by 11.11% the firm and by 56.25% at the portfolio
level.
With respect to the effects of LEV risk factor, the results for
the SRI group are quite different from the Conventional
and Islamic stocks, suggesting that stocks in this category
are less sensitive to the economy-wide leverage risk
factor. Certainly, leverage risk for this type of firms is not
unusually different but perhaps the nature of the business
these firms are involved may make it less susceptive to
economy wide leverage risk. It may also be possible that
during the financial crisis, while socially responsible
investing would have earned positive risk premium with
respect the HML, SRI investors would have earned a
negative risk premium when leverage was employed as
a stock picking strategy. Whether SRI investing produces
a lower return because these stocks are generally less
sensitive to the economy wide risk factors suggests that
these stocks may offer significant diversification benefits.
Overall, further research along these lines would offer
more clues as to why SRI stocks have negative risk
premium for leverage risk.
Partial F-test
Table 18 reports partial F-statistics (across all the three
groups - all stock portfolios), for the aggregate, non-crisis,
and crisis periods in order to test for the significance of
the contribution made by LEV. For the combined stock
portfolios, the partial F statistic is significant in 24 out of
27 portfolios during the aggregate period. During the noncrisis period, the number of significant partial F statistics
is reduced to 17 cases. However, the effect of LEV is
prominent during the crisis period with significant partial
F statistics in all 27 portfolios. For the Islamic stocks, the
Islamic banking and finance – Essays on corporate finance, efficiency and product development
i =1
j =1
σ t2 = W + ∑ a ie 2 t -i + ∑ δ iσ t - j
p
i =1
j =1
σ t2 = W + ∑ a ie 2 t -i + ∑ δ iσ t - j
q
e t |ψ t -1~N(0 , σ t2 ),
e t |ψ t -1~N(0 , σ t2 ),
p
rit - r ft = β 0 + β 1(rmt - r ft ) + β 2 R t ,SMB + β 3 R t ,HML + β 4 R t ,LEV + e t
rit - r ft = β 0 + β 1(rmt - r ft ) + β 2 R t ,SMB + β 3 R t ,HML + e t
q
Model 2
Model 1
Eds. Hatem A. El-Karanshawy et al.
SMB
1108
1158
3.86%
2056
−0.96%
Total
6
1152
2056
0
1115
Positive
0
2076
Negative
Total
7
Negative
2076
Positive
%change in significance
(by model)
Model 2
Model 1
Non-crisis Period
SMB
−3.41%
XMKT
1590
9.88%
1697
Total
1582
0
Negative
8
1447
1697
1757
Total
1439
8
Positive
0
Negative
%change in significance
(by model)
Model 2
1757
Positive
XMKT
HML
5.17%
1362
499
863
1295
378
917
HML
2.48%
1279
482
797
1248
225
1023
LEV
539
80
459
0
LEV
1292
84
1208
Total
Negative
Positive
Total
Negative
Positive
Total
Negative
Positive
Total
Negative
Positive
%change in significance
(by model)
Model 2
Model 1
Non-crisis Period
%change in significance
(by model)
Model 2
Model 1
0.00%
27
0
27
27
0
27
XMKT
0.00%
27
0
27
27
0
27
XMKT
Aggregate Period
Aggregate Period
Model 1
Conventional portfolios: 27
Conventional stocks: 2308
PANEL A
9.52%
23
16
7
21
14
7
SMB
-4.76%
20
18
2
21
17
4
SMB
5.88%
18
3
15
17
0
17
HML
-16.67%
15
3
12
18
0
18
HML
(Continued)
17
2
15
LEV
20
0
20
LEV
where, ri is the return on portfolio i; rf is the return on the risk free asset and rm is the return on the market portfolio. RSMB is the return on the size mimicking portfolio constructed by taking the
simple average of the returns each week of all “small” portfolios minus “big” portfolios. RHML is the return on book to market mimicking portfolio constructed by taking the simple average of
the returns each week of all “high BE/ME” portfolios minus “low BE/ME” portfolios. RLEV is the return on leverage mimicking portfolios constructed by taking the simple average of the returns
each week of all “high leverage” portfolios minus “low leverage portfolios”. All indicated coefficients with (*) are significant at 5% level of significance.
Table 17. Factor loadings by types of firms.
A comparison among Islamic, conventional, and socially responsible stocks
71
72
SMB
1261
420
Total
Total
Negative
Positive
%change in significance
(by period)
%change in significance
(by model)
Model 2
Negative
Positive
1019
Total
1.49%
1019
0
1004
Total
Positive
0
Negative
Negative
1004
Positive
%change in significance
(by model)
Model 2
Model 1
4.71%
645
244
401
616
214
402
8.71%
487
239
248
448
196
252
590
202
388
Total
Negative
Positive
Total
Negative
Positive
%change in significance
(by model)
Model 2
Model 1
Aggregate Period
LEV
231.17%
1785
41
1744
Model 1
Islamic portfolios: 1161
HML
-56.31%
-42.68%
595
115
480
1038
12
LEV
0
Aggregate
Period
SMB
HML
1026
Crisis period
Islamic stocks: 1161
XMKT
8.89%
-79.57%
%change in significance
(by period)
PANEL B
0.24%
14.75%
Total
736
328
Negative
525
92
1258
996
262
Positive
366
76
Negative
Total
290
XMKT
Positive
%change in significance
(by model)
Model 2
Model 1
Crisis period
Table 17. (Continued)
0.00%
27
0
27
27
0
27
XMKT
-100.00%
0.00%
0
0
0
10
0
10
XMKT
11.76%
19
7
12
17
5
12
SMB
-17.39%
18.75%
19
14
5
16
16
0
SMB
HML
5.00%
21
9
12
20
8
12
HML
-44.44%
-62.96%
10
1
9
27
0
27
14
4
10
LEV
58.82%
27
0
27
LEV
Bhatt and Sultan
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Eds. Hatem A. El-Karanshawy et al.
7.10%
−0.78%
9.84%
−87.60%
%change in significance
(by period)
558
69
489
556
0.36%
135
Total
508
48
16.38%
92
43
Positive
116
Total
Negative
83
33
Positive
Negative
XMKT
%change in significance
(by model)
Model 2
Model 1
Crisis period
SMB
528
508
0.18%
1089
Total
−56.44%
−9.80%
230
97
133
255
64
191
HML
263
254
265
493
0
254
512
234
Negative
1087
Total
276
259
HML
1089
0
Negative
236
SMB
Positive
1087
XMKT
Positive
%change in significance
(by model)
Model 2
Model 1
Non-crisis Period
98.91%
911
38
873
LEV
458
280
178
LEV
Total
Negative
Positive
Total
Negative
Positive
Total
Negative
Positive
Total
Negative
Positive
%change in significance
(by period)
%change in significance
(by model)
Model 2
Model 1
Crisis period
%change in significance
(by model)
Model 2
Model 1
Non-crisis Period
XMKT
-92.59%
0.00%
2
0
2
0
0
0
XMKT
0.00%
27
0
27
27
0
27
SMB
−44.44%
−9.09%
10
0
10
11
11
0
SMB
5.88%
18
7
11
17
7
10
HML
−60.00%
−20.00%
8
2
6
10
3
7
HML
−4.76%
20
8
12
21
8
13
(Continued)
73.33%
26
0
26
LEV
15
11
4
LEV
A comparison among Islamic, conventional, and socially responsible stocks
73
74
224
Total
237
Total
0.85%
237
0
Positive
235
Total
Negative
235
0
Positive
XMKT
Negative
%change in significance
(by model)
Model 2
Model 1
Non-crisis Period
0.00%
224
0
Positive
224
Total
Negative
224
0
Positive
XMKT
Negative
%change in significance
(by model)
Model 2
Model 1
Aggregate Period
SRI stocks: 238
PANEL C
Table 17. (Continued)
140
3.70%
−0.63%
7
133
135
11
124
HML
4.37%
191
0
191
183
0
183
HML
159
83
76
160
82
78
SMB
-0.67%
148
69
79
149
69
80
SMB
45
27
18
0
LEV
63
51
12
LEV
Positive
Total
Negative
Positive
Total
Negative
Positive
Total
Negative
Positive
Total
Negative
%change in significance
(by model)
Model 2
Model 1
Non-crisis Period
%change in significance
(by model)
Model 2
Model 1
Aggregate Period
SRI Portfolios: 27
0.00%
27
0
27
27
0
27
XMKT
0.00%
27
0
27
27
0
27
XMKT
10.53%
21
10
11
19
10
9
SMB
10.53%
21
10
11
19
10
9
SMB
0.00%
20
3
17
20
3
17
HML
9.52%
23
2
21
21
1
20
HML
16
8
8
LEV
17
9
8
LEV
Bhatt and Sultan
Islamic banking and finance – Essays on corporate finance, efficiency and product development
−56.25%
−80.95%
−11.11%
%change in significance
(by period)
−100.00%
35.00%
7
0.00%
0.00%
4
0.00%
Total
56.43%
−66.04%
−89.03%
%change in significance
(by period)
−1.35%
40
219
12.50%
44.44%
26
Total
Eds. Hatem A. El-Karanshawy et al.
%change in significance
(by model)
54
12
0
15
14
Negative
222
219
18
Total
12
Model 2
Positive
48
%change in significance
(by model)
0
27
0
2
Negative
0
0
7
27
27
4
2
Total
0
Overall, the regression results suggest that while the
sensitivities of the portfolio returns to the FF factors are
significant during the aggregate and the non-crisis periods,
there are important changes in the sign and significance of
these factors during the crisis period. Their significance also
weakens with the introduction of leverage as a risk factor,
almost to the tune of being subsumed by the leverage risk
factor. The effects of the market factor are persistent before
the crisis period but surprisingly became insignificant
during the crisis period. Leverage factor is consistently
significant across all the periods and its effect is more
prominent during the crisis period due to the greater debt
exposure of the firms and higher macroeconomic risk. The
results further support the conclusions drawn in the earlier
tables.
0
results are quite strong. The number of cases F-statistics is
significant is 17 (aggregate period), 15 (non-crisis period),
and 27 (crisis period). Finally, we find that the SRI stocks
only portfolios are not sensitive to the leverage risk during
the credit crisis.
39
28
Model 2
Positive
0
1
0
12
10
222
36
Negative
HML
SMB
8
XMKT
Positive
Model 1
Crisis period
Negative
0
27
SMB
3
XMKT
LEV
Model 1
Crisis period
Positive
0
HML
LEV
A comparison among Islamic, conventional, and socially responsible stocks
Leverage risk factor for US stocks
A potential shortcoming of the preceding results is
due to the fact that our previous samples include
stocks traded globally and may not accurately quantify
the effects of the credit crisis on the US market. We
therefore conduct another experiment using US
stocks only. This additional exercise is carried out
by excluding all non-US stocks, creating traditional
Fama-French factors, and adding our newly created
LEV factor to represent financial distress. In addition,
we also estimate GARCH regressions to demonstrate
that our risk factors represent macroeconomic shocks
as well. While the results are not presented to save
space, we can summarize the results as follows. We
find that for predicting US industrial production, the
following variables are statistically significant: SMB,
HML, XMKT, HML (t-2), and LEV (t-2). For predicting
US unemployment rate, variables such as SMB, HML,
LEV and several interaction variables on LEV (for 2008
credit crisis period) are significant at various lag length.
We find that several interaction variables using LEV are
significant in predicting the credit spread and term
spread. Finally, dummy variables on LEV are highly
significant in affecting changes in the US inflation
rate. Overall, it is safe to conclude that the risk factors
contain adequate information on the US economy.
Next, GARCH regressions are estimated. For the aggregate
period (Jan2000 – April 2009), out of 27 portfolios, SMB
has 18 (6) positive (negative) coefficients. HML has 16
(8) positive (negative) coefficients. In contrast, for Model
2 (where we add LEV), the results are as follows: SMB
has 18 positive and 4 negative coefficients, HML has
17 positive and 9 negative coefficients, and LEV has 6
positive and 17 negative coefficients. For the non-crisis
period (January 2000 – June 2007), SMB has 18 (7)
positive (negative) coefficients. HML has 13 (10) positive
(negative) coefficients. In contrast, for Model 2, SMB
has 18 positive and 7 negative coefficients, HML has
14 positive and 9 negative coefficients, and LEV has 6
positive and 18 negative coefficients. In all cases, there
is a significant increase in R2 when we add LEV in the
model.
75
Bhatt and Sultan
Table 18. Partial f-statistics testing for the significance of contribution made by the LEV factor.
Old Model
New Model
rit - r ft = β 0 + β 1(rmt - r ft ) + β 2 R t ,SMB + β 3 R t ,HML + e t
rit - r ft = β 0 + β 1(rmt - r ft ) + β 2 R t ,SMB + β 3 R t ,HML + β 4 R t ,LEV + e t
e t |ψ t -1~N(0 , σ ),
e t |ψ t -1~N(0 , σ t2 ),
2
t
q
q
p
p
σ t2 = W + ∑ a ie 2 t -i + ∑ δ iσ t - j
σ t2 = W + ∑ a ie 2 t -i + ∑ δ iσ t - j
i =1
j =1
where, ri is the return on portfolio i; rf is the return on the risk free asset and rm is the return on the market portfolio. RSMB is the return
on the size mimicking portfolio constructed by taking the simple average of the returns each week of all “small” portfolios minus “big”
portfolios. RHML is the return on book to market mimicking portfolio constructed by taking the simple average of the returns each week
of all “high BE/ME” portfolios minus “low BE/ME” portfolios. RLEV is the return on leverage mimicking portfolios constructed by taking
the simple average of the returns each week of all “high leverage” portfolios minus “low leverage portfolios”. All indicated coefficients
with (*) are significant at 5% level of significance. Partial f-statistics and the p-values test for the significance in the contribution of
R-square made by the new model (which includes the LEV factor). GARCH models are estimated using the Bollerslev-Wooldridge
corrections to the standard errors. Model 1 excludes LEV. Model 2 includes LEV. Coefficients of the GARCH variance equations are not
reported to conserve space. They are available upon request.
i =1
Portfolio
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
j =1
Conventional Portfolios
Islamic Portfolios
SRI Portfolios
Aggregate Non-crisis Crisis
period
period
period
Aggregate Non-crisis Crisis
period
period
period
Aggregate Non-crisis Crisis
period
period
period
Partial
Partial
Partial
Partial
Partial
Partial
Partial
Partial
Partial
f-statistic f-statistic f-statistic f-statistic f-statistic f-statistic f-statistic f-statistic f-statistic
21.91*
34.89*
59.14*
4.67*
8.03*
28.87*
6.48*
12.42*
73.34*
39.51*
37.85*
107.88*
21.39*
44.05*
57.86*
5.55*
36.48*
49.86*
0.84
27.10*
108.17*
-8.32
92.18*
98.98*
-5.79
25.65*
258.76*
-1.27
5.64*
17.21*
-0.71
-0.21
4.78*
0.03
1.32
45.25*
0.63
7.25*
39.23*
1.59
9.55*
14.22*
2.44
4.02*
19.90*
-2.49
6.44
29.41*
-5.24
24.24*
29.60*
13.44*
2.07
81.47*
69.32*
83.46*
119.34*
25.16*
51.10*
95.20*
21.11*
80.78*
110.37*
71.80*
74.64*
96.55*
35.39*
82.74*
98.65*
29.10*
109.43*
103.38*
65.33*
72.09*
69.69*
48.33*
79.91*
45.80*
20.00*
77.86*
134.57*
-9.88
16.99*
92.84*
-3.19
7.80*
16.85*
-5.00
-10.27
11.15*
-8.71
-3.90
12.21*
-9.71
7.34*
36.39*
-1.82
4.21*
30.52*
-15.51
35.40*
26.46*
-10.31
22.92*
36.11*
54.37*
19.86*
88.06*
For the crisis period (July 2007- April 2009), SMB
has 20 positive coefficients. HML has 17 (8) positive
(negative) coefficients. For Model 2, SMB has 19 positive
and 1 negative coefficients, HML has 17 positive and
8 negative coefficients, and LEV has 9 positive and 5
negative coefficients. As noted earlier, there is a marked
improvement in the regression R2 when LEV is added.
76
40.84*
-0.38
53.73*
6.05*
0.02
-0.50
13.93*
2.37
0.00
21.96*
5.45*
0.02
6.82*
-0.35
10.23*
14.17*
0.32
2.30
40.79*
6.89*
2.18
2.29
5.23*
4.99*
108.31*
0.54
47.13*
20.16*
45.00*
51.58*
24.07*
20.48*
43.11*
17.95*
36.12*
102.05*
38.52*
34.93*
47.64*
22.37*
35.52*
57.74*
15.63*
38.63*
74.60*
45.68*
44.07*
48.00*
10.15*
39.53*
72.64*
3.51
43.08*
92.43*
43.73*
0.86
1.26
29.93*
-1.26
-3.29
39.16*
3.40
-16.20
17.58*
2.31
-3.58
16.78*
4.54*
-8.10
16.09*
4.81*
-6.02
45.41*
3.41
-7.28
22.20*
0.15
-1.92
53.43*
-2.53
3.03
24.27*
0.74
37.32*
18.52*
0.02
20.11*
22.08*
0.46
16.59*
8.50*
0.13
13.58*
1.64
0.87
7.48*
-0.36
1.16
15.96*
39.12*
-2.41
16.48*
6.17*
0.56
0.55
71.05*
1.10
53.30*
2.56
-2.26
1.22
0.12
-1.37
-2.12
4.96*
-1.58
6.18*
0.08
-0.48
0.42
2.30
-2.03
-0.94
1.87
-0.16
4.00*
-0.74
-1.16
-0.76
-0.59
-0.66
1.66
2.78
-0.97
-0.37
Finally, we estimated the partial F-statistics to measure
the marginal significance of LEV in the model. Similar to
the results for global stocks, we find that LEV contributes
to improving the overall significance of the model.
In all three periods (aggregate, non-crisis and turbulent),
the partial F-statistics is significant in majority of the
cases.
Islamic banking and finance – Essays on corporate finance, efficiency and product development
A comparison among Islamic, conventional, and socially responsible stocks
5. Conclusions
Fama and French (1993) note that the traditional FF factors
SMB and HML are good proxies for the underlying distress
risk of the firms. As of today, a comparison of how well SMB
and HML explain stock returns across good times and bad is
missing from the literature. In particular, an investigation
into whether the economy wide leverage factor replicates
the underlying economic fundamentals and contributes to
systematic risk especially during bad times is still abstruse;
and we attempt to unravel this puzzle in the present study.
Our hypothesis is that, compared to conventional stocks
with high leverage, we would expect SC stocks to have lower
sensitivity to risk factors, as well as lower risk premium.
This finding would significantly reaffirm the notion that
excessive leverage and engaging in economic activities that
are not consistent with the principles of Islamic transactions
can destroy economic and social values, especially during
falling market environment.
Using weekly data on stock returns for 3704 firms, we test
for the significance of the factors constructed on the basis
of size, book to market equity, and leverage. We find that
the significance of the market factor is drastically reduced
during the recent crisis while the explanatory power of the
Fama-French factors, SMB and HML is reduced considerably.
In contrast, leverage risk factor performs considerably well
across all there periods, especially well during the financial
crisis, in capturing systemic risk in the economy. Its addition
to the model is directly correlated with the reduction of
the economic and statistical significance of the traditional
Fama-French factors.
The main result of this paper is that the effects of leverage
risk are robust to heterogeneity of the firms in the sample.
To show that, we perform cross-sectional regressions across
three distinct categories of stocks i.e. Conventional, Islamic,
and SRI stocks. First, as indicated in the earlier section, excess
market returns play a leading role in explaining the cross
section of expected returns prior to the crisis period, but the
effects of the market factor consistently phased out across
all the three categories of stocks during the crisis period.
The effects of the leverage factor are consistently significant
(except in the case of the socially responsible investing stocks)
throughout; however leverage factor gains momentum
during the crisis period and has a significant effect on the
cross-section of expected returns on stocks and portfolios.
The sensitivities of stock returns to the Fama-French factors
are lower after the introduction of the leverage factor.
In a nutshell, the contribution of leverage risk to asset
pricing has been quite strong. The results indicate that
leverage based risk factor can explain a substantial portion
of the cross-section of stock returns across financial and
non-financial stocks, as well as, various categories of stocks
including conventional, Islamic, and SRI stocks. These
results have powerful implications for asset management
using various types of stocks and also during periods of
great uncertainties.
Notes
1.Sharia compliant stocks are household names in
mostly developed countries. Surprisingly, only few
stocks with enough liquidity and strong balance sheet
Eds. Hatem A. El-Karanshawy et al.
data from the emerging and Muslim countries are
included in the Dow Jones Islamic Index.
2.The leverage measure which we are using is the
market value of debt to market value of assets and not
book value of debt to market value of assets. Both debt
to equity and debt to assets are measures of capital
structure of a company reflecting the amount of fixed
liabilities. The only difference being that debt to equity
ratio is more specific to the overall capital used in the
company while debt to assets ratio is a much broader
measure.
3. The authors used weekly data to examine the relative
performance of investing in three different types of
stocks –conventional, Islamic, and SRI stocks. Both
in sample (Jan 2000-June 2007) and out of sample
(July 2007-April 2009) mean-variance optimization
indicated a portfolio with Islamic stocks generated
significantly larger Sharp ratios. The authors claim
that a low credit market exposure of Islamic stocks
was largely responsible for the relative superior
performance. The results are robust even when
financial and real estate companies are removed from
the sample.
4. A combination of these two leverage factors produces
the book to market ratio. See Fama and French (1992).
5.The crisis had a major impact in September and
October 2008 when there was a huge withdrawal
of $144.5 billion from the money market. Major
institutions like Lehman Brothers, Bear Stearns,
Merrill Lynch, Fannie Mae, Freddie Mac and AIG had
to bear the brunt of high debt market exposure.
6.Accounting leverage is defined as assets/(assetsliabilities).
7. Source: http://neutralobserver.blogspot.com/2008/
11/understanding-financial-crisis-leverage.html
8. This part of the discussion has been adapted from
“Leverage 101: The Real Cause of Financial Crisis”,
Sept. 25, 2008, extracted from http://seekingalpha.
com/article/97299-leverage-101-the-real-cause-ofthe-financial-crisis
9.Source:
http://www.huffingtonpost.com/niallferguson/beyond-the-age-of-leverag_b_163872.html
10. Source: http://sg.biz.yahoo.com/080625/67/4hbq2.
html
11. Source: http://sg.biz.yahoo.com/080625/67/4hbq2.
html
12.Source:
http://www.financialweek.com/article/
20080624/REG/705337846/-1/FWDAILYALERT01
13.As stated in a research note by James Lee, vice
chairman of J.P.Morgan Chase – extracted from
http://www.financialweek.com/article/20080624/
REG/705337846/-1/FWDAILYALERT01
14.As stated by Bob Chapman, “Upsurge of Global
Leveraged Speculation: The Financial Crisis is
notover”, Global Research, November 6, 2009 –
extracted from http://www.globalresearch.ca/index.
php?context = va&aid = 15959
15.This is consistent with the portfolio formation
procedure as suggested in Fama and French (1993).
However, for the purpose of firm specific analysis, we
consider all stocks.
16. To avoid complications, we restrict the 3-month T-bill
return to zero for the months of December 2008 and
January, 2009 when intraday return on T-bills was
often negative.
77
Bhatt and Sultan
17.Source:
http://www.lombardodier.com/annexes/
23056/23074/Investment_Strategy_Bulletin_06.10.09.
pdf, “Is de-leveraging an obstacle to recovery?”
18.http://www.wikinvest.com/concept/2007_Credit_
Crunch
19. We split the periods to specifically test the impact of
LEV factor during the non-crisis and the crisis period.
Given the fact the overleveraging leads to increased
risk exposure in the economy, we believe that this
part of the systematic risk was not captured fully by
the traditional FF factors. This leads us to conjecture
that LEV factor has more direct implications for the
performance of the stocks during the recent credit
crisis and hence we expect the LEV factor to exhibit
stronger effects during this period.
20. See Table 5.
21. According to the study, managers choose leverage on
the basis of the private information about the future
growth prospects and hence, the financial health of
the firm.
22. Furthermore, a firm with low leverage (having low
Tobin’s q and insignificant growth opportunities)
are harder hit during distress periods as compared to
firms with higher leverage ratios (with major growth
prospects and positive NPV projects). This also
explains a negative relationship between credit spread
and firm’s leverage.
23.Source: “How leverage can increase a company’s
return on equity”, Putnam Spectral Funds, extracted
from: http://www.putnam.com/spectrum/return-onequity.htm
24.However the risk substantially increases with the
excessive use of debt since the firm is under a pressure
to service its debt on a regular basis. In addition,
during economic distress, the assumption that debt is
available at a lower cost may not hold true. The recent
credit crisis of 2007 presents plentiful evidence where
debt became costly. In fact, a firm which undertakes
risky projects may not enjoy the low cost of debt
because the riskiness of its operations may require the
debt holders to be paid a higher interest.
25. It is believed that financial firms exhibit different
characteristics as compared to non-financial firms
and hence show different sensitivities to the risk
factors. For instance, high leverage for a financial
firm has different implication as compared to a nonfinancial firm with high debt levels. This further
rationalizes the idea of conducting a robustness
check by separating out financial firms from the
sample.
26. Firm-specific regressions are not reported to conserve
space. They are available on request.
27.We thank Dow Jones for providing us with the
proprietary list of stocks classified as conventional,
Islamic and SRI stocks.
28 Specifically, the debt ratio (short-term plus long-term
debt as a percent of market capitalization) must not
exceed 33%, interest income should not represent
more than 5% of total revenue, the ratio of accounts
receivables to total assets does not exceed 45%, and
the ratio of cash and interest bearing securities to
market capitalization does not exceed 33%. See Dow
Jones website for more.
78
References
Abel, AB. 1988. Stock prices under time-varying dividend
risk: An exact solution in an infinite-horizon general
equilibrium model. Journal of Monetary Economics.
22:375–393.
Abel, AB. 1999. Risk premia and term premia in general
equilibrium. Journal of Monetary Economics. 43:3–33.
Arditti, FD. 1967. Risk and the required return on equity.
Journal of Finance, 22:19–36.
Armitrage, S. 2005. The cost of capital: Intermediate
theory. Cambridge University Press.
Bhandari, LC. 1988. Debt/Equity ratio and expected
common stock returns: Empirical evidence. Journal of
Finance. 43:507–528.
Black, F. 1972. Capital market equilibrium with restricted
borrowing. Journal of Business. 45:444–455.
Black, F. 1990. Mean reversion and consumption smoothing.
Review of Financial Studies. 3:107–114.
Bris, A, Koskinen, Y. 2002. Corporate leverage and currency
crises. Journal of Financial Economics. 632, 275–310.
Chan, KC, Chen, N. 1991. Structural and return
characteristics of small and large firms. Journal of
Finance. 46, 1467–1484.
Chou, PH, Ko KC, Lin S. 2010. Do relative leverage and
relative distress really explain size and book-to-market
anomalies? Journal of Financial Markets. 13, 77–100.
Cochrane, J. 2001. Asset Pricing. University Press. New
Jersey.
Connor, G, Korajczyk, R. 1989. An intertemporal
equilibrium beta pricing model. Review of Financial
Studies. 2:373–392.
Daniel, K, Titman S. 1997. Evidence on the characteristics
of cross-sectional variation in stock returns. Journal of
Finance. 52, 1–33.
Davis, JL, Fama, E, French K. 2000. Characteristics,
covariances, and average returns: 1929 to 1997. Journal
of Finance. 55:389–406.
Dempsey, M. 2009. The Fama and French three-factor
model and leverage: compatibility with Modigliani
and Miller propositions. Investment Management and
Financial Innovations. 6:48–53.
Dimitrov, V, Jain, P. 2006. The value relevance of changes
in financial leverage. Working Paper. http://ssrn.com/
abstract = 708281.
Dow Jones & Company. (2009, February 27). Dow Jones
Indexes. Retrieved August 9, 2009, from Dow Jones &
Company: http://www.djindexes.com/mdsidx/index.
cfm?event = showIslamicStats#fund.
Dow Jones & Company. (2009, February 28). Dow Jones
Sustainability Index World Components. SAM Indexes.
Retrieved September 8, 2009, from Dow Jones &
Company: http://www.sustainability-index.com/djsi_
protected/djsi_world/components/SAM_DJSIWorld_
Components.pdf.
Islamic banking and finance – Essays on corporate finance, efficiency and product development
A comparison among Islamic, conventional, and socially responsible stocks
Dow Jones & Company. (2009, March 2). Guide to the Dow
Jones Islamic Market Indexes. Dow Jones & Company.
Dow Jones Sustainability World Index. SAM Indexes.
(2009, February 1). Dow Jones Sustainability World
Index. SAM Indexes. Retrieved September 8, 2009, from
Dow Jones & Company: http://www.sustainabilityindex.com/djsi_pdf/publications/Factsheets/SAM_
IndexesMonthly_DJSIWorld.pdf.
Elton, EJ, Gruber, MJ, Agrawal, D, Mann, C. 2001.
Explaining the rate spread on corporate bonds. Journal
of Finance. 247–277.
Eom, KS, Park JH. 2008. Evidence on the three-factor and
characteristics models: Korea, SSRN: http://ssrn.com/
abstract = 1329664.
Fama, E, French, K. 1993. Common risk factors in the
returns on stocks and bonds. Journal of Financial
Economics. 33:3–56.
Fama, E, French, K. 1992. The cross-section of expected
stock returns. Journal of Finance. 47:427–465.
Fama, E, French, K. 1995. Size and book-to-market factors
in earnings and returns. Journal of Finance. 50:131–155.
Fama, E, French K. 1996. Multifactor explanations of asset
pricing anomalies. Journal of Finance. 51: 55–84.
Fama, EF, French K. 1998. Value versus growth: The
international evidence. Journal of Finance. 53:1975–1999.
Fama, EF. 1981. Stock returns, real activity, inflation and
money. American Economic Review. 71, 545–565.
Fama, EF, Jensen M. 1998. Separation of ownership and
control. Michael C, Jensen Foundations of Organizational
Strategy, Harvard University Press.
Fama, E, Schwert, WG. 1977. Asset returns and inflation.
Journal of Financial Economics. 5:115–146.
Ferguson, M, Shockley R. 2003. Equilibrium anomalies.
The Journal of Finance. 58(6):2549–2580.
Gomes, JF, Schmid L. 2009. Levered returns. Forthcoming.
Journal of Finance.
Jensen, M, Meckling WH. 1976. Theory of the firm:
Managerial behavior, agency costs and ownership
structure. Journal of Financial Economics. 3:305–360.
Jensen, M. 1986. The agency costs of free cash flow:
Corporate finance and takeovers. American Economic
Review. 76:323–329.
Johnson, T, 2004. Forecast dispersion and the cross
section of expected returns. Journal of Finance.
59:1957–1978.
Eds. Hatem A. El-Karanshawy et al.
Kandel, S, Stambaugh, RF. 1990. Expectations and
volatility of consumption and asset returns. Review of
Financial Studies. 3:207–232.
Lakonishok, J, Shleifer AR, Vishny. 1994. Contrarian
investment, extrapolation, and risk. Journal of Finance.
49:1541–1578.
Lally, M. 2004. The Fama and French model, leverage, and
the Modigliani-Miller propositions. Journal of Financial
Research. 27:341–349.
Lang, L, Ofek, E, Stulz, R. 1996. Leverage, Investment and
firm growth. Journal of Financial Economics. 40:3–29.
Liew, J, Vassalou, M. 2000. Can book-to-market, size
and momentum be risk factors that predict economic
growth. Journal of Financial Economics. 57:221–245.
Luoma, GA, Spiller, EA, Jr. 2002. Financial accounting
return on investment and financial leverage. Journal of
Accounting Education. 202:131–138.
Milly, M, Sultan, J. 2009. Portfolio diversification during
financial crisis: An analysis of Islamic asset allocation
strategy. Bentley University Working Paper.
Modigliani, F, Miller, M. 1958. The cost of capital,
corporation finance and the theory of investment.
American Economic Review. 48:261–297.
Penman, SS, Richardson, Irem, T. 2007. The book-to-price
effect in stock returns: Accounting for Leverage. Journal
of Accounting Research. 45:427–467.
Petkova, R. 2006. Do the Fama-French factors proxy for
innovations in predictive variables? Journal of Finance.
61:581–612.
Report on Socially Responsible Investing Trends in the
United States. 10 Year Review. Washington D.C.: Social
Investment Forum. 2004.
Report on Socially Responsible Investing Trends in the United
States. Washington D.C.: Social Investment Forum.
2008.
Ron, YWH, Strange, R, Jenifer, P. 2008. Corporate financial
leverage and asset pricing in the Hong Kong market.
International Business Review. 17(1):1–7.
Sivaprasad, S, Muradoglu, YG. 2009. An Empirical test on
leverage and stock returns. University of Westminster
Working Paper.
Vassalou, M, Yuhang X. 2004. Default risk in equity returns.
Journal of Finance. April: 831–868.
Vassalou, M. 2003. News related to future GDP growth
as a risk factor in equity returns. Journal of Financial
Economics. 68:47–73.
79
Is Shariah-compliant investment universally
sustainable? A comparative study
Mehdi Sadeghi
Department of Applied Finance and Actuarial Studies, Macquarie University, Sydney-Australia,
T: (61) 2–98508527, E: [email protected]
Abstract - The current paper reports the outcome of investigating the sustainability and efficiency
of Shariah–compliant investment from the global and cross-country perspectives. Our findings,
thus far, suggest that global Shariah compliant sustainable shares performed slightly better than
global sustainable shares in general, during 2006-2011, and Shariah compliant shares performed
substantially better than global market during the same period. The superior performances of
Islamic market indexes suggest that Shariah compliant investment is more resilient and sustainable
compared to their counterparts in the long term. Further evidence from our cross country study
suggests that, Shariah compliant investments perform better than the market on whole in Muslim
countries, and worse than the market in predominately non Muslim countries. These findings have
important implications for investors, regulators, customers, and Islamic financial institutions.
Keywords: Shariah-compliant sustainable investment, Shariah-compliant investment, index addition
and deletion, event study, abnormal returns, liquidity changes
JEL Classification: G14, G15
1. Introduction
The market for Islamic financial services is growing at
an impressive rate, reaffirming its position as one of
the most dynamic sectors in international finance. The
Islamic ­
­
finance industry enjoyed a compound annual
growth rate for 2006–2009 of 28%1. The current value of
Shariah compliant assets managed worldwide, according
to the International Monetary Fund (IMF) estimates, now
tops USD 1 trillion. The value of these assets is forecasted
to hit US$1.6 trillion by 2013. This growth represents
a major achievement, as well as new challenges for
investors, regulators, customers, and also Islamic financial
institutions ­themselves.
The biggest share of Islamic financial belongs to the
Islamic banks. The S&P2 report indicate that the assets of
top 500 Islamic banks in 2008 was $639bn, and grew by
28.6 percent to $822 bn in 2009. There are also Shariahcompliant investment funds within Islamic financial system
that cover a wide range of sectors including real estate,
equities, infrastructure, and energy. According to Lipper
data for 2010, 586 Islamic funds were in operation with
$37bn of assets under management, with a bias towards
equity funds (303), mixed asset (101), money markets
(77), and sukuk funds (77).
With the recent troubles in the global economy, finance
industry has been looking at Islamic contracts as the
possible means of preventing such meltdowns from ever
materializing again. Another area that has received more
in-depth media coverage is the field of sustainability.
Recent changes in the world of investment have made asset
owners and managers increasingly aware of the potential
risk and value impact of environmental, social, and
governance (ESG) factors, on an investment profile. There
are arguments in financial literature in favour of both areas
as safer approaches, and less vulnerable to questionable
financial transactions, which may have led to the global
recession beginning in 2008. These arguments have been
substantiated by some empirical findings that suggest some
Islamic financial institutions and companies focused on
sustainability have been more resilient to financial crisis.
For instance, Hasan and Dridi (2010), report that Islamic
banks have been more resilient than conventional banks
during recent global financial crises. This view was also
corroborated by external rating agencies’ reassessment
of Islamic banks’ risk, which was generally found to be
Cite this chapter as: Sadeghi M (2015). Is Shariah-compliant investment universally sustainable? A comparative study.
In H A El-Karanshawy et al. (Eds.), Islamic banking and finance – Essays on corporate finance, efficiency and product
development. Doha, Qatar: Bloomsbury Qatar Foundation
Developing Inclusive and Sustainable Economic and Financial Systems
Sadeghi
more favourable than—or similar to—that of conventional
banks (with the exception of UAE) (ibid). Some studies
also suggest that companies with a strong commitment to
sustainability have outperformed their industry averages
by 17%3.
But are Islamic finance and sustainability finance compa­
tible? What’s really involved in incorporating sustainability
criteria and Islamic principles into investment decisions?
Can they make a material difference to investment
performance? We start answering these questions by
highlighting similarities and differences between these
two. Islamic finance and socially responsible investing
(SRI) approaches have a lot in common with respect to
the screening process, and criteria used for stock selection.
Sustainability, on the other hand, goes above and beyond
SRI by considering positive screens, promoting investment
in companies with best practices. According to World
Economic Forum Report (2011) “Sustainable investing
is an investment approach that integrates long-term
environmental, social and governance (ESG) criteria
into investment and ownership decision-making with the
objective of generating superior risk-adjusted financial
returns”4. As the financial crisis receded into a period
of uncertainty in the past two years, recognition that
sustainability, corporate governance and transparency are
important factors in portfolio management has emerged.
This is a fundamental shift away from the ideological
and political corner of SRI to the real performance of
sustainability.
Some researches assert that Islamic finance holistic and
dynamic perception of SRI is more effective in taking into
consideration the reality and ever-changing circumstances
of societies in contrast to Western humanistic theories.
They conclude that corporations operation on a piety-based
business paradigm acknowledge their social responsibility
to their workers, managers, other corporations, customers,
and society as a whole more significantly (Dusuki and
Abdullah, 2007). However, regardless of their similarities,
and theoretical arguments in support of one or another,
sustainability and Shariah-compliant investments are
assessed on the basis of long-term trends in yield, profitability,
and efficiency in use of limited financial resources.
In January 2006, Dow Jones Indexes launched the world’s
first Dow Jones Islamic Market Sustainability Index. This
index merges Islamic investing principles and sustaina­
bility criteria by combining the methodology of Dow Jones
Islamic Market Indexes5 and Dow Jones Sustainability
Indexes. To be included in the index, companies must
be components of both the Dow Jones Islamic Market
Index and the Dow Jones Sustainability World Index.
Linking Shariah compliant investment performance to
sustainability is, perhaps, the most effective way to highlight
the importance of ESG governing factors to Islamic finance.
The time series data provided by Dow Jones Indexes is an
invaluable resource to help us investigate whether Islamic
finance is a sustainable practice in the long term.
Current paper is a progress reports on our ongoing long term
research objective of testing the efficiency and sustainability
of Shariah compliant investment opportunities around
the world. We have used time series data on Dow Jones
82
Islamic Market Index and Dow Jones Islamic Market
Sustainability Indexes and their constituents to see if there
is any significant difference between the performances of
these indexes with Dow Jones Global Stock Market Index.
We also investigate whether there is any significant change
in the efficiency and liquidity of market following Islamic
index addition and deletion events.
This study is important for several reasons. First, although
Shariah-compliant investment is similar to SRI, an area
that has already attracted a great deal of research interest,
certain differences is evident in the screening procedures
that make Shariah-compliant investment different. For
instance, some Islamic funds do not exclude weapons
manufacturers but they do exclude conventional banks,
while SRI funds normally exclude weapon manufacturing
firms and do not exclude banks. As another difference,
concerns about environmental issues are not as important
in screening Shariah-compliant companies as they are for
SRI funds. Furthermore, Shariah-compliant companies are
subject to certain financial ratio tests that are not relevant
to conventional SRI companies6.
Second, Miller-Modigliani capital structure theory
contemplates that in an imperfect capital market with
corporate taxes, companies can increase their assets’ value
by increasing their leverage. Given that Shariah-compliant
companies are constrained by their level of borrowing, it
would be interesting to investigate how this constraint can
affect their value.
Third, finance theory based on the efficient market
hypothesis (EMH) considers shares with identical risk and
return as perfect substitutes for each other. This makes
market demand for securities elastic and horizontal. Since
Shariah-compliant equities are not a perfect substitute
for the conventional equities, their demand may not be
horizontal. This can bring about a different outcome to the
study of a Shariah-compliant index revision.
Fourth, Islamic screening criteria reduce the number of
available shares to invest. It is claimed by critics that the
reduction of the investment universe through screening will
reduce the performance. Similar counterarguments have
been raised regarding sustainability criteria (Freidman,
1996). It would be interesting to investigate how this
constraint can affect Shriah-compliant portfolios.
Finally, academic research on the performance of Shariahcompliant investments is rare, and to the best of our
knowledge, no similar study on the impacts of the Shariahcompliant index revisions has been conducted before.
Our results, thus far suggest that global Shariah compliant
sustainable shares perform worse than global Shariah
compliant shares in the long term. However, they both
perform better than global stock market as a whole.
Further evidence from individual countries suggests that,
Shariah compliant investments perform better than the
market in Muslim countries, and worse than the market in
predominately non Muslim world. The rest of this study is
organized as follows: Section II is allocated to a short review
of research background. We outline our methodology,
data and hypothesis development in Section III. Empirical
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Is Shariah-compliant investment universally sustainable? A comparative study
findings are discussed in Section IV. Section V articulates
our conclusions, and describes the limits of our study.
2. Research background and literature
review
Research background
We started our study with an investigation of the market
performance and liquidity of Shariah-compliant Index
(SI) portfolio following its introduction by Bursa Malaysia.
Malaysia has one of the largest Islamic fund markets in the
world. It had 155 unit trusts and mutual funds at the end
of June 2010 with a total volume of about RM22.69 billion.
Our findings show that, overall, introduction of SI had
a positive impact on the financial performance and the
liquidity of included shares in this country7.
As time series data on Shariah compliant indexes become
more readily available for other parts of Muslim world
through index providers, such as Dow Jones Islamic
Market Index8, we decided to extend our study to the
MENA (Middle East and North Africa) market in the second
stage of our study. MENA region is another important
hub in Islamic finance, with large market and appropriate
financial infrastructure. Constrained by the availability of
times series data, we used event study methodology and
the improved models of liquidity measures, first to index
addition to equity markets in Qatar, Kuwait, Oman, and
UAE. Our findings showed an even stronger result than for
Malaysia in that market reacts positively to the introduction
of Shariah compliant shares in these countries. This was
reflected in short and long-term market performance
and the improvement in the liquidity of shares9. One of
the limitations of recent study was the small number of
companies in our sample. To test the robustness of findings
with larger samples, we extended our investigation to
Jordan and Egypt. Our results overwhelmingly supported
the robustness of our earlier findings of countries in the
Gulf region10.
Overall, our research on seven markets in Islamic countries
in showed that investors’ reaction to the introduction of
Shariah-compliant shares is positive. This is reflected in
improvement in the share price and market liquidity up to
150 days following the index addition. The positive outcome
for six countries in MENA region is especially important
because they were found from the data that became
available by Dow Jones Indexes immediately following the
start of financial crisis, suggesting that Shariah compliant
investments in Islamic countries has been more resilient to
financial crisis than conventional investments.
In addition to Muslim countries, Islamic finance is practiced
outside the Muslim world without ties to any particular
jurisdiction. Shariah compliant investments are defined
according to certain norms and conditions that can be
applied anywhere in the world where there is a market and
people who wish to engage in financing transactions in a
manner which is consistent with Shariah law. This progress
is specially facilitated by a form of reverse financial
engineering that reconstructs conventional financial
products into Shariah compliant instruments. This
innovation has significantly increased Muslims investments
Eds. Hatem A. El-Karanshawy et al.
in Shariah compliant companies in non-Muslim countries
around the world11. In the case of equities, the differences
between Shariah compliant shares and their conventional
forms are even less significant, only requires screenings.
This screening process is similar to the screening of Socially
Responsible Investing (SRI) instruments.
In the third stage of our research we decided to investigate
Shariah-compliant index addition and deletion to
predominantly non-Muslim countries, starting with
Australia as the first sample. Australia’s skills in complex
financial engineering and experience in infrastructure,
resources, property and agriculture provide her with
a unique opportunity to develop Shariah-compliant
investments. This country also has easy access to rapidly
growing Islamic financial markets with over a billion in
population to accommodate their demand12.
A through presentation of our findings on all eight countries
studied so far is too long to report here. In order to show the
contrasting nature of market reaction to Index addition and
deletion events in predominately Muslim and non-Muslim
countries, we report the report the results on two sample
countries of Egypt and Australia in section IV.
Literature review
From a theoretical perspective, there are two explanations
for the effects of stock additions to an index: demand-based
and information-based. The demand-based explanation
sees index changes as information-free events. For example,
Shleifer (1986), by employing the downward-sloping
demand curve hypothesis, showed that the price effects
following index changes are due to the demand from index
tracking. These effects can be temporary or permanent.
The temporary effect is explained by the price pressure
hypothesis, predicting a reversal of initial price increases
in the long run (Harris and Gurel, 1986). The permanent
effect is explained by the imperfect-substitute hypothesis,
which assumes that there would be no price reversal, as the
new price reflects changes in the distribution of security
holdings in equilibrium13.
Information-based explanations include the information
hypothesis and the liquidity hypothesis. Unlike the demandbased explanations, information-based explanations as­
sume that index changes are not information-free events.
Some studies, such those by Dhillon and Johnson (1991)
and Jain (1987), support the information hypothesis: they
showed that the addition of a stock to the index conveys
favorable news about the firm’s prospects and a permanent
price increase can result following this event. Amihud
and Mendelson (1986), Beneish and Whaley (1996), and
Hegde and McDermott (2003) contended that the price
reactions can be explained by changes in market liquidity.
According to the liquidity hypothesis, the price increase
at index inclusion is caused by the increased liquidity due
to the greater visibility of the shares, greater interest from
institutional investors, higher trading volume, and lower
bid-ask spreads. Amihud and Mendelson (1986) suggested
that the increase in stock liquidity is positively related to
the firm’s value through a reduction in the cost of capital.
Previous studies, such as Harris and Gurel (1986), and
83
Sadeghi
Hegde and McDermott (2003) reported liquidity increases
following index additions.
The topic of Shariah-compliant index revision is important
from two perspectives. First, the nature of companies’
activities and their capital structure makes them Shariah
compatible in the first place. Second, changes in investors’
demand result in subsequent market price reactions,
according to our earlier discussion. For example, reduction
in the level of debt in the capital structure can make a
company Shariah-compliant, bringing about an increase
in the demand from Muslims and higher share prices if
demand is not fully elastic. At the same time, the lower
level of debt may move the capital structure of the company
to a suboptimal level, at a higher cost of capital than in
equilibrium. This may send negative signals to the market
when shares are added to a Shariah-compliant index.
As a result, it is possible that the interaction of opposing
market forces on index revision will bring about different
outcomes compared with the effects of conventional index
additions. Therefore, it is not possible to predict clearly
how the performance and liquidity of shares included in
or excluded from the DJIM index will change, as it largely
depends on how the net effects of the influential factors are
revealed through our empirical investigation.
3. Data and methodology
To determine the impact of additions to and deletions from
the DJIM index, we applied several measures of both short
and long-term price and liquidity performance. We applied
standard event study methodology to find the initial stock
price reaction of firms when an announcement of an
index change was made. We also applied several liquidity
measures to investigate the magnitude and direction of
liquidity changes following the index revision. Data for this
research has been collected through Dow Jones Indexes
and Bloomberg.
Price effect
Our event-study methodology calculates the abnormal
returns. An abnormal return is the difference between
the realized return observed from the market and the
benchmark return. The return to the market portfolio is
estimated via both ordinary least square (OLS) and Scholes
and William (1977) procedures. The latter method is
usually used when stocks do not trade at the same level of
frequency as the market index and OLS may produce biased
beta estimates. This problem is exacerbated for infrequently
or thinly traded stocks as the sampling interval is reduced14.
The advantages of these models are that they control for the
effect of market movements through the market portfolio,
and also allow for an individual security’s responsiveness
as measured by beta. Return on the All Ordinaries index
was used as a proxy for the market rate of return.
We defined the event date as the day that a stock was added
to or deleted from the DJIM index. For each event, the return
time series data were divided into an estimation period
and an event window. The estimation time series data are
used to calculate the benchmark parameters, and the event
window period is used for computing prediction errors
based on the estimated parameters. Abnormal returns are
represented by the prediction errors. The abnormal returns
84
during the event windows can be interpreted as a measure
of the effect of the event on the value of the firms, which is
reflected in their share price.
Our event window extended from 10 days before to 25 days
after the event. This asymmetric event window was chosen
to examine the extended effect of excess returns in the
post-event period15.
The normal returns of stocks are the expected returns
if there are no events. The normal returns are estimated
over a period of time outside the event window (Peterson,
1989). For applications in which the determinants of the
normal return are expected to change due to the event,
the estimation period can fall on both sides of the event
window. This period commences 125 trading days before
and ends 125 trading days after the event dates, excluding
the event period of day −10 to Day 25. As a result, the
estimation period consists of Day −135 to Day −11 and Day
26 to Day 150. We did not allow the event period to overlap
with the estimation period, to avoid biasing the parameter
estimates in the direction of the event effect.
The following section describes the event study
methodology that we used in our study. MacKinlay (1997),
and Kothari and Warner (2004) have provided a survey
of event study methods, and we follow their papers to
describe the models here.
Liquidity effect
Market liquidity is an elusive concept and difficult to
measure. In this study, we use six proxies to evaluate changes
in market liquidity during post-event periods, compared to
the corresponding control periods. The large number of
tests helps to confirm the robustness of our findings and
reduces the chance of making wrong inferences.
These liquidity proxies include: 1) quoted spread, as the
simple difference between bid and ask prices; 2) percentage
spread, as the quoted spread normalized by the midpoint of
the bid and ask prices; 3) changes in the volume of trade as
the daily average of the transaction size, normalized on the
average volume of trade in the control period; 4) changes
in volatility, measured by the standard deviation of returns;
5) the Amivest liquidity ratio, as the average ratio of share
volume to absolute return over all days with non zero
returns; and 6) the proportion of zero daily returns. Zero
daily return is related to trading speed because the days
with zero return indicate delays or difficulties in executing
an order, interrupting the continuity of trading.
In calculating the percentage bid-ask spread and change
in the volume of trade, we largely follow Hegde and
McDermott (2003). Changes in the volume of trade are
directly related, and changes in the bid-ask spreads and
volatility are inversely related to the market liquidity.
It is important to note that an increase in the volume
accompanied by an increase in volatility can actually
impede market liquidity. The Amivest liquidity ratio is
estimated according to Amihud and Mendelson (2002).
This ratio measures the ability of a share to absorb changes
in trading volume without any significant change in
share price. Change in this variable is directly related to
the liquidity. In estimating the proportion of zero daily
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Is Shariah-compliant investment universally sustainable? A comparative study
45
Jones Islamic Market Sustainability Index out performs Dow
Jones Global Sustainability Index by less than 1% during
this period. However, Dow Jones Islamic Market stock Index
shows a much higher return of 22% compare to Dow Jones
World stock Market Index. The superior performances of
Islamic Market indexes suggest that Shariah compliant
investment is more resilient and sustainable compare to
their counterparts within the family of Dow Jones Indexes
in the long term.
DJIMI
DJWSI
40
35
30
30
25
25
20
20
15
15
10
Cross country findings
1/1/11
3/1/10
0
5/1/09
0
1/1/06
5
1/1/06
9/1/06
5/1/07
1/1/08
9/1/08
5/1/09
1/1/10
9/1/10
5/1/11
5
7/1/08
10
9/1/07
35
DJIMSI
DJGSI
11/1/06
40
Figure 1. The Performance of Dow Jones Islamic Market
Sustainability Index (DJIMSI), Dow Jones Islamic Market
Index (DJIMI), Dow Jones Sustainability Index (DJGSI),
and Dow Jones World Stock Index (DJWSI) during
1/1/2006 to 1/5/2011.
returns, we follow Bekaret et al. (2004), as they found it
a reasonable proxy for a liquidity measure to use in their
study. Change in this variable is inversely related to the
market liquidity.
4. Results
Global findings
Figure 1 presents the Performance of Dow Jones Islamic
Market Sustainability Index (DJIMSI), Dow Jones Islamic
Market Index (DJIMI), Dow Jones Global Sustainability
Index (DJGSI), and Dow Jones World Stock Index (DJWSI)
during 1/1/2006 to 1/5/2011. Our findings show that Dow
A through presentation of our findings for all eight
countries studied so far is too long to report here. The
Results reported here is only from Egypt and Australia, in
order to show the contrasting nature of market reaction to
Index addition and deletion in two predominately Muslim
and non Muslim country.
Price effect
Table 1 presents the estimated CARs for index additions in
the pre- and post-event periods for Egypt. The coefficient
for CARs, accumulated during the period (−10, 0), is
-1.71%. However, it is not statistically significant at the
conventional levels. The CARs coefficient estimated over
the shorter period (-5, 0) increases to 2.41% and becomes
statistically significant at the 0.05 level. CARs coefficients
for Day 0 (the event day) and for (0, 5) increase further
to 2.85% and 3.44%, respectively, and become highly
significant at the 0.01 level. CARs for (0, 15) drop to 2.69%
and remain statistically significant at the 0.01 level. CAR
coefficients increase further to 6.31% during (0, 30) and
remain statistically significant at the 0.01 level.
The prolonged effects of the index additions on CARs
in Table 1 indicate that these events are likely to contain
information, thus sending signals about the features of the
index additions to the market. To test this hypothesis, we
compared the cumulative returns (CRs) for the added firms
with the cumulative return for the market over the period
(-10, 150)16.
Table1. Cumulative abnormal returns and relevant statistics for stock additions to the DJIM
index in Egypt.
This table presents the cumulative abnormal returns (CARs) around the index addition for the 25 Egyptian firms
in our sample. Results are presented for the windows (-10, 0), (-5, 0), (0, 0), (0, +5), (0, 15), and (0, 30), where day 0
represents the addition date. The Generalized Sign Z-test is a test with the null hypothesis that the fraction of positive cumulative returns is the same as in the estimation period. The Positive/Negative column reflects how many
firms had positive cumulative abnormal returns in the window. The symbols $, *, **, and *** denote statistical
significance at the 10%, 5%, 1% and 0.1% levels, respectively, using a 1-tail test. The symbols), >, etc., correspond to
$,* and show the significance and direction of the Generalized Sign-Z test.
Scholes-Williams Market Model
Intervals
MCARs
t-Statistics
(-10, 0)
(-5, 0)
(0, 0)
(0, +5)
(0, +15)
(0, +25)
1.71%
2.41%
2.85%
3.44%
2.69%
6.31%
1.12
1.75*
6.36***
2.63**
1.72*
2.69**
Eds. Hatem A. El-Karanshawy et al.
Generalized
Sign Z-test
Positive/
Negative
1.15
1.55$
4.35***
2.74**
2.35**
1.95*
15/10
16/9)
23/2 >>>
19/6 >>
18/7 >>
17/8 >
85
Sadeghi
4
Market Index
Shares
3.5
3
2.5
2
1.5
1
0.5
0
–0.5 1
10
19
28
37
46
55
64
73
82
91 100 109 118 127 136 145 154
Figure 2. Cumulative firm return and market return around day −10 to Day 150
egyptian stocks addition to DJIM index.
40
35
Market Index
Shares
30
25
20
15
10
5
0
–5 1
10
19
28
37
46
55
64
73
82
91 100 109 118 127 136 145
154
Figure 3. Risk adjusted cumulative firm return and market return around day −10
to day 150 Egyptian stocks addition to DJIM index
Figure 2 illustrates CRs for the portfolio of added stocks,
compared with the market CRs during (−10, 150) for Egypt,
showing the shares’ superior performance of 352% gain,
compared with less than 48% for the market by Day 150.
Figure 3 compares the performance of the same variables
on a risk-adjusted basis, calculated using the Sharpe Ratio.
According to this figure, the Sharpe Ratio for the shares
shows a value of 34 compared with a ratio of 1.4 for the
market.
Table 2 and Table 3 present mean cumulative abnormal
returns (CARs) for the added and the deleted Australian
firms, respectively. To test the robustness of our findings,
we have used both the single-factor and Scholes-Williams
market models as the benchmarks for estimating normal
return. Our results show that the magnitudes of CARs and
the level of their statistical significance from the application
of the two methods are similar. Nevertheless, we report
and discuss the results from the Scholes-Williams model
to avoid non-synchronous trading bias, as a considerable
proportion of shares included in this study are likely to
trade less frequently.
Table 2 presents the estimated CARs for index additions
in the pre- and post-event periods. The coefficient for CARs
86
accumulated during Day −10 to Day 0, is −1.22%, and is
statistically significant at the 0.05 level. When the CARs
coefficient is estimated over the shorter interval of Day
−5 to Day 0, it increases slightly to −1.18% and remains
statistically significant at the 0.05 level. CARs for Day 0
(the event day) and Day 0 to Day 5 increase to −0.32% and
0.21%, respectively. However, they are not significantly
different from zero at the conventional levels.
CARs for the intervals Day 0 to Day 10 and Day 0 to Day
25 decline continuously, dropping to −4.04% and become
highly significant at the 0.01 level. CARs for the entire
window (Day −10 to Day 25) is −4.94% and significant
at the 0.01 level. The temporary upward trend in CARs
around the event day may have been caused by the positive
reactions of some Muslim investors to the additions news.
However, this reaction was perhaps not strong enough to
fully offset a negative response from the market as a whole.
The coefficients for generalised sign tests are consistent
with the coefficients for t-statistics, although they are not
as strongly significant as the later ones. It is mainly the
coefficient for Day 0 to Day 25 and the entire event window
(Day −10 to Day 25) that are statistically significant at the
conventional level, indicating that the significance of our
findings is robust to both parametric and non-parametric
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Is Shariah-compliant investment universally sustainable? A comparative study
Table 2. Mean cumulative abnormal return and relevant statistics for stock additions to the DJIM index in Australia.
This table presents the mean cumulative abnormal returns (CARs) around the index addition for the 117 firms in our sample. Results are presented
for the windows (-10, 0), (-5, 0), (0, 0), (0, +5), (0, +10), (0, 15), (0, 25), and (-10, 25), where day 0 represents the addition date. The third column
is the precision-weighted cumulative mean abnormal return. The generalized sign Z is a test of the null hypothesis that the fraction of positive
cumulative returns is the same as in the estimation period. The symbols $, * and ** denote statistical significance at the 10%, 5%, and 1% levels,
respectively, using a 1-tail test.
Scholes-Williams Market Model
Intervals
(-10, 0)
(-5, 0)
(0, 0)
(0, +10)
(0, +15)
(0, +25)
(-10, +25)
Cumulative
average abnormal
return (CAAR)
Precisionweighted
CAAR
tstatistics
Generalized
sign Z-test
-1.06%
-0.78%
-0.15%
-0.36%
-0.65%
-2.03%
-2.94%
-1.70*
-2.24*
-1.16
-1.24
-1.55$
-2.71**
-3.02**
-0.65
-1.21
-0.47
-1.21
-1.39
-1.58$
-1.76*
-1.22%
-1.18%
-0.32%
-1.11%
-1.62%
-4.04%
-4.94%
Table 3. Mean cumulative abnormal return and relevant statistics for Australian stock deletions from the DJIM index.
This table presents the mean cumulative abnormal returns (CARs) around the index deletion for the 87 firms in our sample. Results
are presented for the windows (-10, 0), (-5, 0), (0, 0), (0, +5), (0, +10), (0, 15), (0, 25), and (-10, 25), where day 0 represents the
addition date. The third column is the precision-weighted cumulative mean abnormal return. The Generalized Sign Z is a test of the
null hypothesis that the fraction of positive cumulative returns is the same as in the estimation period. The symbols $, * and ** denote
statistical significance at the 10%, 5%, and 1% levels, respectively, using a 1-tail test.
Scholes-Williams market model
Intervals
(-10, 0)
(-5, 0)
(0, 0)
(0, +5)
(0, +10)
(0, +15)
(0, +25)
(-10, +25)
Cumulative
average abnormal
return (CAAR)
1.57%
2.11%
0.47%
0.04%
5.34%
6.05%
7.45%
8.55%
tests. Our findings are also consistent with the results of
Clarke and Russell’s (2008) study on Socially Responsible
Investing (SRI): they found significant negative CARs for
DS400 additions that persisted at least 30 days after the
events.
Table 3 presents the estimated CARs for index deletion in
the pre- and post-event periods for Australian shares. The
coefficient for CARs, accumulated during Day −10 to Day 0,
is −1.57% and is marginally significant at the 0.10 level. The
CARs coefficient estimated over the shorter interval of Day
−5 to Day 0, increases to 2.11% and becomes statistically
significant at the 0.05 level. This coefficient for Day 0 (the
event day) is 0.47% and statistically significant at the 0.10
level. CARs for Day 0 to Day 5 is negative and statistically
insignificant at the conventional levels. This coefficient
quickly rises to 5.34% during the interval Day 0 to Day
10, and becomes statistically significant at the 0.05 level.
Eds. Hatem A. El-Karanshawy et al.
Precisionweighted
CAAR
tstatistics
1.36%
1.62%
0.41%
-0.04%
2.24%
2.68%
3.82%
4.77%
1.29$
2.51**
1.35$
-0.05
2.28*
2.33**
2.82**
3.02**
Generalized
sign Z-test
0.11
2.60*
1.19
-0.31
1.40$
1.83*
2.05*
2.06*
CARs increases further to 6.05% during the interval Day
0 to day 15, and to 7.45% during the interval Day 0 to Day
25, respectively. Both coefficients remain highly significant
at the 0.01 level. CARs for the entire window (Day −10
to Day 25) is 8.55% and significant at the 0.01 level. The
temporary downward trend in CARs after the event day
may have been caused by the negative reactions of Muslim
investors to the deletion news. However, this reaction did
not seem to be strong enough to fully offset the positive
response from the market as a whole.
Results in Table 2 and Table 3 show CARs of up to 25 days
after additions and deletion, respectively. Some studies in
the literature, such as one by Nesbitt (1994), suggest that
the value of socially responsible investing may be more
apparent in the long-run. To examine whether DJIM Index
additions and deletions have any prolonged information
effects on shares, we compared the cumulative returns
87
Sadeghi
0
–0.02 1
10 19 28 37 46 55 64 73 82 91 100 109 118 127 136 145 154
–0.04
–0.06
–0.08
–0.1
–0.12
Market Index
Shares
–0.14
–0.16
Figure 4. Cumulative return on a portfolio of added Australian shares, compared with cumulative return on the market
for the 160-day period from Day −10 to Day 150 around the event day.
0
–10 1
10 19 28 37 46 55 64 73 82 91 100 109 118 127 136 145 154
–20
–30
–40
–50
–60
Market Index
Shares
–70
–80
–90
Figure 5. Cumulative risk adjusted return on a portfolio of added Australian shares, compared with risk-adjusted
cumulative return on the market for the 160-day period from Day −10 to Day 150 around the event day. Risk-adjusted
returns are estimated according to the Sharpe performance index.
Table 4. Measures of liquidity changes from pre- to post-stock additions to the DJIM index in Egypt.
This table presents the change of a variety of liquidity measures around the index addition day for an equally weighted portfolio of
25 Egyptian firms in our sample. Results are presented for the windows (1, 25), (1, 50), (1, 100), and (1, 150), compared with the
control periods (-35, -10), (-60, -10), (-110, -10), and (-160, -10), respectively. The bid-ask mean difference represents the difference
between average liquidity measures in each interval compared with the corresponding interval in the control period. The symbols $, *, **,
and *** denote statistical significance at the 10%, 5%, 1%, and 0.1% levels, respectively, using a 1-tail test.
Intervals
Liquidity measures
(1, 25) vs.
(-35, -10)
Standard Deviation (SD)
SD (control period)
SD change
Relative bid-ask spread
Relative bid-ask spread
(control period)
Bid-ask mean difference
Average daily volume
Average daily volume
(control period)
Average daily volume change
Amivest liquidity measure
Amivest liquidity measure
(Control period)
Amivest liquidity measure
change
88
(1, 50) vs.
(-60, -10)
1.05%
1.49%
–0.44%
1.53%
0.80%
0.85%
2.56%
–1.71%***
1.36%
0.71%
0.73%
40.24
25
0.65%
53.97
50
60.95%*
13.85
13.29
7.94%
13.84
13.32
0.56***
0.52***
(1, 100) vs.
(-110, -10)
0.89%
2.05%
–1.16%***
1.35%
0.55%
0.80%
130.73
100
30.73%**
13.68
13.81
–0.13$
(1, 150) vs.
(-160, -10)
0.85%
1.80%
–0.95%***
1.36%
0.45%
0.91%
202.37
150
34.91%***
13.69
13.88
–0.19**
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Is Shariah-compliant investment universally sustainable? A comparative study
Table 5. Measures of liquidity changes from pre- to post- Australian stock additions to the DJIM index.
This table presents the change of a variety of liquidity measures around the index addition for an equally weighted portfolio of 117 firms in
our sample. Results are presented for the intervals in days (1–25), (1–50), (1–100), and (1–150), compared with the control periods (-35
to 10), (-60 to -10), (-110 to -10), and (-160 to -10), respectively. Day 0 represents the addition date. The mean difference represents
the difference between average liquidity measures in each interval compared with the corresponding interval in the control period. The
symbols $, *, **, and *** denote statistical significance at the 10%, 5%, and 1%, and 0.1% levels, respectively, using a 1-tail test.
Intervals Liquidity
measures
Absolute bid-ask spread
Absolute bid ask (control period)
Absolute bid-ask mean difference
Relative bid-ask spread
Relative bid-ask (control period)
Relative bid-ask mean difference
Average volatility (SD)
Average volatility (SD)
(control period)
Average volatility ratio
Average daily volume
Average daily volume
(control period)
Average volume difference
Zero-return
Zero-return (control period)
Zero-return mean difference
Amivest liquidity measure
Amivest liquidity measure
(Control period)
Amivest liquidity measure
mean difference
Day 1 to 25
(-35 to -10)
Day 1 to 100
(-110 to -10)
Day 1 to 150
(-160 to -10)
6.02¢
5.61¢
0.41¢*
0.43%
0.47%
–0.04%
3.60%
3.45%
6.17¢
5.63¢
0.54¢**
0.45%
0.46%
–0.01%
3.51%
3.15%
6.24¢
6.14¢
0.10¢
0.49%
0.47%
0.02%$
3.61%
2.87%
0.15%
27.10
25.00
0.36%**
54.15
50.00
0.74%***
106.35
100.00
0.75%***
157.56
150.00
8.40%
12.12%
8%
4.12%***
12.44
12.53
8.30%$
11.33%
7.84%
3.49***
12.38
12.50
6.35%*
12.79%
12.77%
0.02%
12.40
12.46
5.04%*
13.15%
12.64%
0.51%
12.33
12.44
–0.09***
–0.12***
–0.06***
–0.11***
(CRs) for the added and deleted firms with cumulative
return for the market over the period from Day −10 to
Day 15017.
Figure 4 and Figure 5 provide long-term evidence of
negative market reaction to the index addition. Figure 4
illustrates CRs for the portfolio of added stocks, compared
with the market CRs during Day −10 to Day 150, showing
the market’s superior performance of −7.8% compared
with −13.1% for the shares by Day 150. Figure 5 compares
the performance of the same variables on a risk-adjusted
basis, calculated according to the Sharpe performance
index (SPI). According to this figure, SPI for the market
shows a figure of −42.9% compared with the SPI of −70.8%
for the shares.
Table 4 provides evidence of changes in liquidity
measures for Egypt. The results show a decline in the
standard deviation of returns between 0.95% and 1.71%,
accompanied by an increase in the volume of trade from
30.73% to 60.95%. Amivest liquidity measure changes also
suggest an increase in the market liquidity over the short to
medium term and a decline over the medium to long term.
The coefficients for changes in the bid-ask spread is positive;
however, they are not statistically significant. Overall, there
is more evidence for improvement in the liquidity of the
Egyptian stock market than for decline.
Eds. Hatem A. El-Karanshawy et al.
Day 1 to 50
(-60 to -10)
6.48¢
5.84¢
0.64¢***
0.51%
0.46%
0.05%**
3.66%
2.91%
5. Concluding remarks
Current paper reports the outcome investigating
the sustainability and efficiency of Shariah –compliant
investment from the global and cross-country perspectives.
Our findings, thus far, suggest that Dow Jones Islamic
Market Sustainability Index out performs Dow Jones Global
Sustainability Index by less than 1% during 1/1/2006–
1/5/2011. However, Dow Jones Islamic Market stock Index
shows a much higher return of 22% compare to Dow Jones
World Stock Market Index during the same period. The
superior performances of Islamic Market indexes suggest
that, relative to their counterparts within the family of
Dow Jones Indexes, Shariah compliant investments are
generally more resilient and sustainable in the long term.
In the cross country component of our study, we used data
from eight countries (only one is reported in this paper)
and an event study methodology to estimate cumulative
abnormal returns in the days surrounding index additions
and deletions for testing the price effects of market reaction.
We also used several liquidity measures; including the bid–
ask spread, the Amivest liquidity ratio, standard deviation
of returns, and volume of trade to estimate changes in
the liquidity of the added shares around these events. Our
results show that stock prices respond positively to index
additions for Muslim countries and negatively for non
Muslim countries, both in the short and long terms. Further
89
Sadeghi
evidence from non Muslim countries suggests that stock
market react positively to index deletions.
Observing negative abnormal return for index additions,
and positive abnormal return for index deletions in
Australia suggests that market in this country perceives
these events as a value destroying, and value adding
exercises, respectively. This view is in line with Friedman
(1996) agency theory, perceiving any effort by companies
to go beyond maximising their profit as a burden on their
return. These opposing reactions can also be explained
by differences in both fundamental and socio-cultural
factors in Muslim vs. non Muslim countries. For instance, a
company in the West world can become Shariah compliant
by chance, or by force, not necessarily by choice. A low
debt/equity ratio in the capital structure of companies
in Western countries can make them Shariah-compliant.
However, this may occur, perhaps, due to their inability to
borrow money if they are relatively small. While a company
in a Muslim world may intentionally borrow less to comply
with Shariah- principles.
These findings have important implications for the
development and growth of Islamic finance around the
world. For example, if Western countries plan to promote
themselves as a centre for Islamic finance, they need
to overcome certain impediments to be successful. This
includes reduction in psychological barriers, as well as
revision in taxation laws and non-taxation regularities to
ensure that they do not inhibit the development of Islamic
finance. There is also need for a trained work force in
financial sector (education in Islamic economics, finance,
banking, insurance, accountancy, and business law), and
ability to market Islamic financial products overseas once
they are developed.
Notes
1. HSBC Report, Islamic Banking and Finance Summit,
Reuters’ Offices, Dubai, 2009.
2. S&P Press Release, 1st February 2010.
3. Daniel Mahler, A.T. Kearney, Inc. Report, titled Green
Winners: The Performance of Sustainability-focused
Companies in the Financial Crisis, 2009.
4. Transition Towards Sustainable Investing, World
Economic Forum White paper, 2011.
5. DJIM Indices were introduced in 1999 as the
benchmarks
to
represent
Shariah-compliant
portfolios.
6. Socially responsible fixed-income securities are found
in conventional financial markets, while, at least in
theory, they are banned by Shariah.
7. Refer to Sadeghi (2008) for more details.
8. Companies from Islamic countries were added to Dow
Jones Islamic market Index in 2009 for the first time.
9. Refer to Sadeghi (2010a) for more details.
10. Refer to Sadeghi (2010b) for more details.
11.This is a pragmatic compromise, rather than an ideal
situation from an Islamic perspective.
12.The DJIM index constituents are screened from
around the globe and are mostly located in nonMuslim countries. For instance, from 2403 DJIM index
constituents on 30th November 2009, 2204 originated
in the non-Muslim world, especially in the West.
90
13.Refer to Beneish and Whaley (1996), Lynch and
Mendenhall (1997), Kaul et al. (2000), and Wurgler
and Zhuravskaya (2002) for more details.
14.The frequency of trading declines with the reduction
in the sampling interval.
15. This allowed for slow responses from overseas Muslims
that might cause delays in the market reaction to the
index revision.
16. We believe that if index inclusion contains information,
this information must have been reflected in share
prices earlier than the event day and should extend
for some time afterwards. As a result, we have used
a sample of data that extends from 10 days before to
150 days after the event.
17. We believe that if index inclusion and exclusion contain
information, this information must have been reflected
in share prices earlier than the event day and should
extend for some time afterwards. As a result, we have
used a sample of data that extends from 10 days before
to 150 days after the event.
References
Amihud Y, Mendelson M. 1986. Asset pricing and the bidask spread. Journal of Financial Economics. 17:223–49.
Amihud Y. 2002. Illiquidity and stock returns: Cross-section
and time-series effects. Journal of Financial Markets.
5:31–56.
Asyraf Wajdi Dusuki AW, Abdullah NI. 2007. Maqasid
al-Shari`ah, Maslahah, and Corporate Social
Responsibility. The American Journal of Islamic Social
Sciences. 24(1):25–45.
Becchetti LL, Ciciretti I, Hasan. 2009. Corporate social
responsibility and shareholder’s value: An empirical
analysis. Bank of Finland Research Discussion Papers. 1.
Beneish M, Whaley R. 1996. An anatomy of the “S&P 500
Game: The effects of changing the rules. Journal of
Finance. 51:1909–30.
Boehmer E, Poulsen A. 1991. Event-study methodology
under conditions of event-induced variance. Journal of
Financial Economics. 30:253–72.
Chen HG, Noronha VS. 2004. The price response to
S&P 500 Index additions and deletions: Evidence of
asymmetry and a new explanation. Journal of Finance.
59:1901–29.
Clarke S, Russell W. 2008. Stock price effects associated
with socially responsible index constituent changes.
Paper presented at Midwest Finance Association
Conference. San Antonio, Texas, USA.
Dhillon U, Johnson H. 1991. Changes in the Standard and
Poors 500 list. Journal of Business. 64:75–85.
Friedman M. 1996. The social responsibility of business
is to increase profits. In: Schwartz MS (Ed.). Beyond
Integrity: A Judeo-Christian Approach. Zondervan
Publishing House. Grand Rapids.
Harris L, Gurel E. 1986. Price and volume effects associated
with changes in the S&P 500: New evidence for
the existence of price pressures. Journal of Finance.
41:815–30.
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Is Shariah-compliant investment universally sustainable? A comparative study
Hasan M, Dridi J. 2010. Islamic Banking: Put to the Test.
Finance & Development. 47(4):45–47.
Hegde S, McDermott J. 2003. The liquidity effects of
revisions to the S&P 500 Index: An empirical analysis.
Journal of Financial Markets. 6:413–59.
HSBC Report. 2009. Islamic Banking and Finance Summit,
Reuters’ Offices, Dubai.
Jain P. 1987. The effect on stock price of inclusion in or
exclusion from the S&P 500. Financial Analysts Journal.
43:58–65.
Karpoff J. 1987. The relation between price changes and
trading volume: A survey. Journal of Financial and
Quantitative Analysis. 22:109–26.
Kothari S, Warner J. 2004. Econometrics of event studies,
Working Paper. (Tuck School of Business at Dartmouth,
Hanover, USA).
Merton R. 1987. A simple model of capital market
equilibrium with incomplete information. Journal of
Finance. 42:483–510.
Price Waterhouse Coopers International Limited Report.
2009. Shariah-compliant funds: A whole new world of
investment.
Sadeghi M. 2011a. Investment opportunities and stock
liquidity: evidence from DJIM index additions in the
Gulf States. Investment Management and Financial
Innovations. 8(1):50–59.
Sadeghi M. 2011b. Shariah-compliant Investment and
Shareholders’ Value: An Empirical Investigation. Global
Economy and Finance Journal. 4(1):44–61.
Sadeghi M. 2012. Are Faithful Investors Rewarded by
the Market Place? Evidence from Australian Shariahcompliant Equities. Qualitative Research in Financial
Markets, forthcoming.
Sadeghi M. 2008. Financial Performance of ShariahCompliant Investment: Evidence from Malaysian Stock
Market. International Research Journal of Finance and
Economics. 20:15–26.
Nesbitt S. 1994. Long-term rewards from shareholder
activism: a study of the CalPERS effect. Journal of
Applied Corporate Finance. 6:75–80.
Eds. Hatem A. El-Karanshawy et al.
91
The nexus between economic freedom and Islamic
bank performance: Empirical evidence from the
MENA banking sectors
Fadzlan Sufian1, Muhamed Zulkhibri, Abdul Majid2
Universiti Islam Antarabangsa Malaysia.
Economic Research and Policy Dept., Islamic Development Bank. Mailing address: 9th Floor, Economic Research and
Policy Department, Islamic Development Bank, P.O Box 5925, Jeddah, 21432, Saudi Arabia, e-mails:mzulkhibri@isdb.
org; [email protected], Tel: +966-2-646-6533; Fax: +966-2-506047132
1
2
All findings, interpretations, and conclusions are solely of the authors’ opinion and do not necessarily represent the
views of the institutions.
Abstract - The present study provides new empirical evidence on the impact of economic freedom
on Islamic banks’ performance. The empirical analysis focuses on Islamic banks operating in the
MENA banking sectors during the period 2000–2008. We find that the larger, more diversified,
and better capitalized Islamic banks tend to be relatively more profitable, while credit risk and
expense preference behaviour seem to exert negative impact. The findings suggest that greater
financial freedom positively influence the profitability of Islamic banks operating in the MENA
banking sectors. Interestingly, the impact of monetary freedom is negative implying that higher
(lower) monetary policy independence reduces (increases) Islamic banks’ profitability, providing
support to the benefits of government interventions.
Keywords: economic freedom, Islamic banks, profitability, panel regression analysis, MENA
JEL Classification: G21; G28
1. Introduction
Islamic banking is a relatively recent addition to the global
financial markets. Its conventional brick and mortar root can
be traced back to the early 1960s when Myt Ghamar Bank
was formed in Egypt in 1963. Between 1963 and 1971 the
bank provided Muslims with a place to deposit their savings
in accordance to the Syari’a principles1. Despite its humble
beginning, Islamic banks have blossomed throughout the
world and are looked upon as a viable alternative system
which has many things to offer.
Although it was initially developed to fulfill the needs
of Muslims, Islamic banking has now gained universal
acceptance. According to El-Qorchi (2005), the number
of Islamic financial institutions increased from a single
institution in 1975 to approximately 486 financial
institutions operating in more than 75 countries
worldwide2. Total assets of Islamic financial institutions are
estimated at US$250 billion and are tipped to be growing
at 15% per year, three times the rate of conventional banks.
The rapid growth rate confirms the growing importance of
Islamic banking and finance in the global financial markets.
The Islamic banks operate in markets characterized by
competition-inhibiting government regulation and in a
protected banking environment. Islamic banking, being a
participatory type of banking system, has entered on the
global banking market in full force. In recent years, market
conditions in Islamic banking have undergone extensive
changes from both the demand and supply sides. On the
demand side, customers have become more sophisticated,
value-oriented, and price sensitive, while on the supply
side, the globalization of financial markets has been
accompanied by governmental deregulation, financial
innovation, and automation.
These two factors have resulted in an increase in the number
of competitors, cost reductions, and profit declines. The
revolution in information technology, mainly in internet
banking has enabled the larger financial institutions to
penetrate markets and to increase their market share
within both national and overseas markets by providing
competitive products at lower prices. Furthermore,
Islamic equity-type financial instruments are competing
with conventional banking products and now face strong
Cite this chapter as: Sufian F, Zulkhibri M, Majid A (2015). The nexus between economic freedom and Islamic bank
performance: empirical evidence from the MENA banking sectors. In H A El-Karanshawy et al. (Eds.), Islamic banking
and finance – Essays on corporate finance, efficiency and product development. Doha, Qatar: Bloomsbury Qatar
Foundation
Developing Inclusive and Sustainable Economic and Financial Systems
Sufian et al.
competition from both banks and non-bank financial
institutions. This also accentuates competition within the
financial services industry.
It is reasonable to assume that these developments posed
great challenges to Islamic banks as the environment in which
they operates in has changed rapidly. This could sensibly
have an impact on the determinants of their performance.
Despite considerable development of the Islamic banking
sector, empirical works on Islamic banks’ performance is
still in its infancy. The knowledge of the underlying factors
which influences the Islamic banking sector’s performance
is essential given the growing importance of Islamic banking
and finance in the global financial markets. It is therefore
essential not only for the managers of the Islamic banks,
but for numerous stakeholders such as the central banks,
bankers associations, governments, and other financial
authorities to help them identify and formulate policies
to improve the performance of the Islamic banking sector,
particularly in the MENA region3.
On the perspective of economic freedom, economic theories
suggest that economic freedom tend to affect incentives,
productive effort, and the effectiveness of resource use.4
Economists and economic historians have argued that
since the time of Adam Smith, central ingredients for
economic progress are the freedom to choose and supply
resources, competition in business, trade with others, and
secure property rights (North and Thomas, 1973). Within
the context of the MENA region, it can be observed from
Table 1 that the region has achieved modest improvement
in economic freedom during the year 20105. It can be seen
from Table 1 that Bahrain retained the top ranking within
the region and managed to be ranked in the world Top 10,
while Qatar ranks in the world top 30.
The ongoing transformations of innovative and reformoriented states such as Bahrain, Qatar, Kuwait, and Oman
may pave the way for a more robust and dynamic regional
economic growth in the region. On different scale, Jordan
and Oman registered the highest gains in economic
freedom and Qatar’s improvement to 70.5, moved it
from the category of “moderately free” to “mostly free”,
while Syria’s improvement lifted its designation from
“repressed” economy to “mostly unfree”. However, no
other MENA countries are rated as having “mostly free”
economies. Nearly half of the region falls into the “mostly
unfree” category and two countries, namely Libya and
Iran, ranked among the world’s most repressed economies.
The institutional problems, such as lack of investment and
financial freedom and weak systems for protecting property
rights and preventing corruptions continue to degrade the
region’s overall economic freedom and economic potential.
The purpose of the present paper is to extend the earlier
works on the performance of the Islamic banking sector in
the MENA region and to establish empirical evidence on the
impact of economic freedom. The paper also investigates to
what extent the performance of Islamic banks is influenced
by internal factors (i.e. bank-specific characteristics) and
to what extent by external factors (i.e. macroeconomic
conditions and economic freedom). Although studies on
economic freedom is vast in the literature (e.g. Heckelman
and Knack, 2009; Altman, 2008; Powell, 2003; Adkins
et al. 2002; De Haan and Sturm, 2000; Heckelman and
94
Stroup, 2000; Heckelman, 2000; De Haan and Siermann,
1998), these studies have mainly examined the impact
of economic freedom on economic growth. On the other
hand, virtually nothing has been published to examine the
impact of economic freedom on the performance of the
conventional or Islamic banking sectors. This limitation
is somewhat surprising given the importance of bank
lending in promoting economic growth and development
(e.g. Ben Naceur and Ghazouani, 2007; Beck and Levine,
2004; Rajan and Zingales, 1998) and given the impact that
economic freedom is likely to have on the banking sector.
The paper is divided into five sections. The following section
presents the literature review. Section 3 describes the data,
sources, and empirical settings. In section 4 we present the
results and finally, section 5 concludes.
2. Review of the literature
The empirical evidence on the performance of the
conventional banking sectors is extensive. To date, the
numerous studies have mainly focused on the U.S. banking
sector (e.g. DeYoung and Rice, 2004; Stiroh and Rumble,
2006; Hirtle and Stiroh, 2007; Tregenna, 2009) and the
banking sectors of the western and developed countries
(e.g. Williams, 2003; Pasiouras and Kosmidou, 2007;
Kosmidou et al. 2007; Hawtrey and Liang, 2008; Kosmidou,
2008; Kosmidou and Zopounidis, 2008; Athanasoglou et al.
2008; Albertazzi and Gambacorta, 2008; Kasman et al.
2010). On the other hand, empirical works on the Islamic
banking sector is still in its infancy. Typically, studies on
Islamic bank performance have focused on theoretical
issues and the empirical works have relied mainly on
the analysis of descriptive statistics rather than rigorous
statistical estimation (El-Gamal and Inanoglu, 2005).
Hussein (2003) provides an analysis of the cost efficiency
features of Islamic banks in Sudan. By using the stochastic
cost frontier approach, he estimates cost efficiency for
a sample of 17 banks over the period 1990 and 2000.
The results show large variations in the cost efficiency
of Sudanese banks with the foreign owned banks being
the most efficient, while the state owned banks being the
most cost inefficient. The empirical findings suggest that
the small banks are relatively more efficient compared
to their large bank counterparts. In addition, banks with
a higher proportion of musharakah and mudharabah
finance relative to total assets tend to exhibit efficiency
advantages.
In another study on the Sudanese Islamic banking sector,
Hassan and Hussein (2003) examine the efficiency of the
Sudanese banking system during the period of 1992 and
2000. They employed a variety of parametric (cost and
profit efficiencies) and non-parametric Data Envelopment
Analysis (DEA) methods to a panel of 17 Sudanese banks.
They found that the average cost and profit efficiencies
under the parametric method were 55% and 50%
respectively, while it was 23% under the non-parametric
method. During the period under study, they suggest that
the Sudanese banking system has exhibited 37% allocative
efficiency and 60% technical efficiency, suggesting that the
overall cost inefficiency of the Sudanese Islamic banks were
mainly due to technical (managerially related) rather than
allocative (regulatory).
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Eds. Hatem A. El-Karanshawy et al.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
10
27
34
38
43
47
54
61
89
93
96
100
127
132
140
171
173
64.9
60.1
59.6
59.1
58.5
54.2
52.4
51.3
42.1
38.6
77.7
70.5
69.8
68.9
68.5
67.8
66.2
77.4
70.3
69.4
65.8
66.1
67.3
86.1
64.4
57.5
75.7
64.5
80.2
73.7
69.4
55.9
69.4
20.0
1.4
1.5
2.1
2.8
0.8
0.5
2.0
-2.8
0.6
0.4
0.1
-0.5
-0.2
-4.5
1.9
-1.3
-1.6
81.6
80.5
75.8
74.0
53.5
81.6
72.8
65.4
44.8
85.0
82.8
82.4
83.6
78.8
87.8
82.6
82.2
99.9
91.0
67.8
89.6
73.7
83.2
83.5
84.6
81.1
80.3
99.8
99.8
98.5
92.7
62.3
99.9
99.4
69.7
64.9
74.6
65.3
77.6
44.5
62.4
85.3
76.0
44.5
80.2
78.1
68.1
60.9
44.8
79.1
74.6
69.3
77.7
76.5
60.8
77.3
82.3
75.4
69.7
60.7
71.0
74.0
71.9
69.5
81.4
78.4
76.5
64.3
55.0
60.0
65.0
65.0
35.0
45.0
20.0
20.0
0.0
10.0
75.0
45.0
55.0
70.0
80.0
35.0
40.0
50.0
60.0
60.0
50.0
30.0
30.0
30.0
20.0
10.0
20.0
80.0
50.0
60.0
60.0
70.0
50.0
50.0
50.0
25.0
40.0
40.0
50.0
30.0
30.0
30.0
10.0
10.0
60.0
70.0
50.0
55.0
70.0
50.0
45.0
41.0
25.0
33.0
28.0
42.0
21.0
28.0
26.0
18.0
25.0
51.0
70.0
55.0
50.0
61.0
65.0
43.0
67.9
59.0
27.2
53.6
65.7
50.9
52.9
55.8
50.7
20.0
97.0
67.0
89.1
74.2
64.3
72.4
77.0
Note: Each one of the 10 freedom is graded using a 0 to 100 scale, where 100 represents the maximum freedom. A score of 100 signifies an economic environment or
set of policies that is most conducive to economic freedom. Many of the 10 freedoms are based on quantitative data that are converted directly into a score.
Source: The Heritage Foundation.
Bahrain
Qatar
Oman
Jordan
Israel
UAE
Saudi
Arabia
Kuwait
Lebanon
Morocco
Egypt
Tunisia
Yemen
Algeria
Syria
Iran
Libya
Country
2011 Change
Freedom
World Regional Overall from Business Trade
Fiscal Government Monetary Investment Financial Property
from
Labour
Rank Ranking Score
2010 Freedom Freedom Freedom Spending
Freedom Freedom Freedom Rights Corruption Freedom
Table 1. Economic freedom index for MENA.
The nexus between economic freedom and Islamic bank performance
95
Sufian et al.
El-Gamal and Inanoglu (2004) employ the stochastic
frontier approach to estimate the cost efficiency of Turkish
banks over the period 1990–2000. The study compared the
cost efficiencies of 49 conventional banks with four Islamic
special finance houses (SFHs). The Islamic firms comprised
around 3% of the Turkish banking market. Overall, they
suggest that the Islamic financial institutions to be the
most efficient. This could be explained by their emphasis
on Islamic asset-based financing which led to low nonperforming loans ratios.
The study by Hassan (2006) is among the few performed
to examine the efficiency of Islamic banks in a cross-country
setting. He employs both the parametric (Stochastic
Frontier Approach) and non-parametric (Data Envelopment
Analysis) methods to examine the efficiency of banks in the
sample. The findings indicate that during the period 1993–
2001, Islamic banks have exhibited a relatively higher
profit efficiency compared to cost efficiency. He suggests
that the main source of inefficiency is allocative rather than
technical. The results indicate that the overall inefficiency
was output related. The results indicate that on average
the Islamic banking industry is relatively less efficient
compared to their conventional counterparts.
While the above outlines the literature that employs advanced
modelling techniques to evaluate Islamic banks’ performance,
one should also note that there is a growing body of literature
that covers the general performance features of Islamic banks.
Such studies include those by Hassan and Bashir (2003)
who look at the determinants of Islamic banks’ performance
and show that Islamic banks to be just as efficient as their
conventional bank peers if one uses standard accounting
measures such as the cost-to-income ratio. Other studies
that followed similar approach are those by Sarker (1999)
who examines the performance and operational efficiency of
Bangladeshi Islamic banks, while Bashir (1999) investigates
the risk and profitability of two Sudanese banks.
3. Data and methodology
The present study employs an unbalanced annual bank level
data of all Islamic banks operating in the MENA countries
covering the period 2000–2008. The financial statements
of Islamic banks operating in the MENA banking sectors
are collected from the Bankscope database of Bureau
van Dijk’s company. The macroeconomic variables are
retrieved from the IMF Financial Statistics (IFS) and the
World Bank World Development Indicator (WDI) databases
while economic freedom variables are extracted from The
Heritage Foundation.
Measure of performance
Following Ben Naceur and Goaied (2008), Kosmidou
(2008), and Abbasoglu et al. (2007) among others, the
dependent variable used in this study is Return on Assets
(ROA). ROA shows the profit earned per dollar of assets
and most importantly, reflects management ability to
utilize banks financial and real investment resources
to generate profits (Hassan and Bashir, 2003). For any
bank, ROA depends on the bank’s policy decisions as well
as other uncontrollable factors relating to the economy
and government regulations. Rivard and Thomas (1997)
suggest that bank profitability is best measured by ROA,
since it is not distorted by high equity multipliers and
represents a better measure of the ability of firms to
generate returns on its portfolio of assets.
Internal determinants
The bank specific variables included in the regression
models are LLP/TL (loans loss provisions divided by total
loans), EQASS (book value of stockholders’ equity as a
fraction of total assets), NIE/TA (total overhead expenses
divided by total assets), LOANS/TA (total loans divided by
total assets), and LNTA (log of total assets).
Bashir (1999) and Bashir (2001) performed regression
analyses to examine the underlying determinants of Islamic
banks’ performance. By employing bank level data from
the Middle East, the results indicate that the performance
of banks, in terms of profits, is mostly generated from
overhead, customer short-term funding, and non-interest
earning assets. Furthermore, Bashir (2001) claimed that
since deposits in Islamic banks are treated as shares, reserves
held by banks propagate negative impacts such as reducing
the amount of funds available for investment. In essence,
the findings from this literature are that Islamic banks are
at least as efficient as their conventional bank counterparts
and in most cases are relatively more efficient.
The ratio of loan loss provisions to total loans (LLP/TL) is
incorporated as an independent variable in the regression
analysis as a proxy of credit risk. The coefficient of the LLP/
TL variable is expected to enter the regression models with
a negative sign. In this vein, Miller and Noulas (1997)
point out that the greater the exposure of banks to high
risk loans, the higher would be the accumulation of unpaid
loans and profitability would be lower. Miller and Noulas
(1997) suggest that decline in loan loss provisions are in
many instances the primary catalyst for increases in profit
margins. Furthermore, Thakor (1987) also suggests that
the level of loan loss provisions is an indication of the bank’s
asset quality and signals changes in future performance.
The above literature reveals the following research gaps.
First, the majority of these studies have concentrated on
the conventional banking sectors and the banking sectors
of the western and developed countries. Second, empirical
evidence on the developing countries banking sectors,
particularly the Islamic banking sectors are relatively scarce.
Finally, virtually nothing has been published to examine
the impact of economic freedom on the Islamic banking
sector. In light of these knowledge gaps, the present paper
provides new empirical evidence on the impact of economic
freedom on the performance of Islamic banks operating in
the MENA countries banking sectors.
The EQASS variable is included in the regression models
to examine the relationship between profitability and bank
capitalization. Strong capital structure is essential for banks in
developing economies, since it provides additional strength to
withstand financial crises and increased safety for depositors
during unstable macroeconomic conditions. Furthermore,
lower capital ratios in banking imply higher leverage and risk
and therefore greater borrowing costs. Thus, the profitability
level should be higher for the better capitalized bank.
96
The ratio of non-interest expenses over total assets, NIE/
TA, is used to provide information on the variations of bank
Islamic banking and finance – Essays on corporate finance, efficiency and product development
The nexus between economic freedom and Islamic bank performance
operating costs. The variable represents total amount of
wages and salaries, as well as the costs of running branch
office facilities. The relationship between the NIE/TA
variable and profitability levels is expected to be negative,
because the more productive and efficient banks should be
able to keep their operating costs low. Furthermore, the
usage of new electronic technology, like ATMs and other
automated means of delivering services, may have caused
expenses on wages to fall (as capital is substituted for
labor).
An important decision that the managers of Islamic
banks must take refers to the liquidity management and
specifically to the measurement of their needs related to the
process of deposits and loans. For that reason, the ratio of
total loans to total assets (LOANS/TA) is used as a measure
of liquidity. Higher figures denote lower liquidity. Without
the required liquidity and funding to meet obligations,
a bank may fail. Thus, in order to avoid insolvency
problems, banks often hold liquid assets, which can be
easily converted to cash. However, liquid assets are usually
associated with lower rates of return. It would therefore
reasonable to expect higher liquidity to be associated with
lower bank profitability.
The LNTA variable is included in the regression models as
a proxy of size to capture for the possible cost advantages
associated with size (economies of scale). In the literature,
mixed relationships are found between size and profitability,
while in some cases a U-shaped relationship is observed.
LNTA is also used to control for cost differences related
to bank size and for the greater ability of the large bank
to diversify. In essence, LNTA may lead to positive effects
on bank profitability if there are significant economies of
scale. On the other hand, if increased diversification leads
to higher risks, the variable may exhibit negative effects.
External determinants
If analysis is done in a static setting, they may fail to capture
developments in the regulatory environment and in the
marketplace, which may have changed the underlying
production technology and the associated production
functions. Furthermore, different banking forms could
demonstrate different reactions to environmental changes.
Hence, the change in the financial landscape and structure,
etc., may vary across banking groups (Saunders et al.
1990; Button and Weyman-Jones, 1992; Berger, 1995). To
measure the relationship between economic and market
conditions and Islamic banks’ performance, LNGDP, INFL,
CR3, and Z-SCORE variables are used.
Gross domestic product (GDP) is among the most
commonly used macroeconomic indicator to measure total
economic activity within an economy. The GDP is expected
to influence numerous factors relating to the supply and
demand for loans and deposits. Favourable economic
conditions will affect positively on the demand and supply
of banking services, but will have either positive or negative
influence on bank profitability levels.
Another important macroeconomic condition which may
affect both the costs and revenues of banks is the inflation
rate (INFL). Staikouras and Wood (2003) points out that
inflation may have direct effects i.e. increase in the price
Eds. Hatem A. El-Karanshawy et al.
of labour and indirect effects i.e. changes in interest rates
and asset prices on the profitability of banks. Perry (1992)
suggests that the effects of inflation on bank performance
depend on whether the inflation is anticipated or
unanticipated. In the anticipated case, the profit rates
are adjusted accordingly resulting in revenues to increase
faster than costs subsequently positive impact on bank
profitability. On the other hand, in the unanticipated
case, banks may be slow to adjust their interest rates
resulting in a faster increase of bank costs compared to
bank revenues and consequently negative effects on bank
profitability6.
To examine the impact of concentration on Islamic banks’
performance, the CR3 variable is introduced in the
regression models. The CR3 ratio is calculated as the total
assets held by the three largest banks in the country. The
variable is used to examine the impact of asset concentration
in the national banking sector on the profitability of Islamic
banks. The Structure-Conduct-Performance (SCP) theory
posits that banks in a highly concentrated market tend to
collude and therefore earn monopoly profits (Molyneux et
al. 1996). Berger (1995) points out that the relationship
between bank concentration and performance in the U.S.
depends critically on what other factors are held constant.
According to the industrial organization literature, a
positive impact is expected under both collusion and
efficiency views (Goddard et al. 2001).
The Z-Score (Z-SCORE) variable is used as a proxy of
bank soundness. The index measures how many standard
deviations a bank is away from exhausting its capital base
(a distance-to-default measure). The Z-Score is a popular
measure of soundness because it combines banks’ buffers
(capital and profits) with the risks they face in a way that
is grounded in theory (Cihak et al. 2009). A higher Z-Score
implies a lower probability of insolvency, providing a more
direct measure of soundness than, for example, simple
leverage measures (Cihak et al. 2009). This index combines
in a single indicator: (i) profitability, given by a period
average return on assets (ROA); leverage measure, given by
the period average equity-to-asset ratio (K) (equity here is
defined as total equity from the balance sheet of a bank); and
return volatility, given by the period standard deviation of
+K
ROA (Vol. (ROA)) i.e. Z = VolROA
.( ROA ) where ROA (profitability)
is a period average of ROA, K (leverage measure) is the
period average equity-to-asset ratio, and Vol. (ROA) is the
return volatility given by the period standard deviation of
ROA. A higher (lower) Z-SCORE indicates lower (higher)
risk (De Nicolo et al. 2003).
Economic freedom measurements
In simple terms, economic freedom is a conceptual
measure of the private ownership and market allocation
of resources, in lieu of government ownership and
control. Expressing the sentiment of many, including the
originators of the economic freedom index, Berggren
(2003) defines economic freedom as “the degree to which
an economy is a market economy—that is, the degree to
which it entails the possibility of entering into voluntary
contracts within the framework of a stable and predictable
rule of law that upholds contracts and protects private
property, with a limited degree of interventionism in the
form of government ownership, regulations, and taxes”.
97
Sufian et al.
In regression model 2, OVER_FREE is introduced to examine
the impact of overall economic freedom on the performance
of the Islamic banks operating in the MENA banking
sectors. OVER_FREE is the overall economic freedom
index and is defined by multiple rights and liberties. The
index uses 10 specific freedoms, namely Business freedom,
Trade freedom, Fiscal freedom, Government size, Monetary
freedom, Investment freedom, Financial freedom, Property
rights, Labor freedom, and Freedom from corruption.
et al. (2009) suggest that potential endogeneity could be
a problem when assessing bank profitability determinants.
For instance, the more profitable banks may have sufficient
resources to provision for their non-performing loans. The
more profitable banks may also find it easier to increase
their customer base via successful advertising campaigns
and could hire the most skilled personnel, and therefore
enhances their profitability levels (Garcia-Herrero et al.
2009).
Besides the overall economic freedom index, we have
selected three other indices which are closely related to
the financial sector. These include BUSI_FREE, MONE_
FREE, and FINA_FREE indices. BUSI_FREE is the business
freedom index. The index measures how free entrepreneurs
are to start businesses, how easy it is to obtain licenses, and
the ease of closing a business. Impediments to any of these
three activities are deterrents to businesses and therefore
to job creations. MONE_FREE is the monetary freedom
index. The index combines a measure of price stability with
an assessment of price controls. Both inflation and price
control distorts market activity. Price stability without
microeconomic intervention is an ideal state of a free market.
FINA_FREE is the financial freedom index. The index is a
measure of banking security as well as independence from
government’s control. State ownership of banks and other
financial institutions such as insurer and capital markets is
an inefficient burden and political favoritism has no place
in a free capital market. All these indices have 0 to 100
scales, where 100 represents maximum freedom. A score
of 100 signifies an economic environment, or set of policies
that is most conducive to economic freedom.
Arellano and Bond (1991) proposed an efficient
Generalized Methods of Moment (GMM) estimator that
uses instruments of which the validity is based on the
orthogonality between the lagged values of the dependent
variable and the errors. The technique eliminates the
unobserved bank heterogeneity by estimating the equation
in first-differences and to control for possible endogeneity
problem by using the model’s variables lagged by one
or more periods as instruments. We employ the GMM
estimator as proposed by Arellano and Bond (1991) to
ensure efficiency and consistency of the estimations.
Therefore, a dynamic GMM model is adopted via the
inclusion of a lagged dependent variable among the
regressors to capture the persistence of bank profitability
over time reflecting impediments to market competition,
informational opacity, and/or sensitivity to regional/
macroeconomic shocks (Berger et al. 2000).
Finally, CORR_FREE is introduced in regression model
6 to assess the impact corruption on the performance of
Islamic banks. CORR_FREE is the freedom from corruption
index. The index is based on quantitative data that assess
the perception of corruption in the business environment,
including levels of governmental, legal, judicial, and
administrative corruption. Similar to the BUSI_FREE,
MONE_FREE, and FINA_FREE indices, the CORR_FREE
index also takes a value of between 0 and 100, where
100 represent the maximum freedom. Table 1 contains
the summary statistics of the variables used to proxy
profitability and its determinants.
where i = 1, 2, …, N (number of firms) and t = 1, 2, …,
T (time period). In the specification, pit denotes the
profitability of bank i at time t; pt−1 indicates a one period
lagged profitability; Xi,t is vector exogenous bank-specific
regressors: Mt is a vector of country specific variables; Ei is
a vector of country specific economic freedom variables; mt
is a time fixed effect; vi is an unobserved banks’ fixed effect;
ei,t is a serially uncorrelated error term.
Econometric specification
Since the panel data cover many heterogenous banks
and time periods, the possible correlation between the
regressors and bank-specific effects, the endogeneity of
regressors with respect to idiosyncratic shock and the
heteroscedasticity of the disturbance term (idioscyncratic
shock) would result in a biased and inconsistent estimation
with Ordinary Least Square (OLS) estimation technique.
The OLS estimator would result in an upward estimate of
the coefficient while the within-group estimator would
be downward biased (Blundell et al. 1992). A natural
technique for dealing with variable that are correlated with
the error term is to instrument them.
Berger et al. (2000) points out that bank profitability
tend to persist over time reflecting impediments to market
competition, informational opacity, and sensitivity to
macroeconomic shocks. Furthermore, Garcia-Herrero
98
The baseline model is formulated as follows:
π it = a + λπ i ,t-1 + ∑ b X it + ∑γ M t + ∑δ E t + µ t + υ i + e it (1)
We use several tests proposed by Arellano and Bond
(1991) to check whether the instruments are properly
chosen and the assumptions underlying the model holds.
Our estimations rely on the fact that the disturbances
follow an MA(1) process and there is no second order
autocorrelation (m2) together with Sargan/Hansen tests of
over-identifying restriction (J-test) to examine the validity
of the instruments used in the regression models.
Extending Eq. (1) to reflect the variables as described in
Table 2, the baseline model is formulated as follows:
ROA jt = b 0 + b 1 LOANS / TA jt + b 2 LNTA jt
+ b 3 LLP / TL jt + b 4 NIE / TA jt + b 5 EQASS jt
+ b 6 LNGDPt + b 7 INFLt + b 8 CR3t + b 9 Z - SCORE t
+ b 9 OVER _ FREE t + b 10 BUSI _ FREE t
+ b 11 MONE _ FREE t + b 12 FINA _ FREE t
(2)
+ b 13CORR _ FREE t + e jt
Table 3 provides information on the degree of correlation
between the explanatory variables used in the panel
regression analysis. In general, the matrix shows that the
Islamic banking and finance – Essays on corporate finance, efficiency and product development
A proxy measure of bank profitability measured as the return on average
total assets of the bank in year t.
ROA
Eds. Hatem A. El-Karanshawy et al.
A measure of bank’s capital strength in year t, calculated as equity/total
assets. High capital asset ratio is assumed to be indicator of low leverage
and therefore lower risk.
Calculated as non-interest expense/ total assets and provides information
on the efficiency of the management regarding expenses relative to the
assets in year t. Higher ratios imply a less efficient management.
A measure of liquidity, calculated as total loans/ total assets. The ratio
indicates what percentage of the assets of the bank is tied up in loans in
year t.
The natural logarithm of the accounting value of the total assets of the
bank in year t.
EQASS
NIE/TA
LOANS/TA
LNTA
The three largest banks asset concentration ratio.
The Z-Score index. Is used as a proxy measure of the banking sector’s risk to
default.
CR3
Z-SCORE
Overall economic freedom is defined by multiple rights and liberties can
be quantified as an index of less abstract components. The index uses 10
specific freedoms, some as composites of even further detailed and
quantifiable components.
Economic Freedom
The rate of inflation.
INFL
OVER_FREE
Natural logarithm of gross domestic products.
LNGDP
External Factors
Loan loss provisions/ total loans. An indicator of credit risk, which shows
how much a bank is provisioning in year t relative to its total loans.
LLP/TL
Internal Factors
Independent
Description
Variable
Table 2. Descriptive of the variables used in the regression models.
60.464
11.027
0.730
2.246
6.057
8.117
48.583
3.755
21.227
8.321
2.577
Mean
12.575
6.832
0.151
2.154
4.237
2.666
23.706
3.247
23.458
14.13
3.798
Std. Dev.
(Continued)
Heritage Foundation
(www.heritage.org/index)
IMF
International
Financial Statistics
IMF
International
Financial Statistics
IMF
International
Financial Statistics
IMF
International
Financial Statistics
BankScope
BankScope
BankScope
BankScope
BankScope
BankScope
Sources/Database
The nexus between economic freedom and Islamic bank performance
99
100
46.134
Freedom from corruption is based on quantitative data that assess the
perception of corruption in the business environment, including levels of
governmental legal, judicial, and administrative corruption.
CORR_FREE
Heritage Foundation
(www.heritage.org/index)
47.354
Financial freedom is a measure of banking security as well as independence
from government control. State ownership of banks and other financial
institutions such as insurer and capital markets is an inefficient burden,
and political favoritism has no place in a free capital market.
FINA_FREE
24.140
Heritage Foundation
(www.heritage.org/index)
26.329
Heritage Foundation
(www.heritage.org/index)
75.829
Monetary freedom combines a measure of price stability with an assessment
of price controls. Both inflation and price controls distort market activity.
Price stability without microeconomic intervention is the ideal state for the
free market.
MONE_FREE
12.318
64.890
4. Empirical findings
Business freedom measures how free entrepreneurs are to start businesses,
how easy it is to obtain licenses, and the ease of closing a business.
Impediments to any of these three activities are deterrents to business
and therefore to job creation.
correlation between the bank specific variables are not
strong, implying that multicollinearity problems are not
severe. Kennedy (2008) points out that multicollinearity
is a problem when the correlation is above 0.80 which
is not the case here. However, it is worth noting that
the LNGDP variable is highly correlated to most of the
economic freedom variables. To address this concern, we
have also estimated all regression models by excluding
the macroeconomic variables. Furthermore, due to the
high correlation between the economic freedom variables,
the regression models are estimated by including the
each economic freedom indicator at a time, rather than
estimating all economic freedom variables concurrently.
BUSI_FREE
Table 2. (Continued)
14.763
Heritage Foundation
(www.heritage.org/index)
Sufian et al.
The regression results focusing on the relationship
between bank profitability and the explanatory variables
are presented in Table 4. The reliability of our econometric
methodology depends critically on the validity of the
instruments, which can be evaluated with Sargan’s test
of overidentifying restrictions, asymptotically distributed
as c2 in the number of restrictions. A rejection of the null
hypothesis that instruments are orthogonal to the errors
would indicate that the estimates are not consistent (Baum
et al. 2010)7. We also present test statistics for the first and
second order serial correlations in the error process. In a
dynamic panel data context, second order serial correlation
should not be present if the instruments are appropriately
uncorrelated with the errors (Baum et al. 2010).
It can be observed from Table 4 that for all the estimated
models, the Sargan statistics for overidentifying restrictions
and the Arrelano–Bond AR(2) tests shows that our
instruments are appropriately orthogonal to the error and
no second order serial correlation is detected respectively.
Furthermore, the highly significant of the lagged dependent
variable’s coefficient confirms the dynamic character of
the model specification, thus justifying the use of dynamic
panel data model estimation.
Concerning the liquidity results, the empirical findings
suggest a negative sign of the coefficient of the LOANS/
TA in the baseline regression model. As higher (lower)
figures of the ratio denote lower (higher) liquidity levels,
the results imply that the relatively less (more) liquid banks
tend to exhibit higher (lower) profitability levels. On the
other hand, the empirical findings also suggest that the
coefficient of the variable is positive when we control for
overall economic freedom, business freedom, financial
freedom, and freedom from corruption. However, the results
should be interpreted with caution since the coefficient of
the variable is not significant at any conventional levels in
any of the regression models estimated.
It can be observed from Table 4 that the coefficient of the
LNTA variable entered the baseline regression model with a
negative sign and is statistically significant when we control
for economic and financial market conditions lending
support to Spathis et al. (2002), Dogan and Fausten (2003),
and Kosmidou (2008). Moreover, the earlier studies have
concluded that marginal cost savings could be achieved by
increasing the size of the banking firm, especially as markets
develop (Berger et al. 1987; Boyd and Runkle, 1993; Miller
and Noulas, 1997; Athanasoglou et al. 2008). In this vein,
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Eds. Hatem A. El-Karanshawy et al.
1.000
0.493**
1.000
EQASS
INFL
-0.052
-0.141*
-0.326**
0.109
0.184**
0.076
0.046
-0.161**
1.000
LNGDP
0.048
-0.369** -0.293** -0.169* -0.237** -0.097
0.407** -0.026
-0.555** -0.479** -0.432** 0.007
1.000
0.077
-0.204** -0.265** -0.068
-0.129
1.000
0.193
0.025
0.036
-0.015
1.000
0.670** 0.619** -0.092
1.000
0.754** -0.130*
1.000
0.142*
1.000
LNTA
Z-SCORE
LOANS/
TA
CR3
NIE/TA
0.044
0.484**
0.148*
-0.032
-0.624**
-0.841**
-0.324**
-0.084
0.000
1.000
OVER_
FREE
0.095
0.514**
0.238**
-0.094
-0.636**
-0.759**
-0.094
0.002
-0.142
0.831**
1.000
BUSI_
FREE
0.160*
0.419**
0.058
-0.027
-0.622**
-0.791
-0.569**
-0.100
0.045
0.815**
0.639**
1.000
MONE_
FREE
Note: The table presents the results from Pearson correlation coefficients. ** and * indicates significance at 1% and 5% levels, respectively.
LLP/TL
EQASS
NIE/TA
LOANS/TA
LNTA
LNGDP
INFL
CR3
Z-SCORE
OVER_FREE
BUSI_FREE
MONE_FREE
FINA_FREE
CORR_FREE
LLP/
TL
0.031
0.496**
0.202**
-0.117
-0.713**
-0.847**
-0.135*
0.181**
-0.263**
0.846**
0.794**
0.681**
1.000
FINA_
FREE
0.059
0.387**
0.052
0.067
-0.476**
-0.657**
-0.140*
-0.177**
0.220**
0.869**
0.732**
0.618**
0.608**
1.000
CORR_
FREE
The notation used in the table below is defined as follows: LOANS/TA is used as a proxy measure of loans intensity, calculated as total loans divided by total assets; LNTA is
a proxy measure of size, calculated as a natural logarithm of total bank assets; LLP/TL is a measure of bank risk calculated as the ratio of total loan loss provisions divided
by total loans; NII/TA is a measure of bank diversification towards non interest income, calculated as total non-interest income divided by total assets; NIE/TA is a proxy
measure for costs, calculated as non-interest expenses divided by total assets; EQASS is a measure of capitalization, calculated as book value of shareholders equity as a
fraction of total assets; LNGDP is natural log of gross domestic products; INFL is the rate of inflation; OVER_FREE is the overall economic freedom index; BUSI_FREE is
the business freedom index; MONE_FREE is the monetary freedom index; FINA_FREE is the financial freedom index; CORR_FREE is the freedom from corruption index.
Table 3. Correlation matrix for the explanatory variables.
The nexus between economic freedom and Islamic bank performance
101
Sufian et al.
Eichengreen and Gibson (2001) suggest that the effect of a
growing bank’s size on performance may be positive up to
a certain limit. Beyond this point the effect of size could be
negative due to bureaucratic and other reasons.
As expected, the impact of credit risk (LLP/TL) is negative
(statistically significant at the 10% level) suggesting that
Islamic banks with higher credit risk tend to exhibit lower
profitability levels. The results imply that Islamic banks
should focus more on credit risk management, which
has been proven to be problematic in the recent past.
Serious banking problems have arisen from the failure
of financial institutions to recognize impaired assets and
create reserves to write off these assets. An immense help
towards smoothing these anomalies would be provided by
improving the transparency of the banking sector, which in
turn will assist banks to evaluate credit risk more effectively
and avoid problems associated with hazardous exposure.
Similarly, the empirical findings seem to suggest that
expense preference behaviour measured by NIE/TA has
consistently exhibit a negative relationship. The finding is in
consonance with the bad management hypothesis of Berger
and DeYoung (1997). Low measure of efficiency is a signal
of poor senior management practices, which apply to inputusage and day-to-day operations. Clearly, efficient cost
management is a prerequisite to improve the profitability
of Islamic banks operating in the MENA banking sectors.
Furthermore, most of the MENA countries banking sectors
have not reached the maturity level required to link quality
effects from increased spending to higher earnings.
Referring to the impact of capitalization, it can be observed
from Table 4 that EQASS exhibits a positive relationship.
The result is consistent with the previous studies by
among others Isik and Hassan (2003), Goddard et al.
(2004), and Kosmidou (2008) providing support to the
argument that the well capitalized banks face lower costs
of going bankrupt, thus lowers their cost of funding or
that they have lower needs for external funding resulting
in a higher profitability level. Nevertheless, strong capital
structure is essential for banks in emerging economies
since it provides additional strength to withstand financial
crises and increased safety for depositors during unstable
macroeconomic conditions (Sufian, 2009). However,
it should be noted that the coefficient of the variable is
not significant at any conventional levels in any of the
regression models estimated.
The empirical findings seem to suggest that LNGDP has
positive and significant impact on the profitability of Islamic
banks operating in the MENA countries, lending support
to the association between economic growth and banking
sector’s performance. The high economic growth could
have encouraged Islamic banks to lend more and improve
the quality of their assets. The demand for financial services
tends to grow as economies expand and societies become
wealthier. Likewise, it can be observed from Table 4 that the
coefficient of the INFL variable exhibits a positive sign in the
baseline regression model, implying that during the period
under study the levels of inflation have been anticipated
by Islamic banks operating in the MENA banking sectors.
This allows bank managements the opportunity to adjust
the profit rates accordingly and consequently earn higher
profitability.
102
Turning to the impact of banking sector’s concentration,
it can be observed from Table 4 that the coefficient of the
three banks concentration ratio (CR_3) has consistently
exhibit a positive sign and becomes statistically significant
when we control for freedom from corruption (CORR_
FREE). Within the context of the MENA Islamic banking
sector, the empirical findings clearly lend support to the
SCP hypothesis. The SCP hypothesis states that banks in a
highly concentrated market tend to collude and therefore
earn monopoly profits (Short, 1979; Gilbert, 1984;
Molyneux et al. 1996). It can be observed from Table 4 that
the impact of banking sector risk (Z-SCORE) is positive
and highly significant. The result is in consonance with
the findings of among others Boyd and De Nicolo (2006)
lending support to the stringent capital requirements
of Basel II. From the policymaking point of view, the
empirical findings calls for a more effective policymaker’s
role in reducing excessive bank risk exposures and at
the same time to induce more efficient risk management
practices by Islamic banks operating in the MENA banking
sectors.
Does greater economic freedoms foster bank
performance?
To address the issue whether economic freedom matters in
determining the performance of Islamic banks operating
in the MENA banking sectors, we re-estimate Eq. (2) to
include the economic freedom indices variables discussed
in Section 3. The results are presented in columns 3 to 7
of Table 4. As observed, the empirical findings presented
in column 3 of Table 4 suggest that the coefficient of
the overall economic freedom (OVER_FREE) variable
is negative, but is not statistically significant at any
conventional levels.
Concerning the impact of business freedom (BUSI_FREE)
on the profitability of Islamic banks, the empirical findings
presented in column 4 of Table 4 indicate that the coefficient
of the BUSI_FREE variable is positive. The results imply that
the greater ability to start, operate, and close businesses
fosters the performance of Islamic banks. Clearly, the
greater ability to set up new businesses in the MENA
countries is a prerequisite to improve the performance of
the Islamic banking sector.
Referring to the impact of monetary freedom (MONE_
FREE), it is interesting to note that the coefficient of the
variable is negative. If anything could be delved, the
empirical findings indicate that higher (lower) government
intervention in the market increases (reduces) the
profitability of Islamic banks operating in the MENA banking
sectors. A stable and reliable monetary policy is crucial to
business environment, as it may help firms and societies to
make investment, savings, and other long-term plans. High
inflation rates not only confiscate wealth, but also distort
pricing, misallocate resources, and raise the cost of doing
business. Furthermore, the value of a country’s currency
largely depends on the monetary policy of its government.
A monetary policy that endeavors price stability and puts
inflation at bay, enables firms to rely on the market prices
for their future investments plans.
As expected, the coefficient of the financial freedom
(FINA_FREE) variable entered the regression model with
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Eds. Hatem A. El-Karanshawy et al.
0.126
(1.19)
-0.014
(-0.84)
-0.130
(-0.68)
-0.056*
(-1.67)
-0.599**
(-2.22)
0.025
(0.90)
ROA t-1
Z-SCORE
CR3
INFL
LNGDP
EQASS
NIE/TA
LLP/TL
LNTA
LOANS/TA
5.073*
(1.68)
CONSTANT
(1)
0.591**
(2.01)
0.026
(0.29)
9.487
(1.40)
0.181**
(2.20)
0.155*
(1.79)
-0.016
(-0.72)
-0.871***
(-2.60)
-0.026
(-0.97)
-0.807***
(-3.31)
0.009
(0.27)
6.879
(0.42)
-1.291
(-0.20)
0.152*
(1.75)
0.004
(0.15)
-0.248
(-0.42)
0.004
(0.13)
-0.665*
(-1.74)
0.012
(0.29)
-0.771
(-0.91)
-0.019
(-0.19)
11.918
(1.37)
0.280**
(2.47)
0.428
(0.81)
-0.006
(-0.07)
9.478
(1.14)
0.250**
(2.50)
Economic Conditions
0.119*
(1.65)
0.004
(0.19)
0.256
(0.47)
-0.009
(-0.34)
-0.529*
(-1.62)
0.040
(1.08)
(4)
-12.792
(-0.98)
Bank Characteristics
(3)
(2)
ONE STEP SYS-GMM
-0.015
(-0.04)
-0.028
(-0.32)
8.989
(1.12)
0.251**
(2.38)
0.143**
(2.32)
-0.000
(-0.01)
-0.337
(-0.55)
-0.002
(-0.09)
-0.736**
(-2.38)
0.012
(0.28)
5.575
(0.41)
(5)
1.100*
(1.71)
-0.112
(-0.86)
0.014
(0.00)
0.275***
(2.60)
(Continued)
-0.593
(-1.23)
-0.019
(-0.26)
16.262*
(1.87)
0.280**
(2.53)
0.115
(1.43)
0.025
(0.83)
-0.059
(-0.09)
0.006
(0.15)
-0.641*
(-1.86)
0.063
(1.49)
-4.259
(-0.47)
-9.484
(-1.28)
0.107**
(1.96)
0.002
(0.09)
-0.268
(-0.45)
0.001
(0.04)
-0.570*
(-1.80)
0.009
(0.23)
(7)
(6)
The notation used in the table below is defined as follows: LOANS/TA is used as a proxy measure of loans intensity, calculated as total loans divided by total assets; LNTA is a proxy measure
of size, calculated as a natural logarithm of total bank assets; LLP/TL is a measure of bank risk calculated as the ratio of total loan loss provisions divided by total loans; NII/TA is a measure
of bank diversification towards non interest income, calculated as total non-interest income divided by total assets; NIE/TA is a proxy measure for costs, calculated as non-interest expenses
divided by total assets; EQASS is a measure of capitalization, calculated as book value of shareholders equity as a fraction of total assets; LNGDP is natural log of gross domestic products; INFL
is the inflation rate; OVER_FREE is the overall economic freedom index; BUSI_FREE is the business freedom index; MONE_FREE is the monetary freedom index; FINA_FREE is the financial
freedom index; CORR_FREE is the freedom from corruption index.
ROA jt = b 0 + b 1 LOANS / TA jt + b 2 LNTA jt + b 3 LLP / TL jt + b 4 NIE / TA jt + b 5 EQASS jt + b 6 LNGDPt + b 7 INFLt + b 8 CR3t + b 9 Z - SCORE t + b 9 OVER _ FREE t
+ b 10 BUSI _ FREE t + b 11 MONE _ FREE t + b 12 FINA _ FREE t + b 13 CORR _ FREE t + e jt
Table 4. Panel Generalized Methods of Moments (GMM) regression results.
The nexus between economic freedom and Islamic bank performance
103
Sufian et al.
183.61***
0.334
0.693
0.598
94
In order to check for the robustness of the results, we carry
out several sensitivity analyses. First, in light of Holmes
et al. (2008) arguments, we remove all the macroeconomic
and market conditions variables from the regression
models and repeat Eq. (2). The regression results are
presented in Table 5. All in all, it can be observed that the
coefficients of the baseline regression models stay mostly
the same: the sign and the order of magnitude remained
similar and significant as in the baseline regression models.
As observed, the empirical findings suggest that the
coefficient of the OVER_FREE, BUSI_FREE, and MONE_
FREE entered the regression models with a negative sign,
but are insignificant. From column 4 of Table 5 it can be
observed that the coefficient of the financial freedom
(FINA_FREE) retains its positive sign and is significant. On
a similar vein, the empirical findings suggest that CORR_
FREE exhibits the same negative sign.
Values in parentheses are z-statistics.
***, **, and * indicates significance at 1, 5, and 10% levels, respectively.
999.69***
0.182
0.530
0.987
94
137.76***
0.192
0.773
0.878
111
38.41***
0.893
0.457
0.231
148
Wald χ2
AR(1) p-value
AR(2) p-value
Sargan p-value
No. of Observationst-1
CORR_FREE
FINA_FREE
MONE_FREE
BUSI_FREE
OVER_FREE
Robustness checks
1169.36***
0.171
0.966
0.759
94
-0.124
(-1.20)
0.078
(0.71)
-0.264
(-1.16)
Economic Freedom
Table 4. (Continued)
104
Finally, it is observed from column 6 of Table 4 that the
coefficient of the freedom from corruption (CORR_FREE)
variable exhibits a negative sign (statistically significant
at the 1% level). The empirical findings from this study
clearly suggest that corruption (e.g. corruption in the
business environment, including levels of governmental,
legal, judicial, and administrative) has significant negative
impact on the profitability of Islamic banks operating in the
MENA banking sectors.
543.96***
0.241
0.985
0.858
94
0.127**
(1.97)
-0.173***
(-2.58)
237.74***
0.027
0.243
0.999
94
a statistically significant positive sign, suggesting that
banking security as well as independence from government
control exerts positive impact on Islamic banks’ profitability.
The more banks are controlled by the government, the less
free they are to engage in essential financial activities that
facilitate private sector led economic growth.
Second, it is also interesting to examine the persistence of
the explanatory variables over time. To do so, we lag all
the explanatory variables by one period and repeat Eq.
(2). The results are presented in Table 6. As can be seen,
the coefficients of the baseline variables remain stable as
in the baseline regression model. It is also worth noting
that the coefficient of the LNTA variable is now significant
in four out of the six models estimated, while LLP/TL has
consistently exhibit negative and significant impact on
Islamic banks’ profitability levels. The empirical findings
suggest that the impact of capitalization (EQASS) is
positive in all of the regression models estimated.
It is also interesting to note that the impact of overall
economic freedom (OVER_FREE) is now positive and
significant. The empirical findings seem to support the
notion that economic freedom is a key to the creation
of an environment that allows a virtuous cycle of
entrepreneurship, innovation, and sustained economic
growth and development to flourish. Furthermore,
economies with higher levels of economic freedom are
likely to enjoy higher living standards (Holmes et al.
2008). Holmes et al. (2008) points out that a higher level
of economic freedom is associated with a higher level of
per capita GDP. They also suggest that countries which
increase their levels of freedom tend to experience faster
growth rates and the freest economies also have lower rates
of unemployment and inflation. However, the coefficient
Islamic banking and finance – Essays on corporate finance, efficiency and product development
The nexus between economic freedom and Islamic bank performance
Table 5. Panel Generalized Methods of Moments (GMM) regression results.
ROA jt = β 0 + β 1 LOANS / TA jt + β 2 LNTA jt + β 3 LLP / TL jt + β 4 NIE / TA jt + β 5 EQASS jt + β 6 LNGDPt + β 7 INFLt + β 8CR 3 t
+ β 9 Z - SCORE t + β 9OVER _ FREE t + β 10 BUSI _ FREE t + β 11 MONE _ FREE t + β 12 FINA _ FREE t + β 13CORR _ FREE t + e jt
The notation used in the table below is defined as follows: LOANS/TA is used as a proxy measure of loans intensity, calculated as total
loans divided by total assets; LNTA is a proxy measure of size, calculated as a natural logarithm of total bank assets; LLP/TL is a measure
of bank risk calculated as the ratio of total loan loss provisions divided by total loans; NII/TA is a measure of bank diversification towards
non interest income, calculated as total non-interest income divided by total assets; NIE/TA is a proxy measure for costs, calculated as
non-interest expenses divided by total assets; EQASS is a measure of capitalization, calculated as book value of shareholders equity as a
fraction of total assets; LNGDP is natural log of gross domestic products; INFL is the inflation rate; OVER_FREE is the overall economic
freedom index; BUSI_FREE is the business freedom index; MONE_FREE is the monetary freedom index; FINA_FREE is the financial
freedom index; CORR_FREE is the freedom from corruption index.
ONE STEP SYS-GMM
CONSTANT
(1)
(2)
4.095
(0.45)
4.410
(0.92)
(3)
(4)
(5)
10.278
(1.45)
-5.241
(-1.23)
6.055
(1.44)
0.163***
(3.04)
-0.001
(-0.05)
-0.132
(-0.48)
-0.038
(-1.37)
-0.911**
(-2.27)
0.043*
(1.77)
0.111**
(2.21)
0.012
(0.85)
0.544**
(2.07)
-0.017
(-0.55)
-0.725*
(-1.76)
0.057***
(2.60)
0.194***
(3.07)
0.009
(0.48)
-0.064
(-0.27)
-0.040
(-1.42)
-0.975**
(-2.07)
0.068***
(3.44)
Bank Characteristics
ROA t-1
LOANS/TA
LNTA
LLP/TL
NIE/TA
EQASS
0.151***
(2.98)
-0.001
(-0.05)
0.048
(0.16)
-0.041
(-1.26)
-0.827*
(-1.71)
0.050**
(2.50)
0.140**
(2.44)
-0.002
(-0.13)
0.022
(0.11)
-0.042
(-1.33)
-0.823**
(-2.06)
0.050**
(2.13)
Economic Freedom
OVER_FREE
BUSI_FREE
-0.024
(-0.26)
MONE_FREE
-0.023
(-0.73)
FINA_FREE
-0.075
(-1.39)
CORR_FREE
Wald χ2
AR(1) p-value
AR(2) p-value
Sargan p-value
No. of Observationst-1
55.82
0.556
0.470
0.126
129
59.30
0.584
0.558
0.128
129
66.63
0.654
0.334
0.148
129
0.052**
(2.40)
50.43
0.934
0.859
0.151
129
-0.058*
(-1.62)
93.00
0.505
0.422
0.681
129
Values in parentheses are z-statistics.
***, **, and * indicates significance at 1, 5, and 10% levels, respectively.
of the variable is insignificant. Similarly, the empirical
findings seem to suggest that the coefficient of the BUSI_
FREE is significantly related to the profitability of Islamic
banks operating in the MENA banking sectors. However, it
can also be observed from columns 5 and 6 of Table 6 that
financial freedom and freedom from corruption loses their
explanatory power.
Eds. Hatem A. El-Karanshawy et al.
Third, we restrict our sample to banks with more than
three years of observations. All in all, the results remain
qualitatively similar in terms of directions and significance
levels. Finally, we address the effects of outliers in the
sample by excluding the top and bottom 1% of the
sample. The results continued to remain robust in terms of
directions and significance levels.8
105
Sufian et al.
Table 6. Panel Generalized Methods of Moments (GMM) regression results.
ROA jt = β 0 + β 1 LOANS / TA jt + β 2 LNTA jt + β 3 LLP / TL jt + β 4 NIE / TA jt + β 5 EQASS jt + β 6 LNGDPt
+ β 7 INFLt + β 8CR3t + β 9 Z - SCORE t + β 9OVER _ FREE t + β 10 BUSI _ FREE t + β 11 MONE _ FREE t
+ β 12 FINA _ FREE t + β 13CORR _ FREE t + e jt
The notation used in the table below is defined as follows: LOANS/TA is used as a proxy measure of loans intensity, calculated as total
loans divided by total assets; LNTA is a proxy measure of size, calculated as a natural logarithm of total bank assets; LLP/TL is a measure
of bank risk calculated as the ratio of total loan loss provisions divided by total loans; NII/TA is a measure of bank diversification towards
non interest income, calculated as total non-interest income divided by total assets; NIE/TA is a proxy measure for costs, calculated as
non-interest expenses divided by total assets; EQASS is a measure of capitalization, calculated as book value of shareholders equity as a
fraction of total assets; LNGDP is natural log of gross domestic products; INFL is the inflation rate; OVER_FREE is the overall economic
freedom index; BUSI_FREE is the business freedom index; MONE_FREE is the monetary freedom index; FINA_FREE is the financial
freedom index; CORR_FREE is the freedom from corruption index.
ONE STEP SYS-GMM
(1)
CONSTANT
1.290
(0.58)
(2)
(3)
-10.061*
(-1.73)
-6.400**
(-2.14)
(4)
1.873
(0.21)
(5)
-4.750
(-1.25)
(6)
0.559
(0.21)
Bank Characteristics
ROA t-1
ROA t-2
LOANS/TA
LOANS/TAt-1
LNTA
LNTAt-1
LLP/TL
LLP/TL t-1
NIE/TA
NIE/TA t-1
EQASS
EQASS t-1
0.530***
(3.34)
-0.156**
(-1.93)
0.017
(0.57)
-0.027
(-0.89)
-2.492*
(-1.64)
2.582*
(1.63)
-0.078***
(-2.84)
-0.019
(-0.67)
-0.375
(-1.09)
0.269
(0.82)
0.077**
(2.42)
0.009
(0.24)
0.405***
(2.59)
-0.182***
(-3.45)
0.006
(0.22)
-0.003
(-0.12)
-1.465
(-0.98)
1.910
(1.26)
-0.066**
(-2.35)
-0.013
(-0.73)
-0.044
(-0.16)
-0.020
(-0.05)
0.076***
(2.89)
0.027
(0.77)
0.492**
(2.55)
-0.199***
(-3.46)
0.024
(0.70)
-0.019
(-0.64)
-2.223*
(-1.84)
2.649**
(2.05)
-0.061**
(-2.31)
-0.026
(-1.35)
-0.178
(-0.56)
0.172
(0.39)
0.074**
(2.54)
0.030
(0.81)
0.456***
(2.75)
-0.167***
(-2.67)
0.012
(0.45)
-0.011
(-0.43)
-2.686*
(-1.84)
2.854**
(1.98)
-0.076***
(-3.53)
-0.014
(-0.61)
-0.096
(-0.30)
0.032
(0.07)
0.087***
(2.98)
0.013
(0.38)
0.384**
(2.14)
-0.149**
(-2.23)
0.012
(0.51)
-0.003
(-0.11)
-1.361
(-0.62)
1.814
(0.87)
-0.069***
(-3.87)
-0.001
(-0.04)
-0.090
(-0.31)
-0.071
(-0.15)
0.085***
(3.29)
0.022
(0.68)
0.484***
(3.11)
-0.160**
(-2.31)
0.015
(0.52)
-0.011
(-0.40)
-3.135*
(-1.77)
3.334*
(1.85)
-0.082***
(-4.38)
-0.014
(-0.55)
-0.124
(-0.37)
0.101
(0.24)
0.087***
(2.97)
0.022
(0.56)
Economic Freedom
OVER_FREE
BUSI_FREE
MONE_FREE
FINA_FREE
CORR_FREE
0.113*
(1.61)
0.057**
(2.03)
-0.026
(-0.34)
0.031
(1.06)
-0.023
(-0.80)
(Continued)
106
Islamic banking and finance – Essays on corporate finance, efficiency and product development
The nexus between economic freedom and Islamic bank performance
Table 6. (Continued)
Wald χ2
AR(1) p-value
AR(2) p-value
Sargan p-value
No. of Observationst-2
567.29***
0.049
0.465
0.219
111
1685.09***
0.057
0.152
0.229
98
2902.11***
0.047
0.202
0.306
98
3870.54***
0.035
0.097
0.061
98
1471.55***
0.099
0.378
0.049
98
1070.45***
0.038
0.181
0.114
98
Values in parentheses are z-statistics.
***, **, and * indicates significance at 1, 5, and 10% levels, respectively.
5. Concluding remarks
By using an unbalanced bank level panel data, the study
attempts to examine the impact of economic freedom on
the performance of Islamic banks operating in the MENA
banking sectors during the period 2000–2008. We find that
the larger, more diversified, and better capitalized banks are
relatively more profitable. The empirical findings seem to
suggest that efficient cost management is a prerequisite to
improve the profitability of Islamic banks operating in the
MENA banking sectors. Similarly, we find that higher credit
risk has negative and significant influence on the profitability
of Islamic banks operating in the MENA banking sectors. The
results suggest economic conditions exert negative impact on
Islamic banks’ profitability levels when we control for overall
economic freedom, monetary freedom, and freedom from
corruption. We also find that the level of inflation has positive
impact when we control for monetary and financial freedom.
The findings from this study seem to suggest that greater
financial and business freedom exerts positive impacts on
the profitability of Islamic banks operating in the MENA
banking sectors. The positive coefficient of the financial
freedom variable indicate that higher (lower) freedom on
the activities that Islamic banks could undertake increases
(reduces) their profitability, which is consistent with the
view that less regulatory control allows banks to engage in
various activities enabling banks to exploit economies of
scale and scope and generate income from non-traditional
sources. Furthermore, higher freedom on entrepreneurs
to start businesses is conducive to job creation and
consequently increases Islamic banks’ profitability.
Interestingly, the impact of monetary freedom is negative
implying that higher (lower) monetary policy independence
reduces (increases) Islamic banks’ profitability, providing
support to the benefits of government interventions. In
essence, although price stability without intervention is an
ideal state for a free market, the empirical findings from this
study clearly lend support to the benefits of government
interventions in the markets.
The findings of this study present considerable policy
relevance. In view of the increasing competition
attributed to the more liberalized banking sector, bank
managements as well as the policymakers will be more
inclined to find ways to obtain the optimal utilization
of capacities as well as making the best use of their
resources, so that these resources are not wasted during
the production of banking products and services. The
findings pointed to the need for bankers to choose flexible
operating environment and economic system favouring
the rapid development of a vibrant banking sector
to maximize their performance. From the regulatory
perspective, the performance of the Islamic banking
Eds. Hatem A. El-Karanshawy et al.
sector will be based on their operating performance.
Therefore, policy direction is expected to point towards
enhancing the resilience and performance of the banking
institutions with the aim of intensifying the robustness
and stability of the Islamic banking sector.
Within MENA countries, enhancing economic freedom is of
an importance policy if the region is to attract more financial
investments, improve weak financial infrastructure and
enhance the banking system performance. The current state
of the financial sector in MENA which mainly controlled
by state owned banks and dominate banking activities (up
to 95% of assets in several countries in the MENA region)
resulting in poor services, high costs, and weak financing of
new investments and trade. As the MENA region competes for
economic benefits for its citizens in the new global economy,
it is important that the policy makers in these countries
to improve their quality of governance and transparency;
to promote a legal system that protects shareholders and
creditors rights; and to enhance their economic freedom.
Notes
1. The basic tenets and principles of Islamic banking
are built upon the avoidance of usury (riba’) and
the prohibition of impermissible activities as clearly
mentioned in the Quran, the Islam’s holy book and
the traditions of Prophet Muhammad (sunnah):
“Believers! Do not consume riba’, doubling and
redoubling…” (3.130); “God has made buying and
selling lawful and riba’ unlawful…” (2:274).
2. The estimates of the number of Islamic financial
institutions vary considerably between institutions.
For instance, the International Monetary Fund
(IMF) estimate that the number of Islamic financial
institutions has increased to more than 300, while the
Association of Islamic Banking Institutions Malaysia
(AIBIM) suggests that there are around 486 Islamic
financial institutions around the world.
3. MENA stands for Middle East and North Africa region
comprises of Algeria, Bahrain, Egypt, Iran, Iraq, Israel,
Jordan, Kuwait, Lebanon, Libya, Morocco, Oman, Qatar,
Saudi Arabia, Syria, Sudan, Tunisia, UAE, and Yemen.
4. Islam has laid down some principles and prescribed
certain limits for the economic activity of man so
that the entire pattern of production, exchange, and
distribution of wealth may conform to the Islamic
standard of justice and equity.
5. The Economic Freedom Index is released by The
Heritage Foundation and The Wall Street Journal.
6. Islamic banks income must not be uncontaminated by
usury (riba’). Thus, in the case of the Islamic banking
sector, it is reasonable to assume that the interest rate
to be the profit rate.
107
Sufian et al.
7. Following Garcia-Herrero et al. (2009) among others,
we instrument for all regressors. The macroeconomic
characteristics are treated as exogenous (see among
others Baum et al. 2010).
8. To conserve space, we do not report the regression
results in the paper but are available upon request.
References
Abbasoglu, O.F., Aysan, A.F. and Gunes, A. (2007)
“Concentration,
Competition,
Efficiency,
and
Profitability of the Turkish Banking Sector in the PostCrisis Period,” Banks and Bank Systems, 2(3), 106–115.
Adkins, L., Moomaw, R. and Savvides, A. (2002)
“Institutions, Freedom and Technical Efficiency,”
Southern Economic Journal, 69(1), 92–108.
Albertazzi, U. and Gambacorta, L. (2009) “Bank Profitability
and the Business Cycle,” Journal of Financial Stability,
5(4), 393–309.
Altman, M. (2008) “How Much Economic Freedom is
Necessary for Economic Growth? Theory and Evidence,”
Economics Bulletin, 15(2), 1–20.
Athanasoglou, P.P., Brissimis, S.N. and Delis, M.D. (2008)
“Bank Specific, Industry Specific and Macroeconomic
Determinants of Bank Profitability,” Journal of International
Financial Markets, Institutions and Money, 18(2), 121–136.
Bashir, A.H.M. (1999) “Risk and Profitability Measures
in Islamic Banks: The Case of Two Sudanese Banks,”
Islamic Economic Studies, 6(2), 1–24.
Bashir, A.H.M. (2001) “Assessing the Performance of
Islamic Banks: Some Evidence from the Middle East,”
paper presented at the American Economic Association
Annual Meeting, New Orleans, Louisiana.
Baum, C.F., Caglayan, M. and Talavera, O. (2010) “Parlia­
mentary Election Cycles and the Turkish Banking Sector,”
Journal of Banking and Finance, 34(11), 2709–2719.
Ben Naceur, S. and Goaied, M. (2008) “The Determinants
of Commercial Bank Interest Margin and Profitability:
Evidence from Tunisia,” Frontiers in Finance and
Economics, 5(1), 106–130.
Berger, A.N. (1995) “The Relationship Between Capital
and Earnings in Banking,” Journal of Money, Credit and
Banking, 27(2), 432–456.
Berger, A.N. and DeYoung, R. (1997) “Problem Loans
and Cost Efficiency in Commercial Banks,” Journal of
Banking and Finance, 21(6), 849–870.
Berger, A.N. and Humphrey, D.B. (1997) “Efficiency
of Financial Institutions: International Survey and
Directions for Future Research,” European Journal of
Operational Research, 98(2), 175–212.
Berger, A.N., Bonime, S.D., Covitz, D.M. and Hancock, D.
(2000) “Why Are Bank Profits So Persistent? The Roles
of Product Market Competition, Information Opacity
and Regional Macroeconomic Shocks,” Journal of
Banking and Finance, 24(7), 1203–1235.
Berger, A.N., Hanweck, G.A. and Humphrey, D.B. (1987)
“Competitive Viability in Banking: Scale Scope
and Product Mix Economies,” Journal of Monetary
Economics, 20(3), 501–520.
108
Berggren, N. (2003) “The Benefits of Economic Freedom: A
Survey,” Independent Review, 8(2), 193–211.
Blundell, R. and Bond, S. (1998) “Initial Conditions and
Moment Restrictions in Dynamic Panel Data Models,”
Journal of Econometrics, 87(1), 115–143.
Boyd, J. and Runkle, D. (1993) “Size and Performance of
Banking Firms: Testing the Predictions Theory,” Journal
of Monetary Economics, 31(1), 47–67.
Boyd, J.H. and De Nicolo, G. (2006) “The Theory of Bank
Risk Taking and Competition Revisited,” The Journal of
Finance, 60(3), 1329–1343.
Button, K.J. and Weyman-Jones, T.G. (1992) “Ownership
Structure, Institutional Organization and Measured
X-Efficiency,” The American Economic Review, 82(2),
439–445.
ihák, M., Maechler, A., Schaeck, K. and Stolz, S. (2009)
“Who Disciplines Bank Managers?” Working Paper,
International Monetary Fund.
De Haan, J. and Siermann, C.L.J. (1998) “Further Evidence
on the Relationship between Economic Freedom and
Economic Growth,” Public Choice, 95(3–4), 363–380.
De Haan, J. and Sturm, J.E. (2000) “On the Relationship
between Economic Freedom and Economic Growth,”
European Journal of Political Economy, 16(2), 215–241.
DeNicolo, G., Bartholomew, P., Zaman, J. and Zephirin, M.
(2003) “Bank Consolidation, Internationalization, and
Conglomeration: Trends and Implications for Financial
Risk,” Financial Markets, Institutions & Instruments,
13(4), 173–217.
DeYoung, R. and Rice, T. (2004) “Non-Interest Income
and Financial Performance at US Commercial Banks,”
Financial Review, 39(1), 101–127.
Dogan, E. and Fausten, D.F. (2003) “Productivity and
Technical Change in Malaysian Banking: 1989–1998,”
Asia-Pacific Financial Markets, 10(2–3), 205–237.
Eichengreen, B. and Gibson, H.D. (2001) “Greek Banking
at the Dawn of the New Millennium,” Discussion Paper,
Center of Economic and Policy Research.
El Qorchi, M. (2005) “Islamic Finance Gears Up”, Finance
and Development, 42(4), 46–49.
El-Gamal and Inanoglu (2004) “Islamic Banking in Turkey:
Boon or Bane for the Financial Sector,” Proceedings of
the Fifth Harvard University Forum on Islamic Finance,
Cambridge: Center for Middle Eastern Studies, Harvard
University.
El-Gamal, M.A. and Inanoglu, H. (2005) “Efficiency and
Unobserved Heterogeneity in Turkish Banking” Journal
of Applied Econometrics, 20(5), 641–664.
Garcia-Herrero, A., Gavila, S. and Santabarbara, D. (2009)
“What Explains the Low Profitability of Chinese Banks?”
Journal of Banking and Finance, 33(11), 2080–2092.
Gilbert, R. (1984), “Bank Market Structure and Competition
– A Survey,” Journal of Money, Credit and Banking,
16(4), 617–645.
Goddard, J., Molyneux, P. and J. Wilson (2004) “Dynamic
of Growth and Profitability in Banking,” Journal of
Money, Credit and Banking, 36(6), 1069–1090
Islamic banking and finance – Essays on corporate finance, efficiency and product development
The nexus between economic freedom and Islamic bank performance
Goddard, J., Molyneux, P. and Wilson, J.O.S. (2001)
European Banking: Efficiency, Technology and Growth.
New York: John Wiley and Sons.
Miller, S.M. and Noulas, A. (1997) “Portfolio Mix and
Large Bank Profitability in the USA,” Applied Economics,
29(4), 505–512
Hassan, M.K and Hussein, K.A. (2003) “Static and Dynamic
Efficiency in the Sudanese Banking System,” Review of
Islamic Economics, 14, 5–48.
Molyneux, P., Altunbas, Y. and Gardener, E.P.M. (1996)
Efficiency in European Banking. John Wiley & Sons,
Chichester.
Hassan, M.K. (2006), “The X-Efficiency of Islamic Banks,”
Islamic Economic Studies, 13(1–2), 49–77.
North, D., Thomas, R.P., (1973) The Rise of the Western
World: A New Economic History. Cambridge: Cambridge
University Press.
Hassan, M.K. and Bashir, A.H.M. (2003) “Determinants of
Islamic Banking Profitability,” Paper Presented at the
10th ERF Annual Conference, 16th–18th December,
Morocco.
Hawtrey, K. and Liang, H. (2008) “Bank Interest Margins
in OECD Countries,” The North American Journal of
Economics and Finance, 19(3), 249–260.
Heckelman, J.C. (2000) “Economic Freedom and Economic
Growth: A Short-Run Causal Investigation,” Journal of
Applied Economics, 3(1), 71–91.
Pasiouras, F. and Kosmidou, K. (2007) “Factors Influencing
the Profitability of Domestic and Foreign Commercial
Banks in the European Union,” Research in International
Business and Finance, 21(2), 222–237.
Perry, P. (1992) “Do Banks Gain or Lose from Inflation,”
Journal of Retail Banking, 14(2), 25–40.
Powell, B. (2003) “Economic Freedom and Growth: The
Case of the Celtic Tiger,” Cato Journal, 22(3), 431–448.
Heckelman, J.C. and Knack, S. (2009) “Aid, Economic
Freedom and Growth,” Contemporary Economic Policy,
27(1), 46–53.
Rivard, R.J. and Thomas, C.R. (1997) “The Effect of
Interstate Banking on Large Bank Holding Company
Profitability and Risk,” Journal of Economics and
Business, 49(1), 61–76.
Heckelman, J.C. and Stroup, M.D. (2000) “Which
Economic Freedoms Contribute to Growth?” KYKLOS,
53(4), 527-544.
Sarker, M.A.A. (1999) “Islamic Banking in Bangladesh:
Performance, Problems, and Prospects,” International
Journal of Islamic Financial Services, 1(3), 15–36.
Hirtle, B.J. and Stiroh, K.J. (2007) “The Return to Retail
and the Performance of US Banks,” Journal of Banking
and Finance, 31(4), 1101–1133.
Saunders, A., Strock, E., and Travlos, N.G. (1990)
“Ownership Structure, Deregulation, and Bank Risk
Taking,” Journal of Finance, 45(2), 643–654.
Holmes, K.R., Feulner, E.J., and O’Grady, M.A. (2008) 2008
Index of Economic Freedom. The Heritage Foundation:
Washington, D.C.
Short, B.K. (1979) “The Relation Between Commercial
Bank Profit Rate and Banking Concentration in Canada,
Western Europe and Japan,” Journal of Banking and
Finance, 3(3), 209–219.
Hussein, K.A. (2003) “Operational Efficiency in Islamic
Banking: The Sudanese Experience,” Working Paper,
Islamic Research and Training Institute (IRTI), Islamic
Development Bank.
Isik, I. and Hassan, M.K. (2003) “Efficiency, Ownership and
Market Structure, Corporate Control and Governance
in the Turkish Banking Industry,” Journal of Business
Finance and Accounting, 30(9–10), 1363–1421.
Kasman, A., Tunc, G., Vardar, G. and Okan, B. (2010)
“Consolidation and Commercial Bank Net Interest
Margins: Evidence from the Old and New European
Union Members and Candidate Countries,” Economic
Modelling, 27(3), 648–655.
Kennedy, P. (2008) A Guide to Econometrics. Blackwell
Publishing: Malden, Massachusetts.
Kosmidou, K. (2008) “The Determinants of Banks’ Profits in
Greece during the Period of EU Financial Integration,”
Managerial Finance, 34(3), 146–159.
Kosmidou, K. and Zopounidis, C. (2008) “Measurement
of Bank Performance in Greece,” South Eastern Europe
Journal of Economics, 6(1), 79–95.
Kosmidou, K., Pasiouras, F. and Tsaklanganos, A.
(2007) “Domestic and Multinational Determinants
of Foreign Bank Profits: The Case of Greek Banks
Operating Abroad,” Journal of Multinational Financial
Management, 17(2–3), 1–15.
Eds. Hatem A. El-Karanshawy et al.
Spathis, C., Kosmidou, K. and Doumpos, M. (2002)
“Assessing Profitability Factors in the Greek Banking
System: A Multicriteria Methodology,” International
Transactions in Operational Research, 9(5), 517–530.
Staikouras, C. and Wood, G. (2003) “The Determinants
of Bank Profitability in Europe,” Paper presented at
the European Applied Business Research Conference,
Venice, 9–13 June.
Stiroh, K.J. and Rumble, A. (2006) “The Dark Side of
Diversification: The Case of US Financial Holding
Companies,” Journal of Banking and Finance, 30(8),
2131–2161.
Sufian, F. (2009) “Determinants of Bank Efficiency During
Unstable Macroeconomic Environment: Empirical
Evidence from Malaysia,” Research in International
Business and Finance, 23(1), 54–77.
Thakor, A. (1987) “Discussion”, Journal of Finance, 42(3),
661–663.
Tregenna, F. (2009) “The Fat Years: The Structure and
Profitability of the US Banking Sector in the Pre-Crisis
Period,” Cambridge Journal of Economics, 33(4), 609–632.
Williams, B. (2003) “Domestic and International
Determinants of Bank Profits: Foreign Banks in
Australia,” Journal of Banking and Finance, 27(6),
1185–1210.
109
Efficiency of performance of banks in the
Gulf region before, during and after crises
(financial and political)
Abdel Latef Anouze
Department of Management and Marketing, College of Business and Economics, Qatar University, E: [email protected]
Abstract - Only a few cross-country empirical studies have been conducted to measure the perfor­
mance of commercial banks especially before, during, and after crises (financial or political). This study
makes an attempt to fill the gap in the literature by investigating the impacts of crises on Gulf Corporate
Council (GCC) commercial banks’ performance over the period 1997-2007. The rationale behind this
selection is that the GCC countries within this period witnessed two major crises: a political crisis (the
second Gulf war) and a financial crisis (the current global crisis). Clearly, it is important that a manager
recognizes the best bank policy in the face of each crisis that could help both bankers and regulators
in managing these crises. Also, the banking system within GCC countries comprises two different
operating banking systems, Islamic and conventional. As both are operating in similar environments, it
is of interest to examine whether one can make judgments concerning the success of their competitive
strategies, and other management-determined factors by using performance measures.
Two different evaluation methods are computed to measure bank performance: data envelopment
analysis (DEA), and classification and regression tree (CART). The overall results show that
conventional banks perform well during a political crisis, whereas Islamic banks performed better
during the financial crisis. However, this difference is not statistically significant, which means that
GCC commercial banks can be equally competitive when it comes to technical efficiency. Also, there
is no statistically significant relationship between bank geographical location and its efficiency score.
Moreover, the results confirm that large and small size GCC commercial banks are more efficient
than medium-sized banks. Out of the 24 environmental factors included in the study to investigate
the relationship between environmental factors (internal and external) and bank performance,
only 15 factors are considered to be important in predicting fully-efficient banks.
Keywords: data envelopment analysis, classification and regression tree, bank performance, Islamic
bank, GCC countries
1. Introduction
In the light of the on-going international financial crisis,
and the large costs generated for national and international
financial systems, it is essential to assess the performance of
the financial sector in order to avoid the financial disaster
becoming more complicated. Assessing banks’ performance
would help managers examine the success of managerial
decisions that they have taken before, during and after the
crisis; to better understand their management effectiveness,
and it would provide them with valuable reference for
improving their performance. Also, such assessment would
help managers to measure the success of these decisions
compared with those made by their counterparts during
same period. On the other hand, it also helps policy
makers to develop a strong and healthy environment for
the banking sector by examining the impact of economic
and financial reforms that have taken place. Meanwhile,
investors want to see how well a bank is performing before
potentially investing in it. A high stock price alone is not
a sufficient measure to use; they have to see how well a
bank is performing too. Therefore, if a bank is to survive
and succeed, it should learn the status of its efficiency and
how it compares with counterparts in same country or
other countries. Hence, to identify appropriate financial
Cite this chapter as: Anouze A L (2015). Efficiency of performance of banks in the Gulf region before, during and after
crises (financial and political). In H A El-Karanshawy et al. (Eds.), Islamic banking and finance – Essays on corporate
finance, efficiency and product development. Doha, Qatar: Bloomsbury Qatar Foundation
Developing Inclusive and Sustainable Economic and Financial Systems
Anouze
decisions that will achieve better allocation of financial
resources in a more efficient and effective manner, it is
important to assess bank performance at the country and/
or international level. A number of international empirical
studies have been conducted to measure the performance of
the banking sector before, during, and after crises (Mercan
et al. 2003; Jeon and Miller 2004 and 2005). However,
all of these studies among others were carried out prior
to the current global financial crisis. Therefore, this study
attempts to fill the gap in the literature by assessing GCC
commercial banks’ performance before, during and after
the crises to guide bank managers and other stakeholders,
such as policy makers and investors, in their decisions.
There is a substantial body of literature discussing different
methods applied to evaluate the performance of banks
(e.g., Anouze 2010; Fethi and Pasiouras 2010; Berger
and Humphrey 1997). Reviewing 130 studies of the
efficiency of financial institutions, Berger and Humphrey
(1997) classified these methods according to the technical
approach employed into parametric, such as the stochastic
frontier approach (SFA), and nonparametric, such as data
envelopment analysis (DEA). Application of these methods
alone to evaluate banks’ performance determines efficiency
scores but gives no details of factors related to inefficiency,
especially if these factors are in the form of non-numeric
variables such as the operating style of the banking sector
(Emrouznejad and Anouze 2010). This study proposes a
comprehensive performance evaluation framework based
on managerial, financial, and macroeconomic indicators
to measure and predict banks’ performance. It allows
exploration and discovery of meaningful, previously
hidden information from given data. It integrates Data
Envelopment Analysis (DEA) with the Classification and
Regression Tree (CART) technique. DEA is a nonparametric
method for measuring the performance of Decision Making
Units (DMUs) such as banks, hospitals, universities, or
services. It groups data into inputs and outputs to produce
a productive efficiency frontier against which an individual
bank or the banks of an entire country can be benchmarked.
Input variables within the DEA context are resources to be
minimized, while output variables are product or services
to be maximized in order to achieve a high efficiency score.
The DEA efficiency score is a relative measure, which is
derived for each bank from the DEA based on the quality of
transforming the inputs into outputs. CART, on other hand,
is a nonparametric data-mining technique which allows
meaningful information to be explored and discovered from
a given data set. Unlike the DEA model, in which each case
needs to be compared, CART produces results that can easily
be applied to determine the efficiency of a bank. A unique
feature of CART is that it illustrates the data in the form of a
decision tree so that the results can be presented in the form
of diagrams that are easy to understand. Integration of the
two techniques would help stakeholders to assess, predict
and identify the banks that are most likely to be troubling
or, on the other hand, outperforming. Hence, stakeholders
would have an overall understanding of banks’ performance
and, consequently, better improvement policies could be
developed for unsuccessful banks.
2. Literature review: Banking performance
Although there is a huge volume of published research on
banking efficiency, little effort has been made to conduct
112
studies of the impact of financial or political crises on
banking performance. To our knowledge, this is the first
study to explore the combined effect of financial and
political crises on banking performance. However, our
work contributes and relates closely to several branches of
literature on bank performance, including studies of the
impact of the financial crisis, bank health, and financial
regulations on banking performance.
Few research studies have explored the impacts of the
current financial crisis on of bank performance. Xiao
(2009) used qualitative and quantitative tools to examine
the performance of French banks during 2006–2008.
The findings showed that French banks were not immune
but proved relatively elastic to the global financial crisis.
Beltratti and Stulz (2009) studied the bank stock return
across the world during the period from the beginning of
July 2007 to the end of December 2008; they found that
large banks with more deposit financing at the end of 2006
exhibited significantly higher returns during the crisis.
Cornett, McNutt and Tehranian (2010) analyzed the internal
corporate governance mechanisms and the performance
of US banks before and during the financial crisis; they
found that the largest banks faced the largest losses during
the crisis. Dietrich and Wanzenried (2011) examined
how bank-specific characteristics, industry-specific, and
macroeconomic factors affected the profitability of Swiss
commercial banks over the period from 1999 to 2009; their
results provide some evidence that the financial crisis did
have a significant impact on banks’ profitability.
These studies, among others, used a regression analysis
and limited bank performance to a single indicator—such
as profit, capital, and deposit to assets ratio—to measure
bank performance during or after the crisis. Although
regression analysis is a useful tool, it tells nothing about
how to improve the performance, nor which is the best
practice during or after the crisis. Also, it only counts for
a single indicator, whereas banks could aim to maximize
more than one indicator during their financial transactions.
Furthermore, none of these studies have investigated the
performance of the GCC banking sector during the financial
crisis.
Other researchers paid particular attention to the impact
of financial regulation on bank performance. Policy
makers introduce such regulations to develop a healthy
environment that increases competition and improves
banking sector efficiency. Although are numerous studies
have examined the impact of financial regulations on banks’
performance, the overall impact of financial regulation is
ambiguous. Huang, Hsiao, Cheng and Change (2008);
Brissimis, Delis and Papankiolaou (2008); KoutsomanoliFilippaki, Margaritis and Staikouras (2009); Hsiao, Chang,
Cianci and Huang (2010); and Zhao, Casu and Ferrari
(2010), among others, found that deregulation improves
banking performance and stimulates competition in the
financial market. In contrast, findings from other studies
such as those of Fukuyama and Weber (2002); Halkos and
Salamouris (2004); Park and Weber (2006): and Fu and
Heffernan (2009) show a decline in bank efficiency during
a period of financial reform.
Finally, others researchers studied the real effects on bank
performance of deterioration in bank health or competition
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Efficiency of performance of banks in the Gulf region before, during and after crises (financial and political)
during the financial crisis. Almeida, Campello, Laranjeira,
and Weisbenner (2009) and Duchin, Ozbas and Sensoy
(2010) studied the effect of the recent financial crisis on
corporate investment. Their results show that corporate
investment declined significantly following the onset of
the crisis. Berger and Bouwman (2010) examined the
effect of pre-crisis bank capital ratios on banks’ ability to
survive financial crises, market shares, and profitability
during the crises. Their findings show that capital helps
banks of all sizes during banking crises; possession of
higher capital helped banks to increase their probability
of survival, market shares, and profitability. Gryglewicz
(2011) studied the impact of both liquidity and solvency
concerns on corporate finance, and showed how changes
in solvency affect liquidity, and also how liquidity concerns
affect solvency through capital structure choice.
These studies, among others, provide a comprehensive
examination of the effects of a financial crisis on bank
efficiency (see table A1 in appendix for summary of previous
literature). Evaluation of this literature that presented in
table helped us to select the appropriate variables to be
included in our analysis. However, the impact of the crisis on
bank performance has not yet been fully analyzed. Moreover,
most of the reviewed research made use of statistical models
to study the impact of the financial crisis on firms. Statistical
models make some assumptions about statistical distribution
or propensities of the data, but most financial data do
not meet the statistical requirements of certain statistical
models; also, statistical tests are sensitive to sample size.
On the other hand, an operations research technique makes
no assumptions about statistical distribution, and it is more
accurate when testing complex or large samples (Demyank
and Hasan 2010). Therefore, use of such a technique to study
the impacts of the financial crisis on banking performance
tends to be more appropriate in practical situations.
3. Methodology
Data Envelopment Analysis (DEA)
DEA is a nonparametric relative performance evaluation
method developed by Charnes, Cooper and Rhodes in
(1978) considering constant return to scale (CRS). The CRS
model compares banks’ performance based only on overall
efficiency assuming constant returns to scale; however, it
ignores the fact that different banks could be operating
at different scales. To overcome this drawback Banker,
Charnes and Cooper (1984) introduced the variable returns
to scale (VRS) model that is similar to the CRS Model, but it
ensures that an efficient bank is only benchmarked against
banks of similar size, while in the CRS model a bank may be
benchmarked against banks which are substantially larger
or smaller than it. Subsequently, the original DEA models
(CRS and VRS) have undergone many modifications and
developments. Most of these developments occurred when
the deficiencies of the original model were exposed during
its application to solving real life problems (Thompson,
Singleton, Thrall and Smith 1986).
To introduce the DEA-VRS model, assume there are n banks
(j = 1,…, n) using m inputs (xij i = 1,…m) and producing
s outputs (yrj, j = 1,…s). DEA measures the technical
efficiency of bank j0 compared with n peer group of banks,
as illustrated in model 1a and 1b.
Eds. Hatem A. El-Karanshawy et al.
Model 1a: Standard Input
Oriented DEA – VRS
Model 1b: Standard Output
Oriented DEA – VRS
Min θ
subject
to
Max φ
subject
to
n
∑l x
j =1
n
j
ij
≤ θ x ij0 ; ∀i
n
∑l x
j =1
n
j
ij
≤ x ij0 ; ∀i
∑l
j
y rj ≥ y rj0 ; ∀r
∑l
j
y rj ≥ φ y rj0 ; ∀r
∑l
j
=1
∑l
j
=1
j =1
n
j =1
l j ≥ 0; ∀j, θ free
j =1
n
j =1
l j ≥ 0; ∀j,φ free
To reach to the CRS from model 1a and 1b one can remove
n
the following constraint from the above model ∑ j=1 l = 1
j
In the input and output oriented DEA models (model 1a and
1b, respectively) bankj0 is assessed under variable returns
to scale, where the efficiency of bankj0 is the optimal value
of ? in Model 1a and 1/? in Model 1b (Thanassoulis 2001).
One of the key concerns when we have a variable that takes
positive values for some banks and negative values for
others is that its absolute value should rise or fall for the
bank to improve its performance, depending on whether
the bank concerned has a positive or negative value on that
variable (Emrouznejad, Anouze, and Thanassoulis 2010a).
For example, in the case of an output variable, if the bank
has a positive value (profit) the output should rise to
improve further but it should fall in absolute value as long as
it continues to be negative (loss). To overcome this problem
Emrouznejad and Anouze (2009) and Emrouznejad et al.
(2010a and 2010b) treated each variable that has positive
value for some banks and negative for others as consisting
of the sum of two variables, and proposed a semi-oriented
radial model (SORM).
Classification and Regression Tree (CART)
CART is the commonly used decision tree in data mining
that was developed by Breiman, Friedman, Olshen and
Stone (1984) and further improved by Ripley (1996).
In principle, CART is similar to regression analysis since
both are used for prediction. However, CART has some
advantages over the regression model:
• A model generated by a CART is easier to understand
and relatively simple to interpret for non-statisticians
(Breiman et al. 1984; Torgo 1997; Han and Kamber
2001)
• There are no assumptions to be made regarding the
underlying distribution of values of the predictor
variables as it is a nonparametric technique
• CART can handle numerical data, as well as categorical,
with either ordinal or non-ordinal structure.
These are important features of CART as they will
eliminate analyst time which would otherwise be spent
determining whether variables are normally distributed
and making transformations if they are not; specifically, it
is important to use CART with DEA since DEA scores are
skewed to one side.
113
Anouze
To validate the results generated by CART, the dataset
is partitioned into two datasets, training and validation
(Han and Kamber 2001). The data then go into two major
phases of process: growth and pruning (Kim and Koehler
1995). In the growth phase, CART constructs a tree from
the training dataset. In this phase, either each leaf node
is associated with a single class, or further partitioning of
the given leaf would result in the number of cases in one
or both subsequent nodes being below some specified
threshold. In the pruning phase the CART generated in
the growth phase is improved in order to avoid over-fitting.
Also in this phase, the CART result is evaluated against the
validation dataset in order to generate a sub-tree with the
lowest error rate.
There are several criteria for measuring CART results. The
predictive accuracy of a CART is commonly measured by
R-squared (average squared error); however, simplicity
and stability are also important measures for a CART.
Simplicity refers to the interpretability of the CART and is
often based on the number of leaves in the CART. Stability
of a CART refers to obtaining similar results for the training
and validation datasets. One way to assess the stability of
the CART can be by comparing the predicted mean value of
the target variable (based on the training dataset) and the
corresponding value for the validation dataset for each rule
of the CART (Han and Kamber 2001).
Proposed methodology (DEA with CART)
Figure 1 illustrates the proposed analysis, that is, DEA
and CART. The DEA stage is to compute the efficiency
score of each bank using DEA. Accordingly, the banks are
categorized into two groups: efficient banks (target = 1)
and inefficient banks (target = 0). In the CART stage the
classified efficiency score (0 or 1) is used as the target of
CART while the environmental (explanatory) variables
is used as an input. However, an accurate CART requires
a large dataset, whereas our sample was limited to 60
banks. Therefore, a new stage was introduced before
the CART stage to increase the original dataset using the
bootstrapping technique. Thus, we randomly selected 60
banks (by replacement) and repeated this sampling 61
times to achieve 3660 banks, so ensuring better accuracy
on the predicted CART results. The 3660 banks were
divided into the two datasets, training and validation, by
the ratio of 7:3 (Zhou and Jiang 2003; Emrouznejad and
Anouze 2010).
Data description and analysis
Banking industries in Gulf state countries
The early banking sector in the GCC countries experienced
much foreign ownership primarily by British banks with
branches extending across all six GCC countries. Local
banks were uncommon as there was insufficient experience.
Subsequently, governments adopted central banking
systems to strength local banks and to eliminate foreign
involvement. Today there are 68 local banks operating in
GCC countries. These banks can be grouped according
to their operating style (mode of running financial
transactions) into two groups: Islamic and conventional
banks. Unlike conventional banks, Islamic banks run their
financial transactions free of interest (i.e., no interest
rate is taken or given against any financial transaction).
Among the 68 local banks, 18 are Islamic banks and 50 are
conventional banks.
Figure 2 illustrates the share of Islamic and conventional
banking assets within each country. Saudi Arabia is the
largest investor in the GCC, holding 32% of the total bank
assets, with nine conventional banks and two Islamic
banks, and had total assets of US$ 239,095 million in
2007. The UAE, with fifteen conventional and five Islamic
banks, and total assets of US$ 224,542 million is the
second largest investor in the area. Bahrain follows, with
nine conventional and six Islamic banks and total assets of
US$ 108,307 million, along with Kuwait, which has seven
conventional and three Islamic banks and total assets of
US$ 108,174 million. Qatar is in fourth position, with four
conventional and two Islamic banks and total assets of US$
56,429 million, which represents only 7% of the total GCC
assets. Finally, Oman has only six conventional banks and
total assets of US$22,259 million, this representing only
3% of the total assets. Although our study aimed to include
all GCC commercial banks, eight banks are excluded on
the basis of lack of availability of data, the remainder
comprising 48 conventional and 12 Islamic banks.
Figure 1. Integrated data envelopment analysis and classification and regression tree.
114
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Efficiency of performance of banks in the Gulf region before, during and after crises (financial and political)
Figure 2. GCC commercial banks: Share of assets, (2007 data).
Table 1. Descriptive analysis of input and output variables (2007).
Inputs/Outputs
Variables
Mean
Std Dev
Min
Max
Inputs US$M
Fixed Assets
Non-earning assets
Deposits
Investments
Loans
Off-balance sheet
Net profit
7.28
21.86
424.11
226.01
256.26
166.87
8.70
24.16
55.08
940.28
525.43
531.37
423.91
21.52
0.03
0.00
0.00
0.00
1.27
0.00
-289.01
413.34
609.61
11,161.00
5,766
7,528.63
4,619.70
195.97
Outputs US$M
Data Description
Selection of proper input and output variables to define and
measure bank performance is always an extremely important
decision (Mercan et al. 2003). It is especially so when using
DEA, as different results may obtain from different sets of
variables. Traditional bank behaviour theories described
banks as accepting deposits from households, and making
loans to investors (Diamond and Dybvig 1983; Diamond
1984; Gorton and Winton 2003). Yet, the change in bank
involvement in markets and bank behaviour during crises
requires a new theory of financial intermediation. In
this study, the input variables include fixed assets, nonearning assets, and deposit, while the output variables are
investments, loans, off-balance sheet, and net profit.
The selected input and output variables varied over
the study period. It can be seen from table 1 that the
minimum value of fixed assets—which is one of the
inputs—is US$ 0.03 million whereas the maximum value
is US$ 413.34 million, with average US$ 7.28 million
and standard deviation of US$ 24.16 million. Similarly
for other variables, for example, the net profit, the
minimum net profit (loss) is US$ −289.01 million and
the maximum value is US$ 195.97 million, with average
of US$ 8.70 million and standard deviation of US$
21.52 million. Given the long time period analysed,
such variation would be expected; nonetheless, since
DEA models are sensitive to observations it is likely that
significant levels of variation would also be found in
banks’ performance.
Eds. Hatem A. El-Karanshawy et al.
Data Envelopment Analysis (DEA)
Overall performance of GCC commercial banks
To study bank performance before, during and after the
crises one grand-frontier (common-frontier) is computed
for all banks in all countries. The grand-frontier provides
a trend in the efficiency of banks, which would not be
available if we computed the efficiency of banks using a
separate frontier for each year. The approach employed,
therefore, provides variations in the efficiency of banks
over both time and space. This comparison across time and
countries is on the same principle as the global frontier
used by Portela and Thanassoulis (2010). A VRS outputoriented model is used to measure banks’ efficiency,
since the CRS model is not possible in technologies
where negative data can exist (Portela, Thanassoulis
and Simpson 2004). The efficiency score obtained for
all GCC commercial banks at the individual bank level is
aggregated to obtain the annual average efficiency scores
of all banks, and this is thenaggregated at country level
and at operating style level.
For better capturing of bank performance during the
crises, the study period (1998–2007) is divided into four
periods:
1. Before the political crisis (second Gulf war, 1998–2002).
2. Political crisis (2003).
3. After political crisis (2004–2006).
4. During the financial crisis (2007).
115
Anouze
It can be seen from able 2 that the overall average efficiency
score is 85.6% for all banks (60 banks); this suggests that
with the same level of inputs and by adopting best practices,
GCC commercial banks can, on average, increase their
outputs by 14.4% (i.e., 100–85.6%). However, the potential
increment in outputs from adopting best practices varies
from bank to bank. In general, GCC commercial banks have
the scope of producing 1.17 times (i.e., 1/0.856) as much
outputs from the same level of inputs.
The literature on technical efficiency provides no consensus
on how efficiency in banking varies through time in response
to market forces (Berger 1993). However, since the study
period covers a long and turbulent time (including the
second Gulf war in 2003 and the 2007 financial crisis),
it is expected that the political and financial crises will
dominate the market forces.
It can also be seen from table 2 that, of the 60 commercial
banks covered in this study, there are ten banks which are
fully-efficient over the entire study period. The overall
results show relatively low average efficiency scores;
nevertheless, it is possible to detect a slight improvement
in the efficiency levels over the study period (+2.2%
between 1998 and 2004). In general, the table shows
that the technical efficiency remains relatively stable over
the period 1998–2003, then improved a little to reach
its highest level (92%) during 2004, while the period
2005–2007 witnessed a volatility of the efficiency score,
reach 79% at the end of the period. The year 2005, that
is, two years after the Gulf war (political crisis) exhibits
a decreased technical efficiency (77%) across all banks
studied. It seems that, over time, banks were wasting more
resources on average, relative to best practice technical
frontiers for the industry.
To find out whether the efficiency scores show a particular
trend during the period 1998–2007, the question is
whether the mean efficiency score increased since 1998. In
fact, Figure 3 shows that the trend of mean efficiency scores
decreased over time. It moved in the same direction over
the period 1998–2002 (before the political crisis), then
declined a little to reach 86% during the second Gulf war.
It fluctuated over the study period 2004–2006, reaching
its highest level in 2004, and deteriorating to the lowest
efficiency level in 2005–2006. The mean efficiency score
further declined in 2007 (the year of the financial crisis)
to reach 79%. Although 2004seemed to be an atypical
year, it is important to note that the performance of GCC
commercial banks varied over the study period.
Table 2. Summary of banks’ technical efficiency.
Bank Code
Efficiency score
Average
No of efficient banks
1998
1999
2000
2001
2002
2003
2004
2005
2006
89
27
88
30
89
26
88
26
87
27
86
27
92
29
77
24
81
26
2007 Average
79
26
85.6
10
Figure 3. Technical efficiency of GCC banks during study period (1998–2007).
116
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Efficiency of performance of banks in the Gulf region before, during and after crises (financial and political)
Table 3. Statistical descriptive of average overall
technical efficiency.
Mean
Std Dev
1998–
2002
2003
2004–
2006
2007
88.02
14.04
86.20
16.50
83.37
20.03
79.30
24.60
Another appropriate way to study the trend is by looking at
the mean and the standard deviation of technical efficiency.
If the banking markets of the GCC became more alike during
the ten-year period under consideration, an increase in
mean technical efficiency and a decrease in the spread of
technical efficiency would be expected. Table 3 shows that
the exact mean technical efficiency was relatively stable for
the period 1998–2003, and then reached its highest level
in 2004. The lowest efficiency score was exhibited during
the 2005, which is two years after the second Gulf crisis,
then fluctuated below the average for the last two years.
The standard deviation was relatively stable for the period
1998–2003, and then reached its lowest level in 2004. The
standard deviation tends to be low when average technical
efficiency is high, and vice versa. These results strongly
support the view that traditional efficiency techniques
based on pooled frontier efficiency scores tend to estimate
the actual efficiency levels of each bank.
Islamic and conventional banks performance
before and during the crises
To compare commercial bank performance based on their
operating style, whether Islamic or conventional, the
efficiency score of all banks at the individual bank level is
aggregated at the operating style level to obtain the annual
average efficiency scores of Islamic and conventional banks,
as illustrated in the figure below.
Figure 4 shows that the Islamic banks outperformed the
conventional banks for the first four years (1998–2001),
and thereafter their performance declined. It reached its
lowest level of the study period (78.6%) by 2003 (second
Gulf war). The performance improved to reach 88% by
2004; however, it was still below the performance of the
conventional banks. Subsequently, the Islamic banks
appeared to be ahead of the GCC commercial banks, with
an average efficiency score of around 89.3%.
For further analysis and comparison between the
performance of Islamic and conventional banks over the
study period, a Mann-Whitney rank sum test was applied.
The Mann-Whitney test, which is an alternative to the
independent group t-test, is a nonparametric (distributionfree) test for testing whether the number of times scores
from one sample are ranked significantly higher than
scores from another, unrelated, sample. Similar to many
non-parametric tests, it uses the ranks of the data rather
than their raw values to calculate the statistic. For this test,
the efficiency score is considered as the group variable and
the bank operating style as the test variable.
The results of the Mann-Whitney test reveal that there is no
significance difference in bank efficiency performance due
to differences in operating style. Hence, the null hypothesis
that the two efficiency scores have the same value of median
is rejected at the 5% level of significance.
Figure 4. Performance of Islamic and conventional banks before and during the financial crisis.
Eds. Hatem A. El-Karanshawy et al.
117
Anouze
Table 4. Mann-Whitney test for 2007 results.
Table 5. Results of Kruskal-Wallis test.
Bank Type
Bank
Location
Sample Mean Mann-Whitney Z-value
Size
Rank U
Islamic
12
Conventional 48
29.6 245.5
34.04
-0.82
Efficiency of commercial banks across GCC
countries
To measure commercial bank technical efficiency across
countries the efficiency score for all banks is aggregated
at country level to get the annual average efficiency scores
for each country. Figure 5 shows that the Kuwaiti banks
outperform other countries banks before and during the
second Gulf crisis. Thereafter, Kuwaiti banks decline,
becoming the worst performers of all GCC countries,
and then become even worse during the financial crisis.
Although there is tough competition between Saudi and UAE
commercial banks as they appear to be following the same
pattern before and during the second Gulf war, UAE banks’
performance deteriorated during and after the financial
crisis.
Qatari banks performed badly before and during the
second Gulf war; however, performance increased rapidly
after the crisis, but declined again during the financial
crisis. The performance of Bahraini and Omani banks
followed the same behaviour before, during and after the
second Gulf war, whereas the Omani banks outperformed
all other GCC commercial banks during the financial
crisis.
For further investigation of the efficiency score across
GCC countries, we adopted the Kruskal-Wallis rank test
(Sueyoshi and Aoki 2001) to examine whether or not scores
vary among countries. The Kruskal Wallis X^2 statistics
are 6.952 (p = 0.224), meaning that there is no statistically
significant relationship between the geographical location
of a bank and its efficiency scores.
Previous analyses have been directed mainly at bank
managers; however, regulators may require different
Bahrain
Kuwait
Oman
Qatar
Saudi Arabia
UAE
N
11
9
5
6
9
20
Mean
Rank
x2
34.32
19.94
39.20
32.00
36.00
28.05
6.952
d.f.
Asymp.
Sig.
5
0.224
information in order to assist them in developing a strong
and healthy environment. Similarly, investors want to know
where to invest their money in a way that will maximize
their return.
Classification and Regression Tree (CART) analysis
The first stage results show the differences in
inefficiency among banks in the six countries. In this
stage an efficient score is treated as a target variable,
while the internal and external environmental factors
are considered as predictors for the CART algorithm.
These factors were identified from the related literature
and include economical, financial and political factors.
Data of 24 factors was collected and tested to determine
the appropriate factors to include in the CART analysis.
Correlation tests showed a high correlation between
numbers of factors. For example, number of branches
and number of employees were highly correlated so we
included only a number of branches to reflect the size
of banks. Also, the price/book value and price earnings
ratio was highly correlated so we included only a price/
book value factor to reflect the size of the stock market
price for each bank. Therefore, fifteen factors were
considered as input factors for the CART algorithm (see
table 6).
All factors as an input of CART algorithm
We built different CART models with a different selection
of input factors for CART with the efficiency score as
target. First, we included all factors as inputs and efficiency
classification as output. Figure 6 shows the importance
of variables. The fifteen environmental factors were
considered to be important in predicting the fully-efficient
banks; only seven of these factors are considered as primary
splitters for the decision tree. Assets structure is the most
important factor (100%), followed by financial strength
(92%) and ROA (91%), whereas operating style, population
density, size, and support rating have low importance. This
suggests that banks should give more importance to their
assets structure as it is one of the most important factors for
the efficiency of banks.
Figure 7 shows the predicated accuracy of the generated
tree:
Figure 5. Bank performance across GCC countries
before and during the financial crisis.
118
Out of 3,660 cases, 1586 cases are actually efficient and
predicted to be efficient; and 2074 cases are inefficient
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Efficiency of performance of banks in the Gulf region before, during and after crises (financial and political)
Table 6. Statistical description of environmental factors.
Variable
Descriptive Statistics
Variable type
Establish Date
Country
Inflation
Population Density
Operating Style
Internal Growth
GDP Growth
Bank Size
Return on Assets (ROA)
Return on Equity (ROE)
Financial Strength
Support Rating
Loan to Deposit Ratio
Market Share
Asset Structure
Categorical
Categorical
Numerical
Categorical
Categorical
Numerical
Numerical
Categorical
Numerical
Numerical
Numerical
Categorical
Numerical
Numerical
Numerical
Minimum
Maximum
1
1.00
3.60
0.70
1.00
0.27
1.90
1
-2.53
-34.18
1.00
1.00
28.50
0.00
0.02
5
6.00
14.00
23.60
2.00
45.15
8.40
3
8.28
33.37
13.00
4.00
1,904.35
8.44
3,534.00
Mean
Std. Deviation
8.61
4.71
14.93
6.34
8.74
1.98
2.76
17.79
7.90
1.53
8.86
4.36
138.76
1.67
209.70
263.59
1.80
518.82
Established date: Banks are grouped according to their established date into 5 groups to capture the age affect: group
5 banks established before 1960; group 4 (1960–1970); group 3 (1970–1980); group 2 (1980–1990) and group 1
(1990–2000). It is expected to have strong positive relationship between bank performance and the established date; the
older are the more efficient.
Country: Although, GCC countries mostly have the same regime, it is expected to have a variation in efficiency score
according to the bank geographical location due to differences in each country regulations.
Inflation: is an indicator of macroeconomic stability, and is directly related to the interest rate levels and, thus, interest
expense and revenue.
Population density: is measured as a ratio of country population to the GCC countries total populations. It is believed that
banks in heavily populated countries are more likely to operate closer to their optimal size than banks in less populated
country. Hence it is easier for bank management to sustain higher efficiency levels in heavily populated areas than in less
populated.
Operating style: to capture the efficiency of Islamic rule and regulations.
Internal growth rate: is calculated as the percentage of retained profits of the year on the equity at the beginning of the
year.
Bank size: is measured by the bank total assets, which classified into three groups hence, the larger banks (with total
assets more than US $15,000 Million), medium size (with total assets between US $5,000 – 15,000 Million) and small
size (total assets less than US $5,000 Million).
Profitability ratios: we measure this variable using return on assets (ROA) and return on equity (ROE).
Financial strength rating: it provides an opinion of a bank’s intrinsic safety, soundness and risk profile (Arab banking and
finance, 2007). It takes a scale from AAA (extremely strong finance and highly attractive operating environment) to D
(extremely weak financial condition and untenable position).
Support rating: it assesses the possibility that the bank will receive enough financial assistance from the government or
private owners in the event of difficulties to enable them to meet their financial obligations. It takes a scale from 1 (very
likely) to 5 (very unlikely) (Arab banking & finance, 2007).
Loan/Deposit: loan-to-deposit ratio is a measure of the extent to which banks are able to transform deposits into loans. It
is mainly used to measure the loan and deposit fund utilization of banks.
Market Share: is the ratio of total deposit of each bank to total deposit of all banks.
Asset structure: is the ratio of tangible assets to the total assets.
and predicted to be so. This means that the accuracy in
predicting the efficient and inefficient banks is 100%,
which represents a high level of confidence. Certain of
the rules extracted for efficient and inefficient banks are
as follows:
4. Rules for efficient banks
Banks are efficient (total of 1586 cases) if:
Eds. Hatem A. El-Karanshawy et al.
1. Financial strength is greater than or equal 4.0, ROA is
greater than or equal to 2.59, and country is less than
4 (122 cases).
2. Financial strength is greater than or equal 4.0, ROA is
greater than or equal to 2.59, country is greater than
or equal to 4, and internal growth is greater than or
equal to 4 (61 cases).
3.Financial strength is greater than or equal to 4.0,
ROA is less than 2.59, internal growth is greater than
119
Anouze
Figure 6. Factors importance in predicting fully-efficient banks.
or equal to 5.66, and established date is greater than
or equal 4 (100 cases).
5. Rules for inefficient banks
Banks are inefficient if:
1. Financial strength is greater than or equal 4.0, ROA is
greater than or equal to 2.59, country is greater than or
equal to 4, and the internal growth is less than 4.44.
2. Financial strength is greater than or equal 4.0, ROA is
less than 2.59, and internal growth is less than 5.66
(854 cases).
External Factors as a Single Input of CART
Algorithm
To investigate the impact of the economic and political factors
(external) on bank performance, CART is drawn by including
only the external factors. All the external environmental
factors are considered to be important in setting the rules for
fully-efficient banks. Operating style and established date are
the most important factor, followed by inflation (89.14%).
Support rating and GDP growth seems to have medium
importance whereas country and total population density
have low. The predictive accuracy of the generated tree is
92%, which represents a high level of confidence. The rules
of efficient and inefficient banks that extracted as follow:
Figure 7. Predicated accuracy of tree.
120
6. Rules for efficient banks
Banks are efficient if:
1. Established date is greater than or equal to 5, GDP
growth is less than 7.95%, inflation is less than 5.72,
country less than or equal to 4, support rating is
greater than or equal to 2.5 but less than or equal to
3.5, and operating style is 1 (61 cases).
2. Established date is greater than or equal to 5, GDP
growth is less than 7.95%, inflation is less than 5.72,
country less than or equal to 4, support rating is greater
than or equal to 2.5 but less than or equal to 3.5, and
operating style is 2 (61 cases that represent 16.7%).
3. Established date is greater than or equal to 5, GDP
growth is less than 7.95%, inflation is less than 5.72,
country less than or equal to 4, support rating is
greater than or equal to 2.5 but less than or equal to
3.5, and operating style is 2 (183 cases).
7. Results and Discussion
The overall technical efficiency for all GCC commercial
banks over the study period is 85.6%. It reaches its highest
level in 2004, which is one year after the second Gulf
crisis. The reason behind this unexpected improvement
in performance could be due to the injection of more
money into the market through policy makers and
regulators deciding to produce more oil in order to avoid
failure of the banking sector or bankruptcy after the Gulf
crisis. Therefore, the banking sector performed well,
until the regulators stopped the injection of funds, when
performance declined to reach its lowest level over the
study period. It is worth noting that the performance of the
banking sector in countries like Saudi Arabia (the largest
oil producer), Qatar (the largest gas producer), and Oman
all improved after the second Gulf war. The performance of
banks in all GCC countries deteriorated during the financial
crisis, except for Omani commercial banks, which reached
their highest performance level during the crisis.
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Efficiency of performance of banks in the Gulf region before, during and after crises (financial and political)
The highest average efficiency score is for the Saudi banks,
at around 89.8%, followed by banks of UAE, which have
an efficiency score of 86.3%. There seems to be tight
competition between Omani and Bahraini commercial
banks, which have average efficiency scores of 85.7% and
85.1%, respectively. Banks operating in Qatar are the least
efficient banks, with a score of around 81.3%.
Although, not really comparable as they differ in terms of
frontier, inputs and output variables, and the study period,
these results are in the line with the research of Al Shammari
(2003), who found that the banks of Saudi Arabia and UAE
are ahead of those in the other GCC countries, while Qatar
and Bahrain have the poorest performing banks.
When the GCC commercial banks efficiency scores are
compared with those of their counterparts in other countries
(e.g., Singapore banks −95%; Japan −87%; Germany −92%;
Peru −98%), the results show that, on average, those of the
GCC banks are lower. Nevertheless, the results are relatively
similar to the average efficiency for banks in industrial
countries like France (84.3%), US (83%), UK (83.9%),
Spain (82–84%), or other developed countries such as
Lebanon (84%) and China (85%) (Ariss 2008; Avkiran
2009; Burki and Niazi 2009; Emrouznejad and Anouze
2010; Emrouznejad and Anouze 2009; Hermes and Nhung
2008; Huang et al. 2010; Ismail, Davidson and Frank 2009;
Koetter 2008)
The results suggest that, even though it is possible to detect a
slight improvement in the overall efficiency scores, there are
marked insignificant differences in bank efficiency levels across
GCC countries. Islamic banks seemed to be more affected by
the Gulf war than were conventional banks, whereas, during
the international financial crisis, Islamic banks seemed to be
the more resistant. This could be due to the level of involvement
of the banks in the international financial institutes: Islamic
banks might have less involved than conventional ones and,
hence, they were less affected. Also, it could be due to the
differences in the relation between bank and clients, which
is based on profit-loss sharing in Islamic banks and based on
fixed rate (interest rate) in conventional banks, so the latter
made less profit compared with Islamic banks
Assets structure, followed by financial strength and ROA
were most important, whereas, operating style, population
density, size, and support rating were found to be of low
importance. Considering only the external factors, the set
of efficiency rules that allow the prediction that a bank is
fully-efficient indicate that it should be old, and operate in a
country with high GDP growth and a lower level of inflation.
Such rules benefit regulators or policy makers in their quest
to establish a healthy environment that will help their
banking sector to achieve a high level of efficiency as well as
be a regional financial hub. Managers could also benefit from
this analysis in working to improve their bank performance.
DEA produces information to guide an improvement policy
for inefficient banks, and any such improvement may result
in them being considered fully-efficient banks. Furthermore,
investors will find such results in their interest as they will
want to invest their money in such a way as to maximize
returns. Therefore, managers, policy makers, investors and
researchers are encouraged to use the proposed methodology
to gain more information about the performance of the
Eds. Hatem A. El-Karanshawy et al.
banking sector and to establish a set of rules for the efficient
operation of banks.
8. Conclusion
This paper investigates the performance of banks in the Gulf
states before, during and after crises (political and financial).
The study period (1998–2007) includes two crises: the second
Gulf crisis (2003) and the global financial crisis (2007).
This period allowed us to take look deeply into each bank’s
performance under two different situations. The results show
that the overall technical efficiency of all GCC commercial
banks is relatively stable over time. The commercial banks of
Saudi Arabia appear to be ahead of the GCC countries, followed
by banks of UAE, whereas Qatar has the least efficient banks.
However, there is no reason to believe that bank performance
differs from a statistical perspective according to location.
Also, different regulations (if any) that have been put in place
within GCC countries during the crises have had more or
less the same impact on banks’ performance. Furthermore,
conventional banks performed better during the second
Gulf crisis, whereas it was the Islamic banks that performed
better during the global financial crisis. Nevertheless, from a
statistical perspective,Islamic and conventional banks rank
more or less are same.
Out of the 24 environmental factors, fifteen were tested
and considered to be important, and only seven of them
are viewed as primary splitters for the decision tree. Assets
structure is the most important factor, followed by financial
strength and ROA. The operating style, population density,
size and support rating all have low importance. Testing only
for the external environmental factors, operating style and
established date are the most important factors, whereas
country and total population density have low importance.
Finally, this study contributes to the theory in developing
a comprehensive framework for measuring bank
performance, and in identifying the most important factors
that improve bank performance. The study also makes a
practical contribution as it is the first to assess the impact of
financial and political crises on banks of the Gulf states, and
it provides useful information for banks managers, investors
and policy makers for tracking banks’ efficiencies in order
to maintain a sustainable growing sector, and in providing
early warning signals of a bank that is potentially at risk.
The results of this study are limited to the selected banks
and study period. Researchers are therefore encouraged
to study the performance of the GCC banking sector after
the current global financial crisis; also, to compare the
performance of Islamic and conventional banks as the
different financial tools used by each of them may lead to
differences in performance.
References
Al Shammari, S., Structure-conduct-performance and effici­
ency in Gulf Cooperation Council (GCC) banking markets,
Unpublished PhD thesis, Wales University, 2003.
Almeida, H., Campello, M., Laranjeira, B., and Weisbenner,
S., Corporate debt maturity and the real effects of
the 2007 credit crisis, Unpublished working paper,
University of Illinois, 2009.
121
Anouze
Anouze, AL., Evaluating productive efficiency: comparative
study of commercial banks in Gulf countries, Unpublished
PhD thesis, Aston University, 2010.
Das, A., and Ghosh, S., Financial Deregulation and Profit
Efficiency: A Nonparametric Analysis of Indian Banks.
Journal of Economics and Business, 61, 2009. 509–528.
Ariss, R., Financial liberalization and bank efficiency:
evidence from post-war Lebanon, Applied Financial
Economics, 18, 2008. 931–946.
Demyank, Y., and Hasan, I., Financial crisis and bank
failures: a review of prediction methods, Omega, 38,
2010.315–324.
Avkiran, N., Removing the impact of environment with
units-invariant efficient frontier analysis: An illustrative
case study with intertemporal panel data, Omega, 37,
2009. 535–544.
Diamond, D., and Dybvig, P., Bank runs, deposit insurance and
liquidity, Journal of Political Economy, 91, 1983. 401–419.
Banker, R., Chang, H., and Lee, S-Y., Differential impact of
Korean banking system reforms on bank productivity,
Journal of Banking and Finance, 34, 2010. 1450–1460.
Banker, R., Charnes, A., and Cooper, W., Some models
for estimating technical and scale inefficiencies in data
envelopment analysis, Management Science, 30, 1984.
1078–92.
Beltratti, A., Stulz, R., Why did some banks perform better
during the credit crisis? a cross-country study of the
impact of governance and regulation, ECGI’s Finance
Working Paper No. 254/2009, 2009.
Berger, A., and Humphrey, D., Efficiency of financial
institutions: International survey and directions for
future research, European Journal of Operational
Research, 98, 1997. 175–212.
Berger, A., Distribution-free estimates of efficiency in
the U.S. banking system and tests of the standard
distributional assumptions, Journal of Productivity
Analysis, 4, 1993. 261–292.
Breiman, L., Friedman, J., Olshen, R. and Stone, C.,
Classification and regression trees, Pacific Grove,
Wadsworth-Monterey, USA, 1984.
Brissimis, S. Delis, M. and Papankiolaou, N., Exploring the
nexus between banking sector reform and performance:
Evidence from newly acceded EU countries, Journal of
Banking & Finance, 32, 2008. 2674–2683.
Burki, Abid and Niazi, G.S., Impact of financial reforms on
efficiency of state-owned, private and foreign banks in
Pakistan, Applied Economics, 1–14 URL: http://dx.doi.
org/10.1080/00036840802112315, 2009.
Casu, B., and Girardone, C., integration and efficiency
convergence in EU banking markets, Omega, 38, 2010.
260–267.
Charnes, A., Cooper, W., and Rhodes, E., Measuring the
efficiency of decision making units, EJOR, 2, 1978.
429–444.
Chiou, C-C., Effects of financial holding company act on
bank efficiency and productivity, Neurocomputing, 72,
2009. 3490–3506.
Chiu, Y-H., and Chen, Y-C., The analysis of Taiwanese bank
efficiency: incorporating both external environment,
Economic Modelling, 26, 2009. 456–463.
Cornett, M., McNutt, J. and Tehranian, H., The financial
crisis, internal corporate governance, and the
performance of publicly-traded U.S. bank holding
companies, unpublished working paper, Boston College,
2010.
122
Diamond, D., Financial intermediation and delegated
monitoring, Review of Economic Studies, 51, 1984.
393–414.
Dietrich, A., Wanzenried, G., Determinants of bank
profitability before and during the crisis: Evidence
from Switzerland, J. Int. Financ. Markets Inst. Money,
doi:10.1016/j.intfin.2010.11.002, 2011.
Duchin, R., Ozbas, O., and Sensoy, B., Costly external
finance, corporate investment, and the subprime
mortgage credit crisis, Journal of Financial Economics,
97, 2010. 418–435.
Emrouznejad, A, Anouze, AL, and Thanassoulis, E., A semioriented radial measure for measuring the efficiency of
decision making units with negative data, using DEA,
EJOR, 200, 2010a. 297–304.
Emrouznejad, A. and Anouze, AL., A note on the modeling
the efficiency of top Arab banks, Expert Systems with
Applications, 36, 2009. 5741–5744.
Emrouznejad, A. and Anouze, AL., Data envelopment
analysis with classification and regression tree: a case of
banking efficiency, Expert Systems, 27, 2010. 231–246.
Emrouznejad, A., Amin, G., Thanassoulis, E. and Anouze,
AL., On the boundedness of the SORM DEA models with
negative data, EJOR, 206, 2010b. 265–268.
Fethi, M. and Pasiouras, F., Assessing bank performance
with operational research and Artificial Intelligence
techniques: A survey, EJOR, 204, 2010. 189–198.
Fu, X. and Heffernan, S., The effects of reform on China’s
bank structure and performance, Journal of Banking
and Finance, 33, 2009. 39–52.
Fukuyama, H. and Weber, W., Estimating output allocative
efficiency and productivity change: application to
Japanese banks, European Journal of Operational
Research, 137, 2002. 177–190.
Fukuyama, H., and Weber, W., A Slacks-based inefficiency
measure for a two-stage system with bad outputs,
Omega, 38, 2010. 398–409.
Gorton, G., and Winton, A. Financial intermediation, In G.
Constantinides, M. Harris, and R. (Stulz, Handbook of
the Economics of Finance North Holland-Amsterdam,
2003. 431–552.
Grifell-Tatjé, E., Profit, Productivity and distribution:
differences across organizational forms—The case of
Spanish banks, Socio-Economic Planning Sciences, 38,
2010. 1–12.
Gryglewicz, S., A theory of corporate financial decisions
with liquidity and solvency concerns, Journal of
Financial Economics, 99, 2011. 365–384.
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Efficiency of performance of banks in the Gulf region before, during and after crises (financial and political)
Halkos, G. and Salamouris, D., Efficiency measurement
of the Greek commercial banks with the use of
financial ratios: a data envelopment analysis approach,
Management Accounting Research, 15, 2004. 201–24.
Han, J and Kamber, M., Data Mining Concepts and Techniques,
San Francisco: Morgan Kaufmann Publishers, 2001.
Hermes, N. and Nhung, T., The impact of financial
liberalization on bank efficiency: evidence from Latin
America and Asia, Applied Economics, URL: http://
dx.doi.org/10.1080/00036840802112448, 2008.
Hsiao, H-C, Chang, H, Cianci, A, and Huang, L-H., First
financial restructuring and operating efficiency:
Evidence from Taiwanese commercial banks, Journal of
Banking & Finance, 34 (2010), pp. 1461–1471.
Hsiao, H-C., Chang, H., Cianci, A., and Huang, L.-H.,
First financial restructuring and operating efficiency:
evidence from Taiwanese commercial banks, Journal of
Banking and Finance, 27 (2010), pp. 1461–1471
Huang, L-H., Hsiao, H-C., Cheng, M-A., and Chang, S-J., Effects
of financial reform on productivity change, Industrial
Management & Data Systems, 108, 2008. 867–886.
Huang, T-H, Liao, Y-T and Chiang, L-C., An examination
on the cost efficiency of the banking industry under
multiple output prices’ uncertainty, Applied Economics,
42, 2010. 1169–1182.
Ismail A, Davidson, I, and Frank, R., Operating performance
of European bank mergers, The Service Industries
Journal, 29, 2009. 345–366.
Jeon, Y. and Miller, M., Performance of domestic and
foreign banks: The case of Korea and the Asian financial
crisis, Global Economic Review, 34, 2005. 145–165.
Jeon, Y. and Miller, M., The effect of the Asian financial
crisis on the performance of Korean nationwide banks,
Applied Financial Economics, 14, 2004. 351–360.
Kim, H. and Koehler, G., Theory and Practice of Decision
Tree Induction, Omega, 23, 1995. 637–652.
Koetter, M., The stability of bank efficiency rankings
when risk preferences and objectives are different, The
European Journal of Finance, 14, 2008. 115–135.
Koutsomanoli-Filippaki, A., Margaritis, D., and Staikouras,
C., Efficiency and productivity growth in the banking
industry of Central and Eastern Europe, Journal of
Banking & Finance, 33, 2009. 557–567.
Lozano-Vivas, A., and Pastor, T., Do performance and
environmental conditions act as barriers for crossborder banking in Europe? Omega, 38, 2010. 275–282.
Mercan, M., Reisman, A., Yolalan, R. and Emel, A., The
effect of scale and mode of ownership on the financial
performance of the Turkish banking sector: results of a
DEA-based analysis, Socio-Economic Planning Sciences,
37, 2003. 185–202.
Eds. Hatem A. El-Karanshawy et al.
Park, K. and Weber, W., A note of efficiency and productivity
growth in the Korean banking industry, 1992–2002,
Journal of Banking & Finance, 30, 2006. 2371–86.
Portela S, Thanassoulis, E, and Simpson, G., A Directional
distance approach to deal with negative data in DEA:
an application to bank branches, JORS, 55, 2004.
1111–1121.
Portela, S. and Thanassoulis, E., Malmquist-type indices in
the presence of negative data: An application to bank
branches, Journal of Banking & Finance, 34, 2010.
1472–1483.
Ray, S., and Das, A., Distribution of Cost and Profit
Efficiency: Evidence from Indian Banking, European
Journal of Operational Research, 201, 2010. 297–307.
Ripley, B., Pattern recognition and neural networks,
Cambridge University Press, Cambridge, 1996.
Shleifer, A. and Vishny, R., Unstable banking, Journal of
Financial Economics, 97, 2010. 306–318.
Siriopoulos, C., and Tziogkidis, P., How do Greek banking
institutions react after significant events? – A DEA
Approach, Omega, 38, 2010. 294–308.
Staub, R., Souza, G., and Tabak, B., Evolution of bank
efficiency in Brazil: A DEA approach, European Journal
of Operational Research, 202, 2010. 204–213.
Sueyoshi T. and Aoki, S., A use of a nonparametric statistic
for DEA frontier shift; the Kruskal and Wallis rank test,
Omega, 29, 2001. 1–18.
Thanassoulis, E., Introduction to the Theory and application
of data envelopment analysis: a foundation text with
integrated software, Kluwer Academic Publisher,
Massachusetts, USA, 2001.
Thompson, R. Singleton, F. Thrall, R. and Smith, B.,
Comparative site evaluations for locating a high-energy
physics lab in Texas, Interfaces, 16, 1986. 35–49.
Torgo, L. (1997), Functional Models for regression tree
leaves, Proceedings of the Fourteenth International
Conference on Machine Learning (ICML 1997),
Nashville, Tennessee, USA, July 8–12, 1997.
Xiao, Y., French Banks Amid the Global Financial Crisis,
IMF Working Paper, WP/09/201, 2009).
Zhao, T., Casu, B., and Ferrari, A., The impact of regulatory
reforms on cost structure, ownership and competition
in Indian banking, Journal of Banking & Finance, 34,
2010. 246–254.
Zhou, Z. and Jiang, Y., Medical diagnosis with C4.5 Rule
preceded by artificial Neural network ensemble, IEEE
Transactions on information Technology in Biomedicine,
7, 200337–42.
123
Anouze
Appendix
Table A-1. Summary of selected studies.
Study
Country
Study
Period
Approach
Inputs
Outputs
Banker,
Chang, &
Lee (2010)
Casu &
Girardone,
(2010)
Korea
1995–2005
Intermediation
(i) interest expense and
(ii) other operating expense
European
Countries
1997–2003
Intermediation
Chiou
(2009)
Taiwan
1999–2004
Intermediation
Chiu &
Chen (2009)
Taiwan
2002–2004
Intermediation
Das &
Ghosh
(2009)
India
1992–2004
Intermediation
Fukuyama &
Weber
(2010)
Japan
2000–2006
Production and
intermediation
(i) Personnel expenses,
(ii) other administrative
expenses,
(iii) interest paid,
(iv) non-interest expenses.
(i) staff,
(ii) fix asset,
(iii) bank deposits (including
current deposits, savings
deposits, time deposits, check
deposits, & other deposits), &
(iv) salary expense.
(i) Number of employees,
(ii) total deposits,
(iii) fixed assets
(i) deposits,
(ii) labor,
(iii) capital/fixed assets
(iv) equity
1st stage:
(i) labor,
(ii) physical capital,
(iii) financial capital
2nd stage:
(i) deposits
(i) interest revenue, &
(ii) other operating
revenue
(i) total loans and
(ii) other earning assets.
Grifell-Tatjé
(2010)
Spain
1994–2004
Intermediation
Hsiao et al.
(2010)
Taiwan
2000–2005
Intermediation
LozanoVivas &
Pastor
(2010)
Ray & Das
(2010)
European
Countries
2004
Production
India
1996–2006
Intermediation
Siriopoulos &
Tziogkidis
(2010)
Greece
1995–2003
Intermediation
Staub,
Souza, &
Tabak
(2010)
Brazil
2000–2007
Intermediation
124
(i) Provision of loan
services (business &
individual loans),
(ii) investments
(iii) interest revenue and
(iv) non-investment
revenue.
(i) Total amount of loans,
(ii) total investment,
(iii) non-interest revenue.
(i) Loans & advances,
(ii) investments,
(iii) other income.
1st stage:
(i) deposits
2nd stage:
(i) loans, and
(ii) securities investments,
and
(iii) other business
activities.
(i) financial expense
(i) Real operating profit
from intermediation activities, (interest on deposits,
loans, labor expense)
(ii) real gross loan and
financial income, and
(iii) average value of loans &
financial investments.
(i) interest expenses,
(i) interest revenue,
(ii) non-interest expenses, and (ii) non-interest
revenue, and
(iii) total deposits.
(iii) total loans.
(i) labour,
(i) loans, and
(ii) funds and
(ii) other earning assets
(iii) physical capital
(i) deposits,
(ii) labor,
(iii) capital/fixed assets
(iv) equity & reserves
(i) Personnel expenses,
(ii) provisions,
(iii) operational expenses.
(i) Labor,
(ii) capital, &
(iii) purchased funds.
(i) investments,
(ii) earning advances, and
(iii) other income
(i) Financial claims,
(ii) operational income,
(iii) net income before
taxes.
(i) outputs,
(ii) loans and
(iii) investments,
Islamic banking and finance – Essays on corporate finance, efficiency and product development
The relationship between Islamic bank efficiency
and stock market performance: Evidence from
GCC countries
Samir Srairi1, Imen Kouki2, Nizar Harrathi3
Associate Professor, Finance, Faculty of Law, Economics and Management of Jendouba, (LAREQUAD), Tunisia,
E-mail: [email protected]
2
Assistant Professor, Finance, Higher Institute of Management, (LAREQUAD), Tunisia, E-mail: [email protected]
3
Assistant Professor of Quantitative Methods, Faculty of Economic Sciences and Management of Nabeul,
(LAREQUAD), Tunisia
1
Abstract - Using data envelopment analysis (DEA), this paper estimates the efficiency of 25 Islamic
banks operating in Gulf Cooperation Council (GCC) countries during the period 2003–2009.
It also examines the relationship between the efficiency of Islamic banks and the performance of
their stock. The results suggest that efficiency measures, particularly technical and pure technical
efficiency, have increased over the period of study but remain low as compared to conventional
banks. The inefficiency of Islamic banks can be attributed to pure technical inefficiency rather than
to scale inefficiency. We also find that large and small banks are more efficient than medium banks
in terms of overall technical efficiency. Furthermore, the empirical findings show that both technical
and pure technical efficiency changes are positively related to share returns, while changes in scale
efficiency have no impact on stock performance. Finally, the regression also indicates a significant
and positive association between market return and the book-to-market equity ratio with share
prices.
Keywords: banking, technical efficiency, stock performance, Islamic banks, data envelopment
analysis, GCC countries
1. Introduction
An Islamic bank is an institution that mobilizes and invests
financial resources according to Shariah. Islamic banking
transactions are based on six basic principles: prohibition
of interest, risk sharing, money as potential capital,
prohibition of speculative behaviour, sanctity of contracts,
and Shariah approved activities (Iqbal 1997).
Islamic banking, which started to operate from the 1960s,
exists today in all regions of the world, particularly in the
Middle East and Southeast Asia. According to the report of
the Blominginvest bank, which was established in February
2009, more than 390 Islamic financial institutions are
spread across 75 countries with total assets estimated
to be close to $1 trillion by 2010. The rating agency,
Moody’s Investors’ Service, forecast that Islamic bank
assets worldwide will reach $4 trillion within five years.
The Islamic financial system is considered to be one of
the fastest growing financial and economic sectors in the
world. During the last decade, the Islamic banking industry
has grown at a remarkable pace, at 20–30% per year being
three times the rate for conventional banks. According to
many reports, the rapid and continued growth of Islamic
banking is driven by multiple factors such as: increasing
demand from a large number of Muslims; increasing oil
wealth of Muslim countries; low banking penetration in
Muslim majority nations; increasing demand from nonMuslim customers and countries; and the support of
government and regulatory bodies for the development
and promotion of Islamic banking.
Furthermore, the Islamic financial system has been less
affected than the traditional system by the latest economic
and financial crisis (2008), due mainly to its profit-loss
sharing principle, and also because of its strict prohibition
of investments in risky instruments, such as toxic assets and
derivatives. In addition, according to an IMF survey (2010)
and Chapra (2009), Islamic banks have contributed to
Cite this chapter as: Srairi S, Kouki I, Harrathi N (2015). The relationship between Islamic bank efficiency and stock
market performance: Evidence from GCC countries. In H A El-Karanshawy et al. (Eds.), Islamic banking and finance –
Essays on corporate finance, efficiency and product development. Doha, Qatar: Bloomsbury Qatar Foundation
Developing Inclusive and Sustainable Economic and Financial Systems
Srairi et al.
financial and economic stability during the global financial
crisis. The strong performance of Islamic banks over recent
years has encouraged several universal banks in developed
countries to add Islamic products to their conventional
banking industry, through Islamic banks windows or
Islamic banking subsidiaries.
In view of the rapid growth of Islamic banks, several issues
are revealed about the performance of these financial
institutions. In addition, as Islamic banking was introduced
as a parallel system of conventional banks in the majority of
countries, the performance of the new form of banking may
have an impact on the soundness and stability of the banking
system as a whole (Mariani 2010). Moreover, the last
economic and financial crisis has turned the focus towards
Islamic financial institutions which, according to many
sources, have showed stronger resilience than conventional
banks (e.g., Moody’s; IMF working paper 2010). Despite
the strong position of Islamic banks, several studies (Iqbal
2007; Iqbal and Van Greuning 2007) have identified
weaknesses and vulnerabilities among Islamic banks in the
areas of risk management (operational risk; weak internal
control processes) and human resource issues (quality
of management; technical expertise; professionalism).
Therefore, it will be interesting to analyse the performance
of Islamic banks during the last decade in order to provide
some guidelines for managers, investors and policy makers
to improve the efficiency of these banks and to formulate
managerial strategies and public policies. Therefore, the
aim of this study is to investigate the efficiency of Islamic
banks operating in Gulf Council Cooperation (GCC)
countries during the period 2003–2009, and to examine
the relationship between the efficiency of Islamic banks
and the performance of their stock. To our knowledge, this
is the first study which analyses the relationship between
efficiency and share performance in the context of Islamic
banks in GCC countries.
To gain a better understanding of the Islamic banking sector
in GCC countries, our analysis is conducted in two steps.
First, by employing Data Envelopment Analysis (DEA) as
a non-parametric approach, we estimate the technical
efficiency of 25 GCC Islamic banks under the profit-oriented
method which defines cost variables as inputs, and revenue
variables as outputs. In addition, to analyse the sources of
inefficiency of these banks, we calculated pure technical
efficiency and scale efficiency as two components of
technical efficiency. We chose a period of six years between
2003 and 2009 in order to investigate the evolution of the
efficiency of Islamic banks over time. Moreover, in this
study we attempt to compare the efficiency measures of
Islamic financial institutions according to their size in terms
of total assets. Following several studies concerning the
conventional banking industry (e.g., Haddad et al. 2010;
Pasiouras 2008; Beccali et al. 2006), in the second stage of
this paper we investigat the potential association between
Islamic banks’ efficiency and their share prices. To meet this
objective, we regress annual stock returns calculated as the
sum of daily share returns on efficiency scores obtained in
the first step, adding some control variables.
This paper presents some interesting points compared
with some other studies on Islamic banking efficiency
in GCC countries. First, our sample comprises more than
90% of GCC Islamic banks assets, which makes it the most
126
comprehensive database on the GCC Islamic banking
industry. Also, to the best of our knowledge, this is the
first study that relates the efficiency of Islamic banks in
GCC countries to their stock prices. Finally, our paper also
attempts to study the impact of the recent economic and
financial crisis on the performance of GCC Islamic banks,
and compares the efficiency of large, medium and small
banks.
2. Literature review
Two streams of literature are discussed in this study, the
first concerning the efficiency of Islamic banks, the second
being relevant to the relationship between bank efficiency
and share performance.
Studies on Islamic bank efficiency
While there is wide discussion in the literature on bank
efficiency within the conventional bank sector, particularly
for the developing countries and, to a smaller degree, the
transition economies, the work on Islamic banks remains
limited. Even with the development of the Islamic banking
sector in several regions of the world, few studies have
evaluated the efficiency of the new form of banking, and
noneconcern the relationship between bank efficiency and
share performance.
According to Bashir (2007) and Sufian et al. (2008), the
majority of studies on Islamic banks have focused on
the concept issues describing the underlying principles
(Al-Omar and Iqbal 2000; Zahar and Hassan 2001; Lewis,
2008) and performance measures using the traditional
financial ratios of these type of banks (Bashir 2001;
Olson and Zoubi 2008; Srairi 2009). A few studies have
utilized frontier analysis techniques rather than traditional
methods to estimate the efficiency of Islamic banks. Using
both the stochastic frontier approach (SFA) and the DEA
models, Hassan (2007) estimated a variety of parametric
techniques (cost, profit efficiency, and productivity)to a
panel of 43 Islamic banks operating in 22 countries during
the period 1993–2001. He found that Islamic banks are
relatively more efficient in generating profits compared
with control costs. In fact, the score of profit efficiency was
found to be about 84%, while for cost the efficiency was
only 74%. The results also indicated that the major source
of inefficiency was allocative inefficiency rather than
technical inefficiency.
Mokhtar et al. (2008) used a non-parametric DEA technique
and an intermediation approach to estimate the technical
and cost efficiency of the fully-fledged Islamic banks as
well as Islamic windows in Malaysia from 1997 to 2003.
The main results of the study revealed that, although the
fully-fledged Islamic banks were more efficient than the
Islamic windows, the two types of Islamic banks were still
less efficient than the conventional banks. This finding also
showed that the average efficiency of the overall Islamic
banking sector increased over the survey period.
Employing the DEA model, Sufian et al. (2008) examined
the technical efficiency and its components (pure technical
efficiency and scale efficiency) of 37 Islamic banks
operating in 16 MENA and Asian countries during the
period 2001–2006. The results suggest that pure technical
Islamic banking and finance – Essays on corporate finance, efficiency and product development
The relationship between Islamic bank efficiency and stock market performance: Evidence from GCC countries
inefficiency dominated scale inefficiency of Islamic banks
during all years except for the year 2006. On the other hand,
the authors found that the MENA Islamic banks exhibited
higher technical efficiency compared to their Asian Islamic
bank counterparts.
A more recent study concerning GCC countries was
conducted by Srairi (2010), who employed a SFA model
with country-specific environment variables and estimated
the cost and profit efficiency of 71 commercial banks
during the period 1999–2007. The empirical results
indicated that, on average, the conventional banks are
more efficient in terms of cost and profit than the Islamic
banks. This study also revealed that both conventional and
Islamic banks in Arab Gulf countries are relatively more
efficient in generating profits than in controlling costs.
Bank efficiency and share performance
While there is an extensive literature examining several
issues on bank efficiency, such as the impact of liberalization
on the efficiency of banks (e.g., Chen et al. 2005; Das and
Ghosh 2006; Paul and Kourouche 2008), the sources
of bank inefficiency (e.g., Grigorian and Manole 2006;
Pasiouras 2008; Sufian 2009), the comparison of the
efficiency of banks according to country (e.g., Fries and
Taci 2005; Kasman and Yildirim 2006; Inui et al. 2008),
ownership structure (e.g. Isik and Hassan 2003; Bonin
et al. 2005; Kyj and Isik 2008), and comparison of type
of bank (foreign and domestic: Havrylchyk 2006; new
and old: Canhoto and Dermine 2003; conventional and
Islamic: Srairi 2010), only a limited number of papers have
investigated the impact of the efficiency of banks on stock
performance, and none of these papers have concerned
Islamic banks. The relationship between the efficiency of
banks and stock performance within the conventional
banking sector has been studied both on the basis of an
individual country and for a cross-section of countries.
Haddad et al. (2010) estimated the monthly efficiency and
productivity of 24 listed Indonesian banks and their market
performance using the non-parametricSlack-Based Model
(SBM) approach over the period January 2006 to July
2007. They found that the stock market values of the banks
were in accordance with their performance. The results
also indicated a positive correlation between the index of
the Indonesian stock exchange (JCI) and bank efficiency.
On the other hand, the findings suggest that Indonesian
banks with foreign ownership tend to be less efficient than
their domestic counterparts.
Using both DEA and SFA methods, Xiang and Shamsudding
(2009) calculated the technical, cost and profit efficiency
of nine publicly-listed Australian banks over the period
1997–2007, and analyzed the potential link between these
efficiency scores and stock returns. They observed that
an improvement in cost and profit efficiency, calculated
using the SFA model, increased bank stock performance.
However, the DEA efficiency scores were uncorrelated with
stock returns.
Pasiouras et al. (2008) examined the association
between the efficiency of ten Greek banks and their share
performance between 2000 and 2005. The authors used the
DEA technique (profit-oriented approach) and computed
Eds. Hatem A. El-Karanshawy et al.
three efficiency levels: technical efficiency under constant
returns to scale (CRS); technical efficiency under variable
returns to scale (VRS); and scale efficiency. The results
indicated that annual changes in technical efficiency (under
CRS or VRS) were positively related to stock returns, while
changes in scale efficiency had an insignificant impact on
share performance. Erdem and Erdem (2008) used a DEA
with intermediation approach, and found no association
between stock price returns and change in economic
efficiency for Turkish banks.
Across international financial markets, Beccali et al. (2006)
used both SFA and DEA approaches to estimate cost
efficiency for a sample of banks operating in five European
countries (France, Germany, Italy, Spain and United
Kingdom) in the year 2000. The results suggested that
the change in the prices of bank shares reflects percentage
changes in cost efficiency, particularly those derived from
DEA. More recently, Liadaki and Gaganis (2010), who
employed a larger sample (15 EU countries and 171 banks)
and a longer time period (2002–2006) than Becalli et al.
(2006), estimated the cost and the profit efficiency by using
the SFA model and taking into account the macroeconomic
and other country-specific characteristics. The main result
of this study showed higher profit inefficiency (21%) than
cost inefficiency (10%). This means that European banks
are more efficient in controlling costs than in generating
profits. However, Srairi (2010) found that profit efficiency
scores are more informative to shareholders and investors
in Gulf Arab countries. In fact, changes in profit efficiency
have a positive and significant effect on stock returns, while
there is no association between changes in cost efficiency
and stock returns.
3. Methodology and data
In this study, we employ a three-stage procedure to analyse
the efficiency of Islamic banks and the relation to share
price performance:
1.A non-parametric approach (DEA technique) is used
to estimate efficiency scores with an input-oriented
model.
2. Annual stock returns are calculated on the basis of daily
share returns in order to measure the share performance
for each bank.
3.The relationship between bank efficiency and stock
performance is examined by regressing the annual
return on stock against the yearly change of efficiency
levels.
DEA model
From the literature, it is apparent that two models are used
to examine the efficiency of banks. Parametric techniques,
such as Stochastic Frontier Analysis (SFA), Thick Frontier
Approach (TFA), and Distribution Free Approach (DFA),
use econometric tools and specify the function form for
the cost or profit function. On the contrary, non-parametric
approaches, such as Data Envelopment Analysis (DEA)
and Free Disposable Hull Analysis (FDHA), do not
make an assumption concerning the functional form of
the frontier, and use a linear program to calculate the
efficiency level. The small size of our sample pushed us to
adopt the DEA technique, which was first introduced by
127
Srairi et al.
Charnes et al. (1978). According to Avkiran (1999), DEA is
thought to work well with fewer data, fewer assumptions,
and limited sample sizes. Furthermore, DEA does not
require any specification of the functional form on the data
to construct the production frontier, and the distribution
forms of errors (Bauer et al. 1998). However, DEA has
some limitations. This technique is very sensitive to
outlying observations, and all deviations from the frontier
indicate inefficiency (Havrylchyk 2006). Moreover, the
DEA approach does not allow for any error in the data and,
in consequence, it may overstate the true levels of relative
inefficiency for some entities (Drake and Hall 2003; Berger
and Mester 1997). Despite its limitations, we propose that
DEA is a robust tool for examining the efficiency of Islamic
banks in GCC countries.
assumption (TE = PTE*SE). To calculate these efficiency
scores, we employed the software DEAP version 2.1
developed by Coelli (1996).
DEA is a deterministic model that can be used to examine the
relative efficiency of a number of entities (decision-making
units: DMUs) in the sample having the same multiple inputs
and multiple outputs. To calculate the efficiency scores, a
linear programming model is solved for each bank. The DEA
model measures the efficiency of each DMU relative to all
other DMUs, with the simple restriction that all DMUs lay
on, or below, the efficiency frontier (Das and Ghosh 2006).
If a DMU lies on the frontier, it is referred to as an “efficient
unit”. Otherwise, it is DEA-inefficient. The value of the
efficiency score for each DMU is ranged between zero and
one. To define the best practice frontier, DEA can run under
either constant returns to scale (CRS), or variable returns
to scale (VRS). The main difference between these two
models is the treatment of returns to scale. The VRS model,
which was defined by Banker et al. (1984), compares each
bank only with other banks operating in the same region
of return to scale (banks of similar size). However, the CRS
assumption is only justifiable when all banks are operating
at an optimal scale. It means that a rise in inputs results
in a proportionate rise in outputs. On the other hand, a
DEA model can be constructed using an input-orientation
(minimizing inputs) or output-orientation (maximizing
outputs) approach.
Following recent studies on bank efficiency (e.g., Drake et al.
2006; Pasiouras 2008; Sturm and Williams 2004), in this
study we adopt the profit-oriented approach. This method
focuses on revenues as well as costs. It also has the advantage
of allowing a better understanding of the strategies used by
banks to respond to the changes in environment. Accordingly,
three inputs and two outputs are selected to estimate
efficiency levels. Hence, the vector of inputs comprises:
employee expenses (x1), other operating expenses (x2) and
loan loss provisions (x3). The vector of outputs includes two
variables: net interest income (y1 = interest income- interest
expense) and other operating income (y2).
The input-orientation approach is defined as the ability
of the bank to obtain a given level of outputs by utilizing
a minimum combination of inputs; the opposite approach
analyzes the ability of banks to produce the maximum
level of outputs, given the current level of inputs (Cooper
et al. 2000). In this study, we adopt an input-oriented DEA
technique because of the expressed interest of the Islamic
banking sector in more control costs. Many studies (e.g.,
Archer and Abdel-Karim 2002; Kamaruddin et al. 2008)
conclude that the cost of funds and labour in Islamic banks
is higher compared with those in conventional banks.
The DEA approach permits calculation for each bank of
the overall technical efficiency (TE) and its two components,
pure technical efficiency (PTE) and scale efficiency (SE).
PTE, also called “managerial efficiency”, represents the
failure of the bank to extract the maximum output from its
adopted input level and, hence, it relates to the ability of the
manager to utilize the firm’s given resources (Drake and Hall
2003; Pasiouras 2008). SE, another indicator of efficiency,
measures the proportional reduction in input usage if the
bank can operate at a point where the production exhibits
CRS (Kyj and Isik 2008). It can be computed by dividing
TE under the assumption of CRS to the TE under the VRS
128
Specification of inputs and outputs
To estimate the efficiency frontier using the DEA techni­
que, we needed measures of inputs and outputs. In the
literature, there has been little consensus over which
inputs and outputs should be used with the DEA model
and how they could be measured (Berger and Humphrey
1992). Consequently, several approaches are used in
bank efficiency studies: the production approach, the
intermediation approach, the operating approach, and the
profit approach.
Bank efficiency and share performance
Once the efficiency scores (TE, PTE, SE) and the annual
share returns are computed,in the third stage of this study
we examine the impact of the efficiency of Islamic banks
on performance (e.g., Liadaki and Gaganis 2010; Sufian
and Abdul-Majid 2009; Erdem and Erdem 2008). The
relationship is checked using the following linear model:
RSit = a + b1CEit + b2MRjt + b3BSFit + eit(1)
where RSit is the annual return on bank i’s stock in year t.
CEit represents the annual percentage change in bank
efficiency and includes the technical (TE, model 1) or pure
technical (PTE, model 2) or scale efficiency (SE, model 3)
for bank i in year t. MRjt is the market return for the banking
sector j in year t, and BSFit concerns some specific factors
and includes two variables, LTAit, which is the size of
bank i in year t measured as the natural logarithm of total
assets, and BMit, which is the book-to-market equity ratio
calculated as the ratio of the book value of a bank’s equity
to its market value. The a intercept represents the constant
of the model, bi is the parameters to be estimated and eit is
the disturbance term calculated as follows:
eit = uit + vi
Since we have a panel regression combining cross section
and time series data, we estimate this model by using a
fixed effect model (ni which represents bank specific effect
is fixed over time) and a random effect model (in the case
ni is considered as an error term). The fixed effect model is
tested by the Fisher (F) test, while the random effect model
is examined by the Lagrange Multiplier (LM) test. If the
null hypothesis of heteroscedasticity residual variance
is rejected, the ordinary least square (OLS) regression
Islamic banking and finance – Essays on corporate finance, efficiency and product development
The relationship between Islamic bank efficiency and stock market performance: Evidence from GCC countries
Table 1. Summary statistics of dataset used in the study (average values).
2003
2004
2005
2006
2007
2008
23.47
19.66
19.95
103.92
27.86
25.21
21.26
19.05
116.50
38.62
32.7
36.47
17.71
168.35
77.86
48.35
43.77
13.16
200.29
113.12
67.63
52.78
18.92
272.66
157.13
73.89
58.83
61.73
275.72
132.67
70.27
61.22
120.47
262.78
95.82
Panel B: control variables and stock returnb
– Total assets (US$ Millions)
2707
– Book-to-market equity
–
– Annual stock return
–
3065
2.15
44.82
3835
0.90
67.84
5200
1.35
-27.33
7124
1.45
16.56
9062
2.23
-81.25
9739
1.53
-15.33
Panel A: inputs and outputsa
– Employee expenses (x1)
– Other operating expenses (x2)
– Loan loss provision (x3)
– Net interest income (y1)
– Other operating income (y2)
2009
= variables in US$ million; b = all variables are in percentages, except where indicated.
a
is favored. To choose between these two models, we
calculated the Hausman test (H).
Data
Our sample comprises 25 Islamic banks operating in five
Gulf Arab countries (GCC) with six banks in Bahrain, eight
banks in Kuwait, two banks in Qatar, two banks in Saudi
Arabia, and seven banks in the United Arab Emirates, over
the period 2003–2009. The choice of region is justified
for many reasons: first, the GCC countries, which comprise
six states (Bahrain, Kuwait, Qatar, Saudi Arabia, the United
Arab Emirates, and Oman) hold the largest share (about
61.6%) of Islamic bank’ assets in the world ($263 billion
in 2008). Saudi Islamic banks occupy the first place in
terms of GCC Shariah-compliant assets (35%), followed by
Kuwait (24%), the United Arab Emirates (19%), Bahrain
(14%), and Qatar (8%). During the last decade, the Islamic
banking sector in GCC countries had achieved strong
growth in term of total assets (over 35%). Also, since 2002,
the GCC region has been in a relatively strong position
(7% growth between 2002 and 2008) and is expected to
continue at the same pace and to launch huge projects of
more than $1 trillion during the next decade. Finally, while
the GCC states provide opportunities in many sectors and
offer ample liquidity in the banking sector, Islamic banks
are expected to further diversify their products and services
and so attract a wider clientele. In addition, the Islamic
financial system will continue to spread to investment
banking, project finance, capital markets, insurance,
wealth management and micro-finance (Iqbal 2007).
The annual data of Islamic banks (financial statements)
used to calculate the efficiency scores are collected from
Bankscope Database of Bureau Van Dijk’s Company. The
daily stock prices and market index are obtained from
Datastream. Since Gulf countries have different currencies,
all the annual financial values are converted into US dollars
using appropriate average exchange rates for each year.
Also, to ensure comparability of data across countries, all
values are deflated to the year 2003 using each country’s
consumer price index (CPI).
Table 1 summarises the mean of inputs and outputs
employed in the DEA model and also presents the average
value of stock returns and control variables used in the
Eds. Hatem A. El-Karanshawy et al.
regression over 2003 to 2009. The table shows a great
increase of all inputs and outputs during the period of study.
In fact, we note that employee expenses, the other operating
expenses, the net interest income, and other operating
income have risen about 200%, 211%, 153%, and 243%,
respectively. The loan loss provision was constant during
2003–2007 and grew rapidly during the two last years of
the study period (2008 and 2009). It is interesting to note
that the crisis did not have the same effect on Islamic banks
as is reported for conventional banks (Blominvest bank
report 2009), the income of Islamic banks exhibiting only a
small decrease of 4%. Finally, we note an increase of more
than 25% of the average rate of assets.
4. Empirical results
The analysis of the empirical findings on the efficiency of
Islamic banks in GCC countries is structured in two main
parts. First, we estimate the overall technical efficiency and
its components, measured by DEA method, and evaluate
its evolution over time. Further, we attempt to examine
the efficiency of Islamic banks according to their size. In
the second part, we extend the analysis by examining the
relationship between efficiency scores of Islamic banks and
their share performance
DEA efficiency measures
In this section, we examine the efficiency scores of Islamic
banks calculated under the profit-oriented approach and
obtained by the DEA technique. In order to analyse the
evolution of the efficiency of Islamic banks between 2003
and 2009, we chose to construct a common frontier for
all banks in the sample; the implicit assumption was of
an absence of technical change during the period of study.
In this approach, the efficiency of each bank observed
in different years is estimated in relation to a common
benchmark technology (Canhoto and Dermine 2003).
Table 2 provides a summary of annual means of efficiency
indexes over 2003–2009 classified by year (panel A) and
by size (panel B). As can be seen from this table, overall
technical efficiency scores exhibit an upward trend from
2003 to 2009. The mean of TE varies from 61.2% (2003)
to 68.5% (2009) with an average equal to 65.5%. This
result appears to show an improvement of the efficiency of
129
Srairi et al.
Table 2. Efficiency scores by year and size (average values).
TE
Panel A: by year
2003
2004
2005
2006
2007
2008
2009
Panel B: by size
Small banks
Medium banks
Large banks
Overall
PTE
SE
Mean
SD
Mean
SD
Mean
SD
0.612
0.643
0.650
0.642
0.671
0.681
0.685
0.147
0.195
0.141
0.112
0.115
0.162
0.085
0.718
0.738
0.751
0.778
0.799
0.813
0.817
0.136
0.149
0.143
0.141
0.106
0.128
0.138
0.855
0.864
0.883
0.839
0.847
0.838
0.852
0.116
0.138
0.200
0.150
0.131
0.127
0.128
0.669
0.653
0.686
0.156
0.118
0.159
0.676
0.762
0.779
0.153
0.123
0.147
0.990
0.840
0.885
0.157
0.140
0.142
0.655
0.140
0.773
0.137
0.855
0.144
TE = technical efficiency; PTE = pure technical efficiency; SE = scale efficiency.
Islamic banks during the period of study. Indeed, efficiency
scores, particularly TE and PTE, increased by 12% and 13%
on average, respectively, while scale efficiency remained
constant. However, during 2008 and 2009, these measures
are constant but slightly changed and increased by 1.5%
and 1.7%, respectively, compared to 2007. It is apparent
that the last financial and economic crisis has affected the
performance of Islamic banks, but to a lesser extent than
for conventional banks. According to Hasan and Dridi
(2010), “the initial impact of the crisis on Islamic Banks’
profitability in 2008 was limited. However, with the impact
of the crisis moving to the real economy, Islamic Banks
in some countries faced larger losses compared to their
conventional peers”.
Despite the increase in efficiency of Islamic banks between
2003 and 2009, the average of the input waste is large
and equal to 34.5%. Therefore, there is still room for
improvement in the performance of these banks through
more efficient use of resources. Indeed, the efficiency scores
of Islamic banks in GCC countries are low compared not
only to conventional banks (Srairi, 2010; Rosly and Abu
Baker 2003) but also to Islamic banks in other countries. For
instance, Kamaruddin et al. (2008) found that the average
of technical efficiency of the Malaysian Islamic banks is
93% for the period 1998–2004. In a recent study of Islamic
banks in MENA and Asian countries, Sufian et al. (2008)
found that Islamic banks in Indonesia during the period
2001–2006 are the most efficient from the Asian region,
exhibiting a mean technical efficiency of 92.3%. However,
several studies (e.g., Mohammed et al. 2008; Hassan et al.
2009) suggested that there are no significant differences
between the overall efficiency results of conventional
compared with Islamic banks.
The decomposition of overall technical efficiency into PTE
and SE components provides information on the source
of technical inefficiency. Table 2 reveals that the pooled
means for PTE and SE during the period analyzed are
of 77.3% and 85.5%, respectively. The result shows that
130
the inefficiency in Islamic banks could be attributed to
pure technical inefficiency (29.3%) rather than to scale
inefficiency (17%). It means that Islamic banks in GCC
countries are managerially inefficient in controlling costs
but manage their inputs efficiently. This finding of the
dominant impact of managerial inefficiency over scale
inefficiency is also reported in other studies, for example,
Sufian et al. (2008) for Islamic banks in MENA and Asian
countries; Kyj and Isik (2008) for the Ukrainian banking
industry; and Zaim (1995) for Turkish banks. According
to several studies (e.g., Bashir 2007; Iqbal 2007), the
inefficiencies in Islamic banks can also be attributed
to many other causes such as: limited number of shortterm instruments; shortage of products for medium and
long term maturities; portfolios of Islamic banks being
concentrated on equity and non-interest based financing,
and especially focused on trade financing; small size
of banks; weak management; and lack of proper riskmonitoring systems.
Furthermore, we attempt in this study to identify the
nature of scale inefficiency, which can be due to increasing
returns to scale (IRS) or decreasing returns to scale (DRS).
Table 3 displays statistics for the number of banks in the
different categories of scale economies, and also presents
the returns to scale of banks classified by size. According to
the figures in this table, only 19% of Islamic banks operate
at their optimal scale (CRS) and the majority of banks are
scale-inefficient (58% at DRS and 23% at IRS). It is also
interesting to note that the share of the banks experiencing
economies of scale (IRS) and diseconomies of scale (DRS)
are relatively constant during the sample period. The
results confirm those shown in Table 2, relative to the
stability of scale efficiency of Islamic banks over the period
of study. Panel B of Table 3 also indicates that the majority
of Islamic small banks (83%) exhibited IRS (53%) or CRS
(30%), while the medium and large banks operated at DRS
(80%). It means that increasing the activities and size of
Islamic small banks may bring significant cost savings and,
in consequence, improve the technical efficiency of these
Islamic banking and finance – Essays on corporate finance, efficiency and product development
The relationship between Islamic bank efficiency and stock market performance: Evidence from GCC countries
Table 3. Return to scale in Islamic banks by year and size.
DRS
Years
Panel A: by year
2003
2004
2005
2006
2007
2008
2009
Total
Panel B: by size
Small banks
Medium banks
Large banks
IRS
CRS
Nb. of banks
% share
Nb. of banks
% share
Nb. of banks
% share
Total
of
banks
15
15
14
13
12
15
15
99
62
62
58
52
48
60
60
58
6
5
6
6
7
5
5
40
25
21
25
24
28
20
20
23
3
4
4
6
6
5
5
33
13
17
17
24
24
20
20
19
24
24
24
25
25
25
25
172
10
44
45
17
79
80
32
3
5
53
5
9
18
9
6
30
16
11
60
56
56
DRS: decreasing returns to scale; IRS: increasing returns to scale; CRS: constant returns to scale.
banks, in contrast to the case of expansion by the medium
and large banks. A similar finding has been made for other
countries such as Singapore (Rezvanian and Mehdian
2002), Turkey (Isik and Hassan 2002) and India (Rezvanian
et al. 2008).
In order to compare the efficiency scores of banks according
to their size, we categorized the sample banks into three
groups based on their total assets, with an approximate
number of banks in each category. The first group comprises
nine small banks with an asset size of less than $3 billion.
The second group includes medium banks (eight banks)
whose assets are between $3 and $5 billion, while, the last
group comprises large banks (eight banks) whose assets
exceed $5 billion.
In terms of overall technical efficiency, panel B of Table 2
shows that large (68.6%) and small (66.9%) banks are
the most efficient, while the medium banks presented the
lowest mean TE of 65.3%. This is consistent with several
studies which reported a significant positive association
between size and efficiency (e.g., Drake and Hall 2003;
Chen et al. 2005; Pasiouras 2008; Srairi, 2010). Large
banks present some advantages over small and medium
banks. According to Kyj and Isik (2008), “large banks
may be able to hire a better management team, utilize
better technology, be located in larger, more competitive
markets, and have more diversified loan portfolio. Large
banks, thus, may have lower default risk, and lower
borrowing costs”. However, other studies found a negative
(e.g., Christopoulis et al. 2002; Bonin et al. 2005) or no
significant (e.g., Berger and Hannam 1998; Girardone
et al. 2004) relationship between size and efficiency. On
the other hand, the result indicates that large (77.9%)
and medium (76.2%) banks are more pure technically
efficient than small banks (67.6%). However the latter
display a superior measure for scale efficiency, this being
10.5% and 15% higher than for medium and large banks,
respectively. Consequently, it seems that Islamic small
banks need more improvement in terms of managerial
Eds. Hatem A. El-Karanshawy et al.
practice, while Islamic medium banks need to increase
their scale efficiency.
Efficiency and share performance
To assess the relationship between the efficiency of
Islamic banks and their share prices, we regress annual
stock price returns on annual percentage change of
efficiency scores, derived from DEA analysis, with other
explanatory variables. Models 1, 2 and 3 in Table 4
present the regression results estimated by the fixedeffect model for technical, pure technical, and scale
efficiency changes respectively. The results indicate that
both technical and pure technical efficiency changes
have a positive and statistically significant (1% for
TE and 5% for PTE) effect on stock returns. Indeed,
the share prices of Islamic banks respond positively
towards improvement in managerial efficiency. Hence,
it seems that information regarding the efficiency of
banks is reflected in the stock prices of banks. In fact,
in an efficient market, share prices incorporate all
publicly available information (Fama 1970). Thus,
according to Beccalli et al. (2006) and others, efficient
banks can better improve their share price performance
than inefficient banks. So, our results are in line
with several studies in other countries which found
a positive association between technical efficiency
change and share performance (e.g., Pasiouras et al.
2008 for Greek banks; Xiang and Shamsudding 2009
for Australian banks; Sufian and Abdul Majid 2009 for
China banks). However, other researches (e.g., Liadaki
and Gaganis 2010 for European banks; Ioannidis et al.
2008 for Asian and Latin American banks; Chu and
Lim 1998 for Singapore banks) show that changes in
stock returns reflect changes only in profit efficiency
rather than in cost efficiency. According to Liadaki and
Gaganis (2010), these results can be explained by the
fact that rational shareholders and investors are more
interested by the profit of banks as an indicator of the
future dividends. Moreover, cost efficiency reflects the
131
Srairi et al.
Table 4. Regression results of equation (1).
RSit = a + b1CEit + b2MRit + b3BSFit + eit
RSit is the annual return on bank i’s stock in year t. CEit represents the annual percentage change in bank efficiency and
includes the technical (TE, model N°1) or pure technical (PTE, model N°2) or scale efficiency (SE, model N°3) for bank
i in year t. MRjt is the market return for banking sector j in year t and BSFit concerns some specific factors and includes
two variables: LTAit is the size of bank i in year t measured as the natural logarithm of total assets and BMit is the book-tomarket equity ratio calculated as the ratio of the book value of a bank’s equity to its market value.
Variables
Model 1
Model 2
Model 3
Constant: α
2.15
(2.75)*
2.03
(2.62)*
2.11
(2.70)*
3.14
(2.30)**
–
–
–
–
–
7.07
(2.61)*
–
0.84
(9.05)*
0.71
(1.61)
7.31
(2.55)**
0.617
38.66*
125
48.3
385.2
36.1
0.86
(8.88)*
0.56
(0.91)
7.35
(2.49)**
0.597
35.29*
125
46.4
391.3
33.8
Annual change in efficiency scores
Technical Efficiency: TE
Pure Technical Efficiency: PTE
Scale Efficiency: SE
Control Variables
Market return: MR
Size of bank: LTA
Book to market equity ratio: BM
Adjusted R2
F value
Nb. Observations
F value
LM
Hausman test
0.83
-0.83
0.87
(8.96)*
0.51
(1.60)
7.12
(2.41)**
0.499
35.60*
125
43.2
380.4
31.2
T-statistics are between parentheses; *, and ** indicate statistical significance at 1%, and 5% respectively.
capability of managers but it is not directly observed in
the stock market.
From Table 4 (model 3), it is also noted that the estimated
coefficient of scale efficiency change is positive but it is not
statistically significant. It means that scale efficiency does
not have any impact on a bank’s share returns. This finding
is also confirmed by the coefficient of bank size which is
insignificant in all of the regression models. A similar result
was also found by Pasiouras et al. (2008) and Sufian and
Abdul Majid (2009).
With regard to the control variables and their influence
on stock returns, Table 4 indicates that market return in
all models has the expected sign and a significant power
to explain the variation in stock prices. This result, which
is consistent with previous studies (e.g., Xiang and
Shamsuddin 2009; Erdem and S.Erdem 2008), shows that
stock price returns of Islamic banks are positively related to
the overall performance of the market. On the other hand,
the association between the ratio of book-to-market value
(BM) and share performance is positive and significant
at the 5% level for all models. However, our results are
different from the study of Xiang and Shamsuddin (2009)
concerning Australian banks which found a negative sign
132
of BM, implying a possibility of market expectation of
systematic risk.
5. Discussion
Many policy implications and recommendations can be
derived from the results of this paper. First, since Islamic
banks in GCC countries exhibited a lower level of efficiency
compared to conventional banks, it is necessary for these
institutions to promote and enhance their functioning in
several areas (Bashir 2007; Iqbal 2007). Islamic banks are
still operating with a limited number of instruments for the
short-term, and there is a shortage of products for mediumto long-term maturities. In this regard, Islamic banks have to
offer new products and modes of finance that enhance risk
management and portfolio diversification. Due to limited
size and resources, Islamic banks are unable to reap the
benefits of economies of scale and are also unable to afford
high cost management information systems to assess and
monitor risks. Accordingly, Islamic banks have to perform
strategic alliances with other Islamic financial institutions,
and collaborate with conventional banks which are more
sophisticated in terms of financial engineering. Further,
to better control and reduce their costs, Islamic banks
need to invest more in technology, to develop innovative
Islamic banking and finance – Essays on corporate finance, efficiency and product development
The relationship between Islamic bank efficiency and stock market performance: Evidence from GCC countries
methods in terms of risk management, and to increase
the efficiency of their staff by investing in training and
development. The results also show that there has been an
improvement in efficiency of Islamic banks over the period
of liberalization in Gulf countries. Therefore, authorities in
this region should continue to reinforce financial reforms,
increase economic integration between countries, and
undertake constructive policy actions to develop Islamic
capital markets which help to integrate Islamic financial
institutions into regional and international financial
systems. Finally, while there is a positive association
between the performance of Islamic banks and their stock
price returns, it appears that efficiency measures contain
important and helpful information which could be used by
managers of banks, shareholders and investors.
6. Conclusions
Islamic banking is viewed as competitive and an alternative
to the conventional banking system in many states of the
world, particularly in GCC and some Asian countries. In
addition, during the last decade, Islamic banking assets
have been growing at a faster pace (an average annual
growth of 20%) than the overall banking system, with the
expectation that it will play an increasing important role
in the coming years. Moreover, the Islamic financial system
has proved to be the least affected by the last economic
and financial crisis. In the light of these considerations, it
is important to assess and analyse how Islamic banks have
performed during the past few years.
In the present study, we estimate the efficiency of
25 GCC Islamic banks over the period 2003–2009. By
using a non-parametric DEA technique, under the profit
oriented approach, we calculate technical, pure technical
and scale efficiencies to study the evolution of these
efficiency measures across time and to analyse the size
efficiency relationship. Additionally, this paper attempts
to investigate the influence of the performance of Islamic
banks in terms of efficiency on their stock prices. Several
important findings emerge from this present study. The
results indicate that the average technical efficiency was
equal to 66% and that there was a rising trend for both TE
and PTE, suggesting that Islamic banks in GCC countries
improved their efficiency during the survey period. This
was the period where the processes of liberalization of
the GCC financial system were realised at an accelerated
pace. Overall, we also find that inefficiency in Islamic banks
is attributed mainly to pure technical inefficiency (29%)
rather than scale inefficiency (17%). Thus, it seems that
Islamic banks are managerially inefficient in controlling
their costs and their inputs. It is interesting to note that the
majority of Islamic banks are scale-inefficient and are either
small- or medium-sized, which implies that these banks
can achieve cost savings and improve their efficiency by
increasing their size and scale of operations. Furthermore,
our findings regarding the impact of size on the efficiency
of Islamic banks suggest that, while large banks are more
managerially and technically efficient than small banks,
they are also less scale-efficient than the smaller banks. In
terms of pure technical efficiency, large-sized Islamic banks
seem also to be the most efficient ones, followed by the
medium banks. In this regards, it appears that small banks
need to improve their managerial practices, while medium
banks have to increase their scale efficiency.
Eds. Hatem A. El-Karanshawy et al.
Using the efficiency scores of Islamic banks, we analysed
the link between efficiency change and stock returns.
The results derived from the fixed effect model show that
percentage changes in the prices of bank stocks reflect
percentage changes in both technical and pure technical
efficiency. However, we do not find any significant
relationship between scale efficiency and stock returns.
Thus, our results seem to support the argument that
stock returns respond positively towards improvement in
managerial efficiency, but do not react towards changes
in scale efficiency (Sufian and Abdul Majid 2009). Hence,
the efficiency of a bank’s operation provides significant
information about its share price performance, which is not
explained by market movements.
One implication of the findings is that managerially-efficient
banks should be more profitable and therefore generate
greater shareholder returns. This is in line with the efficient
market theory that, in an efficient market, a change in cost
efficiency should be incorporated in the price formation
process. Finally, the study also revealed that market return
and ratio of book-to-market value have a positive impact on
stock returns.
References
Al-Omar, F.A., Iqbal, M. Some strategic suggestions for
Islamic banking in the 21st century, Review of Islamic
Economics, 9, 2000. 37–56.
Archer S, Abdel-Karim R. Islamic Finance. Euromoney
Books, UK. 2002.
Avkiran, N. K The evidence on efficiency gains: The role
of mergers and the benefits to the public, Journal of
Banking and Finance, 23, 1999. 991–1013.
Bashir, A.H.M. Assessing the performance of Islamic banks:
Some evidence from the Middle East, in American
Economic Association Annual Meeting, New Orleans,
USA, 2001.
Bashir, A.H.M. Islamic banks participation, concentration
and profitability: Evidence from MENA countries.
Working Paper 0402, 2007.
Banker, R. Charnes, A. Cooper, W “Some models for
estimating technical and scale inefficiencies in data
envelopment analysis”. Management Science, 1984.
1078–1092.
Bauer, P. W. Berger, A. N. Ferrier, G. D. Humphrey, D.
B. Consistency conditions for regulatory analysis of
financial institutions: A comparison of frontier efficiency
methods, Journal of Economics and Business, 50,
(1998). 85–114.
Becalli, E. Casu, B. Giraradone, C. Efficiency and stock
performance in European banking, Journal of Business
Finance and Accounting, 33, 2006. 245–262.
Berger, A. N., Humphrey, D. B. Measurement and Efficiency
issues in commercial banking, in Z. Griliches (Ed.)
Measurement issues in the service sectors, National
Bureau of Economic research, University of Chicago
Press, 1992. 245–279.
Berger, A. N. Hannam, T. H. The efficiency cost of market
power in the banking industry: A test of the quiet life and
133
Srairi et al.
related hypotheses, Review of Economics and Statistics,
80, 1998. 454–465.
Berger, A. N. Mester, L. J. Inside the black box: What explains
differences in the efficiencies of financial institutions,
Journal of Banking and Finance, 21, 1997. 895–947.
Blominvest Bank Islamic banking in the MENA region. 2009.
Bonin J. P., Hassan I., Wachtel P. Bank performance,
efficiency and ownership in transition countries. Journal
of Banking and Finance, 29, 2005.31–53.
Canhoto, A. Dermine, J. A note on banking efficiency in
Portugal: New vs. old banks, Journal of Banking and
Finance, 27, 2003. 2087–2098.
Chapra, M. U. The global financial crisis: Can Islamic
finance help, NewHorizon, January-March, 170, 2009.
Charnes, A. Cooper, W. Lewin, A. Seiford, L. Data
envelopment analysis: Theory, methodology, and
application, Kluwer Academic Publishers, Boston, 1994.
Charnes, A. Cooper, W. Rhodes, E. Measuring the
efficiency of decision making units, European Journal
of Operational Research, 2, 1978. 429–444.
Fields, J. A. Murphy, N. B. Tirtiroglu, D. An International
comparison of scale economies in banking: Evidence
from Turkey, Journal of Financial Services Research, 4,
1993. 157–168.
Fries S, Taci A. Cost efficiency of banks in transition: Evidence
from 289 banks in 15 post-communist countries. Journal
of Banking and Finance, 29, 2005. 55–81.
Girardone C, Molyneux P, Gardener E P M. Analyzing
the determinants of bank efficiency: the case of Italian
banks, Applied Economics, 36, 2004. 215–227.
Grigorian, D. A. Manole, V. Determinants of commercial
bank performance in transition: An application of data
envelopment analysis, Comparative Economic Studies,
48, 2006. 497–522.
Hadad, M. Hall, M. Kenjegalieva, K. Santoso, W. Satria,
R. Simper, R. Bank efficiency and stock Market
performance: An analysis of listed Indonesian banks,
Review Quantitative Finance and Accounting,
DOI.10.1007/s11156–010–0192–1, 2010.
Hassan, K. M. The X-efficiency of Islamic banks, Islamic
Economic Studies, 2, 2007. 49–77.
Chen, X. Skully, M. Brown, K. Banking efficiency in China:
Application of DEA to pre- and post-deregulation
era: 1993–2000, China Economic Review, 16, 2005.
229–245.
Hasan, M. Dridi, J The effects of the global crises on
Islamic and conventional banks: A comparative study,
International Monetary fund, Working Paper /10/201,
2010.
Christopoulos, D.K., Lolos, S.E.G. Tsionas, E.G. Efficiency
of Greek banking system in view of the EMU: A
heteroscedastic stochastic frontier approach, Journal
of policy model, 24, 2002. 813–829.
Hassan, T. Mohamed, S. Bader M. Efficiency of conventional
versus Islamic banks: Evidence from the Middle East,
International Journal of Islamic and Middle Eastern
Finance and Management, 2, 2009. 46–65.
Chu, S. Lim, G. Share performance and profit efficiency
of banks in an oligopolistic market: Evidence from
Singapore, Journal of Multinational Financial
Management, 8, 1998. 155–168.
Havrylchyk, O. Efficiency of the Polish banking industry:
Foreign versus domestic banks, Journal of Banking and
Finance, 30, 2006. 1975–1996.
Coelli, T. A guide to DEAP version 2.1, A data envelopment
analysis, (Computer Program), CEPA, Working Paper
96/08, University of New England, Armidale, 1996.
Available at: www.une.edu.au/econometrics/cepa.htm.
Cooper, W. W. Seiford, L. M. Tone, K. Data envelopment
analysis. Kluwer Academic publishers, Boston, 2000.
Das, A. Ghosh, M. S. Financial Deregulation and efficiency:
An empirical analysis of Indian banks during the post
reform period, Review of Financial Economics, 15,
2006. 193–221.
Drake, L. Hall, M. J. B. Efficiency in Japanese banking,
Journal of Banking and Finance, 27, 2003. 891–917.
Drake, L. Hall, M. J. B. Simper, R. The impact of
macroeconomic and regulatory factors on bank
efficiency: A non-parametric analysis of Hong Kong’s
banking system, Journal of Banking and Finance, 30,
2006. 1443–1466.
Erdem C, Erdem M.S. Turkish banking efficiency and its
relation to stock performance. Applied Economics
Letters, 15, 2008. 207–211.
Fama, E F. Efficient capital markets: A review of theory and
empirical work. Journal of Finance, 25, 1970. 383–417.
Fare, R. Grosskopf, S. Lovell, C.A The measurement of
efficiency of production, Kluwer Academic Publishers,
Boston, 1985.
134
Inui, T. Park, J. Hyun-Han, S International comparison
of Japanese and Korean banking efficiency: Comments
and discussion, Seoul Journal of Economics, 21, 2008.
1–16.
International Monetary Fund Survey Islamic banks: More
resilient to crisis, Washington, D.C., 2010.
Ioannidis, C. Molyneux, P. Pasiouras, F. The relationship
between bank efficiency and stock returns: Evidence
from Asia and Latin American, University of Bath,
School of Management, Working paper 2008.10, 2008.
Iqbal, Z. Islamic financial systems,
Development, 34, 1997. 42–45.
Finance
and
Iqbal, Z. Challenges facing Islamic financial industry,
Journal of Islamic Economics, Banking and Finance, 3,
2007. 1–14.
Iqbal, Z. Van Greuning, H Analyzing banking risk for
Islamic banks, World Bank Publications, Washinghton,
USA, 2007.
Isik, I. Hassan, K. M. Efficiency, ownership and market
structure, corporate control and governance in Turkish
banking Industry, Journal of Business Finance and
Accounting, 30, 2003. 1363–1421.
Kamaruddin B H, Safa M S, Mohd R. Assessing production
efficiency of Islamic banks and conventional bank
Islamic windows in Malaysia. International Business
Management Research, 1, 2008. 31–48.
Islamic banking and finance – Essays on corporate finance, efficiency and product development
The relationship between Islamic bank efficiency and stock market performance: Evidence from GCC countries
Kasman A, Yildirim C. Cost and profit efficiency in transition
banking: the case of new EU members. Applied
Economics, 38, 2006. 1079–1090.
Kyj, L. Isik, I. Bank x-efficiency in Ukraine: An analysis
of service characteristics and ownership, Journal of
Economics and Business, 60, 2008. 369–793.
Liadaki, A. Gaganis, C. Efficiency and stock performance
of EU banks: Is there a relationship? Omega, 38, 2010.
254–259.
Lewis, M (2008). In what ways does Islamic banking differ
from conventional finance, Journal of Islamic Economics
Banking and Finance, 4, 9–22.
Mariani, A.M. The input requirements of conventional and
Shariah-compliant banking, The International Journal
of Banking and Finance, 7, 2010. 51–78.
Mohamad, S. Hassan, T. Bader M. Efficiency of conventional
versus Islamic banks: international evidence using
stochastic frontier approach, journal of Islamic
Economics, Banking and Finance, 24, 2008. 107–130.
Mokhtar, H. S. Abdullah, N. Alhabshi, S. M. Efficiency
and competition of Islamic banking in Malaysia,
Humanomics, 24, 2008. 28–48.
Moody’s (2008). Islamic banks in the GCC: A comparative
analysis.
Olson, D. Zoubi, T. Using accounting ratios to distinguish
between Islamic and conventional banks in the GCC
region, The Intermediation Journal of Accounting, 43,
2008. 45–65.
Pasiouras, F Estimating the technical and scale efficiency
of Greek commercial banks: The impact of credit risk,
off-balance sheet activities, and international operation,
Research in International Business and Finance, 22,
2008. 301–318.
Pasiouras, F. Liadaki, A. Zopounidis, C. Bank efficiency
and share performance: Evidence from Greece, Applied
Financial economics, 18, 2008. 1121–1130.
Paul, S. Kourouche, K. Regulatory policy and the efficiency
of the banking sector in Australia, The Australian
Economic Review, 41, 2008. 260–271.
Rezvanian, R. Mehdian, S. (2002). An examination of cost
structure and production performance of commercial
banks in Singapore, Journal of Banking and Finance,
26, 79–98.
Eds. Hatem A. El-Karanshawy et al.
Rezvanian, R., Rao, N., Mehdian S. M. Efficiency change,
technological progress and productivity growth of
private, public and foreign banks in India: Evidence
from the post-liberalization era, Applied Financial
Economics, 18, 2008. 701–713.
Rosly, SA. Abu Baker, MA. Performance of Islamic and
mainstream banks in Malysia, International Journal
Social Economics, 30, 2003. 1249–1265.
Srairi, S. A comparison of the profitability of Islamic and
conventional banks: The case of GCC countries, Bankers,
Markets, Investors, 98, 2009. 16–27.
Srairi, S. Cost and profit efficiency of conventional and
Islamic banks in GCC countries, Journal of Productivity
Analysis, 34, 2010. 45–62.
Sufian, F. Determinants of bank efficiency during unstable
macroeconomic environment: Empirical evidence
from Malaysia, Research in International Business and
Finance, 23, 2009. 54–77.
Sturm, J. E., Williams B. Foreign bank entry, deregulation
and bank efficiency: Lessons from the Australian
experience, Journal of Banking and Finance, 28, 2004.
1775–1799.
Sufian, F. Abdul Majid, M.Z. Bank efficiency and share
prices in China: Empirical evidence from a three-stage
banking model, International Journal of Computational
Economics and Econometrics, 1, 2009. 23–47.
Sufian, F. Noor, M.A Abdul Majid, M.Z. The efficiency of
Islamic banks: Empirical evidence from the MENA and
Asian countries Islamic banking sectors, The Middle
East Business and Economic Review, 20, 20081–19.
Xiang, D. Shamsuddin, A. Efficiency and stock market
performance of Australian banks, paper presented at
the proceedings of the Asian Finance Association 2009
Conference. Brisbane, June 30 to July 3, 2009.
Zahar, T.S. Hassan, K.H. A comparative literature survey
of Islamic finance and banking, Finance Markets
institutions and instruments, 10, 2001155–199.
Zaim, O. The effect of financial liberalization of the
efficiency of Turkish commercial banks, Applied
Financial Economics, 5, 1995. 257–264.
135
Conventional banks versus Islamic banks:
What makes the difference?
Huseyin Aytug1, Huseyin Ozturk2
Department of Economics, University of California at Santa Cruz
[email protected]
1
2
Abstract - This paper investigates the determinants of banking profitability in the Turkish banking
sector between 2003 and 2011. In addition, we calculate the effect of being an Islamic bank on
banking profitability, which allows us to differentiate conventional and Islamic banks. We introduce
the method of propensity score matching to the banking literature in order to estimate the average
treatment effect (ATE) of being an Islamic bank in Turkey where there exists a dual banking system.
The results show that in terms of return on asset (ROA) and return on equity (ROE), being an Islamic
bank does not create any difference. However, being an Islamic bank turns out to have a significant
and negative effect on net interest margin (NIM). These results have many policy implications in the
Turkish banking industry where Islamic banks mimic others to be one of the leading examples.
Keywords: average treatment effect, propensity score matching, Islamic bank, profitability
JEL Classification: C51, G15
1. Introduction
Islamic finance is a source of funding that complies with
Islamic jurisprudence. This source of funding has already
been in practice in countries where the majority of the
population is Muslim; however, the importance of Islamic
finance has prevailed in the global financial system recently.
Although the distinction between Islamic and conventional
finance needs deeper understanding, the main difference
within Islamic finance can be attributed to the Islamic
idealism of creating a moral economy where profits come
from commerce or real transactions not from money
lending or speculative transactions. While conventional
finance helps direct the flow of capital to investment
opportunities that are supposed to provide the highest
return in the market, Islamic finance allegedly seeks socioeconomic optimality. Another principle of Islamic finance
is that investment is expected to produce an optimal socioeconomic outcome in line with Islamic norms.
The globalization has affected the scope and breadth of
Islamic capital markets in a way that its pervasiveness
has reached a global scale. Moreover, the financial crisis,
which has recently erupted and hit the global finance and
economy severely, has paved the way for other means
of financial schemes. Islamic finance, which has stayed
relatively resilient, became a focus of the global financial
system since then.1 The bulk of Islamic funds staying in
financial hubs of many Gulf States and in global financial
networks promises a lucrative and stable source of funding.
Despite its growing popularity, Islamic banking remains a
small part of the total financial sector and will likely remain
so due to lack of penetration in the market and strong
competition that is challenged from the conventional
banking system.
The most explicit distinction of Islamic finance from
conventional finance, where the return is based upon
interest, is the prohibition of riba (interest based lending)
and gharar (speculation or uncertainty). The underlying
contracts of Islamic products differ from their conventional
counterparts in structure as well. For example murabaha (a
sale-based instrument), which is similar to a conventional
loan, involves the purchase of an asset by a bank and
its sale to a client at a cost plus a pre-determined profit.
The structure of a sale has important legal implications
according to the Islamic rules, which extensively dictate the
terms of risk and mutual consent. Other sale-based products
include the financing of commissioned manufacturing or
construction (istinaa) and the forward-sale (salam), and
these types of products require structural differences from
their conventional equivalents. Lease-based instruments
(ijara) are similar to the traditional leasing with certain
distinctions and equity-based financial intermediation,
which is known as mudaraba, takes place through profit
and loss arrangements (Warde, 2010). Another product
that recently became a regular budget financing instrument,
Cite this chapter as: Aytug H, Ozturk H (2015). Conventional banks versus Islamic banks: What makes the difference?
In H A El-Karanshawy et al. (Eds.), Islamic banking and finance – Essays on corporate finance, efficiency and product
development. Doha, Qatar: Bloomsbury Qatar Foundation
Developing Inclusive and Sustainable Economic and Financial Systems
138
30.15
28.40
24.33
38.31
51.89
67.95
60.20
67.69
61.93
30.30
46.11
94.24
94.02
94.26
95.01
95.07
95.82
96.25
96.62
95.94
94.38
95.16
5.76
5.98
5.74
4.99
4.93
4.18
3.75
3.38
4.06
5.62
4.84
12.42
22.66
25.56
14.91
22.24
12.96
2.83
26.97
21.00
18.89
19.94
17.79
26.90
25.37
51.83
54.38
81.50
16.62
48.25
50.19
30.47
40.33
95.98
95.98
96.11
96.11
97.02
97.63
98.51
98.45
97.90
96.00
96.95
4.20
4.02
3.89
3.89
2.98
2.37
1.49
1.55
2.10
4.00
3.05
24.68
17.05
16.14
23.98
15.87
28.42
24.80
24.26
23.34
20.46
21.90
Market
Share:
Loans
CB
Market
Share:
Loans
IB
Growth:
Equity
CB
Growth:
Equity
IB
Market
Share:
Equity
CB
Market
Share:
Equity
IB
Growth:
Assets
CB
Growth:
Assets
IB
31.25
25.21
31.32
37.18
38.68
60.47
43.65
45.21
47.00
31.24
39.12
95.61
95.82
96.08
96.52
96.84
97.35
97.87
98.14
97.55
96.01
96.78
4.39
4.18
3.92
3.48
3.16
2.65
2.13
1.86
2.45
3.99
3.22
2011
2010
2009
2008
2007
2006
2005
2004
2004–2008
2008–2011
2004–2011
Following the Asian financial crises in the late 1990s, the
Turkish economy experienced tremendous volatility, which
caused some consolidation of the special finance houses.
Poor regulation, the accumulation of public debt, and
politically driven lending habits contributed to a severe
financial crisis in Turkey in 2001. This affected all strands of
the domestic banking sector, although conventional banks
Market
Share
CB
Islamic banking in Turkey followed a parallel path to
the history of the Turkish economy. Turkey had a highly
centralized economy, whereby state institutions owned and
managed most important industries until the beginning of
the 1980s (Ozturk et al., 2010). However, the 1980s saw
a period of liberalization of the tightly controlled Turkish
economy. As part of a plan to attract more foreign direct
investment from the Arab Gulf states, a decree passed in
1983 legalized the operation of special finance houses to
provide interest-free banking. These institutions were
highly regulated but they did not gain the same status as
conventional banks, e.g., they were not covered by the
insurance scheme the other banks in the system utilized and
could not invest in government securities by its nature.2
Market
Share
IB
Table 1–2 gives an idea about the market share and the
growth of conventional and Islamic banks between 2004
and 2011. It is striking that the market share of Islamic
banks in assets, equity and loans has grown from 1% to 5%
approximately and the growth rate of these fundamentals
have been higher than the conventional banks. Despite the
fact that the growth rate has declined after the crisis in
2008, it is still higher than its counterparts.
Table 1. Market share and growth in assets, equity and loans of Islamic and commercial banks.
Islamic finance in Turkey is yet a more recent issue. Changes
in domestic financial systems and public sensibilities have
allowed participation banking to gradually become more
visible. They have also emerged resilient in the context of
two periods of economic turmoil: the domestic financial
crisis of 2001 and the global financial crisis of 2008. The
severe banking crisis in 2001 did not have as much inverse
effects on participation banks as the conventional banks.
The contagious effect of the 2001 crisis had a limited effect
on participation banks due to lack of interbank activities
by participation banks. The growing financial capacity of
the religiously conservative public has also been another
factor that made participation banks attractive (Hardy,
2012). These types of banks became the sole option for
those people who are resistant to conventional banking.
The increasing level of associated client portfolios and
deposits have enabled participation banks to reach over 4.5
percent of market share in total assets from nearly 1 percent
in 2001. Recently a new legislation passed to facilitate
Islamic banking in the private sector, and government
officials have indicated interest in issuing sovereign sukuk
(rent certificate), comparable to bonds, for funding central
government budget requirements.
Growth: Growth
Loans
Loans
IB
CB
sukuk, is an asset-based instrument of which tangible
assets are specified according to the type of the issuance.
In comparison with conventional debt instruments,
these products do not pledge fixed income or an interestbased income stream. Yet, it collects returns based on the
collection of lease or sale of certain assets that are specified
beforehand. These types of products are predominantly
used in the Middle East, but their prevalence is visible in East
Asian countries as well as developed European markets.
35.51
22.91
7.30
36.32
28.00
49.86
43.62
45.87
41.84
25.51
33.67
Aytug and Ozturk
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Conventional banks versus Islamic banks: What makes the difference?
Table 2. Banks’ size and market share (*share of assets/**share of equity/***share of loans).
2011-Q4*
Banks
Akbank
Denizbank
Finans Bank
HSBC
ING Bank
Sekerbank
TEB
TR Ziraat
TR Garanti
TR Halk
TR Is
TR Vakiflar
Yapi Kredi
Albaraka Turk
Bank Asya
Kuveyt Turk
Turkiye Finans
2007-Q4*
2011-Q4**
2007-Q4**
2011-Q4***
2007-Q4***
IB/CB Overall IB/CB Overall IB/CB Overall IB/CB Overall IB/CB Overall IB/CB Overall
11.51 10.98
3.10 2.96
3.98 3.80
2.08 1.98
1.81 1.73
1.24 1.18
3.28 3.13
13.84 13.21
12.63 12.05
7.85 7.49
13.93 13.39
7.68 7.33
9.31 8.88
18.65 0.86
30.65 1.41
26.57 1.22
24.12 1.11
12.15 11.75
2.66 2.57
3.72 3.60
2.39 2.31
2.23 2.16
1.08 1.05
2.10 2.03
14.42 13.94
12.04 11.64
7.17 6.93
14.29 13.81
7.56 7.30
8.97 8.67
18.99 0.64
32.21 1.08
19.90 0.67
28.90 0.97
12.68 12.14
2.85 2.73
4.11 3.94
2.02 1.93
1.71 1.63
1.06 1.01
3.04 2.91
9.52 9.11
12.70 12.15
6.24 5.97
12.94 12.39
6.72 6.43
8.45 8.09
16.22 0.69
34.51 1.48
23.22 0.99
26.05 1.12
suffered more than the special finance houses since they
had a much larger role in the overall economy. Although
the crisis had a major negative effect on the entire banking
sector, the special finance houses also suffered. Turkish
holding company Ulker purchased Faisal Finans in 2000,
changing its name to family Finance House. Then, Ihlas
Finans filed for bankruptcy in 2001 as the liquidity crisis in
Turkey reached its peak. This showed that the participation
banks were not immune to crises even tough they
functioned with a different business model. The reasons
for this were twofold; the participation banks were not
decoupled from the whole financial system due to business
connections with other banks and the economic crisis
hit the overall economy. Yet, it is worth mentioning here
that the prohibition of holding public debt protected the
participation banks from a worse shock than what could
have otherwise ensued. In the initial stages of the financial
crisis, Ihlas and other participation banks, that could
have held liquid government securities, were thus not as
greatly affected as conventional banks. However, when the
liquidity crisis hit, Ihlas exposure led it to a collapse, and it
experienced a traditional bank run on its deposits.
The 2001 crisis led to a rehabilitation of Turkey’s financial
system, and the parliament passed a new law in order to
discipline the overall banking system (Law No. 4389). In
addition to strengthening banking regulations and creating
new oversight bodies for conventional banks, Law No. 4389
founded the Union of Private Finance Houses in order to
address common issues among participants and provide
a level of state control for the sector. All special finance
institutions were required to become members of this
association, but they still lacked many of the privileges that
conventional banks had, such as the provision of deposit
insurance. In 2006, banking law No. 5411 officially replaced
the term “special finance institutions” with the name
“participation banking.” Participation Banks Association of
Turkey was established and employed with the unification
of Private Finance Houses. The new law created a savings
Eds. Hatem A. El-Karanshawy et al.
14.43 13.98
1.98 1.92
3.57 3.46
2.75 2.66
1.73 1.68
1.18 1.14
1.24 1.20
9.82 9.52
9.37 9.07
5.96 5.78
14.43 13.98
7.11 6.89
6.67 6.47
22.58 0.70
36.12 1.13
16.44 0.51
24.86 0.77
10.58 9.97 13.20 12.55
3.38 3.18 3.71 3.53
4.56 4.29 5.05 4.81
2.08 1.96 3.33 3.17
2.31 2.18 3.03 2.89
1.28 1.21 1.29 1.23
3.86 3.64 2.45 2.33
10.75 10.13 7.70 7.32
12.62 11.88 13.27 12.62
8.46 7.97 6.46 6.14
13.79 12.99 12.12 11.52
8.63 8.13 8.37 7.96
10.20 9.61 10.17 9.66
17.71 1.03 18.56 0.91
32.06 1.86 29.74 1.46
25.01 1.45 20.37 1.02
25.22 1.47 30.97 1.52
deposit insurance fund for participation banks as well. In
doing so, the insurance scheme began to cover the whole
banking system.
These changes may represent a shifting paradigm in the
level of acceptance for participation banking in Turkey.
Participation banking emerged stronger after each of these
periods of instability. Evidently, participation banking’s
role in the economy will also likely grow as Turkey
considers options for attracting investors from the Gulf
region who are currently highly liquid in terms of capital
and also religiously conservative. Therefore, the recent
developments in Islamic banking bring old debates to
discussion again whether they are really different from the
conventional banks or not.
The current literature on Islamic banking addresses the
issue of whether the distinction between conventional
banks and Islamic banks is only their names or the
distinction also appears in their business model. As a bold
example, Kassim et al. (2009) questions the argument
whether Islamic banks are not susceptible to the interest rate
changes as compared to their counterparts where both of
them operate in tandem. Their research question emanates
from the basic proposition whether they differentiate in
behavior to common macro-economic shocks. We aim to
contribute to the current literature by putting Turkey into
the center and investigating the profitability issue with a
focus on Islamic and conventional bank differentiation.
Although the determinants of profitability in conventional
banks in Turkey have been a subject of some research, there
have been a few studies conducted regarding profitability
of the participation banks in the literature (Macit,
2012). Islamic banking in Turkey is under an interesting
transformation that is reflected in their asset and equity
growth. This transformation opens a new debate: what
are the determinants of this change? Do Islamic banks
really differ from their counterparts? To what extent do
the dynamics associated with the performance of Islamic
139
Aytug and Ozturk
banks differs from their counterparts? These questions
will be addressed in this study.
We use a broad set of data to investigate the determinants of
bank profitability. The bank specific variables that may have
an impact on bank profitability are selected in accordance
with the current literature (e.g., Athanasoglou et al. (2008);
Garca-Herrero et al. (2009); Dietrich and Wanzenried
(2011). The macroeconomic and industry specific variables
that are considered to be influential on bank efficiency are
also included. The recent financial crisis is also considered
by separating the whole period as two sub-periods: pre crisis
and post crisis. This approach is similar to Hasan and Dridi
(2011); Dietrich and Wanzenried (2011), who separate the
whole period as pre crisis and post crisis. One of the main
contributions of this paper is that we introduce the method of
propensity score matching to the banking literature in order
to estimate the average treatment effect (ATE) of being an
Islamic bank where there exists a dual banking system. The
results are compared with the ordinary least squares (OLS)
estimation results. We make use of a unique and up-to-date
database by combining quarterly conventional bank and
participation bank data. The results provide insightful policy
making implications and will be discussed in the upcoming
sections. All in all in our study seeks to examine a series of
questions. First, the issue of being an Islamic bank on bank
profitability will be mainly discussed. This issue deserves
particular attention due to motivation discussed above.
Beyond this main scope, the determinants of profitability of
the Turkish banking industry will be explored. This is also
crucial to study considering the limited number of studies
specific to Turkish banking. Last but not the least, the effect of
recent financial crisis on profitability of banks will be explored.
The recent financial crisis that caused great havoc in the
banking sector of many countries – e.g., many defaults or bailouts have taken place very recently especially in developed
country space – reasonably might affect the profitability.
The remainder of the study is organized as follows: the
second section will explain the data. The data set we compile
is the largest data set as to the best of our knowledge. The
third section will discuss the methodology in some detail.
The ATE methodology and the model specifications will be
discussed in this section. The fourth section will discuss the
results and policy implications. The policy implications and
policy recommendation will be provided together specific
to the Turkish case. The fifth section will conclude.
2. Literature review
In the studies, which investigate the Islamic banking and
conventional banking dualism, the recent findings reveal
that there is not a fundamental difference in terms of
their routine activities. In other words, they show similar
responses to basic impulses, e.g., their profitability measures
respond to market interest rates in a similar way. Ergec
and Arslan (2011) contend that Islamic banks, relying on
interest-free banking, shall not be affected by the interest
rates; however, in concurrence with the previous studies,
the article finds that the Islamic banks in Turkey are visibly
influenced by interest rates. This study differentiates from
Kassim et al. (2009) in one respect. Kassim et al. (2009),
claim that the primary reason why Islamic banking may
become more stable compared to conventional banks is
that they are not affected by the fluctuations on interest
140
rates. In other words, Islamic banks are expected to be more
stable than the conventional banks where Islamic banking
is not influenced by interest rates. Stability in demand for
money holds some positive effects in terms of efficiency in
monetary policies and the financial stability in the system.
On the other hand, Kia and Darrat (2007) refer to two
major reasons why interest-free Islamic banks contribute to
the stability more than the others. A first reason is related
with the demand for money whereas the second one is an
assessment with the balance sheet perspective. Of these
factors determining demand for money, interest rates
appear to be the most effective component for speculation.
Thus, interest free banking reduces stability in the banking
system since Islamic banks shy away from interest rate.
As for the balance sheet perspective, along with changes
in the interest rates, the banks revalue their assets before
liabilities. The loan interest rates respond to a change in the
interest rate much earlier than the savings interest rates.
In this case, the revaluation of balance sheet entries, a key
component for profit maximization, makes the impact of
interest rate changes more vulnerable. However, in the
Islamic banking system, there is no need for revaluation of
balance sheet entries because there is no risk of interest rate.
All in all, for these two primary reasons, it is anticipated
that the banking system, dominated by conventional
banking, is more unstable. To test for the stability issue in
the Islamic banking system, there are also plenty of studies
investigating the country experiences from interest rate
change perspective. All these studies present empirical
findings suggesting that demand for money is more stable in
banking systems where Islamic banks are the key players.3
Since the early works by Short (1979), Kwast and Rose
(1982) and Bourke (1989), a considerable amount of recent
studies have investigated some of the major determinants of
bank profitability. The empirical studies have focused their
analyses either on cross-country evidence or on the banking
system of individual countries. The studies by Molyneux
and Thornton (1992), Demirguc and Huizinga (1999),
Goddard, et al. (2004), Micco, et al. (2005), and Pasiouras
and Kosmidou (2007) investigate a panel data set. Studies
by Berger, et al. (1987), Berger (1995), Michelle (1997),
Bennaceur and Goaied (2008), Athanasoglou, et al. (2008)
and Garca-Herrero, et al. (2009) center their analyses on
single country cases.
Bank profitability is vastly measured as a function of return
on asset (ROA) or return on equity (ROE).
Some works also add net interest margin (NIM) as a
complementary measure that it related with profitability.
The literature classifies determinants as being internal and
external determinants. The internal determinants include
bank-specific variables of which the intrinsic features of
individual banks compose this bloc. The external variables
reflect factors that are expected to affect the profitability
of banking industry although banks are not capable of
controlling them.
In most studies, variables such as bank size, risk, capital ratio
and operational efficiency are used as internal determinants
of banking profitability. Pasiouras and Kosmidou (2007)
find a positive and significant relationship between the size
and the profitability of a bank. This is due to the fact that
larger banks are likely to have a higher degree of product
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Conventional banks versus Islamic banks: What makes the difference?
and loan diversification than smaller banks. Furthermore,
large banks benefit from economies of scale. Berger, et al.
(1987), provide evidence that costs are reduced only slightly
by increasing the size of a bank and those very large banks
often encounter scale inefficiencies. Micco, et al. (2005) find
no correlation between the relative bank size and the return
on assets for banks, i.e., the coefficient is always positive but
never statistically significant. Therefore the impact of being
a big bank in size on profitability is mix.
If it is the risk that is concerned in this literature, Abreu and
Mendes (2002) who examined banks in some European
countries, find that the loans-to-assets ratio, as a proxy for
risk, has a positive impact on the profitability of a bank.
Bourke (1989) and Molyneux and Thornton (1992), among
others, find a negative and significant relationship between
the level of risk and profitability. This result reflects the
fact that banks that are exposed to high risk also have a
higher accumulation of non-performing loans, and nonperforming loans lower the returns of the affected banks.
Another bank feature that is suggested to be effective on
the profitability of banks is the asset composition of the
banks. Empirical evidence by Bourke (1989), Demirguc
and Huizinga (1999), Abreu and Mendes (2002), Goddard,
et al. (2004), Bennaceur and Goaied (2008), Pasiouras and
Kosmidou (2007) and Garca-Herrero, et al. (2009) indicate
that the best performing banks are those that maintain
a high level of equity relative to their assets. The authors
explain this relation with the observation that banks with
higher capital ratios tend to face lower costs of funding
due to lower potential bankruptcy costs. I.e., equity has the
lowest order to be paid during liquidation.
One more bank-specific variable is the ownership of a
bank. Private or state owned banks are the most visited
separation in the literature. This separation has many
insightful findings. For instance, the owner-ship structure
plays an important role in explaining banking profitability.
Micco, et al. (2007) found that the separation of banks
as privately owned or state-owned provides insights in
examining bank performance. According to their results,
state-owned banks operating in developing countries tend
to have a lower profitability, lower margins, and higher
than comparable privately owned banks. In industrialized
countries, however, this relationship is found to be much
weaker. Iannotta, et al. (2007) point out that government
owned banks exhibit a lower profitability than privately
owned banks. Therefore, bank ownership structure is also
important regardless of being developed or developing,
yet the analysis suggests a different degree of significance.
Another strand of bank ownership is being a foreign
bank. The international connection of a bank may have
a significant impact on profitability. This is plausible in
a sense that foreign banks have easy access to a pool of
funds abroad that domestic banks cannot easily reach.
Yet in terms of profitability issues, being a foreign bank is
found to have differing impact on profitability. Demirguc
and Huizinga (1999) suggest a significant relationship, on
the other hand Bourke (1989) and Molyneux and Thornton
(1992) find this relationship insignificant.
Many studies in profitability literature take—such as
central bank interest rate, inflation, GDP growth etc.—
another bloc of variables that affect bank profitability.
Eds. Hatem A. El-Karanshawy et al.
Most studies have shown a positive relationship between
inflation, central bank interest rates, GDP growth, and
bank profitability (e.g., Bourke (1989), Molyneux and
Thornton (1992), Demirguc and Huizinga (1999),
Athanasoglou, et al. (2008), Albertazzi and Gambacorta
(2009)). As per the effects of macroeconomic variables, the
effect of inflation rate on bank profitability depends upon
whether the inflation is anticipated or unanticipated. In the
case of an anticipated inflation bank profits may improve
as the banks may adjust the price of lending according to
the inflation rate. However, an unanticipated inflation
may have negative effects. Bourke (1989) and Molyneux
and Thornton (1992) find that a higher inflation rate is
associated with better profitability indicators. Recently,
Macit (2012) has recorded similar findings.
To measure the effects of market structure on bank
profitability, the structure conduct and performance
(market-power) hypothesis states that increased market
power yields monopoly profits. The inverse relation
between the degree of market concentration and degree
of competition has been the underlying assumption of
the structure conduct performance hypothesis. The bank
concentration is discussed in some studies. The Herfindahl
Hirschman index was used to proxy the level of competition
in the industry (see Dietrich and Wanzenried (2011)).
According to the results of Bourke (1989) and Molyneux
and Thornton (1992), the bank concentration ratio shows
a positive and statistically significant relationship with the
profitability of a bank and is, therefore, consistent with the
structure conduct performance paradigm. In contrast, the
results of Demirguc and Huizinga (1999) and Staikouras
and Wood (2002) indicate a negative but statistically
insignificant relationship between bank concentration and
bank profits. Likewise, the estimations by Berger (1995)
and E.C. and P.C. (2003) contradict the structure-conduct
performance hypothesis. As briefly discussed above, the
determinants of bank profitability can be defined with three
blocks of variables briefly discussed above. The literature
is more or less similar in terms of the data selection. The
variation of data employed in the analysis is rather sparse.
Therefore the existing literature provides a comprehensive
examination of the effects of bank-specific, industry-specific,
and macroeconomic determinants on bank profitability.
In this study, we take being an Islamic bank as the centre
and investigate the effect of being an Islamic bank on
profitability. This contribution is vital in investigating the
on-going transformation in the banking sector of Turkey.
Considering the lack of studies on banking profitability in
the case of Turkey, especially the special attention given to
Islamic banks, this study bridges an important gap. Dietrich
and Wanzenried (2011), underlines the relative scarcity of
literature that discusses the effect of the recent financial
crisis on bank profitability. In our study, we also address
this issue by making pre-crisis and post-crisis analysis. The
data employed in our analysis is in line with the current
literature with small variations that will be discussed in the
next section.
3. Data
There is a well-established set of determinants available
to investigate the profitability of the banking system in
Turkey. To examine the effect of being an Islamic bank
on banking profitability, we rely on our data set in line
141
Aytug and Ozturk
with current literature. The data is gathered from the
quarterly unconsolidated balance sheets of banks that
operated between 1994Q1 and 2011Q4. The balance
sheets are obtained from The Banks Association of
Turkey and Banking Regulation and Supervision Agency.
There are 29 conventional banks and 4 Islamic banks in
1994Q1–2011Q4. Macroeconomic variables are from
Central Bank of Turkey and the Undersecretariat of
Treasury who are responsible for economy management
in Turkey. As per the effects of macroeconomic variables,
the real interest rate on government bonds used to proxy
interest rate. The daily interest rate of government bonds is
not available. Therefore we use one-year T-bill rates at the
data issue. These will proxy for the interest rates for each
and every quarter. The level of foreign exchange rate is the
USD/TRY rate and an increase in exchange rate implies
depreciation in Turkish Lira.
Islamic banks constitute a small portion of the banking
system in Turkey but potential to grow, may be in size rather
than number. Of these four banks, two of them are open to
public and are daily traded in the stock market. Three of
these four banks are foreign and only one bank is domestic.
In terms of bank specific determinants, we look at various
different variables, namely the ratio of equity to total assets,
the ratio of net loans to total assets, log of real assets, and the
ratio of non-performing loans to total loans. To control for
macroeconomic determinants of profitability, we use GDP
growth, level of foreign exchange rates, consumer inflation,
and real interest rate. Turkey has experienced great growth
during the period subject to analysis. Therefore, GDP
growth is expected to have a positive impact on profitability
regardless of being conventional or Islamic.
Industry-wise we do not include concentration measures
in our analysis. The Turkish banking industry constitutes
a competitive market without a dominant group of
banks or single bank. Dietrich and Wanzenried (2011)
take concentration into their analysis. Yet, their focus
is on the Swiss banking industry, where there exists a
huge concentration. As per industry specific variables
we use growth measures. The reason why we employ
growth measures is that the growth in the banking
industry was visible in the Turkish case in the last decade.
The recapitalisation was the major theme of the banking
industry (see Yeldan, 2007).
We define state bank as the base and present three dummies
for private bank, state banks and Islamic banks. In doing so,
we aim at controlling for industry specific effects on bank
profitability. The variables that are used in our analysis are
detailed in Table 3.
Table 4 provides descriptive statistics for our sample
showing the observations, means and standard deviations
of all variables. Observations are divided into two groups
Table 3. Variables used in the Empirical Analysis
Bank-Specific Variables
Market-Specific Variables
Macroeconomic Variables
Dummy Variables
Variable
Notation
Measure
Return on asset
Return on equity
Net interest margin
ROA
ROE
NIM
Capital adequacy
Asset quality
Asset size
ETA
LTA
NPLTA—log (TA)
Liquidity
Deposits
Credit risk
Liquidity risk
Total asset growth
Total equity growth
Total loans growth
GDP growth
Inflation rate
Exchange rate
LQD
DPTA
CR
LR
TAG
TEG
TLG
GDPGR
INF
FX
Real interest rate
Islamic bank dummy
IR
IB
Ownership dummy
PD
Net income/total assets
Net income/total equity
Net interest income/total
assets
Equity/total assets
Loans/total assets
Non-performing loans/total
assets (Natural logarithm of
total assets)
Liquid assets/total assets
Deposits/total assets
Loans loss provision/loans
Loans/deposits
Growth of total assets
Growth of total equity
Growth of total loans
Growth of GDP
Inflation rate
Exchange rate between
TL and $
Real interest rate
1=Islamic banks,
0=Normal banks
1=Privately owned;
0=Publicly owned
1=Foreign owner;
0=Domestic owner
1=After the crisis;
0=Before the crisis
FD
Crisis dummy
142
CD
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Conventional banks versus Islamic banks: What makes the difference?
Table 4. Descriptive Statistics
Conventional Banks
Islamic Banks
Variable
Obs
Mean
Std. Dev.
Min
Max
Obs
Mean
Std. Dev.
Min
Max
ROA
ROE
NIM
ETA
LTA
NPLTA
LOGTA
LQD
DPTA
CR
LR
1044
1044
1044
1044
1044
1044
1044
1044
1044
1044
1044
0.013
0.074
0.030
0.174
0.413
0.023
6.574
0.217
0.513
0.029
1.138
0.027
0.128
0.028
0.140
0.212
0.022
0.890
0.182
0.235
0.146
1.883
–0.176
–1.786
–0.0217
0.037
0.001
0.000
4.346
0.009
0.000
0.000
0.000
0.322
0.465
0.169
0.916
0.847
0.177
8.231
0.812
0.903
3.095
34.532
144
144
144
144
144
144
144
144
144
144
144
0.007
0.058
0.017
0.118
0.694
0.036
6.569
0.080
0.786
0.008
0.897
0.003
0.033
0.009
0.024
0.089
0.022
0.391
0.050
0.068
0.002
0.186
0.001
0.007
0.005
0.073
0.436
0.005
5.664
0.016
0.322
0.001
0.518
0.024
0.236
0.063
0.183
0.835
0.135
7.236
0.235
0.880
0.012
2.012
as conventional banks with 1044 observations and Islamic
banks with 144 observations. Simple inspection of the table
shows that conventional banks are likely to perform better
in profitability since the means of return on asset, return on
equity and net interest margin are higher. For conventional
banks, ROA is 1.3%, ROE is 7.4% and NIM is 3% on average.
For Islamic banks, ROA is 0.7%, ROE is 5.8% and NIM is 1.7%.
On the other hand, credit risk and liquidity risk are lower for
Islamic banks as one may expect this result because of the
risk-sharing principle. While credit risk and liquidity risk are
2.9% and 114% for conventional banks, 0.8% and 89% for
Islamic banks. Another issue that is worth underlining is the
total cumulative asset growth, which is higher for Islamic
banks than conventional banks, being 31.28% and 25.64%
respectively. Correlation matrix for independent variables
is presented in Table 5. Correlations among most of the
variables are quite low, signalling multi-collinearity does not
create significant bias in any of our analyses.
4. Methodology
In the banking literature, it is quite common to use ordin­ary
least square (OLS) methods. The recent studies also employ
a generalized method of moments (GMM) since the profit
persistence is the common feature of banking data. The
main focus of our study is to observe the differing behavior
of Islamic banking in profitability. To observe the differing
behavior, the sample can be split into two sub-samples, or
pooled estimation can be done through assigning a dummy
variable to examine the “being an Islamic bank” effect.
However, one of the main problems in such analysis is the
selection bias or non-random selection that may have been
produced by OLS or GMM estimators. Specifying some
certain banks as Islamic banks and investigating the effect
of being an Islamic bank is a self-selection choice, therefore
it is possible to have self-selection bias.
In order to overcome this possibility, we introduce the
meth­
od of propensity score matching developed by
Rosenbaum and Rubin (1985) to the banking literature
in order to estimate the “average treatment effect” (ATE)
of being an Islamic bank in Turkey where there exists
a dual banking system. The purpose of this method is
to create a control group that is similar to a treatment
group, and the similarity among banks will be assessed
Eds. Hatem A. El-Karanshawy et al.
by calculating the propensity score that is defined as the
conditional probability of receiving a treatment despite the
unavailability of experimental data. In the present context,
while Islamic banks constitute the treatment group,
conventional banks constitute the control group, and being
an Islamic bank is defined as receiving the treatment.
IBjt is defined as a dummy variable whether bank j is an
Islamic bank at time t. While, y 1jt denotes the profitability
of bank j that is an Islamic bank at time, y 0jt denotes the
profitability of bank j at time t that is a conventional bank.
The average treatment effect of being an Islamic bank is
defined as:
τ jt = y 1jt − y 0jt , (1)
( )
( )
If both states of the world, y 1jt and y 0jt , were
observable, the average treatment effect would be
estimated with no trouble. Nevertheless, due to
unobservability, the average treatment effect (τ) would be
equal to the difference of mean outcomes ( y 1 − y 0 ) . Since
either of states of world, y 1jt and y 0jt , are observable,
we use propensity score matching to calculate the average
treatment effect. As argued in Rosenbaum and Rubin
(1985), a vector of covariates, Z, can be used to compare
Islamic and conventional banks. Z is defined as:
( )
( )
y 1jt , y 0jt ⊥ IB|Z Pr ( IB = 1|Z ) ∈(0, 1) , (2)
where ⊥ denotes independence. However it is not possible
to find observations with identical values for all covariates
in Z. In order to eliminate this problem, they suggest using
propensity score matching, which uses probability of bank
pairs that receive the treatment (IB) on the characteristics
of the pair. I estimate the probability of a bank-pair that
receives the treatment (IB) using the following logit
specifications.
(
)
p IB jt = 1 = F ( BSF , MSF , MAC ) , (3)
where BSF is a vector of bank-specific variables includes
capital adequacy, asset quality, credit risk, etc. MSF is a
vector of market specifics variables, including total asset
growth, total loans growth and total equity growth, and
143
144
ROA
ROE
NIM
ETA
LTA
NPLTA
LOGTA
LQD
DPTA
CR
LR
IB
PD
FD
TAGR
TEGR
TLGR
GDPGR
INF
FX
IR
−0.027
−0.076
−0.017
−0.050
−0.028
0.131
0.066
0.084
0.024
0.026
0.029
0.081
−0.254
1.000
0.643
0.246
0.420
−0.189
0.174
0.016
0.136
ROA
1.000
0.167
0.016
0.046
-0.003
0.265
0.078
0.006
0.024
0.136
0.043
0.200
0.098
0.035
0.104
0.034
0.077
0.013
0.034
0.034
ROE
Table 5. Correlation matrix
1.000
0.226
0.037
0.248
0.000
0.113
0.081
0.074
0.004
0.163
0.000
0.033
0.030
0.115
0.059
0.108
0.017
0.009
0.019
NIM
LQD
DPTA
1.000
0.357 −0.409 1.000
−0.113 0.199 −0.197
−0.081 0.036 −0.375
−0.002 −0.253 0.373
−0.421 0.161 0.055
−0.490 0.366 −0.185
0.007 −0.090 0.127
−0.036 −0.062 0.081
−0.038 −0.094 0.058
−0.507
1.000
NPLTA LOGTA
1.000
0.024 1.000
0.379 −0.033
0.549 0.019
0.581 0.000
0.258 0.137
0.020 −0.086
0.415 0.192
0.118 −0.044
−0.164 −0.091
0.141 −0.006
0.050 0.036
0.050 −0.042
LTA
−0.027 −0.252
0.001 0.022
0.066 0.098 0.069 0.014
0.005 −0.220 −0.014 −0.030
−0.067
−0.043 −0.065 −0.019 −0.050 −0.051 −0.006
0.006 −0.256 0.068 −0.199 0.006 −0.026
0.006
−0.063
0.424
0.062
0.137
0.162
0.077
−0.518
1.000
0.464
0.376
0.421
0.424
ETA
LR
IB
−0.066 1.000
−0.051 −0.045 1.000
0.042 −0.082 0.117
−0.071 0.087 −0.011
−0.026 −0.008 0.281
−0.024 −0.018 0.207
−0.021 0.010 0.142
0.033 −0.004 0.000
−0.011 −0.042 0.000
0.051 −0.048 0.000
−0.013 −0.022 0.000
1.000
CR
FD
TAGR
1.0000
0.326 1.0000
0.033 −0.003
1.0000
0.024 −0.002
0.111
0.017 −0.002
0.770
0.000 0.000
0.184
0.000 0.000 −0.308
0.000 0.000 −0.219
0.000 0.000 −0.295
PD
TLGR GDPGR
INF
FX
IR
−0.182 1.000
−0.117 0.501 1.000
0.052 −0.288 0.015 1.0000
−0.103 −0.359 −0.124 0.048 1.000
0.017 −0.242 −0.028 0.824 −0.105 1.000
1.000
TEGR
Aytug and Ozturk
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Conventional banks versus Islamic banks: What makes the difference?
MAC is a vector of macroeconomic variables, including GDP
growth, inflation rate, foreign exchange rate between the
Turkish lira and the US dollar and real interest rate. The logit
model is used here to identify the effect of Islamic banking
on banking profitability. By using the logit model we can
compare banking profitability of conventional and Islamic
banks since both types of banks are similar in terms of their
propensity scores. Next step is to use a matching technique in
order to estimate missing counterfactuals by using obtained
0
0
1
propensity scores, y 1jt , y tjt
or y jt , y tjt . There are
several methods used as matching technique but we use
three of them in our analysis, Nearest-Neighbor Matching,
Stratification Matching and Kernel Matching. Thanks to
these techniques we will be able to observe whether different
matching algorithms result in different treatment effects.
The average treatment effect of Islamic banking is given by
(
(
)
)
τ TT = E {E [ y 1|IB = 1, z ] − E [ y 0|IB = 0, z ]} , (4)
5. Results
To test the relationship between bank profitability and
the bank specific, market specific and macroeconomic
determinants described earlier, we first estimate a linear
regression model in the following form:
ROA = α + β 1 BSFit + β 2 MSFit + β 3 MAC it + ε it (5)
ROE = α + β 1 BSFit + β 2 MSFit + β 3 MAC it + ε it (6)
NIM = α + β 1 BSFit + β 2 MSFit + β 3 MAC it + ε it (7)
Tables 6,7 and 8 present regression results for three
different dependent variables (ROA, ROE and NIM).
In our analysis, the coefficients are estimated for the
entire, pre-crisis and post-crisis periods. For each time
period, we estimate three regression models while the
first one only includes bank specific factors, the second
one includes both bank and market specific factors, and
the third one includes bank-specific, market-specific and
macroeconomics factors. Table 6 depicts the regression
results when ROA is the dependent variable. According
to Table 6, the determinants of profitability vary over
the periods, and being an Islamic bank does not have a
statistically significant effect on ROA except in the third
model for the full sample. Capital adequacy, asset size,
credit risk, liquidity risk, total asset growth, total equity
growth, foreign exchange rate have statistically significant
effects on ROA. The results are in line with the recent
literature (Athanasoglou et al., 2008).
Table 7 shows the results when banking profitability is
measured as return on equity (ROE). It is obvious that
the determinants of profitability do not vary over periods
as much as they do for ROA. However for ROE, being an
Islamic bank never plays a significant role, even though
the coefficients are higher compared to the coefficients in
Table 6. On the other hand, being a private bank is one of
the determinants of profitability over periods. Interestingly,
liquidity risk is an important factor for ROE while it is not
for ROA.
Table 8 treats NIM as a dependent variable and presents
regression results. The determinants of NIM and ROE are
very similar as they are revealed. ETA, LTA and NPLTA
are positively correlated with NIM and ROE. The most
interesting result is being an Islamic bank is negatively
Table 6. Regression results: ROA is the measure of bank profitability
Dependent Variable
ROA
Full sample (1994–2011)
Before the crisis (1994–2007) After the crisis (2008–2011)
Independent
Variables
Model 1 Model 2 Model 3 Model 1 Model 2 Model 3 Model 1 Model 2 Model 3
IB
ETA
LTA
NPLTA
LOGTA
LQD
DPTA
CR
LR
PD
FD
TAGR
TEGR
TIGR
GDPGR
INF
FX
IF
R-SQ
PROB>CHI2 (F test)
−0.0053
0.0661
0.014
−0.0329
0.0028
0.0043
−0.0016
−0.0235
0.0003
−0.0045
0.0003
−0.0081
0.0631
0.0143
−0.0250
0.0027
0.0044
−0.0028
−0.0229
−0.0003
−0.0042
0.0002
0.0579
0.0314
−0.0209
0.18
0.00
0.18
0.00
Eds. Hatem A. El-Karanshawy et al.
−0.0091
0.068
0.0187
−0.0161
0.006
0.0071
−0.0035
−0.0211
−0.0004
−0.0024
0.0025
0.0533
0.0318
−0.0006
−0.0002
0.0003
−0.0086
0.0001
0.19
0.00
0.0098
0.0759
0.0316
0.0025
0.0053
0.0164
0.0082
−0.0059
−0.0004
−0.0069
0.0033
−0.0139
0.0709
0.0325
0.0170
0.0042
0.0155
0.0063
−0.0053
−0.0005
−0.0073
0.0026
0.1075
0.0141
−0.0558
0.14
0.00
0.15
0.00
−0.0140
0.0674
0.0312
0.0128
0.0026
0.0122
0.0056
−0.0040
−0.0005
−0.0082
0.0018
0.0826
0.0118
−0.0188
−0.0016
0.0003
0.0284
−0.0004
0.16
0.00
−0.0021
0.0898
−0.0076
−0.0673
0.0071
0.0043
0.0075
−0.0214
0.0027
0.0007
0.0001
−0.0036
0.0916
−0.0022
−0.1036
0.0079
0.0066
0.0059
−0.0189
0.0028
0.0011
0.0008
0.0642
0.0238
−0.0879
0.39
0.00
0.43
0.00
−0.0033
0.0934
0.0000
−0.1071
0.0097
0.0101
0.0051
−0.0162
0.0031
0.0023
0.0018
0.1019
0.0098
−0.1709
0.0002
0.0003
−0.0155
−0.0001
0.45
0.00
145
Aytug and Ozturk
Table 7. Regression results: ROE is the measure of bank profitability
Dependent Variable
ROA
Independent
Variables
IB
ETA
LTA
NPLTA
LOGTA
LQD
DPTA
CR
LR
PD
FD
TAGR
TEGR
TIGR
GDPGR
INF
FX
IF
R-SQ
PROB>CHI2 (F test)
Full sample (1994–2011)
Model 1 Model 2
−0.0116
0.1389
0.062
−0.5497
0.0233
0.0070
−0.0005
−0.0509
0.0075
−0.0806
0.0104
−0.0256
0.1204
0.0643
−0.5206
0.0228
0.0068
−0.0072
−0.047
−0.0075
−0.0794
0.0099
0.2695
0.1898
−0.1402
0.12
0.00
0.13
0.00
Model 3
−0.0306
0.1424
0.0877
−0.5001
0.0400
0.0188
−0.0113
−0.0369
−0.0077
−0.0694
0.0220
0.2350
0.1901
−0.0217
−0.0011
0.0030
−0.0452
0.0001
0.15
0.00
Before the crisis (1994–2007)
After the crisis (2008–2011)
Model 1
Model 1 Model 2 Model 3
−0.0023
0.2284
0.11778
0.486
0.0422
0.0130
0.0137
−0.0120
−0.0105
−0.0894
0.0354
0.14
0.00
Model 2
−0.0177
0.2026
0.1244
−0.4218
0.038
0.0065
−0.0238
−0.0090
−0.0107
−0.0907
0.0331
0.5297
0.0979
−0.3472
0.15
0.00
Model 3
−0.0220
0.1702
0.1119
−0.4526
0.0298
−0.0181
−0.0301
−0.0015
−0.0115
−0.0936
0.0297
0.3850
0.1059
−0.0486
−0.0099
0.0013
0.0421
−0.0008
0.17
0.00
−0.0170
0.1966
−0.1393
−0.1330
0.0623
0.0248
0.1387
−0.0157
0.0452
0.0004
0.0028
0.23
0.00
−0.0267 −0.0260
0.2042 0.2298
−0.1109 0.0999
−0.3459 −0.2325
0.0694 0.0926
0.0354 0.05655
0.1318 0.1216
−0.0057 0.0169
0.0458 0.0465
0.0057 0.0206
0.0083 0.0225
0.2357 0.3929
0.1601 0.0953
−0.3472 −0.647
0.0003
0.0061
−0.0718
−0.0010
0.26
0.26
0.00
0.00
Table 8. Regression results: NIM is the measure of bank profitability
Dependent Variable
ROA
Independent
Variables
IB
ETA
LTA
NPLTA
LOGTA
LQD
DPTA
CR
LR
PD
FD
TAGR
TEGR
TIGR
GDPGR
INF
FX
IF
R-SQ
PROB>CHI2 (F test)
146
Full sample (1994–2011)
Model 1
Model 2
−0.019
0.0263
0.0165
0.1572
−0.0011
0.0167
0.0084
−0.0163
−0.0004
−0.0021
−0.0027
−0.0238
0.0197
0.0165
0.1651
−0.0017
0.0157
−0.0062
−0.0153
−0.0005
−0.0020
−0.0032
0.1326
0.0472
−0.0685
0.11
0.00
0.14
0.00
Before the crisis (1994–2007) After the crisis (2008–2011)
Model 3 Model 1
−0.0244
0.0229
0.0182
0.1754
0.0002
0.0168
0.0059
−0.0142
0.0005
−0.0008
−0.0018
0.1123
0.0520
−0.0329
−0.0003
0.0001
−0.0006
0.0002
0.17
0.00
Model 2
−0.0110
0.0383
0.0136
0.0977
0.0033
0.0211
0.0026
−0.0077
−0.0006
−0.0004
0.0045
−0.0176
0.0293
0.0139
0.1258
0.0017
0.0203
−0.0007
−0.0065
−0.0007
−0.0001
0.0056
0.1311
0.0401
−0.0510
0.13
0.00
0.15
0.00
Model 3
−0.0185
0.0226
0.0056
0.1313
−0.0026
0.0134
−0.0014
−0.0059
−0.0008
−0.0017
−0.0081
0.1098
0.0453
−0.0128
−0.0025
0.0009
0.0289
−0.0003
0.14
0.00
Model 1
Model 2 Model 3
−0.0229
0.0203
0.0233
0.3258
0.000
0.0231
0.0025
−0.0076
0.0032
0.0003
0.0024
−0.0272
0.0239
0.0355
0.2664
0.0012
0.0259
−0.0003
−0.0025
0.0033
0.0005
0.0038
0.1681
0.0447
−0.185
0.22
0.00
0.30
0.00
−0.0272
0.0239
0.0385
0.2905
0.0033
0.0315
−0.0012
0.0019
0.0038
0.0018
0.0050
0.3036
0.0249
−0.4159
0.0008
0.0023
−0.0190
−0.0001
0.35
0.00
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Conventional banks versus Islamic banks: What makes the difference?
correlated with NIM and it is always significant at a 99%
level for the full sample and after the crisis. Nevertheless,
being an Islamic bank lowers the profitability when it
is measured as net interest margin (NIM). The effect of
Islamic banks was pretty limited or not significant before
the crisis.
Capital adequacy (ETA) has a positive and significant
impact on ROA and ROE in all periods with different
models. The findings imply that banks that have higher
capital adequacy are more profitable. This may be related
to the fact that banks with higher capital adequacy tend
to be more credible and operate with lower costs. The
specific finding for the Turkish case constitutes a different
dimension in explaining the solvency risk (capital
adequacy) on profitability. Pervan, et al. (2012) find that
capital adequacy ratio is negatively related with return in
their analysis. They conclude that higher capital adequacy
implies lower profitability. They point out that a higher
level of bank capital provides safety and over-caution in
the banking business and it reduces profitability in the
Macedonian banking system. Likewise, our results show
that capital adequacy doesn’t have a significant effect on
NIM after the crisis. While it has a positive and significant
effect on NIM for the full sample, the relationship between
capital adequacy and NIM is blurry before and after the
crisis with different models.
In terms of liquidity risk, the findings are quite specific to
the Turkish case and are not in accordance with studies in
the literature (Pervan et al., 2012). While liquidity risk has
a positive and significant effect on ROA and ROE only after
the crisis, it has a negative effect on ROE before the crisis
and for the full sample. The banks with less liquidity risk
are deemed to be more profitable, since the banks that have
higher loan to deposit ratios are expected to produce more
returns due to interest revenues. However, the Turkish
case proves to be different than the theory, and Turkish
banks are more tempted to invest in government assets
that proposed higher yields (Aysan and Ceyhan, 2007).
Therefore, the liquidity risk implies a positive relationship
between liquidity and profitability as it is in some extent
not in line with recent findings.
In the model where ROA is dependent, it is interesting to
note that the coefficient of non-performing loans to asset
reveals a negative relationship with bank profitability
and is statistically significant only after the crisis. The
estimated coefficients are negative for all the samples.
When the ROE is a dependent variable, the coefficients
are still negative, but they are statistically significant only
for the full sample. The empirical finding is in contrast
with the skimping hypothesis of Berger and DeYoung
(1997). Berger and DeYoung (1997) suggest that under
the skimping hypothesis, a bank maximizing the long
run profits may rationally choose to have lower costs
in the short run by skimping on the resources devoted
to underwriting and monitoring loans but bear the
consequences of greater loan performance problems.
Therefore, cost minimization in the short run may
not bring long-term profitability. Yet, the findings are
plausible in Turkey’s case where the non-performing
loans are low. Therefore, holding more funds kept for
potential losses in the future increases banks cost, thus
Eds. Hatem A. El-Karanshawy et al.
diminishing bank profitability. However the effect on NIM
reveals a different picture since the coefficients are always
positive and significant.
The sector specific variables, the asset and equity growth
positively affect profitability in Turkey. In aggregate terms,
as the banking system as a whole scales up, the asset and
equity size, individual banks tend to be more profitable.
This also implies that asset and equity growth is distributed
evenly among banks. Therefore, no specific bank or
banking group dominated the others. In contrary to this,
total loans growth has a negative and significant effect on
profitability.
In both estimated regressions where ROA and ROE is the
dependent variable, asset size is found to have a significant
and positive impact on profitability. Hauner and Peiris
(2005) suggest two potential explanations for this impact.
First, if it relates to market power, large banks should pay
less for their inputs, i.e., lower cost as mentioned in capital
adequacy case. Second, there may be increasing returns to
scale gains through the allocation of fixed costs.
The deposit to total assets ratio has a negative but
insignificant impact on bank profitability. This might be
an indication of the fact that the link between deposit and
lending is not efficiently operated. Turkish banking systems
have suffered from the short term character of its deposit
base. The maturity of deposit is mainly short and cannot
be efficiently converted into lending, i.e., higher income
earnings.
Our results regarding the impact of ownership on
profitability support the findings of Micco, et al. (2007)
and Iannotta, et al. (2007), who point out that state banks
exhibit a lower profitability than privately owned banks.
The case also holds for foreign banks. Foreign banks are
also more profitable than state banks. The findings shed
light on the inefficiency of state banks in the past where
state banks were mandated as the lender of unprofitable
and politically driven projects. For instance, the duty losses
that were one of the main causes of the 2001 banking
crisis were an indication of how state banks operated with
political bias. One of the main purposes of this paper is to
find the average treatment effect (ATE) of being an Islamic
bank. Regression results give us some insights about how
the ATE might look for different measures of profitability.
According to the regression results Islamic banks play an
insignificant role when the measure is ROA and ROE. In
addition to this, Islamic banks lower the profitability when
the measure is NIM.
Estimates of treatment effects are presented in Table 9. As
we mentioned above we use the different algorithms to
calculate the treatment effect for the sake of robustness.
First, the treatment effect proves that there is no relationship
between Islamic banks and ROA. The coefficient is always
zero. Second, Islamic banks are positively correlated
with ROE and the magnitude is higher. However, OLS
coefficients are insignificant. Third, the ATE on NIM
varies over periods. It is always positive before the crisis
and negative after crisis. For the entire period, different
algorithms prove the negative relationship between Islamic
banks and NIM.
147
Aytug and Ozturk
Table 9. Estimates of treatment effect
1994–2011
ROA
Islamic Banking
No. Obs
Treated
Controls
RE
−0.005
1188
144
1044
NM
0
160
144
16
1994–2007
S
0
1080
144
936
OLS
−0.014
660
80
580
1994–2011
ROE
Islamic Banking
No. Obs
Treated
Controls
RE
−0.010
1188
144
1044
NM
0.004
160
144
16
Islamic Banking
No. Obs
Treated
Controls
RE
−0.023
1188
144
1044
NM
−0.003
160
144
16
S
0.003
1080
144
936
OLS
0.005
528
64
464
OLS
−0.006
660
80
580
NM
−0.007
84
80
4
S
−0.004
1080
144
936
This paper provided evidence on the implications of Islamic
banking in Turkey. As a baseline analysis and in parallel
with the literature, the determinants of profitability in
Turkish banking have been found to be familiar with the
previous studies. However, our main difference comes
with the role of being an Islamic bank. In our study
we investigated the effect of being an Islamic bank on
bank performance in terms of various bank profitability
measures. Specifying some certain banks as Islamic bank
and investigating the effect of being an Islamic bank is a
conscious “self-selection” choice, therefore have a potential
to have “self-selection bias”. In order to overcome this
possibility, we introduced the method of propensity score
matching to the banking literature in order to estimate the
average treatment effect (ATE) of being an Islamic bank in
Turkey where there exists a dual banking system. In order
to estimate the average treatment effect propensity score
matching has been conducted. The results are compared
with the OLS results. According to the OLS results, Islamic
banks have a negative relationship with NIM and a positive
relationship with ROA. The effect of being an Islamic bank
on ROE is insignificant on the other hand. Propensity
score matching technique confirms some of OLS results.
The ATE of being an Islamic bank, on ROA and ROE, are
positive, yet it is blurry but negative on NIM. The results
provided evidence of increasing performance of Islamic
banks in the Turkish banking system. Therefore, although
not in total asset size – these banks signal the efficacy of
Islamic banks in the near future.
Several potential research themes emerge as well. Firstly,
the blurry results of being an Islamic bank on NIM have to
be identified in some detail. This discrepancy could provide
further insights about Islamic banks operations if deeply
OLS
−0.018
660
80
580
NM
0.001
84
80
4
NM
0.001
68
64
4
S
0.001
478
62
416
2008–2011
S
0.012
600
4
596
1994–2007
6. Conclusion
148
S
0
600
4
596
1994–2007
1994–2011
NIM
NM
0
84
80
4
2008–2011
OLS
0.017
528
64
464
NM
0.015
68
64
4
S
0.013
478
62
416
2008–2011
S
0.011
600
4
596
OLS
−0.023
528
64
464
NM
S
−0.005 −0.005
68
478
64
62
4
416
investigated. Second, the results of the Turkish case have
to be compared with other country cases by identifying
more proper econometric methodologies as introduced
in this paper. The issue of “self-selection bias” needs to
be tested in other country cases as well. As a continuation
of the concurrent problem of “self-selection bias,” the
profitability of Islamic banks should be compared with the
conventional banks in a synthetic control group framework
introduced by Abadie and Gardeazabal (2003).
Notes
1. See Hasan and Dridi (2010) for differing performance
of conventional and Islamic banks during and after the
financial crisis in countries where conventional banks
and Islamic banks jointly operate.
2.After the decree, initial institutions were Bahrainbased Al Baraka Turk and Saud-based Faisal Finans
Kurumu, which each opened subsidiaries in 1985. The
Kuwaiti-based Kuveyt Turk Kurumu began operations
in 1989. Eventually, special finance houses began
lending with mainly domestic capital, including:
Anadolu Finans (1991), Ihlas Finans (1995) and Asya
Finans (1996).
3.See Darrat (1988), Zuberi (1992), Darrat and
Salaaming (1990), Kia and Darrat (2007).
References
Abadie A, Gardeazabal J. (2003) “The Economic Costs of
Conflict: A Case Study of the Basque Country. American
Economic Review. 93(1):113–132.
Abreu M, Mendes, V. (2002). Commercial bank interest
margins and profitability: evidence from EU countries.
Working Paper, Porto.
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Conventional banks versus Islamic banks: What makes the difference?
Albertazzi U, Gambacorta L. (2009) Bank Profitability
and the Business Cycle. Journal of Financial Stability.
5(4):393–409.
Athanasoglou PP, Brissimis SN, Delis MD. (2008)
Bank-Specific, Industry-Specific and Macroeconomic
Determinants of Bank Profitability. Journal of
International Financial Markets, Institutions and Money.
18(2):121–136.
Hasan M, Dridi J. (2011) The Effects of the Global Crisis on
Islamic and Conventional Banks: A Comparative Study.
Journal of International Commerce, Economics and Policy
(JICEP). 2(02):163–200.
Hauner D, Peiris SJ. (2005) Bank Efficiency and Com­petition
in Low-Income Countries: The Case of Uganda. IMF
Working Papers 05/240. International Monetary Fund.
Aysan AF, Ceyhan SP. (2007) Market Disciplining Role
of Crisis on the Restructuring of the Turkish Banking
Sector. Technical Report.
Iannotta G, Nocera G, Sironi A. (2007) Ownership
Structure, Risk and Performance in the European
Banking Industry. Journal of Banking and Finance.
31(7):2127–2149.
Bennaceur S, Goaied M. (2008) The Determinant of
Commercial Bank Interest Margin and Profitability:
Evidence from Tunisia. Frontiers in Finance and
Economics. 5(1):106–130.
Kassim S, Majid M, Yusof R. (2009) Impact of Monetary
Policy Shocks on the Conventional and Islamic Banks in
Dual Banking System: Evidence from Malaysia. Journal
of Economic Cooperation and Development. 30(1):41–58.
Berger AN. (1995) The Profit-Structure Relationship
in Banking-Tests of Market-Power and EfficientStructure Hypotheses. Journal of Money, Credit and
Banking. 27(2):404–31.
Kia A, Darrat AF. (2007) Modelling Money Demand Under
the Profit-Sharing Banking Scheme: Some Evidence on
Policy Invariance and Long-Run Stability. Global Finance
Journal. 18(1):104–123.
Berger AN, DeYoung R. (1997) Problem Loans and Cost
Efficiency in Commercial Banks. Journal of Banking and
Finance. 21(6):849–870.
Kwast ML, Rose JT. (1982) Pricing, Operating Efficiency,
and Profitability Among Large Commercial Banks.
Journal of Banking and Finance. 6(2):233–254.
Berger AN, Hanweck GA, Humphrey DB. (1987)
Competitive Viability in Banking: Scale, Scope and
Product Mix Economies. Journal of Monetary Economics.
20(3):501–520.
Macit F. (2012) Bank Specific and Macroeco­
no­
mic
Determinants of Profitability: Evidence From Partici­
pation Banks in Turkey. Economics Bulletin. 32(1):
586–595.
Bourke P. (1989) Concentration and Other Determinants
of Bank Profitability in Europe, North America and
Australia. Journal of Banking and Finance. 13(1):65–79.
Micco A, Panizza U, Yaez M. (2005) Bank Ownership and
Performance Does Politics Matter? Working Papers
Central Bank of Chile 356. Central Bank of Chile.
Darrat A, Salaaming M. (1990) Islamic Banking: An Outline
of Some Conceptual and Empirical Aspects. Savings and
Development. 19:185–192.
Micco A, Panizza U, Yanez M. (2007) Bank Ownership and
Performance – Does Politics Matter? Journal of Banking
& Finance. 31(1):219–241.
Darrat AF. (1988) The Islamic Interest-Free Banking
System: Some Empirical Evidence. Applied Economics.
20(3):417–425.
Michelle NC, Wheelock D. (1997) Why Does Bank
Performance Vary Across States? Review. (Mar):27–40.
Demirguc A, Huizinga H. (1999) Determinants of
Commercial Bank Interest Margins and Profitability:
Some International Evidence. World Bank Economic
Review. 13(2):379–408.
Dietrich A, Wanzenried G. (2011) Determinants of Bank
Profitability Before and During the Crisis: Evidence from
Switzerland. Journal of International Financial Markets,
Institutions and Money. 21(3):307–327.
EC M, PC R. (2003) Determinants of Greek Commercial
Banks Profitability, 1989–2000. Spoudai. 53(1):84–94.
Ergec EH, Arslan BG. (2011) Impact of Interest Rates on
Islamic and Conventional Banks: The Case of Turkey.
Garca-Herrero A, Gavil S, Santabrbara D. (2009) What
Explains the Low Profitability of Chinese Banks? Journal
of Banking and Finance. 33(11):2080–2092.
Goddard J, Molyneux P, and Wilson JOS. (2004) Dynamics
of Growth and Profitability in Banking. Journal of
Money, Credit and Banking. 36(6):1069–90.
Hardy L. (2012) The Evolution of Participation Banking
in Turkey. Online Journal on Southwest Asia and Islamic
Civilization. Winter 2012.
Eds. Hatem A. El-Karanshawy et al.
Molyneux P, Thornton J. (1992) Determinants of European
Bank Profitability: A Note. Journal of Banking and
Finance. 16(6):1173–1178.
Ozturk H, Gultekin-Karakas D, Hisarciklilar M. (2010)
The Role of Development Banking in Promoting
Industrialization in Turkey. Region et Developpement.
32:153–178.
Pasiouras F, Kosmidou K. (2007) Factors Influencing
the Profitability of Domestic and Foreign Commercial
Banks in the European Union. Research in International
Business and Finance. 21(2):222–237.
Pervan M, Curak M, Poposki K. (2012) Industrial Con­
centration and Bank Performance in an Emerging
Market: Evidence from Croatia. In: Advances in Finance
and Accounting. Tomas Bata University, Czech Republic.
Rosenbaum PR, Rubin DB. (1985) Constructing a Control
Group Using Multivariate Matched Sampling Methods
that Incorporate the Propensity Score. The American
Statistician. 39(1):33–38.
Short BK. (1979) The Relation Between Commercial Bank
Profit Rates and Banking Concentration in Canada,
Western Europe, and Japan. Journal of Banking and
Finance. 3(3):209–219.
149
Aytug and Ozturk
Staikouras C, Wood G. (2002) The Determinants of
European Bank Profitability. International Business and
Economics Research Journal. 3:57–68.
Yeldan E. (2007) Patterns of Adjustment Under the Age
of Finance: The Case of Turkey as a Peripheral Agent of
Neoliberal Globalization. Technical report.
Warde I. (2010) Islamic Finance in the Global Economy.
2nd Edition. Edinburgh University Press. Edinburgh.
Zuberi HA. (1992). Interest-Free Banking and Economic
Stability. The Pakistan Development Review. 31:1077–1087.
150
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Estimating expected returns on Mudaraba
time deposits of Islamic banks
Zeynep Topaloglu Calkan
Adjunct Assistant Professor, Georgetown University – Qatar, Phone: +974 5568 9211, [email protected]
Abstract - On the deposit side, Islamic banks work on a mudaraba (partnership) contract, where
the depositor and the bank are business partners. While in conventional banks the depositor is
provided with a fixed interest rate, in Islamic banks the depositor can only discover his return when
the investment period is over. This fundamental distinction brings forth a disadvantage for Islamic
banks while competing with their conventional counterparts in the market.
On the other hand, most of the credits extended by Islamic banks follow a murabaha (cost-plus
sale with deferred repayment) contract, and the banks specify profit rate on the credits from the
beginning. Using this information we have developed a forecast model to quote the depositors their
expected returns on mudaraba time deposits within a 95% confidence interval at the beginning of
the investment term. Besides increasing competitive advantage, estimating expected returns will
assist Islamic banks in their risk management and asset-liability management.
Keywords: Islamic banking, estimating returns on Islamic time deposits
1. Introduction
The financial crisis of 2007–2008 triggered an extended
global recession, which still distresses the economic
activities. This crisis also evoked questions about morality
of the current global financial system. The research after
the crisis shows that financial institutions of the West,
with the help regulators and the rating agencies, deceived
their clients and the public.1 Islamic finance and its ethical
principles emerged as an alternative model because most
of the elements that caused and extended the crisis are not
permissible by Shariah rule.2 Although the virtue of the
Islamic financial system was invigorated after the crisis,
the scholars and practitioners of Islamic finance have been
emphasizing its value long before. Considering its merit,
the insignificant size of Islamic financial assets compared
to global financial assets makes us question what can be
and should be improved by practitioners of the area of
Islamic finance to reach its deserved levels. Here, in this
paper, we propose a tool to improve competitiveness of
Islamic banks. This is an attempt to facilitate the process
of reflecting intrinsic value of Islamic finance to statistical
figures.
Except a few countries3 where the entire internal financial
system has been transformed, Islamic banks have to
compete with their conventional counterparts in attracting
customers. Factors like lower penetration4 and fewer
financial instruments place Islamic banks in an unfavorable
position compared to their competitors.
One of the structural factors that cause difficulties for
Islamic bankers is the inability to provide a specific rate of
return to depositors. Depositors are used to knowing the
interest rate and are most of the time comparing different
investment alternatives based on their rate of return. This
becomes challenging especially when banks are interested
in attracting customers to long-term savings accounts.
Depositors of Islamic banks need to wait up until the end
of their investment period in order to observe their return,
whereas in conventional banks depositors know how much
they will receive at the beginning of their investment.
The reason behind this lack of information is the underlying
Shariah contract in Islamic savings accounts. Typically two
types of contracts, wadia (safe-keeping) and mudaraba
(profit-loss sharing), are used in these accounts. Briefly,
wadia is a contract where the bank acts as a safe keeper
to depositor’s funds with a permission to utilize them at
its own risk, and mudaraba is a type of profit-loss sharing
contract where a depositor invests money into a partnership
with the bank, where the bank provides its labor/services
to the partnership in an agreement to share proceedings –
either profit or loss – later on. In both of these contracts it
is forbidden to specify a rate of return in the beginning as
Cite this chapter as: Calkan Z T (2015). Estimating expected returns on Mudaraba time deposits of Islamic banks.
In H A El-Karanshawy et al. (Eds.), Islamic banking and finance – Essays on corporate finance, efficiency and product
development. Doha, Qatar: Bloomsbury Qatar Foundation
Developing Inclusive and Sustainable Economic and Financial Systems
Calkan
it nullifies the essence of these contracts. Besides Shariah
prohibition, in reality it is not possible to know the value of
ingenerated profits/losses.
Currently, the banks that use these two contracts in their
profit sharing investment accounts (PSIA) do not share
a profit rate with their deposit clients. They only share
historical information, mainly previous term’s profit
distribution. In this paper, we argue it is possible for Islamic
banks to provide an anterior rate to their clients without
breaking Shariah rules by sharing already available
information in their accounts.5
Although it is not permitted and possible to specify
the profits in advance, it is still possible to make precise
estimations. The reason behind it lies in the way PSIA funds
are being utilized in the Islamic banks. Scholars argue that
Islamic banks are initially established in order to promote
sharing risk not only on the depositor side but also on the
financing side through profit/loss sharing system. However
due to reasons like mismatch between asset and liability
duration, moral hazard and adverse selection issues, today
the main mode of financing in Islamic finance is not one of
the profit/loss sharing contracts (musharaka or mudaraba)
but one of the trade contracts namely murabaha. In
murabaha contract the Islamic financial institution (IFI)
acts as a tradesman. It purchases the items from the vendor
and sells them to its client with a mark-up in a deferred
payment plan. There are several requirements of murabaha
contract that have to be satisfied by the bank and the client;
however, for our research, the significance of this contract
is its feature to provide ex-ante profit rate. Both parties
know and agree on how much profit will be charged by the
bank on the item well before the transaction takes place.
Most of the time Islamic banks follow conventional loan
rates when quoting their mark-up price since they are in
competition with them.
Therefore even if IFIs cannot specify the profit rate that
they will distribute to PSIA holders, on the financing
side they know how much money they will make when
they sell the murabaha contract. It is possible to use this
knowledge to generate a reliable estimation about their
profit distribution.
In the following sections of the paper the research proceeds
as follows: In the second section, the main discussions
about profit loss sharing accounts in the literature have
been presented. In the third section the methodology to
estimate PSIA return has been developed, and in the last
section the results and conclusion have been shared with
the readers.
2. Literature review
Profit sharing investment accounts (PSIA) have been a topic
of interest among scholars in the field of Islamic finance.
There are mainly two issues, one of which is accounting
treatment and risk management implications of PSIAs,
while the other is the profit smoothing applications of
Islamic Financial Institutions (IFI) against the risk of losing
PSIA investors, in other words displacement risk.
Accounting and Auditing Organization for Islamic Financial
Institutions (AAOIFI) in its Financial Accounting Standard
152
No. 66 defines PSIAs as a new category between liability
and owners’ equity since PSIA investors are not regular
depositors receiving a fixed return but they are profit/
loss-sharing partners. On the other hand PSIA investors
do not have managerial and voting rights as owners of
the bank. Akacem and Gilliam (2002), Kahf (2005),
Sultan (2006), Ayub (2007), Ibrahim (2007), Shubber and
Alzafiri (2008), in agreement with the definition provided
by AAOIFI, states PSIAs are an equity-like instrument
compared to conventional deposits. This attribute of the
PSIAs brings challenges in terms of accounting treatment
of them. Atmeh and Ramadan (2012) critically evaluate
accounting treatment of mudaraba returns by AAOIFI, and
they explain how some of its implications deteriorate the
reliability and fairness of the financial statements and how
they compare AAOIFI standards with IFRS in that respect.
In application we observe not all banks stick to the AAOIFI
standards. Rating Agency Malaysia (RAM) in its Research
Report (October–December 2007) comparing Malaysian
and Middle Eastern IFIs, points out reporting PSIAs as
a liability on the balance sheet as one of the differences
of Malaysia.
Besides accounting treatment, PSIAs equity-like structure
brings out different risk implications compared to
conventional risk management. Since PSIA investors are
supposed to bear the associated loss in these accounts,
capital adequacy calculation, and the approach towards
interest rate risk and credit risk differs in IFIs. Khan and
Ahmed (2001) provide a detailed analysis of the risk
management in the Islamic financial industry. Archer
and Abdel Karim (2009) especially call attention to the
regulatory challenges faced in Europe and North America,
where there is no special regulation for IFIs. Ariffin and
Kassim (2011) through a survey on selected banks uncover
the areas for improvement in the risk management practices
within IFIs.
On the other hand, Sundararajan (2005) with his crosscountry study argues PSIAs are subject to a considerable
amount of smoothing on their return, which in turn implies
the investment risks of the banks that are not fully shared
by the PSIA investors. He challenges whether IFIs really
need a distinct risk treatment compared to conventional
banks.
Second major issue about PSIAs is the return equalization
activities of IFIs. As pointed out by Sundararajan (2005),
and Archer et al. (2010), in order to compete in the
market and avoid withdrawal risk by providing marketrelated returns to PSIAs, IFIs employ a variety of return
smoothing activities. Additionally, in Jordan, Malaysia and
Qatar, the central bank requires IFIs to manage PSIAs in a
way not to reflect losses to the investors and to “smooth”
returns. Therefore due to regulatory requirements or
market pressure, IFIs are driven to use a combination of
methods like conservative investment strategies, profit
equalization reserves (PER), (a reserve account formed
out of profits of PSIA to smooth profits), investment risk
reserves (IRR), (another similar account to cover periodic
losses) or a donation to PSIA holders from the share of the
owners. Besides Sundararajan (2005), many scholars like
Zuobi and Al-Khazali (2007), Taktak et al. (2010), Farook
et al. (2012) confirm that IFIs pursue income-smoothing
activities to compete with market-based deposit interest
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Estimating expected returns on Mudaraba time deposits of Islamic banks
rates. Taktak et al. (2010) demonstrate unlike conventional
banks, IFIs do not use Loan Loss Provisions (LLP) but use
PER and IRR to provide steady returns. On the other hand
Zuobi and Al-Khazali (2007) report GCC banks that use
LLP to smooth their returns.
The evidence for profit distribution management revealed
recently by Farook et al. (2012) represents the strongest
support in the literature. They have utilized an extended
dataset covering 37 banks in 17 countries. They have shown
most IFIs do really manage profit distributions, with IFIs
in Brunei, Malaysia and UAE showing lower average profit
distribution management, while in Bahrain, Indonesia,
Pakistan, Saudi Arabia, IFIs presenting a higher average
profit distribution management. Percentage of Muslim
population, financial development, market concentration,
depositor reliance and age of the Islamic bank are the most
significant factors to answer the question why IFIs engage
in income smoothing.
The existence of income smoothing activities by itself
proves the competitive pressure on Islamic banks. Literature
shows competing with the conventional banks is one of the
most important challenges of IFIs and they utilize various
methods to overcome this challenge. In this study we will
provide a model to forecast PSIA returns in the short-run to
provide a marketing tool for IFIs.
3. Methodology
In order to define PSIA returns, we will use a simplified
version of the framework mandated by the Turkish Banking
Authority to the IFIs in Turkey. The reasons behind choosing
the Turkish framework follow:
value. It represents the participation share of PSIA holder to
the total funds. There is an account value for each individual
account holder. The total of all depositors’ account values
make the account value for the total funds.
After defining the basic terminology, we will now define
the formula to calculate unit value. As stated, unit value
is basically an index to track the performance of PSIAs.
It can be said that by converting PSIAs to an index value,
the Turkish Banking Authority requires Islamic banks to
remove the size effect and concentrate on the performance.
In our study we use the following simplified formula for
unit value:
ut+1 =
uat + Rt - (ct + y t )
at
(2)
where ut+1 represents unit value for time t + 1, the next day,
uat represents unit account value for time t, Rt represents
revenues to PSIA for time t, Ct represents costs to PSIA for
time t, yt represents reserves for time t and at represents
account value for time t.8
Rt, revenues to PSIA, is defined as a combination of profits
accrued to the PSIA account within current business day
plus any annulment of former provisions and/or reserves.
Ct, costs to PSIA, is defined as a combination of provisions
imposed by regulations, plus payments to deposit insurance
fund and loan loss provisions. Revenues and costs are
reflected on the calculation starting with the first day of
the loan. Therefore even if the bank does not receive any
installments, the PSIA account will start recording a profit
when the loan is extended. For the purposes of simplicity
we will assume all loans follow a murabaha contract.9
1. PSIA funds are not mixed with bank equity capital
2. Profit/loss distribution is done on a daily basis
3. Instead of PER and IRR, LLP is used to smooth returns
which makes it comparable to conventional banks
After calculating unit value, we define daily returns to our
PSIA funds with the following formula:
Before stating the formula for PSIA returns, we need to
define some of the terminology that will be used in our
specification.
Unit Value (ut): It is assumed to be equal to 100 on the
first day IFI accepted deposits to its PSIAs. It changes daily
based on the distribution of profit and loss to the account.
We can consider unit value as an index to tell us how our
funds are performing.7 There is one unit value for each day
of operation. The change from one day to another reflects
the percentage of profits or losses made by the bank during
one day.
Unit Account Value (uat): It defines the current total
monetary value of PSIAs for the bank, which can be
calculated by multiplying unit value and account value.
For every individual account holder, unit account value is
equal to the amount of money they have deposited in the
first day. Unit account value will grow each day based on
bank’s performance.
uat = ut × at(1)
Account Value (at): It is the value calculated by dividing
the amount of money deposited to PSIAs by that day’s unit
Eds. Hatem A. El-Karanshawy et al.
rt+1 =
ut+1 − ut
× 100 (3)
ut
In this calculation for unit value, PSIA costs and reserves
are constant percentages of PSIA revenues given by
the regulatory authority. Therefore the only variants in
this formula are unit account value, which represents
the amounts deposited to the bank and PSIA revenues,
representing the returns from the loans extended by the
bank. Unit account value fluctuates as the depositors
withdraw their money or deposit new funds to their
accounts. PSIA revenues increase as new murabaha loans
are extended to the clients. Therefore in our analysis we
will simulate these two variables and the unit value and
daily returns to PSIA funds will adjust.
4. Results and conclusion
Data
In our analysis, we will use the formula generated in
the methodology section to simulate the daily average
PSIA returns of Turkish participation banks.10 Currently
there are four participation banks operating in Turkey.
They publish daily returns to their PSIA funds for the
153
Calkan
Figure 1. Average PSIA returns for 4 Turkish participation banks.
trailing one-month, three months, six months and one
year for different currency classes in their websites. We
have taken four banks’ average data for trailing onemonth (31 days) returns to Turkish Lira PSIA funds for
the period between January 15, 2012, and February 15,
2012. In our simulation we try to estimate returns during
this one-month period. We have chosen one-month
maturity, since these accounts hold the highest amount
of funds compared to other maturities. The date selection
is arbitrary. Turkish lira has been chosen as the selected
currency since the analysis is being done with Turkish
data. Figure 1 presents the path followed by PSIA returns
for our selection.
First simulation
In our simulation we have assumed the following:
Unit account
value
Unit value
Account value
PSIA revenues
PSIA costs
Reserves
At t = 0
ua0 : TL6 billion
u0
a0
R0
C0
y0
: 100
: 60 million
: TL 2 million
: 15% of revenues
: 5% of revenues
At t >0
Endogenous
Endogenous
Assumed constant
Will be estimated
Exogenous
Exogenous
Unit account value represents the deposits collected by
our average participation bank. Since we are using the
data from the beginning of 2012 for TL accounts, we have
referred to TL deposits collected at the end of 2011. Total
value of TL deposits collected has been TL 24.04 billion.11
In our simulation we have taken the average of this value
and used TL 6 billion12 as unit account value.
Unit value is assumed to be 100. Since it is an index to
follow the performance of the funds, instead of its absolute
value, the changes from day to day are significant for our
analysis.
154
Assuming a unit account value of TL 6 billion and unit value
of 100, account value is calculated to be TL 60 million.13
With the separation of unit value and account value, it
is now possible to track changes in deposits in two ways.
Unit value captures the changes due to the investment
returns (profits and losses), and account value captures
the fluctuation in the funds deposited to the bank (deposits
and withdrawals).
The percentages 15% and 5% for PSIA costs and reserves
assumed in our simulation are an approximation based
on the rules imposed by the Turkish regulatory authority.
Turkish regulatory authority requires banks to set aside
reserves under the names of special provision, general
provision, federal deposit insurance premium and
precautionary provision. Based on Turkish monetary and
macroeconomic policy, the rates on these provision expenses
can be altered; a total of 20% is a fair approximation to
the reality. The ratios for PSIA costs and reserves are kept
constant in all simulations since they represent the best
legal approximation.
For PSIA revenues our selection of TL 2 million daily
return is derived from 6.75%, with a starting monthly
PSIA return given in Figure 1. TL 2 million daily revenue
is equal to 0.03% daily return on TL 6 billion pool of
funds. Using 70/30 profit sharing ratio between the
bank and the depositor, it refers to a 6.72%14 monthly
PSIA return.
In our first simulation, we have assumed the account
values to be constant at t = 0 value, meaning that there are
no deposits or withdrawals in PSIA funds. Therefore we
only need to simulate the trend in PSIA revenues. In 2011,
participation banks have extended a total of TL 41.14
billion15 in loans. This will approximately refer to daily TL
40 million16 of loans. The revenue on these loans depends
on the profit rate charged by the bank to the borrower.
There is going to be a negative correlation between the
rate and the loan amount. As the rate on the loan goes
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Estimating expected returns on Mudaraba time deposits of Islamic banks
Figure 2. Estimating the trend in PSIA revenues: Panel A gives the area for mean-standard deviation combination,
Panel B uses 4 sample points from Panel A to construct confidence interval for PSIA returns.
down there will be more demand for loans and vice versa.
Also the profit rate has to be set in accordance with market
conditions. Islamic banks using their internal data can
make analysis on this negative correlation and find out
the relationship between the profit rate and the amount of
loans they can extend in detail. However in our analysis
we will use market interest rate as a reference point since
participation banks are competing with conventional banks
in attracting customers. Average monthly commercial loan
rate for the period January 15th –February 15th 2012
Eds. Hatem A. El-Karanshawy et al.
in Turkey was 1.27%17. Thus, if the bank extends TL 40
million of loans at a rate of 1.27% everyday, PSIA revenues
are increasing around TL 17,000 per day. Of course this is
a rough estimation. Therefore in a way to provide the 95%
confidence interval to PSIA returns curve given in Figure 1,
we have constructed a range for the trend of PSIA returns. If
mean and standard deviation of PSIA revenue change stays
in the area given in Figure 2 Panel A, the bank can provide
a profit rate quote that is within 95% confidence interval of
the actual realized value. In order to demonstrate this, we
155
Calkan
have plotted confidence interval for the actual PSIA return
for four different points in the area given.
Therefore our assumptions in this simulation are as
follows:
The result for the first simulation represents the possibility
of providing a profit rate quotation to the deposit holders
in Islamic banks. As our analysis suggests the banks can do
this with a level of flexibility in their revenue estimations.
Here in our analysis we have used market interest rate and
four banks’ average for other variables. Islamic banks on the
other hand have access to historical data about individual
loans with profit rates, loan values, payment information
and so on. Using this extra information it is even possible
for them to provide a better estimate for PSIA returns.
At t = 0
Unit account ua0 : TL6 billion
value
Unit value
u0 : 100
Account value a0 : 60 million
Second simulation
With the introduction of PSIA funds variation, mean
and standard deviation combinations for revenues have
been updated. Figure 4 Panel A displays the new range
for revenues that makes it possible for the Islamic bank
to provide a profit rate estimate within 95% confidence
interval of the actual value. In Panel B, four different
mean and standard deviation combinations from Panel A
have been used to plot confidence intervals for the return
estimate. Actual value lies within 95% confidence interval
in each of them.
In the second simulation, we release our assumption of
constant account value, in other words now there are
deposits to and withdrawals from PSIA funds. In the first
simulation account value was assumed constant at TL
60 million along the estimation period. Islamic banks
in applying this framework can use their internal data
about the fluctuations in the PSIA funds. However for our
analysis we will use market growth rate as the trend. We
have extracted the trend in TL deposits for participation
banks for a period two months prior to our estimation
period. Weekly deposits data is available at Central Bank’s
website18. Deposits data is equivalent of unit account value
in our framework. However we calculate unit account
value endogenously using equation 1. Therefore we will
use the trend in deposits as an approximation to trend in
account value. Figure 3 displays the path followed by TL
deposits of Participation Banks. As we have been doing all
along, here again we have divided total TL deposits value
by four to reach an average value for one of the four banks
in the market.
PSIA revenues R0
PSIA costs
C0
Reserves
y0
At t >0
Endogenous
Endogenous
Linked to deposit
growth
: TL 2 million
Will be estimated
: 15% of revenues Exogenous
: 5% of revenues Exogenous
This second simulation demonstrates a more accurate
estimate for realized values. We have incorporated
variations in PSIA funds due to new deposits and
withdrawals. Although we have used market growth rate
as an approximation to PSIA variations, recognizing it as
an exogenous variable, in reality banks have some level of
control over their deposits. They are affected from market
developments, but by charging a higher rate they can
attract more deposits and they are always free to refuse
new deposits. We have assumed in our analysis that they
Figure 3. Average TL deposits of participation banks.
156
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Estimating expected returns on Mudaraba time deposits of Islamic banks
Figure 4. Estimating the trend in PSIA revenues: Panel A gives the area for mean-standard deviation combination,
Panel B uses 4 sample points from Panel A to construct confidence interval for PSIA returns.
do not have any control. Figure 5 demonstrates the effect
of change in deposit assumption from one simulation to
another. It plots Panel A of Figure 2 and 4 on top of each
other. As can be seen, when we assume growing deposits,
the bank needs to receive more credit return meaning they
will provide more murabaha credits to the clients. This
figure also shows that Islamic bank can use deposit growth
as another policy tool. The changes in deposit growth will
move the policy region up or down. Also, orange colored
Eds. Hatem A. El-Karanshawy et al.
area provides a more conservative estimate, which suits
both growth assumptions.
In both of the analyses we have considered PSIA revenues
as a policy tool. The bank can change the profit rate they
distribute to deposit holders by increasing or decreasing the
mark-up rate they charge on murabaha credits. Although
the demand for credit depends on the rate they charge,
there is still room for policy making. Therefore, using this
157
Calkan
Figure 5. Changes in revenues.
tool, Islamic banks can also compete with each other and
with conventional banks in attracting deposits, besides
providing a profit rate quotation.
6. Conclusion
In this study we have studied the possibility of providing
an ex-ante return rate quotations to deposit holders at
Islamic banks. Simulation results suggest, it is possible to
offer a reliable forecast within a 95% confidence interval of
upcoming actual PSIA returns to the clients. The benefits
of this study are three-fold. The most important benefit
of providing this information is increasing Islamic banks’
competitive advantage compared to their conventional
counterparties. Secondly, using these simulation techniques
IFIs can improve their fund management by controlling the
amount of funds deposited to PSIAs and setting their markup rates according to their PSIA return targets. Furthermore
these simulation methods can be used to better manage
risks associated with asset liability management and rate of
return.
Appendix: Detailed unit value calculation
The information below is extracted from Annex 1
regulation on the principles and procedures for accepting,
withdrawal of deposits and participation funds as well as
the prescribed deposits, participation funds custody and
receivables in Turkish banking legislation.
UA + R - (C + Y)
A
Details of Numerator
Unit Account Value
C
PSIA Costs (a+b+c+d)
a Special Provision Expenses
PSIA Revenues (a+b+c+d+e)
a
Participation Share of Dividend Incomes
b General Provision Expenses
Deposit Insurance Fund (DIF) Premium
Dividend Incomes Procured from Loans
c Expenses
a.1 Extended Arising from PSIA
a.2 Profit Equivalent to Extended Fund Surplus
d Precautionary Provision Expenses
b
Collections Made from Loans Cancelled
Amounts/ Reserves Allocated from Profit
c
Cancellations of Special Provisions
Y
to be Distributed to PSIA
d
Cancellations of General Provisions
Provision Cancellations Set Aside from Profits
e
to be Distributed to PSIA
Denominator
0
Account Value
New U
UA
R
A
158
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Estimating expected returns on Mudaraba time deposits of Islamic banks
PSIA Revenues (a + b + c + d):
a) Participation Share of Dividend Incomes: Dividend
amount equivalent to extended fund surplus is deducted
from dividend incomes procured from extended loans
arising from participation account. The amount found
is separated on a currency type basis according to its
weight in total participation accounts. The amount
found by multiplying the separated amount by the
ratio of account owner’s participation in profit defines
the amount in dividend income falling to the share of
participation accounts.
1.Dividend Incomes Procured from Extended Loans
Arising from PSIA: This is the dividend income
procured from funds extended arising from PSIA on
currency basis. Whether or not delay funds collected
for those not paid in their maturity among those
funds or dividends deprived of as well as income from
required reserves shall be taken into consideration as
dividend income in the unit value calculation of PSIA
shares are determined in PSIA contracts.
2.Profit equivalent to extended fund surplus: the
amount found by multiplying by the ratio calculated
by dividing the dividend income procured from
loans extended arising from PSIA on a currency basis
to the sum of funds extended, with extended fund
surplus.
b) Collections Made from Loans Cancelled: The amount
falling to the share of participation accounts, from the
collections made concerning cancelled loans from loans
extended arising from PSIA.
c) Cancellation of Special Provisions: The amount relating
to PSIA, among cancelled amounts of special provisions
set aside for loans arising from PSIA classified as
non-performing loans pursuant to the regulation
on principles and procedures for determination of
qualifications of loans and other receivables by banks
and provisions to be set aside.
d) Provision Cancellations Set Aside from Profits to be
Distributed to PSIAs: It is the amount cancelled for
meeting SDIF premium and special and general
provisions of provisions monitored in amounts set aside
from profits to be distributed to PSIAs.
PSIA Costs (a + b + c + d +e):
a) Special Provision Expenses: It is the part fall to the share
of PSIA of general provisions set aside for PSIA emanated
loans classified as NPL pursuant to the regulation on
principles and procedures relating to for determination
of qualifications of loans and other receivables by banks
and provisions to be set aside.
b) General Provision Expenses: It is the part that falls to the
share of participation accounts of general provisions set
aside for PSIA emanated loans pursuant to the regulation
on principles and procedures relating to determination
of qualifications of loans and other receivables by banks
and provisions to be set aside.
c) DIF (Deposit Insurance Fund) Premium Expenses: It is the
part that falls to the share of DIF premium participation
accounts.
d)Precautionary Provision Expenses: It is the amount of
precautionary provision to be used in meeting the part
fall to the share of DIF premium PSIA and special and
Eds. Hatem A. El-Karanshawy et al.
general provisions from the total amount of income
items stated in (b), (c) and (d) sub paragraph of the
PSIA revenues explanation. These provisions set aside
are recorded to the account of amounts set aside from
profit to be distributed to PSIAs included in communiqué
on uniform chart of account and its explanation to be
implemented by participation banks.
Amounts Allocated from Profit to be Distributed to PSIAs:
It is the provision amount allocated within the scope of
the provision of the article 14(3) of the regulation on
principles and procedures relating to determination of
qualifications of loans and other receivables by banks
and provisions to be set aside from profit amounts to be
distributed to participation accounts by calculation date of
unit values.
Notes
1.Wall Street and Financial Crisis: Anatomy of
Financial Collapse, report by US Senate Permanent
Subcommittee on Investigations, 2011, Available
at:
http://www.hsgac.senate.gov//imo/media/
d o c / Fi n a n c i a l _ C r i s i s / Fi n a n c i a l C r i s i s Re p o r t .
pdf?attempt = 2.
2. Kayed and Hassan (2011), Al Mamun and Mia (2012).
3. Iran, Sudan and Pakistan.
4. Average market share of Islamic banks in MENA region
is 14%.
5. Some of the banks might be using this information in
their internal policy making at the moment.
6.AAOIFI Statement of Financial Accounting No. 2:
Concepts of Financial Accounting for Islamic Banking
and Financial Institutions.
7. Turkish Banking Authority’s approach to PSIA return
calculation is similar to return calculation in a fund.
8. See appendix for detailed calculation of unit value.
9.As majority of the loans in the Islamic banking
are provided with murabaha system, we will assume
all loans given follows murabaha contract for the
simplicity.
10.In Turkey, Islamic Financial Institutions are called
participation banks.
11.Data from the website of Participation Banks
Association of Turkey: www.tkbb.org.tr.
12. 24 billion/4 banks = 6 billion.
13. 60 million = 6 billion/100.
14.0.03% × 360days = 9.60% total return; 70% × 9.60% =
6.72% depositors share.
15.Data from the website of Participation Banks
Association of Turkey: www.tkbb.org.tr
16. 41.14 billion/4 = 10.28 billion per bank; 10.28/250
(business days in a year)~40 million per day.
17.From Turkish Central Bank Electronic Data Distri­
bution System.
18.From Turkish Central Bank Electronic Data Distri­
bution System.
References
Akacem M, Gilliam L. (2002) Principles of Islamic Banking:
Debt Versus Equity Financing. Middle East Policy.
9(1):124–138.
159
Calkan
Al-Mamun Md, Mia MAH. (2012) Origin of & Solution
to Global Financial Meltdown: An Islamic View.
International Journal of Business and Management.
(7):12.
Shubber K, Alzafiri E. (2008) Cost of Capital of Islamic
Banking Institutions: An Empirical Study of a Special
Case. International Journal of Islamic and Middle Eastern
Finance and Management. 1(1):10–19.
Archer S, Abdel Karim RA. (2009) Profit-Sharing Investment
Accounts in Islamic Banks: Regulatory Problems and
Possible Solutions. Journal of Banking Regulation.
10:300–306.
Sultan SAM. (2006) An Overview of Accounting Standards
for Islamic Financial Institutions. Finance Bulletin.
Jan-Mac.
Archer S, Abdel Karim RA, Sundararajan V. (2010)
Supervisory, Regulatory, and Capital Adequacy
Implications of Profit-Sharing Investment accounts
in Islamic Finance. Journal of Islamic Accounting and
Business Research. 1(1):10–31.
Ariffin NM, Kassim SH. (2011) Risk Management Practices
and Financial Performance of Islamic Banks: Malaysian
Evidence. Paper presented at The 8th International
Conference on Islamic Economics and Finance, Doha,
December 19–21.
Atmeh MA, Ramadan AH. (2012) A Critique on Accounting
for the Mudarabah Contract. Journal of Islamic
Accounting and Business Research. 3(1):7–19.
Ayub M. (2007) Understanding Islamic Finance. John Wiley
& Sons, Ltd. England.
Farook S, Hassan MK, Clinch G. (2012) Profit Distribution
Management by Islamic Banks: An Empirical
Investigation. The Quarterly Review of Economics and
Finance. 52(3):333–347.
Ibrahim SH. (2007) IFRS vs. AAOIFI: The Clash of Standard?
Munich Personal RePEc Archive (MPRA). No. 12539.
Kahf M. (2005) Basel II: Implications for Islamic Banking.
Paper presented at the 6th International Conference
on Islamic Economics and Banking, Jakarta, November
22–24.
Kayed RN, Hassan MK. (2011) The Global Financial
Crisis and Islamic Finance. Thunderbird Int’l Bus Rev.
53:551–564.
Khan F. (2010) How ‘Islamic’ is Islamic banking? Journal
of Economic Behavior & Organization. 76(3):805–20.
Khan T, Ahmed H. (2001) Risk Management – An
Analysis of Issues in Islamic Financial Industry. Islamic
Development Bank-Islamic Research and Training
Institute. Occasional Paper No. 5. Jeddah.
160
Sundararajan V. (2005) Risk Measurement and Disclosure
in Islamic Finance and the Implication of Profit
Sharing Investment Accounts. Paper presented at the
6th International Conference on Islamic Economics,
Banking and Finance. Jakarta, November 22–24.
Taktak NB, Zouari SBS, Boudriga A. (2010) Do Islamic
Banks Use Loan Loss Provisions to Smooth Their
Results? Journal of Islamic Accounting and Business
Research. 1(2):114–12.
Turkish Banking Legislation: Annex 1 to Regulation on
the Principles and Procedures for Accepting, Withdrawal
of Deposits and Participation Funds as Well as the
Prescribed Deposits, Participation Funds Custody and
Receivables. Available at: <http://www.bddk.org.tr/
WebSitesi/english/Legislation/9656engdepositandpar
ticipation_fundannex_08_06_2010.pdf>. Accessed: 10
May 2013.
Turkish Banking Legislation: Regulation on Procedures
and Principles for Determination of Qualifications of
Loans and other Receivables by Banks and Provisions
to be Set Aside. Available at: <http://www.bddk.org.tr/
WebSitesi/english/Legislation/8836eng_provisions_
to_be_set_aside_13_06_2011.pdf>. Accessed: 10 May
2013.
Turkish Banking Legislation: Communiqué on Uniform
Chart of Account and its Explanation to be Implem­
ented by Participation Banks. Available at: <http://
www.bddk.org.tr/WebSitesi/english/Legisla­t ion
/8802eng_par ticipationbank s_unifor m_c har t_
ofaccounts_08_06_2011.pdf>. Accessed:10 May 2013.
Zoubi TA, Al-Khazali O. (2007) Empirical Testing of the
Loss Provisions of Banks in the GCC region. Managerial
Finance. 33(7):500–511.
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Self-adjusting profit sharing ratios for Musharakah
financing
Volker Nienhaus
Honorary Professor, University of Bochum, ICMA Centre, University of Reading (United Kingdom),
Tel. +49 201 8695750, [email protected]
Abstract - Banks avoid participatory financing due to serious information asymmetries, adverse
selection and moral hazard problems resulting in negative impacts for the return on capital
provided. Even financing instruments with a participatory legal form such as musharakah sukuk
have been stripped of their risk sharing substance and become functional equivalents of interestbearing bonds. Several authors have addressed these issues, but some proposals are applicable only
for (listed) joint stock companies, while others imply Shariah compliance issues. To overcome these
limitations, a “self-adjusting profit sharing ratio” is proposed, based on building blocks found in
AAOIFI Shariah standards for musharakah financing and musharakah sukuk. These building blocks
allow a (surprisingly) wide range of discretionary adjustments of participatory contracts, provided
the contracting parties come to an agreement in re-negotiations of the contractual terms. This
requires an agreement on a fair distribution of profits. What the parties consider a fair distribution
is already known when the contract is initially concluded: It determines the parties’ profit shares
based on their profit expectations at this point in time. The AAOIFI building blocks allow the
structuring of a formula for the profit sharing ratio, which automatically adjusts to changes in the
expected or actual profit. It thus ensures continuously a profit distribution in line with the initially
agreed-upon principles of fairness. The formula can be calibrated such that the financing party gets
under “normal” circumstances a return in line with a predetermined benchmark (e.g., the market
rate of fixed term financings plus a risk mark-up) while the financed party has the advantages of an
“insurance” against losses and unrestricted upside gains. Thus, financing instruments or sukuk with
new risk/return profiles and some participatory elements could be structured so as to overcome the
problems caused by information asymmetries in “pure” PLS financings.
Keywords: Islamic finance, musharakah, profit and loss sharing, information asymmetries
1. Discrepancies between theory and
practice of Islamic finance
Islamic economists consider finance based on profit and
loss sharing (PLS) (participatory finance) as the genuine
Islamic mode of finance and the major factor distinguishing
Islamic from conventional finance. Indeed, an economic
system where PLS is the dominant mode of financing would
have different qualities with regard to efficiency, stability
and distribution compared to a conventional interest- and
debt-based system. However, the practice of Islamic finance
(which was factually Islamic banking until the early 2000s)
was and is very different from this ideal model.
•PLS financing hardly ever takes place in Islamic
banks. Bad experiences of pioneering banks and
theoretical explanations of adverse selection and
moral hazard issues in PLS financing (where the
ratio of profit sharing is fixed in advance and not
changed afterwards) can explain the abstinence
from participatory modes of financing.
• Banks apply the PLS principle only in the deposit
business, i.e., in contracts of mudarabah-based
investment accounts, but even there the practice
was quite different from the model: Fluctuations
of investment returns were not passed on to the
account holders but rather smoothed out by
recourse to reserves (investment risk reserves and
profit equalization reserves) and, in worst cases,
by interest-free loans or even “voluntary” gifts of
the shareholders to the account holders. This was
done to avoid massive withdrawals by disappointed
investment account holders which would have
Cite this chapter as: Nienhaus V (2015). Self-adjusting profit sharing ratios for Musharakah financing.
In H A El-Karanshawy et al. (Eds.), Islamic banking and finance – Essays on corporate finance, efficiency and product
development. Doha, Qatar: Bloomsbury Qatar Foundation
Developing Inclusive and Sustainable Economic and Financial Systems
Nienhaus
created serious problems for the bank. With
ex ante announced “anticipated” returns which were
typically congruent with the realized ex post returns,
and with the coverage of investment accounts by
capital protecting deposit insurance schemes in
some jurisdictions, mudarabah-based investment
accounts got the “look and feel” of conventional
interest-bearing deposits.
• With the growing popularity of sukuk in the 2000s, a
new option for PLS financing emerged in the Islamic
capital market: musharakah sukuk. But again, the
practice converted this participatory instrument into
an Islamic bond with factually fixed returns for the
sukuk holders. The applied techniques are explained
in detail later in this paper.
Unfortunately, Islamic bankers and Shariah scholars never
shared the enthusiasm of Islamic economists for PLS
finance in practice. It is not that Shariah scholars did not
allow PLS arrangements: Mudarabah and musharakah
contracts (and modern derivatives thereof) are explicitly
approved as Shariah compliant. However, the approval was
done in a way which opened the door widely for Islamic
bankers to convert the participatory concepts into close
functional equivalents of conventional interest-based and
factually risk-free modes of financing. In practice, Shariah
scholars have approved the conversion of “equity-based”
sukuk (participatory instruments with a variable return)
into “Islamic bonds” (debt instruments with fixed costs).
The growth of Islamic finance over the last decade was driven
by the inroad of Western financial institutions into this new
and seemingly lucrative segment of the global financial
industry. Many Islamic financial institutions are run by
CEOs and management teams who have been socialized
in conventional finance before they converted to Islamic
finance. In addition, many executives and staff of Islamic
financial institutions were hired from the conventional
sector. Individuals and teams who were successful in
conventional finance before they joined the Islamic finance
industry were familiar with strategies and instruments for
a good or even outstanding performance of their financial
institution. This was probably the reason why they were
lured away for their previous employment, and shareholders
expect a continuation of such a management performance
in the new Islamic environment. Thus, it should not come
as a surprise that Islamic bankers who were socialized in
the conventional system tried to replicate those strategies
and instruments which they had applied successfully in
their previous position in conventional finance.
The Islamic economics literature was often too theoretical,
abstract, macro-oriented or prescriptive to be of much
use for practitioners who were looking for more effective
instruments in commercial, for-profit financial institutions.
The Islamic economists have not been able to convey their
enthusiasm for PLS instruments to Islamic bankers – and
probably also not to Shariah scholars. While participatory
finance is often seen by Islamic economists as the core
of Islamic finance—as sale- and rent-based modes of
finance are mentioned only on the sidelines in their
models—it is exactly the opposite from the perspective
of Shariah scholars: sale- and rent-based contracts have
been elaborated over centuries in extensive detail, while
financing based on profit and loss sharing were sidelined.
162
Islamic economists have recognized that the practice
of Islamic finance deviates substantially from the ideal,
but they did not simply criticize practitioners for this
unfortunate development. They analyzed the reasons
for the observable discrepancy (mainly agency problems
in anonymous markets), and they came forward with a
number of proposals suggesting how to solve the identified
problems and promote participatory finance.
• The second chapter summarizes and comments on a
number of such contributions. Looking at the rapid
growth of musharakah sukuk in the 2000s, it seemed
that at least the Islamic capital market had overcome
the agency problems and moved toward the PLS
ideal. Unfortunately, this was not the case.
•The third chapter takes a closer look at the
structuring of these sukuk and explains how
equity-based instruments could be converted into
functional equivalents of debt instruments with
predetermined costs and capital guarantees. After
a critique of prevailing practices by a prominent
Shariah scholar in 2007, and a resolution of the
Accounting and Auditing Organisation of Islamic
Financial Institutions (AAOIFI) in 2008, the issuance
of musharakah sukuk dropped sharply from then
until today. Despite the recovery of the sukuk market
in recent years, the once dominant form of sukuk has
become marginal by today. This is a deplorable signal
because mudarabah and musharakah sukuk were the
only financing instruments of significant quantitative
weight in the Islamic finance markets which upheld
at least the PLS form.
• The fourth chapter outlines the mechanics of a
musharakah sukuk concept, which is based on the
PLS principle but uses building blocks found in the
Shariah standards of AAOIFI to overcome agency
problems. The concept allows the structuring of a
financial instrument that brings differing commercial
interests of contracting parties in balance and
keeps this balance by a self-adjusting profit sharing
ratio. The automated ratio adjustment takes place
whenever new information on the expected or
actual performance (=profit) of the financed venture
becomes available.
2. The agency problem of participatory
finance in anonymous markets
Participatory finance is a generic term for different forms of
financing on the basis of profit and loss sharing, for example
mudarabah and musharakah bank financing or mudarabah
and musharakah sukuk.
Bacha (1997) sees mudarabah financing as a hybrid of
elements of debt and equity financing. Somehow the
agency problems of both are combined in this hybrid:
• The equity agency problem is that the mudarib has a
strong incentive to “produce” costs which accrue to
him as benefits. This goes on as long as the marginal
utility from fringe benefits or perks exceeds the
corresponding reduction in the mudarib’s share of
the profit. In addition, since mudarabah financing
is for a specified project of an existing firm, the firm
may have various possibilities to shift overhead and
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Self-adjusting profit sharing ratios for Musharakah financing
other costs to the mudarabah project, thus reducing
the profits which have to be shared with the bank
(without reducing the overall profit of the firm).
• The debt agency problem is one of moral hazard:
Debt (or more general: external capital at costs
lower than the return on total capital) can leverage
the return of equity substantially (even if a certain
percentage of the profit goes to the provider of the
external funds). External mudarabah funds have not
only a factual but even a contractual loss absorbing
quality. Consequently, a high leverage by mudarabah
funding would create incentives for a firm to embark
on projects with high profit potentials and high risks
because the upside chances surpass the downside
risk for the equity holders.
Bacha concludes that mudarabah financing has more
agency problems than a pure debt or a pure equity financing
and therefore is an inferior option for the capital providing
party. To overcome the agency problems of mudarabah in its
genuine form, Bacha proposed the introduction of “equity
kickers:” specified events or outcomes to trigger an equityrelated provision in a financing contract. The equity kicker
in a mudarabah contract would be a clause “whereby in the
event of losses in the Mudarabah financed project, the RabUl-Mal absorbs the losses but is ‘reimbursed’ for the amount
of losses thru issuance of new equity by the Mudarib to
him.” (Bacha 1997, 18). By this clause the mudarib transfers
benefits to the rab al-mal so that in the end the rab al-mal’s
will be shielded against “avoidable” losses (caused by profit
reducing management practices or excessive risk taking),
and his risk will decrease. For the mudarib the opportunity
costs of profit compression or high risk ventures would
increase, and this should solve or at least mitigate the
agency problems.
A major drawback of the equity kickers approach is that it
can be applied only for the mudarabah financing of joint
stock companies. Start-ups and small and medium sized
enterprises (SMEs) usually do not have this legal form. But
by far the largest number of enterprises are SMEs, which
play a prominent role for employment, income generation
and poverty reduction in many Muslim (and non-Muslim)
countries.
Another problem is the Shariah compliance of the equity
kickers. Shariah principles prohibit a protection of the
rab al-mal’s capital through a guarantee by the mudarib.
Bacha claims that this is not the case because the equity
kickers are no guarantee against losses. The rab al-mal
“will make losses if the project makes losses – although
it will be much less than under existing Mudarabah.”
Ignoring the problem of a gharar-free determination of
the value of the transferred equity in cases of loss, the fact
remains that the equity has some value and thus is a partial
compensation for the loss. This could be seen as equivalent
to a partial guarantee of mudarabah capital by the mudarib
– which would be a violation of a Shariah stipulation.
But Bacha himself sees another violation of a Shariah
requirement: “The one Shariah requirement that would
not be met by the proposed arrangement is the requirement
that in Mudarabah, the financier should absorb all the
losses. Any proposal that seeks to overcome the problems
of existing Mudarabah would invariably come up against
this injunction.” This injunction is repeated over and again
Eds. Hatem A. El-Karanshawy et al.
in all AAOIFI standards dealing with mudarabah and
musharakah, and later the IFSB took the same position.
A protection against capital losses can only be given by an
independent third party, not by the mudarib.
Ahmed (2002) also addresses agency problems. He
proposes a contractual arrangement that deals with the
moral hazard problem arising from an underreporting
of profits. It shall be overcome by an incentive-compatible
contractual arrangement. The financing bank will have
the right to undertake a costly audit if the profit reported
by the financed firm falls short of the expected profits (on
which the parties have to agree). The auditing expenses
will be shared ex ante by both parties and become a codetermining factor for the calculation of the profit sharing
ratio. But they materialize only if the bank has reason to
undertake the audit. If the audit shows that the firm has
reported correctly, the bank will refund the firm’s share
of the audit costs (as a reward for honesty). If the audit
uncovers false reporting, i.e., if it becomes apparent that the
actual profit is higher than the reported profit, the firm has
to bear the full auditing costs, calculate the bank’s profits
share on the basis of the actual profit and pay a significant
fine as a (additional) penalty for the false reporting. This
contractual arrangement shall reduce (if not eliminate) the
moral hazard incentives for the financed firm.
By resolving the firm’s incentive to swindle, the model
removes only one of several obstacles that deter Islamic
financial institutions to engage more in “true” PLS
financings. If the bank has fixed the profit sharing
ratio, and the correctly reported profits fall short of the
expectations, the bank may not receive the benchmark
of, for example, murabaha or ijarah financings with low
risk and predetermined returns (especially not if the
bank insisted on an audit and has to bear the full auditing
costs). This, however, may be to the advantage of the firm
because the PLS financing costs would be less than the
costs of sale/rent-based modes of financing with low risk
and predetermined returns for the bank. This, by the way,
creates another problem – a kind of an ex ante replacement
of the discouraged ex post moral hazard: the higher the
profit expectations of the bank, the lower the profit sharing
ratio necessary to meet a given benchmark. If the firm
is able to present its project in such a way that the bank
becomes more optimistic about the future profit than the
firm itself, and if the firm’s expectations are correct, then
the results for the firm are similar to those of false reporting
– but without any risk of penalties.
Hasan (2002) presented a model for the determination
of the equilibrium profit share for a bank (as provider of
mudarabah capital to a firm) in a mixed system (= where
the firm has the option of an interest-based loan financing)
under alternative constellations of the externally-set
market rates of interest (which is a kind of benchmark),
a risk premium (to be paid in case of loan financing), the
so-called leverage ratio (= share of externally provided
capital on interest or mudarabah base in the total capital
of the firm), and the total return on capital of the firm.
The equilibrium profit sharing ratio of the bank will be less
than its loss sharing ratio, and it will vary inversely with
the return on total capital and directly with the leverage
ratio. The model clarified some interdependencies
between variables, but it did not explain (without
163
Nienhaus
additional exogenous hypotheses) why the use of mud­
arabah financing is so low or what could be done to
increase its use. Therefore Hasan supplemented his model
with contemplations on a change of Muslims’ attitudes
towards very high-risk aversion in business matters (risk
understood as uncertainty, which cannot be measured and
insured). Since banks are no exceptions to this behavioral
pattern, they are very reluctant to finance risky ventures
in which both the earnings and the repayment of the
provided capital are uncertain. To improve the situation,
Hasan makes two proposals:
• The contracting parties should agree on a clause in
the mudarabah contract “to treat the bank among
the preferential creditors of the borrowing firm in
case of insolvency …
• … alternatively, the bank can be allotted redeemable
preference shares for the money advanced.” (Hasan
2002, 49–50).
It is very questionable whether Shariah could accommodate
these provisions. For example, AAOIFI had dealt with
the Shariah rules and requirements in the Financial
Accounting Standard No. 3 on mudarabah financing and
No. 4 on musharakah financing, both adopted in 1996.
Neither the treatment of the bank as a preferential creditor
nor the allotment of redeemable preference shares, as a
kind of security, are compatible with the restrictive Shariah
principles spelled out by AAOIFI. AAOIFI explicated the
principles in more detail again in the Shariah Standards
No. 12 on musharakah and No. 13 on mudarabah, both
adopted in 2002.
Diaw, Bacha and Lahsasna (2012) address agency
problems of participatory finance not in banking but
for musharakah sukuk. They diagnose a fundamental
incongruence between (formal) requirements of Shariah
contracts and they aim “to reproduce the substance of a
financial instrument that is repugnant to their nature and
to the Islamic paradigm in finance” (p. 45). The proposed
solution takes inspirations from convertible bonds and
develops the idea of equity kickers (Bacha 1997) further.
Basic agency problems arise in mudarabah or musharakah
arrangements (bank financings or sukuk) because the
financing contract covers only a limited period of time.
It is in the interest of the mudarib or the managing party
of a musharakah and the owners of the financed firm to
keep as much of the realised profits within the firm during
the period of the participatory financing. A wide range
of possible measures to compress the reported profit that
has to be shared with the capital provider – from cost
allocation to false reporting – has already been indicated.
These techniques reduce the value (payouts) of sukuk
and increase the value of the firm. This is a problem for
the capital provider if he cannot participate in the increase
of the value of the firm. A solution would be a conversion
clause (which is like an embedded option in the financing
contract): If the sukuk performance falls short of a
benchmark (for example, if the return on the sukuk is less
than a stipulated minimum), the sukuk holders have the
right (but not the obligation) to convert their sukuk into
shares of the financed firm. The exercise of the option
would not only prolong the initially temporary partnership
to an indefinite period but would also grant ownership
rights to the previously silent partner.
164
A weakness of this concept that it is applicable only for
joint stock companies and it requires the readiness of the
actual shareholders to accept a certain dilution of their
ownership rights by the issuance of new or the transfer of
existing stocks to new shareholders. This cannot be taken
for granted, for example, for family-run businesses.
Further, one may challenge the fairness or balance of a
conversion clause for cases where the underperformance
is not due to profit skimming of the financed firm but
to market forces which lead to an unexpectedly poor
performance of the financed project. In such a situation
the conversion option may be felt as an undue and very
severe penalty for something that was beyond the control
of the management and not caused by its “misbehavior.”
The penalty is severe because the conversion gives the
financing party a perpetual claim on parts of all future
profits, which may exceed the amount of the initially
provided capital by a multiple. While this may be very
attractive for the capital provider, it is far from obvious that
it would also be acceptable for the financed firm, especially
if the murabahah or musharakah was only for a short to
medium term.
Finally, a very similar result for overcoming the moral
hazard issues, but without the requirement of a particular
legal form of the financed firm (joint stock company) and
the “penalty character” of the conversion clause (under
adverse market conditions), could be achieved by longer
term or even a perpetual sukuk with a redemption clause.
The redemption clause would be a promise (wa’d) of
the issuer to redeem sukuk certificates at their fair value
on the demand of the sukuk holders. An increase of the
value of the firm would be reflected in the fair value of the
certificates.
It has to be mentioned here that Diaw, Bacha and Lahsasna
(2012) offer another approach for the analysis of equitybased sukuk which is independent from the equity kickers
proposal. It is based on the idea of variable profit sharing
ratios. A few remarks on differences between their analysis
and the model outlined below can be found at the end of
this paper.
3. The lack of participatory financing
in banking and capital market
While the agency problems of participatory modes of
finance can hardly be ignored, it seems that none of the
solutions developed in the academic literature have been
applied in practice. While mudarabah financing is virtually
non-existent in banking, musharakah sukuk expanded at
an extraordinary rate over the first half of the 2000s and
became the most widely used form of sukuk by 2007 (see
chart below).
Profit and loss sharing in banking
The virtual non-existence of participatory (= mudarabah
or musharakah) bank financing may, on the one hand, be
explained by the aforementioned agency problems and an
adverse selection problem that is summarized in the box
below (The “lemon problem” in mudarabah financing).
These issues in their totality have not been solved in
theoretical models.
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Self-adjusting profit sharing ratios for Musharakah financing
The “lemon problem” in mudarabah financing
An entrepreneur can always compare the costs of Shariah-compliant funding based on the expected profit and a
negotiated profit-sharing ratio on the one hand, and the costs of a fixed mark-up sale/rent financing for his project
on the other hand. Less religious-minded entrepreneurs could also add riba-based financing alternatives offered by
conventional banks. The mark-up is determined by competition (within the Islamic banking sector and/or between
Islamic and conventional banks) and becomes the benchmark rate to which the profit sharing ratio has to be adjusted
for a given expected profit. Roughly-speaking, both for the bank and the entrepreneur the mark-up multiplied by the
amount of financing should equal the expected profit multiplied by the respective profit-sharing ratio. If the entrepreneur
were able to convince the bank to become optimistic about his project and to expect a profit that is higher than the
profit he himself expects realistically (without communicating this to the bank), then the bank would agree on a profit
sharing ratio, which is too low (compared with the benchmark or the ratio based on a “realistic” profit expectation).
In this setting, the entrepreneur would benefit from profit sharing financing. The bank has to take into account that in
principle all customers who ask for a mudarabah financing have a strong incentive to present overly optimistic profit
projections. The bank could protect itself to some degree against wrong profit projections by a thorough evaluation
of business plans. But that requires human resources with a profound knowledge of the markets of their customers,
and the experts of the bank should, on average, be better than the entrepreneurs themselves in predicting financial
outcomes of business plans. Expert staff with such qualifications is hard to find, very expensive and probably even
harder to retain (because these employees have all the qualities to become entrepreneurs by themselves). Therefore,
the bank may take recourse to a less expensive protective mechanism, namely a simple “safety margin” on all profitsharing ratios. But this will be anticipated by the entrepreneurs. If an entrepreneur presents a realistic profit projection,
the “safety margin” on the bank’s profit sharing ratio will make the mudarabah financing more expensive for him than
the mark-up financing. Entrepreneurs with good projects may not like to enter into the troubles of debating with the
bank the credibility of their profit projections (in order to eliminate the safety margin). Instead, they prefer fixedcost financing from the outset. In contrast, for entrepreneurs with weak projects, the mark-up financing may be too
expensive, and they have a strong incentive to present an acceptable profit projection to the bank. In the end, the
bank will have more weak than strong projects in its mudarabah portfolio, and it is highly probable that a number of
weak projects will go bust so that the realised profit will fall short of the expectations. To avoid this, it is probably the
best option not to enter into participatory financing at all. The market for mudarabah and musharakah financing will
collapse (or never emerge).
On the other hand, Hasan’s reference to an exaggerated
risk aversion in Muslim countries points to another
issue in Islamic banking: The most widely used Shariah
compliant alternative for interest-bearing savings and term
accounts are mudarabah-based unrestricted investment
accounts. Conceptually, losses from the investment of the
investment account holders’ funds should be passed on to
the investment accounts. The investment account holders
are most probably risk averse, although they have signed a
profit sharing and risk bearing contract: Muslims who were
looking for an alternative for conventional savings and term
deposits which could give them a Shariah compliant return
have hardly an alternative to a mudarabah-based contract,
which exposes their funds to a risk of loss. But to accept
such a contract because no Shariah compliant alternative
is available, does not imply that the account holders would
ever want this risk to materialize. Most probably they
expect from the bank that all conceivable forms of risk
mitigation and risk avoidance are applied. Some Islamic
bankers articulate such expectations very forcefully. They
defend their banks’ policies of virtually risk free mark-up
techniques only and the abandonment of mudarabah or
musharakah in the financing business with their fiduciary
duties towards the risk-averse investment account holders.
Mudarabah and Musharakah Sukuk
In view of the serious agency problems of participatory
financing, the boom of musharakah sukuk is much more
surprising than the lack of mudarabah or musharakah in
bank financing. However, a closer look at the practice of
Eds. Hatem A. El-Karanshawy et al.
the issuance of mudarabah and musharakah sukuk offers
a simple explanation: In spite of their participatory form
and their classification as “equity-based” sukuk, most of
these sukuk have never had a participatory substance.
Hence, they did not suffer from the agency problems
discussed in the academic literature. Instead, most of
the musharakah sukuk were intentionally structured as
functional equivalents of conventional bonds, i.e., as debt
instruments with predetermined returns. On the other
hand, “equity-based” sukuk are very flexible and allow (in
contrast to most other types of sukuk such as murabahah
or ijarah sukuk) the issuing of a security, which is not tied
to the true or beneficial ownership of an existing specific
tangible asset. Instead, mudarabah and musharakah
sukuk create joint ventures for the investment of the sukuk
capital in profit-generating Shariah approved assets, but
these assets must not yet exist when the joint venture is
formed. The assets can be created by the employment of
the sukuk resources (i.e., the money paid by the sukuk
subscribers), and the composition of the assets held by
the joint venture can change over the life of the sukuk.
This flexibility regarding the underlying asset was the
main attraction for practitioners and can explain the rapid
growth of musharakah sukuk. Their popularity was not due
to their equity structure which, in theory, brought them
closer to the Islamic economists’ ideal of participatory
finance. On the contrary, the equity elements, in particular
the possible volatility of returns and the downside risk of a
capital loss, were somewhat disturbing and have effectively
been removed by contractual engineering.
165
Nienhaus
Debt character of equity-based Sukuk:
Capital guarantees
The debt character of equity-based sukuk was achieved
by a (binding) promise of the obligor to repurchase the
sukuk certificates at maturity (or in the event of a default)
at their issuing price, respectively their face value. This
eliminates contractually the risk of a capital loss of the rab
al-mal, and it is a functional equivalent to the guarantee
of the capital of one party by the other. Effectively, losses
are not borne in proportion to the capital contributed to the
venture, or even not borne at all by the capital providers.
An arrangement with such a consequence can hardly meet
the Shariah principle that only risk justifies return. The
principle of loss sharing or loss bearing has been stated over
and again—from classic legal manuals to contemporary
AAOIFI standards (for example, Financial Accounting
Standard No. 4 on musharakah financing, adopted in 1996,
Shariah Standard No. 5 on guarantees, adopted in 2001,
No. 12 on sharika (musharakah), adopted in 2002, and No.
17 on investment sukuk, adopted in 2003).
There is but one escape from the rule of loss bearing by
the capital provider, namely a voluntary guarantee by an
independent third party. The case study of the 2005 IDB
sukuk by Mokhtar (2011, pp. 34–35) reveals that the criteria
for the definition of an “independent third party” can be
very formalistic and limp in economic substance. IDB had
set up for its sukuk issuances the IDB Trust Services Limited
in Jersey, a SPV with an authorized share capital of £10,000
and an issued share capital of £2. All the assets underlying
the sukuk issuances were transferred from IDB to its SPV,
and the prospectus of the 2005 sukuk advertised the fact
that IDB was the unconditional and irrevocable guarantor
of the sukuk issuance. The Shariah Board of IDB declared:
“As it [IDB] is not the issuer of the Trust Certificates and is
not a manager or participant, IDB can enter into contractual
obligations which have the effect of guaranteeing the
Aggregate Nominal Amount of the Trust Certificates and
any Periodic Distribution Amounts in respect of the Trust
Certificates.” Unfortunately, the Shariah resolution does
not disclose whether the Shariah Board considered his
decision in accordance with AAOIFI standards, and if not,
how it would justify its different position.
Predetermined returns for equity-based Sukuk
For achieving predetermined returns, sukuk engineers used
at least three different techniques:
• When actual profits fall short of the expected profits,
sukuk managers often provided interest-free loans
(which should be recovered later) in order to meet
the expectations of the sukuk holders and to beef-up
the payouts to them. AAOIFI made it clear that this
practice is not Shariah compliant: “It is not permissible
for the Manager of Sukuk, whether the manager acts as
Mudarib (investment manager), or Sharik (partner),
or Wakil (agent) for investment, to undertake to offer
loans to Sukuk holders, when actual earnings fall short
of expected earnings. It is permissible, however, to
establish a reserve account for the purpose of covering
such shortfalls to the extent possible, provided the
same is mentioned in the prospectus.”
• Another technique that violates Shariah principles
was applied in some musharakah sukuk with very
166
special purchase undertakings (PUs), which did not
only guarantee the face value of the certificate at
maturity, but factually also the expected profit. These
PUs were not only triggered by the maturity of the
sukuk but also when the venture was not performing
well and the obligor failed to pay the expected profit.
The exercise of the PU obliged the obligor to pay
the outstanding principal plus any so far accrued
but unpaid profit. Mokhtar (2011, p. 33) points out
that “accrued profit is not necessarily actual profit
earned. Profit accrued is the expected profit that is
earned by the investors as time passes by.” Given that
the prospectus indicated an expected profit accrued
over the life of the sukuk, then the early redemption
clause for the PU was effectively a guarantee of a
(minimum) predetermined return. This converts the
substance of an equity certificate into the equivalent
of a conventional bond. It is obvious that this very
special form of a “face value plus accrued profit PU”
violates Shariah principles even more than a “plain
face-value PU” at maturity.
• A more widely used technique for the conversion
of a participatory instrument into a close equivalent
of a bond with (nearly) predetermined returns
for the sukuk holders in “good” years (where the
mudarabah or musharakah generates a return that
meets or surpasses an articulated profit expectation
(=benchmark) were “incentive fees”: The profit share
of the sukuk holders is set, for example, at 99%.
Should the actual profit fall short of the expectation
(benchmark), (nearly) all of the profit goes to the
sukuk holders. But as soon as the actual profit exceeds
the expected profit (which should be the normal
situation), the amount of the actual profit that
exceeds the benchmark is given to the sukuk manager
as an incentive fee for “good management.” Suppose
that the agreed upon expected profit is calculated on
the basis of LIBOR plus a risk factor as the benchmark.
Then this arrangement implies that sukuk holders
will receive under “normal circumstances” the
equivalent of the risk-adjusted market rate of interest.
It may come as a surprise, but such a technique is in
harmony with AAOIFI standards and thus should be
considered Shariah compliant.
In November 2007 the chairman of AAOIFI’s Shariah
Board criticized the prevailing practices that changed
the substance of musharakah sukuk from an equity-based
instrument (as it was conceived) into a functional equivalent
of an interest-bearing bond, and in 2008 AAOIFI issued a
resolution on sukuk. This resolution is a pointed summary
of what was already contained in the AAOIFI standards.
Building blocks of “AAOIFI compliant” Sukuk
However, AAOIFI did not summarize in this resolution
a number of remarkable provisions in AAOIFI standards,
which could be used as “building blocks” for the structuring
of more “AAOIFI compliant” sukuk:
• The profit sharing formula has to be fixed at the
beginning of a mudarabah or a musharakah: “It is a
requirement that the mechanism for distributing profit
must be clearly known in a manner that eliminates
uncertainty and any possibility of dispute.”
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Self-adjusting profit sharing ratios for Musharakah financing
• The profit distribution does not have to be based on a
profit sharing ratio which is a simple percentage. The
ratio can also be based on a more complex formula:
“It is permissible for the partners to agree on the
adoption of any method of allocation of profit, either
permanent or variable, for example, by agreeing that
the percentages of profit shares in the first period are
one set of percentages and in the second period are
another set of percentages, depending on the disparity
of the two periods or the magnitude of the realised
profit. This is allowed provided that using such a
method does not lead to the likelihood of a partner
being precluded from participation in profit.”
• Irrespective of the complexity of the initially accepted
formula, it is “permissible for the parties to change the
ratio of distribution of profit at any time and to define the
duration for which the agreement will remain valid.”
• The changing of the distribution scheme can even be
made when the actual profit of the venture is known
at the end of the life of the sukuk: “The parties may
bilaterally agree to amend the percentages of profitsharing on the date of distribution. Also, a partner may
relinquish, on the date of distribution, a part of the
profit that is due to him in favor of another party.”
• “It is permissible for the issuer or the certificate
holders to adopt permissible methods of managing
risk, of mitigating fluctuation of distributable profits
(profit equalisation reserve), such as establishing
an Islamic insurance fund with contributions of
certificate holders, or by participating in Insurance
(Takaful) by payment of premiums from the income
of the shares of Sukuk holders or through donations
(tabarru’at) made by the Sukuk holders.”
The changing Sukuk landscape
Instead of using these building blocks to modify sukuk
in such a way that they meet the 2008 AAOIFI Shariah
pronouncement on musharakah sukuk, the practice moved
away from equity-based sukuk and switched to ijarah and
murabaha sukuk since 2008.
The trends captured in the graph until 2009 continued: In
2012, the share of mudarabah and musharakah sukuk in
global sukuk issuances (in US$) had declined to 15% while
murabaha and similar sale-based sukuk accounted for 65%
and ijarah sukuk for 16%.
The growing popularity of sale- and rent-based sukuk could
be explained as follows:
• Murabaha sukuk are based on short term debt creating
sale transactions, and a sufficiently large portfolio of
such transactions should provide effectively a builtin protection (albeit not a formal guarantee) for the
capital invested by the sukuk holders and accrued
profit shares.
• Ijarah sukuk have a longer maturity and a significantly
higher market risk that could lead to losses. However,
AAOIFI allows for this type of sukuk a straightforward
guarantee of the capital by a purchase undertaking
of the issuer at face value.
Another reason for the departure from musharakah sukuk
may be that the aforementioned options for more flexible
profit distribution rules and re-negotiations were not
practicable in the business environment in which sukuk
flourished, namely the market of institutional investors
and the interbank market. The market participants are
mostly financial institutions which are hardly interested in
participatory components in their contractual arrangements
because that would imply an additional rate of return risk.
Further, if predetermined returns are the objective, it may
be less complicated and less risky to structure sale- or leasebased deals than to insert more complex profit distribution
formulas into musharakah contracts. Finally, a readjustment
Figure 1. Trend in types of Sukuk issued.
Source: Mokhtar 2011, p. 5
Eds. Hatem A. El-Karanshawy et al.
167
Nienhaus
of the profit sharing ratios at the day of distribution, i.e.,
when the distributable profit is known, is commercially
equivalent to the fixing of the distribution of the profits in
absolute amounts.
More important, a profit sharing ratio agreed upon at the
beginning of the contract will generate one allocation of
amounts (in absolute figures) on the distribution day, and a
revised ratio will generate a different allocation of amounts.
To accept the revised ratio implies a definite gain for one
party and a definite loss for the other. It is extremely unlikely
that banks or institutional investors would voluntarily give
up profits which are legally due to them in favour of another
bank or institutional investor. Against this background
it is not surprising that the additional building blocks for
mudarabah and musharakah sukuk were not utilized in the
typical market environment for this kind of sukuk.
But the situation could be different under a different
market setting and for institutions with somewhat different
objectives. For example, “Modarabah Companies” were
established in the 1980s in Pakistan. They can have a similar
function as an SPV (set up, for example, by a financial holding
company or a bank): As non-banks, they could collect funds
through the issuance of mudarabah sukuk and invest these
funds in profitable projects, for example in the leasing of
equipment to manufacturing firms on the basis of ijarah
contracts or in seed or growth financing on a musharakah
basis. “Initially, there was a desire for the mudarabah sector
to concentrate on funding SMEs in Pakistan, which were at
times neglected by the banks. Unfortunately, due in part to
the profit motive, it has steered more toward medium-sized
enterprises and big-ticket funding, an area where banks
are already involved. Rather than competing with banks,
it may prove a competitive advantage to focus on smaller
enterprises in a microfinance manner, as was originally
envisioned.” (Khwaja 2009, 245).
Non-bank finance companies could utilize for their own
funding sukuk structures. Their sukuk issuances will have
similarities but also marked differences to mudarabah and
musharakah sukuk of investment banks and actors in the
interbank market.
• Finance companies which do not take deposits but
issue sukuk would not need a full banking license
in most jurisdictions, and they would not be subject
to the strict Basel III capital adequacy and liquidity
rules for banks.
•Their sukuk could be subscribed by institutional
investors, but it is also possible to envisage a retailoriented marketing. Malaysia has recently taken
steps towards the creation of a retail sukuk market.
•Institutional investors have already shown their
strong preference for predetermined returns. On
the other hand, it seems reasonable to assume that
also retail investors – similar to investment account
holders – would be risk averse even if they purchase
equity-type securities.
•The profit generating projects of sukuk-funded
finance companies will be smaller than the big ticket
transactions in the actual sukuk market. Small and
medium sized enterprises (SME) could become a
particular target group for the finance companies.
On the one hand, SMEs may be still underserved
168
by commercial banks in many Muslim countries, in
particular in those countries where the transformation
of the economic system is on the political agenda. On
the other hand, SMEs are usually individual or family
run enterprises which do not have the legal form of a
joint stock company and cannot get funding from the
floating of shares.
• Two phases in the life of a SME can be of particular
interest to sukuk-funded finance companies: the
start-up phase and the expansion phase (after
a SME has become established and grows in its
market) with different risk/return profiles for funds
provided. Both could be quite attractive for finance
companies compared with financings in areas where
stiff competition (by banks) has compressed margins.
However, sufficient expertise in the industries of the
financed SMEs is required for the assessment of the
business plans for start-up or expansion.
• The finance companies could provide sale/rentbased financings (murabaha, istisna’, ijarah) to their
target group, but they could also apply equity-based
types of financing, for example mudarabah for startups, musharakah for growth financing. Participatory
modes of finance could generate higher returns, but
they are also associated with higher risks: general
business risks, but also the risks resulting from the
various agency problems outlined above.
Where SMEs are not joint stock companies, the agency
problems in financing cannot be solved by those proposals
which are based on a conversion of temporary external
participatory capital to permanent equity by providing
stocks to mudarabah or musharakah investors. Some SMEs
may have the legal form of joint stock companies, but if
they are owned by individuals or families who do not want
to dilute the ownership structures through the issuance
of new (or transfer of existing) stocks to other parties, they
will hardly opt for finance contracts with equity kickers or
similar conversion clauses.
4. A Musharakah Sukuk with a selfadjusting profit sharing ratio
Since AAOIOFI has provided an interesting but underutilized
toolbox for flexible mudarabah and musharakah sukuk
structures, it is possible to reconcile the main interests of
sukuk holders and financed entrepreneurs, namely:
• the interest of risk-averse sukuk holders in riskmitigated predictable and stable returns,
• the interest of entrepreneurs
• to get some financial relief in “bad times,” i.e.,
to have some risk-sharing elements in financing
arrangements in a year of unexpectedly poor
performance
• to have shared but uncapped upside potentials in
“goods times,” i.e., to avoid a total skimming-off
of profits in years of above-average performance
The basic idea of the musharakah sukuk with a selfadjusting profit sharing ratio is quite simple: If it is
permissible to adjust the profit-sharing ratio at any time
during the life of a mudarabah or musharakah sukuk by
consent of the contracting parties, this must not be done in
a discretionary manner (by separate negotiations after new
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Self-adjusting profit sharing ratios for Musharakah financing
performance information become available). Instead, this
can be automated by a simple formula which links the profit
sharing ratio to the actual performance (in the simplest
case to the actual profit) of the financed project or venture.
To calculate performance-dependent profit sharing ratios for
a musharakah structure between an SPV and an enterprise,
it must be known how much capital is provided externally
by the SPV (respectively the sukuk holders) and how much
internally by the enterprise (originator). At the beginning
of the joint venture, a certain total profit is expected which
has to be distributed to the external and the internal capital
provider. The profit sharing ratio denotes the share of profit
allocated to the external capital provider. This profit sharing
ratio will not be fixed numerically directly but computed by
a formula on which the parties have agreed. This formula is
based on a distribution pattern for the profits about which
the contracting parties have achieved a consensus. It could
be, for example, a fixed relation between the rate of return for
the external and for the internal capital, or it could tie one of
the rates of return to an external benchmark. The distribution
pattern, i.e., a particular relation between the rates of return
or the link of one rate to a benchmark, has to be determined
at the beginning of the joint venture, and this pattern
(= distribution formula) should not be modified afterwards.
A model in which only ratios are agreed upon does not fix
ex ante the absolute value (in currency units) of the profit
for the contracting partners that would violate Shariah
principles. It only determines the relative positions of the
parties. The absolute volumes depend on the realized
profits. To keep the relative positions of the parties in the
agreed-upon proportion, it is necessary that the profit
sharing ratio is adjusted automatically whenever new
information on the expected or actual profit of the joint
venture become available. This is automatically achieved
by a simple formula. For the computation of the adjustable
profit sharing ratios, the following variables are used:
Ke = external capital (provided by the SPV)
Ki = internal capital (provided by the entrepreneur)
K = Ke + Ki = total capital
P = total distributable profit
Pe = s ∙ P = profit allocated to the external capital
Pi = (1 − s) P = profit allocated to the internal capital
s = Pe/P = profit sharing ratio = share of total profits
allocated to the external capital
β = benchmark rate of return
re = Pe/Ke = rate of return for external capital
ri = Pi/Ki = rate of return for internal capital
r = p/K = rate of return on total capital
Next an “objective function” has to be defined. The
contracting parties may discuss various alternatives. For
example, they may consider it as a just distribution that
both parties achieve the same rate on return on their
invested capital. A variant of this approach could be to give
one party a predetermined bonus over the share of the other
partner (for example, as a compensation for management
efforts). Another plausible scenario could be that the
provider of the external capital prefers a profit distribution
which reduces the volatility of his own profit share and
gives him a more stable revenue stream that meets a certain
benchmark; for example, the return from an investment in
fixed-income securities (such as ijarah sukuk); exceeding
Eds. Hatem A. El-Karanshawy et al.
profits will remain with the other party in good years, the
value of its profit share would be reduced in bad years.
The simple model outlined below indicates, inter alia, that
the distribution parameters could also be calibrated with
respect to leverage effects of external capital (provided
the parties agree on such a pattern). These are only a few
examples for a wide range of conceivable arrangements for
the reconciliation of the revealed preferences and interests
of the contracting parties in a participatory finance setting
such as a musharakah (sukuk or bank financing). The
preferences and interests are contractually recorded and
translated into a structure that protects the distributional
preferences of the contracting parties by automatic
alignments of the distribution parameters (in particular
the profit sharing ratio) whenever new performance
information of the project or joint venture become available.
The automatic adjustment obviates or replaces ex post
re-negotiations on redistributions which are permissible
according to AAOIFI, but very difficult to realise when the
gains of one party are the losses of the other party.
The following is an illustration of the basic mechanism
underlying the idea of an adjustable profit sharing ratio.
Assume that the contracting parties had agreed on one of
the following distribution rules (“objective functions” in
the model):
The profit sharing ratio should be such that either
• the return on external capital is equal to the return
on total capital, or
• the return on external capital is the same as the
return on internal capital, or
• the return on external capital equates the benchmark
β, or
• the return on internal capital is a multiple or fraction
α of the return on external capital
The appropriate profit sharing ratios are determined as
follows:
Cases (1) and (2), profit sharing ratio for re = ri = r:
Case (3), profit sharing ratio for re = β:
Case (4), profit sharing ratio for ri = α ∙ re:
A numerical example is given in the following table. It
illustrates the influence of different benchmarks, of a
surcharge on the relative return for one party, and of
different (actual or expected) profits. The table does not
change the relation between external and internal capital,
but the relevance of the capital structure is clearly visible
from the profit sharing formulas above.
The above sample did explain the basic mechanism of the
model only for three simple objective functions. In practice,
contracting parties might agree on more complex formulas
(for example, with caps or equivalents of “incentive fees”).
Admittedly, it would also be possible to define an objective
in such a way that one party receives a fixed amount as
profit share (provided the volume of the total profit is at
least as large as the fixed amount) – which would not be
permitted under Shariah. But discretionary re-negotiations
can achieve the same result. Insofar the automated system
is not better or worse than discretionary practices regarding
a possible “misuse.”
169
Nienhaus
Profit Sharing Ratios for Different Objective Functions
Objective function:
Ke
Ki
K = Ke+Ki
P
s = Pe/P
Pe = s∙P
Pi = (1-s)P
β and α
re = Pe/Ke
ri = Pi/Ki
r = P/K
Assumption
Assumption
Definition
Assumption
Definition
Def. & result
Def. & result
Assumption
Def. & result
Def. & result
Definition
400
100
500
50
re = r ( = ri)
re = β
re = β
ri = αre,
α = 1.1
ri = αre,
α = 0.8
400
100
500
50
400
100
500
50
400
100
500
50
400
100
500
50
400
100
500
50
40
10
48
2
0,120
0,120
0,020
0,100
32
18
0,080
0,080
0,180
0,100
39
11
1,100
0,098
0,108
0,100
42
8
0,800
0,104
0,083
0,100
0,960
0,640
0,784
0,833
0,100
0,100
0,100
0,100
re = r (= ri)
0,800
Solutions:
s = Ke/K
s = β(Ke/P)
re = β
s = 1/(1+αKi/Ke)
ri = αre
Objective function:
Ke
Ki
K = Ke+Ki
P
s = Pe/P
Pe = s⋅P
Pi = (1-s)P
β and α
re = Pe/Ke
ri = Pi/Ki
r = P/K
Assumption
Assumption
Definition
Assumption
Definition
Def. & result
Def. & result
Assumption
Def. & result
Def. & result
Definition
400
100
500
75
re = r ( = ri)
re = β
re = β
ri = αre,
α = 1.1
ri = αre,
α = 0.8
400
100
500
75
400
100
500
75
400
100
500
75
400
100
500
75
400
100
500
75
60
15
48
27
0,120
0,120
0,270
0,150
32
43
0,080
0,080
0,430
0,150
59
16
1,100
0,147
0,162
0,150
63
13
0,800
0,156
0,125
0,150
0,640
0,427
0,784
0,833
0,150
0,150
0,150
0,150
re = r (= ri)
0,800
Solutions:
s = Ke/K
s = β(Ke/P)
s = 1/(1+αKi/Ke)
re = β
ri = αre
It is clear that the profit sharing mechanism will only work as
long as profits are generated. It does not provide an effective
protection against the need of “loss sharing” in an individual
musharakah setting. In the interest of risk mitigation and loss
avoidance, the management of a musharakah sukuk should,
for example, diversify investments by financing not only one
entrepreneur but a number of firms in different markets
with uncorrelated market trends, and it may also invest
some funds in risk-minimized fixed-return instruments.
But risk management in Islamic finance in general is a topic
which goes beyond the scope of this paper.
The idea of a variable profit sharing ratio can also be found
in the paper of Diaw, Bacha and Lahsasna (2012). But their
perspective is a theoretical analysis and not the outline of
an implementable recommendation for contracting parties
170
in a musharakah structure. They present a general equation
that indicates the directions in which different parameters
influence the profit sharing ratio. They do not transform
their general equation in such a way that it would become
an objective function and could be used (after calibration)
in negotiations on a distribution pattern. Instead, Diaw,
Bacha and Lahsasna present a Monte Carlo simulation,
which gives a feeling of implied dependencies and possible
dynamics. In addition, they offer a backtesting, which
shows relations between the average return on capital,
an average benchmark (indicated opportunity costs or
benefits), the average return to sukuk, return to equity and
the profit sharing ratio. However, their input data for this
exercise were taken from mudarabah sukuk, i.e., sukuk
which do not incorporate a profit sharing ratio. Insofar it
is not clear what the results of the backtesting can show.
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Self-adjusting profit sharing ratios for Musharakah financing
It may be read such that it shows (on the basis of observed
returns on equity and returns on sukuk capital) what profit
sharing ratios would have been necessary in each period to
achieve a certain benchmark return.
Diaw, Bacha and Lahsasna do not explicitly integrate
changes of the expected (or actual) profit into their
analysis. Changes of profits and those adjustments of the
profit sharing ratio during the life of a participatory finance
are the main focus here, which are required to sustain the
initially stipulated profit distribution pattern.
Islamic economists have made many attempts to overcome
agency problems in participatory finance in order to make
them more acceptable to market participants, in particular
to providers of funds. This is because many see participatory
finance as the ideal form of Islamic finance and as the core
element of a genuine Islamic financial system. Even if one
would not go that far, participatory modes of finance such
as mudarabah or musharakah could fill a gap, for example,
in the start-up and growth financing of SMEs. In general,
participatory finance gives entrepreneurs some financial
relief in times of an unexpectedly poor performance of their
business. This risk-reduction can enhance their willingness
and ability to ramp up innovations (new products, processes
or technologies) or to enter into new markets. This should
not only generate private profits but also social benefits
in terms of employment and income opportunities. The
model presented here outlines a technique which could
make participatory modes of financing more attractive to
financiers as well as entrepreneurs seeking finance.
5. Conclusion
AAOIFI has allowed a remarkable wide range of “corrective
measures” in mudarabah and musharakah contracts
which are deemed Shariah compliant: The contracting
parties can re-negotiate factually all commercially relevant
aspects of their contracts at any time. A main reason for
such re-negotiations will be new information about the
performance (profit) of the financed project or venture. The
crux of re-negotiations is that they will be successful only if
one party is willing to give up advantages (financial gains)
of the contractual status quo. This voluntary redistribution
is not very likely.
The proposal of a self-adjusting profit sharing ratio obviates
the need for discretionary re-negotiations. It uses elements
of the AAOIFI toolbox to structure a contractual arrangement
that maintains the distribution pattern, which was initially
agreed upon by the contracting parties. It does so by an
automatic adjustment of the profit-sharing ratio whenever
new information on the expected profit becomes available.
This technique facilitates a solution of fundamental agency
problems in participatory finance. It does not require
a particular legal form of the financed firm (such as a
joint stock company) and can be calibrated such that risk
aversion of the financier can be factored in. A musharakah
contract with a self-adjusting profit sharing ratio is incentive
compatible insofar as it does not provide the incentives for
profit compression by underreporting, by allocation of fixed
cost, by fringe benefits, or by shifting profits into periods
after the termination of the musharakah contract. All this
is achieved without the need for a sophisticated accounting
system of the financed enterprise.
Eds. Hatem A. El-Karanshawy et al.
For the financed entrepreneur it is beneficial that
a participatory financing implies a kind of “embedded”
insurance against unexpected downside risks (loss sharing).
However, the loss absorbing qualities of a mudarabah or
musharakah contract will not come for free but will be
reflected in the costs of funds (e.g., by a risk premium
added to a benchmark rate by the financier). On the other
hand, the contract can be calibrated such that both parties
can enjoy upside gains. This is in contrast to the widespread
practice of a complete skimming off of gains by one party
in musharakah sukuk. The financier can trade in upside
opportunities for predictable and stable returns based on a
risk-adjusted benchmark rate.
It is obvious that arrangements based on self-adjusting profit
sharing ratios are no “real” profit and loss sharing contracts
or “full” risk sharing partnerships. Insofar the approach is
not the first best solution in an ideal world. However, under
real-world circumstances first best models do not work,
and the technical and Shariah merits or shortcomings of
the approach should be discussed in relation to other realworld proposals to overcome the inherent agency problems
of PLS or risk sharing arrangements.
References
AAOIFI. (2008) Guidance Statement on Accounting for
Investments and Amendment in FAS 17. Available at:
http://www.aaoifi.com/aaoifi_sb_sukuk_Feb2008_
Eng.pdf.
AAOIFI. (2010a) Accounting, Auditing and Governance
Standards for Islamic Financial Institutions. 1432 H –
2010. Manama: AAOIFI.
AAOIFI. (2010b) Shari’a Standards for Islamic Financial
Institutions. 1432 H – 2010. Manama: AAOIFI.
Abdel-Khaleq AH, Crosby T. (2009) Musharakah Sukuk:
Structure, Legal Framework and Opportunities. In:
Abdulkader T (Ed.). Sukuk. Sweet & Maxwell Asia.
Petaling Jaya. 187–222.
Ahmed H. (2002) Incentive-Compatible Profit-Sharing
Contracts: A Theoretical Treatment. In: Munawar I,
David T. Llewellyn DT (Eds.). Islamic Banking and
Finance – New Perspectives on Profit-Sharing and Risk.
Edward Elgar. Cheltenham. 40–54.
Bacha OI. (1997) Adapting Mudarabah Financing to
Contemporary Realities: A Proposed Financing Structure.
The Journal of Accounting, Commerce and Finance. 1(1).
http://mpra.ub.uni-muenchen.de/12732/.
Casey P. (2012) Regulatory Lessons on Sukuk Financial
Products, an Opinion. In: Ariff M, Munawar I, Shamsher
M (Eds.). The Islamic Debt Market for Sukuk Securities:
The Theory and Practice of Profit Sharing Investment.
Edward Elgar. Northampton, MA. 99–118.
Diaw A, Obiyathulla IB, Ahcene L. (2012) IncentiveCompatible Sukuk Musharakah for Private Sector
Funding. ISRA International Journal of Islamic Finance.
4(1):39–80.
Hasan, Z. (2002) Mudarabah as a Mode of Finance in
Islamic Banking: Theory, Practice and Problems. Middle
East Business and Economic Review. 14(2):41–53. http://
mpra.ub.uni-muenchen.de/2951/.
171
Nienhaus
Ibn R. (1996) The Distinguished Jurist’s Primer, Vol. 2
(Bidayat al-Mujtahid). Translated by Imran Ahsan Khan
Nyazee. Garnet. Reading.
Kapetanovic H, Muhamed B. (2009) Mudharabah Sukuk:
Essential Islamic Contract, Applications and Way
Forward. In Abdulkader T (Ed.). Sukuk. Sweet &
Maxwell Asia. Petaling Jaya. 223–247.
Khwaja M. (2009) Mudharabah Sector Report – Pakistan.
In Abdulkader T (Ed.). Sukuk. Sweet & Maxwell Asia.
Petaling Jaya. 245–246 (= Annex to Kapetanovic and
Becic 2009).
Demetriadies DG, Tyser CR, Effendi IH (Translators).
(2001) The Mejelle – Being an English Translation of
Majallah El-Ahkam-i-Adliya and a Complete Code on
Islamic Civil Law. The Other Press. Kuala Lumpur.
Saeed A, Salah O. (2012) History of Sukuk: Pragmatic and
Idealist Approaches to Structuring Sukuk. In: Ariff M,
Iqbal M, Shamsher M (Eds.) The Islamic Debt Market
for Sukuk Securities: The Theory and Practice of Profit
Sharing Investment. Edward Elgar. Northampton, MA.
42–66.
Mokhtar S. (2011) Application of Wa’ad in Equity Based
Sukuk: Empirical Evidence. ISRA Research Paper 20.
ISRA. Kuala Lumpur.
172
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Indexing government debt to GDP: A risk
sharing mechanism for government financing
in Muslim countries
Syed Aun R. Rizvi1, Shaista Arshad2
INCEIF, Lorong Universiti A, Kuala Lumpur, Malaysia, Phone: +60136145752, Email: [email protected]
IIUM, Institute of Islamic Banking and Finance, Kuala Lumpur, Malaysia, Phone: +60133771370,
Email: [email protected]
1
2
Abstract - Over the past decades much effort and research has gone into establishing a viable set
of Islamic financial institutions. An area of utmost importance, which still has gaping holes, is the
development of instruments for government financing on a global level. Most Muslim countries,
with the exception of a few Gulf countries, are heavily indebted with high reliance on multilateral
financing primarily based on high interest rates. This vicious cycle of interest rates and debt have
stunted the growth of these nations and worsened the conditions of the masses. This paper is
an attempt at introducing the concept of risk sharing instrumentation in the form of GDP-linked
Papers for Muslim governments for their financing through multilateral and supranational bodies.
Without dwelling much on the Shariah technicalities, we attempt to discuss the potential benefits
of transferring government debt in these forms of risk sharing instruments. The authors have
attempted to represent the economic benefits of these risk-sharing mechanisms for a set of four
indebted Muslim countries with empirical proof. Through this paper, we endeavour to initiate a
thought provoking and practical discussion for further development of these instruments for the
betterment of Islamic countries.
Keywords: risk sharing instruments, sovereign debt, GDP-linked paper, Muslim countries
JEL Classification Codes: F34, H63
1. Introduction
Islamic finance has progressed phenomenally over the past
few decades, moving from a small market in the Arab world
to a much larger global presence. Marked with impressive
growth, the presence of Islamic finance is estimated to be
US$1.6 trillion and is expected to grow at the rate of 15%
over the next few years.
The viability of Islamic finance as an alternative was put
to test in the wake of the recent financial crisis, where
in the midst of bank failures and stock market crashes,
institutions practicing Islamic finance managed to stay
afloat and rather provided a buffering effect to the crisis.
Several economists argue that due to the innate laws of
Islamic finance it was able to withstand the financial crisis
much better than its conventional counterpart was.
Yet despite this impressive record, Islamic finance only
accounts for 1% of the global financial system. Some
scholars are of the opinion that this is due to the substantial
indebtedness of Muslim nations. Narrowing in on the
Organization of Islamic Centre (OIC) member countries, a
study revealed 19 from 39 Heavily Indebted Poor Countries
(HIPC) are part of the OIC (IMF, 2010). Furthermore, a mere
6 out of 57 OIC member countries, in the Gulf Cooperation
Council (GCC) are prepared to meet financial shocks of
sorts. The remainder are heavily reliant on borrowing from
the international market to pay off their debts.
Additionally, these countries are dependent on Western
financial intermediaries with high interest rates to meet
their debt requirements, transacting mainly in dollars.
During times of economic instability, liability dollarization
(see Calvo et al., 2003; Calvo and Reinhart, 1999; Durdu
Cite this chapter as: Rizvi S A R, Arshad S (2015). Indexing government debt to GDP: A risk sharing mechanism for
government financing in Muslim countries. In H A El-Karanshawy et al. (Eds.), Islamic banking and finance – Essays on
corporate finance, efficiency and product development. Doha, Qatar: Bloomsbury Qatar Foundation
Developing Inclusive and Sustainable Economic and Financial Systems
Rizvi and Arshad
and Mendoza, 2005) and relying heavily on high interest
borrowings can backfire for these countries, as they are more
likely to have long-term shocks on the economy leading to
higher cost of borrowings. Alternatively, they may need to
restructure or renegotiate the terms of borrowing or lastly,
be forced to employ pro-cyclical policies that would lead to
disastrous economic and social outcomes.
One way to avoid any of the above possibilities is to index
their debt against real economic variables to reduce
financial distress. Several economists advocate the value
of indexing government debt reasoning that it would allow
for an improved risk sharing among debtor countries and
international creditors.
Amongst the key proponents of this instrument is Bailey
(1983) who had initially suggested transferring the sovereign
debt into claims on the exports of the country. Delving
further on this, Krugman (1988) and Froot et al. (1989)
studied the virtues of indexing sovereign debts to variables
that are controlled by the country, partially or completely.
Shiller (1993) first proposed the concept of a true GDPlinked bonds, which fundamentally were perpetual claims
on a fraction of a country’s GDP. Similarly, Borensztein
and Mauro (2004) also sought to review the case for
GDP growth-linked bonds, and Griffith-Jones and Sharma
(2006) in their paper summarize the benefits of introducing
GDP-linked bonds. Working on Shiller’s concept, Obstfeld
and Peri (1998) propositioned members of the European
Union to issue GDP-linked euro denominated securities.
Likewise, Haldane and Quay (1999) argue in favour of
indexing sovereign debt to commodity prices. Dreze (2002)
advocated using GDP-indexed bonds as an integral part of
restarting the debt of the poorest countries strongly.
While the idea of GDP-linked debt has been implemented
to a limited extent, it received renewed impetus after a
wave of financial and debt crisis that flooded emerging
markets in the 1990s. Caballero (2002) has suggested the
issuance of copper indexed bonds during Chile’s financial
upheaval. He further recommends making the returns
of debt instruments dependent on external shock variables
that trigger crises (such as exports, regional economic
growth, etc.) arguing that most emerging markets are
affected by external shocks that they are not well-equipped
to withstand.
Reverting back to our issue at hand, as most of Muslim
nations are severely debt ridden, a fundamental question
arises: Does Islamic finance have the aptitude to provide for
the borrowing and financial needs of Muslim governments?
Much has been written about the development and
innovation of Islamic finance over the years. Mirakhor et al.
(2011 and 2012) believes that risk sharing is the foundation
of Islamic finance and is instrumental in the development
of any financial products within the ambits of the Shariah.
A dearth in the present literature found concerns GDPlinked papers using the principles of Islamic finance. The
potential benefit of using GDP-linked Sukuk was discussed
by Diaw et al. (2012) where he proposed structuring the
instrument based on forward ijarah. Mirakhor (2011)
highlights the benefits of using macro-markets and similar
GDP-linked papers in his works. Similarly, the prospective
benefits from using an equity financed budget for Muslim
174
countries was discussed by Hasan and Siddiqui (1992) who
highlighted that an equity financed deficit is not necessarily
unstable and that the output of the economy can be
enhanced by the government expenditure policies.
Keeping in mind the evident lack of research on this
pertinent topic, this paper aims to assess the benefits of
GDP-linked Sukuk, and tries to empirically show some of
the key benefits these instruments can offer. This paper also
attempts to address some of the concerns from the investor
angle, to provide a case for these instruments as alternative
to sovereign debt for Islamic countries.
The present paper is structured as follows: Succeeding the
introduction is the motivation and justification for the study,
which is followed by an insight into the need for GDP-linked
Sukuks, benefits and fiqhi aspect in Section 3. Section 4
provides some simple empirical results on how the economies
will benefit with these instruments, which is followed
by discussion on the viability for investors in Section 5.
Subsequently, the authors provide concluding remarks.
2. Motivation and justification of study
In this study, the primary focus of the authors is to
emphasis on the benefits of employing GDP-linked papers
for Muslim countries that are emerging but debt ridden.
With the substantial growth of Islamic finance over the
years, it becomes imperative to introduce instruments that
would benefit Muslim economies at large. In particular, the
reliance of Muslim nations on high-interest rate borrowing
has largely been neglected in Islamic finance, with little to
no research available on alternatives.
The motivation for this paper arises from the need of
evidence to support the practicality of linking real economic
variables to sovereign debt for OIC countries. This will
further allow us to examine the efficacy of such instruments
on the economies of Muslim countries.
This paper is amongst the few that have discussed the use
of GDP-linked instruments, primarily discussed by Diaw
and Bacha (2011) who proposed Sukuk structures indexed
with real economic variables. The use of real economic
variable linked instruments, as proposed in this study, aims
to quell instability in Muslim countries and help stabilize
government spending while reducing the impact of interest
borrowings.
Islamic finance theoretically has the capacity to provide
an alternative to these Muslim countries that would allow
them to restructure their debt and continue growing as
an emerging economy. Islamic finance should aim to be
the principal financial system for Muslim countries, hence
raising the need for the development of new instruments
and products to meet the needs of Muslim economies at
both a micro- and macro-level.
This study is novel in this area as it is a humble attempt to
initiate an empirical research-based argument to provide
policy makers and economists an alternative Shariah
compliant instrument as a replacement of the conventional
sovereign debt.
In the current literature in conventional finance this area
has not been studied in detail. Regarding Islamic finance,
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Indexing government debt to GDP: A risk sharing mechanism for government financing in Muslim countries
it is very rarely studied area. But considering the structure
of Muslim economies, which are primarily indebted
economies, the issue of sovereign debt is of immense
importance. The objective of this study is to provide
empirically evidence on the viability and economic benefits
of GDP-linked Sukuks.
3. GDP-linked Sukuk
Need for GDP-linked Sukuk
GDP-linked Sukuk aim to provide a breathing room to
economies when they go into recession and on the other
side provide an automatic slowing-in mechanism when the
economy goes into high growth zones and into overheat.
Analysing the advantages of this instrument, we can find
two sets of clear advantages from an economic point of
view for the Muslim countries.
Firstly, since the payment of profit is linked to the actual
growth of the economy, it can reduce the likelihood of
crises. In comparison to the normal borrowing mechanism
that Muslim economies are subjected to via plain vanilla
sovereign bonds or multilateral agency debts, the payments
are fixed, and any slowdown in economy results in a
ballooning of debt.
In existing literature, there is an immense amount of
evidence on the sovereign debt’s position of dependence
on the economic growth. A glance at the history shows
that slow growth has primarily underlined debt crises,
like the Latin American crisis of 1980’s and the near
complete default of the Heavily Indebted Poor Countries
of 1990’s. In economic theory and its applications, the
ratio of external debt to GDP is a significant predictor of
impending crises. Detragiache and Spilimbergo (2001) in
their study empirically show that an increase of 10% in the
debt/GDP ratio has an impact of nearly 20% increment in
the probability of the crisis. In another study by Easterly
(2001), he argues with proof that a hundred basis point
decline in GDP growth rate is associated with 1.5 times
more debt rescheduling in the fifteen years that follow.
In recent times, Argentina in the global economic system
has flirted with the concept of GDP-linked borrowings.
Although the conventional GDP-linked bonds differ from
the conceptual Sukuk proposed in this paper, the mechanics
of economics can be argued for a benchmark and as an
example. There appears to be opposing views on how much
the Argentinian government benefited from GDP-linked
borrowing, but economic numbers show a stable growth
and recovery for the economy.
Secondly, GDP-linked Sukuk, as an alternative to the
conventional debt, reduce the need of procyclical policies,
which have stunted the growth of most Muslim countries
post crises. It acts as automatic stabilizers in the case of
an economic slowdown and in boom periods. Considering
the example of Pakistan, post 2008stock market crash,
and civil unrest owing to terrorism, the economy slowed
down. Having immense amounts of external debt, which
had a fixed interest payment due, the country had to revert
to an IMF bailout plan, thus incurring another 10 billion
dollar loan to meet its prior commitments. The increase in
debt and the bailout plan came with specific conditions of
fiscal tightening and tax reforms. These policies of fiscal
Eds. Hatem A. El-Karanshawy et al.
tightening and tax reforms and removal of subsidies provide
a further negative impact on growth. A fiscal tightening
by the government would reduce the expenditure in the
country influencing the industrial sector’s growth.
Similarly, an increase in tax and removal of subsidies would
reduce the disposable income of the people, thus reducing
the consumption in the country. Both of these results
end in amplifying the economic downturn. Calvo (2003)
highlights some of these issues in his literature on sudden
stops, where he argues that sovereign debt fails to serve as
a device for inter-temporally smoothing the impact of such
slowdowns.
Key benefits of GDP-linked Sukuk
In this section we try to highlight the key benefits GDPlinked Sukuk offer the issuing country from an economic
perspective:
•In case the economy of the issuing country
experiences a persistent low growth phase, the GDPlinked papers provide a cushion against mounting
debt as in conventional borrowing. Since the payment
due on these papers would vary with the growth
rate, the Debt/ GDP ratio would only increase via a
deficit borrowing of the country. This would lead to
a less probability of default, and provide breathing
room for policy makers to expand fiscally or through
tax cuts to boost the economy.
•Usually emerging countries and HIPC countries
(where most Muslim economies fall) face extreme
difficulty for borrowing in times of crises due to weak
economic numbers. With GDP-linked papers, since
their liabilities would reduce, that would provide room
for counter cyclical policies to boost the economy.
• As Barro (1995) highlights, a mechanism of linking
GDP with the liability of the country can provide ease
of maintaining smoother tax rates and provide basic
infrastructure to the population in times of crises
without increasing tax rates.
• It has been observed, in the case of many Islamic
countries, as well as emerging economies, the
tendency to go into frivolous expenditure in the times
of high economic growth, which may lead to overheat
of the economy followed by a recessionary phase to
bring the economy back to equilibrium level. Amongst
the key Islamic countries, the case of Indonesia and
Malaysia, in the early 1990s is an example. The
economic crash of 1997 affected the region bringing
the growth number to negative in one year after a
decade of double-digit growth. If the government is
using GDP-linked paper, in times of high growth the
payment on Sukuk will be higher as well, which will
keep a check on excessive fiscal expansion.
Any instrument is not viable for financial markets until it
provides benefits to both sides involved in the transaction.
Although in this paper we aim to provide a concept of a
possibility of using GDP-linked papers in Muslim economies,
why it may work is provided in the following passage by
listing the potential benefits to the buyer of this instrument.
• Diversification of investments: The economic cycles
of different countries are far from being correlated
perfectly. Although there exists regional correlation
175
Rizvi and Arshad
to some extent, but global perfectly correlated
economic cycles is a utopian scenario. The investors
in these papers can diversify their portfolio while
keeping their investments in the real sector of the
economy instead of financial papers only.
• An application of GDP-linked papers will provide a
smoothing of the economic growth of the Muslim
and emerging economies, reducing the probability
and frequency of defaults. From an investor
perspective, a default is a high cost of litigation and
restructuring.
• While in developed countries, the stock markets
provide an investment opportunity into the future
prospects of the real sector, in Muslim economies, the
stock markets are still at nascent stage. Issuance of GDPlinked Sukuk will provide an opportunity to investors
to invest in the potential of the real economy.
Insight into structure and Fiqhi issues
The proposal for a GDP-linked Sukuk has been raised in
the corridors of Islamic finance previously in a paper by
Diaw et al. (2011), where they have proposed an Ijarah
based GDP-linked Sukuk for public sector funding and
debt management. The proposal of this Sukuk takes into
account this earlier study and is an attempt to further the
idea. Whereas the previous study proposes an ijarah-linked
paper, we propose a pure musharakah linked paper.
The idea of authors is to issue a musharakah paper by
the government as the wakil of the country, where the
underlying business is the economy as a whole. Keeping in
mind the musharakah nature, where profit is paid on actual
profit is the business, which in this case is the growth of
the economy. This idea is subject to Shariah debate and
approval, but in the understanding of the authors if we
consider the economy as one business, there should not be
a problem in sharing of the risk of the business (economy).
A key insight of this model is that in case the economy goes
to zero growth or negative growth in any year, the investors
would share in the losses, which would mean a reduction in
their principle amount invested in the business (economy).
This sharing in risk of the economy via a musharakah paper
is different from indexation of the debt, since a debt is not
created when a musharakah Sukuk is shared, but is similar
to issuance of equity shares in the business (economy) of
the country.
The issuance of indexation is a much-debated topic under
the Shariah law, but is primarily focused on “debt” and not
much has been discussed and debated on the issue of Equity
shares linked to GDP. Bacha and Mirakhor (2012), provide
a strong argument in favor of development of similar GDPlinked securities, linking it to welfare of Ummah through
the concept of macro markets.
4. Empirical analysis
Empirical evidence for some Muslim economies
The empirical section shows some primary basic economic
numbers to analyze what benefit GDP-linked Sukuk may
provide to the Muslim economies. To make our analysis,
we take into account four countries, Malaysia, Indonesia,
Iran and Turkey. These four countries throughout the
176
Islamic world comprise nearly 60% of the GDP of Islamic
ex GCC countries. These four countries have experienced
economic crashes and have been under external debt for
the past few decades.
For a simple review of the benefits of a GDP-linked
sovereign can be illustrated by taking the example of our
sample countries. Suppose a simple musharakah based
GDP-linked Sukuk was used for all financing of these four
countries from 1985 onwards, including conversion of the
existing debt to this instrument. The country would have to
pay a profit rate of whatever the growth of the underlying
business is – in this case the economy of the country. In any
year where the business (economy) does not grow, there
would be a zero payment of profit.
For the purpose of this example, we make the strong
assumption that the composition of the debt and the changes
in the debt levels does not have any impact on the behavior of
any of the other variables in the economy during 1985–2010.
What repayments the countries would have had to make
and the interest cost savings is illustrated in the Figure 1.
The graphs show, what was the actual interest payment and
debt forgiveness in each year as a representative of the cost
the country actually had to bear due to conventional mode
of financing. In addition, the payments for GDP-linked
Sukuk are shown for the same period and the savings as a
percentage of the national income each year.
Looking at the Malaysian case, the Malaysian economy
after a short decline in the mid-1980s bounced back and
grew at a robust speed for the second half of the eighties
and first part of nineties. This was followed by the infamous
Asian financial crisis of 1997, which brought the ASEAN
economies down to their knees in a matter of a year. With a
couple of years of negative economic growth, the economy
was propped back up with fiscal support policies and
capital controls. Indonesia experienced a similar crisis but
decided to take the multilateral agency bailout plan, and
both economies have experienced a revival in economies
in the last decade.
With liberalization policies in Indonesia, a massive influx
of foreign investment supported by steady infrastructure
development has resulted in high economic growth figures.
Within these two economies, in our example for using
GDP-linked Sukuks, the graphs show the crisis would have
resulted in zero payments during the crisis years, providing
room to the government for fiscal expenditure, to propel the
economy further. Governments using GDP-linked Sukuk
would have obtained a large reduction in the interest bill,
leaving more room to avoid procyclical fiscal measures.
Large interest savings would also have applied during the
sudden slowdown of 2008–2009. In the case of Malaysia,
the average profit payment over the 25-year period amounts
to 5.99% as compared to 6.54% average interest payments
that were paid by the Malaysian government. In addition
to the lower average payment, over the period, Malaysia
would have saved nearly 14% of its national income. In the
case of Indonesia the difference between profit payment on
GDP-linked Sukuk and interest payment is a mere 30 basis
points annually, but the savings as a percentage of national
income would have been 5%.
On the other hand, Turkey has had a unique economic
growth case, where the real GDP has experienced high
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Indexing government debt to GDP: A risk sharing mechanism for government financing in Muslim countries
Malaysia 1985-2010
12%
10%
Savings in National Income
GDP Linked
Interest Paid
12%
10%
8%
6%
6%
4%
4%
2%
2%
0%
0%
–2%
–2%
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
8%
12%
Turkey 1985-2010
10%
Savings in National Income
GDP Linked
Interest Paid
12%
10%
8%
6%
6%
4%
4%
2%
2%
0%
0%
–2%
–2%
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
8%
12%
Indonesia 1985-2010
10%
Savings in National Income
GDP Linked
Interest Paid
12%
10%
8%
6%
6%
4%
4%
2%
2%
0%
0%
–2%
–2%
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
8%
14%
Iran 1985-2010
12%
Savings in National Income
GDP Linked
Interest Paid
14%
12%
10%
8%
8%
6%
6%
4%
4%
2%
2%
0%
0%
–2%
–2%
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
10%
Figure 1. Interest savings over the economic cycle.
Eds. Hatem A. El-Karanshawy et al.
177
Rizvi and Arshad
volatility as compared to other countries. With its strategic
location as the bridging country between Europe and Asia,
Turkey has experienced internal economic crises during
our sample period.
The 1980–88 Iraq-Iran war benefitted the Turkish economy.
At that time, the Turkish economy increased trade with
both countries and became the supply route for Iraqi oil
exports. The end of war in 1988saw a sudden slowdown
in the economy, which stabilized with the economic
policy of import substitution of the then government. The
economy again received a major blow during the Persian
Gulf War, with a nearly $3 billion loss due to reduced trade.
Saudi Arabia, Kuwait, and the United Arab Emirates (UAE)
moved to compensate Turkey for these losses, and by 1992,
the economy again began to grow rapidly, before it was
plunged again into crisis in 1994, owing to government
borrowings. Post crisis, the economy stabilized but
experienced sharp downturns due to military assault in
Iraq, the economic slowdown of Europe and the financial
crisis of 2007. With this volatile nature of the economy,
GDP-linked Sukuk, would have provided savings as a
percentage of national income to the tune of 24% during
this period, while the average payment per annum for GDPlinked paper is 4.77% compared to 6.83% average annual
interest paid by Turkey.
The fourth country in the sample, Iran, with its decade-long
war with Iraq in the 1980s and subsequent international
sanctions owing to its nuclear research has experienced
an economic situation marked by dogged growth, where
it has been impacted by sanctions and restrictions on its
oil export. Iran has the largest Islamic financial assets in
the world, and using the Musharakah based GDP-linked
Sukuk would have benefitted in aggregate savings of 6%
over national income while the average profit payments on
GDP-linked papers would have been 4.07% as compared to
actual interest payments of 5.30% annually that Iran paid
to external lenders.
Debt/GDP stabilization effect
To illustrate the benefits of the GDP-linked Sukuk, in
stabilizing the Debt/GDP ratio of a country we perform a
simple exercise. For our sample countries using the wellestablished identity in fiscal economics:
D 
Dt
= (1 + r - g t ) t-1  - St
Yt
 Yt-1 
where Dt is government debt, Yt is output, st is the primary
surplus as a share of GDP, gt is the growth rate, and r is the
interest/profit rate on the debt.
Using actual data for the time period under consideration,
holding all other values of growth as constant (a very
strong assumption), we simulate the new Debt/GDP levels
for these four economies. Table 1 reports the standard
deviation as a volatility measure of the Debt/GDP ratio
during that period. It is evident that using the GDP-linked
Sukuk would reduce the volatility in the Debt/GDP ratio
for all the countries minus turkey, where it stays similar.
Although the difference in the standard deviation doesn’t
seem huge, but if it is translated in absolute dollar terms, it
goes in excess of a US$1 billion for these countries.
178
Table 1. Standard deviation of Debt/GDP.
Without
GDP-linked
papers
With
GDP-linked
papers
Iran
Indonesia
Malaysia Turkey
0.087
0.267
0.122
0.061
0.084
0.256
0.110
0.061
Tables 2 and 3 represent the maximum and minimum value
of the Debt/GDP ratio for a comparative analysis of actual and
simulated using GDP-linked Sukuk. For our sample countries,
we observe that during the 25 year period under study, the
maximum levels of Debt/GDP would have been substantially
lower for Indonesia and Malaysia by nearly 300 basis and 400
basis respectively. In the case of Turkey and Iran, the maximum
level of Debt/GDP would have been considerably lower as
well. Same trend holds for minimum level of Debt/GDP ratio,
where the minimum levels would have been much less than
the actual Debt/GDP ratios these countries experienced.
Table 2. Maximum value of Debt/GDP.
Iran
Without
GDP-linked
papers
With
GDP-linked
papers
Indonesia Malaysia Turkey
39.11%
158.69%
77.48% 57.62%
38.63%
155.50%
73.43% 55.78%
Table 3. Minimum value of Debt/GDP.
Without
GDP-linked
papers
With
GDP-linked
papers
Iran
Indonesia
Malaysia
Turkey
2.93%
27.57%
29.30%
32.80%
2.67%
28.34%
29.04%
31.35%
The earlier results highlight the potential benefits of having
GDP-linked Sukuk as a replacement of conventional debt.
Nevertheless, the illustrations in earlier tables have been
provided under strong assumption of keeping all factors
constant. In reality, the potential benefits may be large
owing to interacting nature of economic variables. We have
assumed that the primary surplus remained the same as
actual every year in our sample period, but it is important
to note that while using this Sukuk fiscal policy economic
growth and in savings/increments in profit payment may
respond. This would change the primary surplus and thus
may amplify the benefits achieved by these instruments in
stabilizing the Debt/GDP ratio.
Primarily in years of sluggish or negative growth with
debt sustainability as a major concern, governments
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Indexing government debt to GDP: A risk sharing mechanism for government financing in Muslim countries
Table 5. Correlation of GDP growth rates amongst selected few Islamic countries.
Bangladesh
Egypt
Indonesia
Iran
Malaysia
Nigeria
Tunisia
Turkey
Bangladesh
Egypt
Indonesia
Iran
Malaysia
Nigeria
Tunisia
Turkey
1.000
0.369
1.000
(0.125)
0.003
1.000
0.321
0.043
0.101
1.000
(0.127)
0.013
0.778
0.182
1.000
0.070
0.214
0.122
0.304
(0.009)
1.000
0.431
0.336
(0.096)
0.559
0.027
(0.020)
1.000
0.225
0.001
0.177
0.063
0.296
0.074
0.142
1.000
tend to move towards an increase in the primary surplus,
effectively implying that fiscal policy is procyclical. In our
sample countries, this is evident when we analyze the
primary surplus during economic downturns. In the case
of Malaysia and Indonesia it was evident during the Asian
financial crisis when Malaysia with an average deficit of
6% in 1997–98 moved to a 6% surplus in the succeeding
2 years. This is a common occurrence in Islamic countries,
which are primarily liquidity constrained and have to
impose these policies to maintain their credibility in the
international credit market.
5. Viability for investors
By illustrating some of the ways GDP-linked Sukuk
can benefit the Islamic countries, within our sample
countries, the authors do realize that implementation
of these instruments require a much more thorough and
rigorous empirical work. In addition to the fine-tuning of
these instruments, a major obstacle that can be foreseen
is the creation of markets for new instruments globally.
Past experiences with the introduction of innovative
instruments in world credit markets has been met with
skepticism and resistance.
Since GDP-linked Sukuk based on pure musharakah
structure is a novel idea in the Islamic financial landscape,
not much debate has happened on potential obstacles.
Nevertheless, owing to the global economic environment,
these instruments would have to be introduced in the
world financial markets, which are dominated by the
conventional investors. Due to this reason, we borrow
some of the concerns of global investors from literature
of conventional finance that seem to hold valid for these
instruments as well.
A serious question is based on whether international
investors would be willing to directly expose their
investments to volatility of GDP in Muslim and emerging
countries. In the understanding of the authors, international
investors are already exposed to the economic growth
of the country implicitly via the investments in the stock
markets and the standard debt contracts. If a country goes
into a recessionary phase, and debt sustainability levels
are breached, the borrower country is likely to default or
go into a restructuring program, which is a heavy legal
and financial cost. The GDP-linked Sukuk tends to provide
a buffer against the possibility of default and its costs by
providing the investors a steady return that is bound to pay
a higher amount in the long term as economies pick up.
Eds. Hatem A. El-Karanshawy et al.
Secondly, the issue of risk exposure and diversification while
using GDP-linked Sukuk is an implementation issue. Studying
the economic structure of different regions and considering
that Muslim economies are spread far and wide across the
globe, the economic cycles are not perfectly synchronized
and have low or negative correlations. If a mass issuance of
the GDP-linked Sukuk is done by Muslim countries, from
an investor perspective a well-diversified portfolio can
be constructed to hedge against over exposure to a single
business cycle. Table 4, provides the correlation matrix for
GDP growth rate amongst a selected few Islamic countries.
Our sample countries are highlighted in grey. It can be
observed from the specific countries that most of them
have a varying correlation in growth with each other.
Malaysia and Indonesia, being in the same region and
having similar economic fundamentals, tend to have
a high correlation while these two countries with Iran
have a very low correlation. Looking at the selected few
countries, we observe that there are traces of negative
correlation between Islamic countries as well, for instance,
in the case of Bangladesh with Malaysia/Indonesia, and
Nigeria with Malaysia. This bodes well from an investor
perspective for diversification and builds a good case for
these instruments.
6. Conclusion
This study focuses on the predicament of several Muslim
countries that are debt-ridden and struggling to finance
their debt. Most commonly, these countries turn to western
financial intermediaries offering high interest rates to meet
their debt requirements. Unfortunately, the reliance of
Muslim nations on high-interest rate borrowing has largely
been neglected in Islamic finance, with little to no research
available on alternatives.
The development of financial instruments for sovereign
financing has only just begun whereby a few seminal
studies started to show how Shariah compliant instruments
can help alleviate Muslim nations from their current
predicament.
Owing to fact that 51 out 57 OIC member countries are
debt-ridden, the authors propose a solution to this problem
found within the ambit of Islamic finance. Stemming from
conventional finance, the authors recommend the use of
GDP-linked papers using the principles of Musharakah.
The paper highlights the key benefits of using GDP-linked
Sukuk arguing that such an instrument will allow for a
179
Rizvi and Arshad
lower probability of default, which permits counter cyclical
policies to boost the economy. Similarly, it leads to an ease
of smoother tax rates and curbs excessive spending during
times of economic growth.
Barro RJ. (1995) Optimal Debt Management. Working
Paper. NBER, 5327.
Covering Malaysia, Indonesia, Iran and Turkey, the
authors attempt to further emphasis the benefits of GDPlinked Sukuks. In the case of Malaysia, using GDP-linked
sovereign paper would have provided an average profit
payment of 5.99% as compared to a 6.54% interest payment
that was paid by the Malaysian government over the 25year period. This would provide more room for Malaysia
to avoid procyclical fiscal measures. In addition, Malaysia
would have saved nearly 14% of its national income.
Calvo GA, Reinhart CM. (1999) When Capital Inflows Come
to a Sudden Stop: Consequences and Policy Options.
Center of International Economics, Department of
Economics. University of Maryland.
Indonesia, on the other hand, only shows a difference of 30
basis points between the profit payments and interest rate
payments on GDP-linked sovereign paper and borrowings.
Nonetheless, the savings on its national income would
have amounted to 5%. The case of Turkey showed that due
to the volatile nature of the economy, GDP-linked Sukuk
would have provided a savings of around 24% as a fraction
of national income. The average payment per annum for
GDP-linked paper is 4.77% compared to 6.83% average
annual interest paid by Turkey. Lastly, Iran’s saving would
have been 6% over national income while profit payments
would have been only 4.07% compared to the 5.30% that
was paid out to lenders.
The authors further highlight that using GDP-linked Sukuk
would reduce the volatility in the Debt/GDP ratio for
the countries. The maximum value of Debt/GDP ratio was
considerably lower for all four countries, with Malaysia and
Indonesia recording a reduction of nearly 400 bases and
300 bases respectively. Parallel trends apply for minimum
levels of Debt/GDP ratio, where the minimum levels would
have been much less than the actual Debt/GDP ratios that
these countries experienced.
While this study lays out an introductory emphasis
on the benefits of a GDP-linked Sukuk, it is not without
limitations. The implementation of this study has been
restricted to only four countries; further research calls for a
much larger sample study. Furthermore, this study makes
an assumption that all interactions of economic variables
remain the same, while in reality a more detailed analysis
would be better to see how using these instruments will
change the behaviour of growth and other economic
variables.
This preliminary study is an emphasis on the benefits of
Islamic finance as an alternative for government financing.
The empirical evidence provided by this study is promising
and provides an avenue for further research. The plight
of Muslim nations with respect to their debt structure has
largely been ignored in Islamic finance. It is the opinion of
the authors that such instruments are vital for the support
and development of Muslim nations and should be further
researched.
References
Bailey N. (1983) A Safety Net for Foreign Lending. Business
Week. January 10.
180
Caballero RJ. (2002) Coping with Chile’s External
Vulnerability: A Financial Problem. MIT.
Calvo GA, Izquierdo A, Talv E. (2003) Sudden Stops,
The Real Exchange Rate, and Fiscal Sustainability:
Argentina’s Lessons. NBER Working Paper.
Detragiache E, Spilimbergo A. (2001) Crises and Liquidity—
Evidence and Interpretation. IMF Working Papers 01/2.
International Monetary Fund.
Diaw A, Bacha OI, Lahsasna A. (2011) Public Sector Funding
and Debt Management: A Case for GDP-Linked Sukuk.
Paper presented at the 8th International Conference
on Islamic Economics and Finance, 19–21 December,
Doha, Qatar.
Diaw A, Bacha OI, Lahsasna A. (2012) Incentive Compatible
Sukuk Musharakah for Private Sector Funding. ISRA
International Journal of Islamic Finance. 4(1).
Dreze JH. (2002) Loss Reduction and Implicit Deductibles
in Medical Insurance. CORE Discussion Papers 2002005.
Université catholique de Louvain, Center for Operations
Research and Econometrics (CORE).
Durdu CB, Mendoza EG. (2006) Are Asset Price Guarantees
Useful for Explaining Sudden Stops? The Globalization
Hazard-Moral Hazard Trade of Asset Price Guarantees.
IMF Working Paper. WP/06/73.
Easterly W. (2001) The Effect of International Monetary
Fund and World Bank Programs on Poverty. Policy
Research Working Paper Series, 2517. The World Bank.
Froot K, Keen MJ, Stein, J. (1989) LDC Debt: Forgiveness,
Indexation, and Investment Incentives. Journal of
Finance. 44(5):1335–1350.
Haldane A, Quah D. (1999) UK Phillips Curves and
Monetary Policy. Journal of Monetary Economics.
44(2):259–278.
Hasan MA, Siddiqui AN. (1992) Is Equity Financed Budget
Deficit Stable in an Interest Free Economy? Journal of
Islamic Economic Studies. 1(2).
Krugman P. (1988) Financing vs. Forgiving a Debt Overhang.
Journal of Development Economics. 29:253–68.
Obstfeld M, Peri G. (1998) Regional Nonadjustment and
Fiscal Policy: Lessons for EMU. Center for International
and Development Economics Research (CIDER) Working
Papers C98–096. University of California at Berkeley.
Shiller RJ. (1993) Aggregate Income Risks and Hedging
Mechanisms, NBER Working Papers 4396. National
Bureau of Economic Research, Inc.
Zamir I, Abbas M. (2012) An Introduction to Islamic Finance:
Theory and Practice. 2nd Edition. Wiley Publications.
Zamir I, Abbas M, Hossein A, Noureddine K. (2011) Risk
Sharing in Finance: The Islamic Finance Alternative.
Wiley Publications.
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Concept and mathematics of Islamic valuation
and financial engineering
Nadi Serhan Aydin, FRM1, Martin Rainer2
Organization of Islamic Cooperation (OIC), Ankara Centre – SESRIC, Institute of Applied Mathematics,
Middle East Technical University, Ankara, Turkey, Email: [email protected]
2
ENAMEC Institute, Institute of Applied Mathematics, Middle East Technical University, Ankara, Turkey,
Email: [email protected]
1
Abstract - Starting from the fundamental principles, whether a designed contract implies a clean
bay’ rather than ribā is central to our discussion. We argue that ribā and gharar may easily arise
through neglect of risk or inappropriate valuation methods for value and risk of assets and financial
instruments. This possibility, in turn, strongly necessitates an estimation of expected forward
values, market risk, and default risk, which is consistent with Islamic principles of avoiding ribā
and gharar. Based on consistent estimatation of risk and return, the expected costs of risks should
be quantified for all parties to the contract in order to judge, whether a contract is free from usury
(ribā) and evitable risk (gharar) and whether it the remaining inevitable risks are distributed fairly
between the counter-parties, for example between the investor (rabbu l-māl) and the entrepreneur
(mu-ārib) within a mu-āraba contract. Therefore, an unbiased quantitative estimation of the risks
and returns is desirable so as to put Islamic principles to work.
We argue that Islamic principles, in particular the avoidance of riba¯ and gharar, should be
applied to real economic value in the first place, and not a priori to a monetary value in terms of
conventional currency. In order to reconcile monetary value with economic value, we propose a
reference currency linked to an appropriate commodity basket, reflecting the common economic
realities and needs of the respective monetary union. Based on this currency, real economic value
can be computed in analogy with conventional financial engineering methods.
In order to reflect global economic needs and realities, a global reference currency should be linked
to a basket of commodities including in particular the natural resources necessary to ensure both,
sustainable survival of mankind and a sustainable living standard above poverty.
Referring to the recent financial crisis of the European Union, we argue that apart from the common
economic realities and needs within a given socio-political union, such as the OIC countries, also the
different realities and needs should be honoured appropriately. We propose a 3-level construction
of reference currencies, reflecting the economic realities and needs globally – for each region, and
for each country.
We compare conventional financial engineering, based on zero bond numéraires computed from
fixed income forward contracts, with Islamic financial engineering based on numéraires computed
from bay’u l-salam or/and forward contracts on the basis of the reference currencies relevant for
the counter-parties of the contract.
We propose that contract valuation and risk management should be performed on the basis of
Islamic financial engineering rooted on the reference currencies reflecting the economic realities
and needs relevant for the counter parties. Considering the benefits of such a risk management
for social economy, particularly when the Organization of Islamic Cooperation (OIC) countries
are considered, we argue that an implementation of the described Islamic financial engineering
Cite this chapter as: Aydin N S, FRM, Rainer M (2015). Concept and mathematics of Islamic valuation and financial
engineering. In H A El-Karanshawy et al. (Eds.), Islamic banking and finance – Essays on corporate finance, efficiency
and product development. Doha, Qatar: Bloomsbury Qatar Foundation
Developing Inclusive and Sustainable Economic and Financial Systems
Aydin et al.
enables Islamic risk management making transparent expected return and risks and enables their
fair distribution between the counter-parties.
Islamic financial industry following the guideline of such principles of Islamic financial eng­
ineering would be able to contribute to sustainable development by (i) more risk-(and-return)consciousness reflected in the participatory structure of financial contracts, and thus, (ii)
encouraging Islamic financial institutions to innovate financial products consistent with real
implementation Islamic principles, reflecting the real necessities of modern business and economy.
1. Introduction
Recent decades have witnessed an explosion in financial
innovation and engineering of novel contracts. First of all,
from the so-called “conventional banking system” emerged
increasing the need to adapt to the challenges from the
demand side. Secondly, particularly also within the socalled “Islamic finance industry,” the incentive to innovate
around prohibited and disadvantaged transactions has
been unfortunately high. A clash between abused financial
engineering and Islamic principles seems to prevail, not
only since the renaissance of Islamic finance, and not in
the Muslim world only. In medieval Europe canonical law
prohibited usury, i.e., ribā. European merchants, however,
used a combination of contracts called contractum trinius
to circumvent prohibition of usury. The construction
uses put-call parity to synthesize a conventional interestbearing loan. During the recent renaissance of Islamic
finance, this construction prevailed in practices of some socalled “Islamic” banks, when they include into murābaha
contracts a combination of conditions similar to contractum
trinius, thus effectively ensuring and concealing a risk-free
profit, which might also be called “regulatory arbitrage.”
Exceeding considerably the inflation rate, such a so-called
“risk-free return” is nothing else but usury, i.e., ribā. Such
a combination of contracts in order to circumvent Islamic
principles may be considered as product structuring that is
a part of conventional financial engineering.
Here and below, we will be concerned however not so much
with product structuring but instead with quantitative
valuation of financial instruments. Careful computation
of expected value, expected return, and expected risks is a
prerequisite not only for risk management, but also already
for thorough valuation of compliance of contacts with
respect to regulatory requirements including compliance
with Islamic law. A clean product structuring, therefore,
depends on a serious valuation of value and risk as the more
fundamental and more demanding section of financial
engineering, which we will deal with subsequently.
In Europe in particular, the growing public debt of several
European Union (EU) member countries has surfaced
already in several financial crises, which were escalated by
salvation programs for financial institutions. According to
IMF 2010, the public debt of Italy and Greece is significantly
exceeding annual GDP, while others (like Portugal, Ireland,
and Belgium) are following with over 90% of GDP. Continuing
conventional public loan policy, it is likely that European
public debt will exceed all possible salvation funds in the
long run. Comparing the 2008 financial crisis of Iceland
with the 2010 crisis of Greece, some important difference
182
appears: Iceland had its own currency, which helped to
soften the consequences, but Greece is currently trapped
within the EUR monetary union. Recently enterprises and
banks already simulate scenarios for the reintroduction of
the Drachma1 in Greece. The point that was missed, when
Greece entered 2001 for the second time2 into a monetary
union, is that a common currency can be sustainable only
within a political region that is operating sufficiently within
its economic realities and needs. If some important factor
such as economic productivity differs largely among the
participating countries, an artificially introduced monetary
union creates artificial fluxes and processes that may be
in contradiction to the economic reality of some country.
This recently became evident with Greece again. The idea
that a monetary union of a region like Europe is a priori
going to strengthen its political union has proven again to
fail, because in the case of Greece, the EUR did not reflect
Greek economic reality, with economic mismatch finally
resulting in political instability. We will come back to this
point below when we discuss the construction of reference
currencies reflecting real value of economy.
On the background of financial crises in Europe, what
can be learned from Islamic finance? On the Banque de
France conference, March 4, 2011, Kenneth Rogoff of
Harvard University commented: “Western policymakers
and economists often portray Islamic financial systems, with
their emphasis on shared risk and responsibility in lending,
as less efficient than western systems that put no strictures
on debt. Yet one can equally argue that Western financial
intermediation is far too skewed towards debt, and as a
consequence generate many unnecessary risks.”
In the Muslim world, conventional agreed interest loans are
commonly objected for the fact that they avoid a sharing
participation in the default risk related to the purpose of
the loan. For the same reason, transfer of risk by selling
it to another party is often objected. Within the Islamic
finance community, this sometimes appears to create a
far spread impression that a passive attitude towards risk
would be more ethical rather than an active management
of the risks involved within a project and related contracts.
We would like however to argue that both, transparency
about all risk involved, and clear agreements how they are
to be shared should be part of any contract. Also we believe
that for a single agent, entrepreneur or investor, it should
be legitimate to use his portfolio of contracts with different
counter-parties in order to minimize his own exposures
to different types of risks. In fact, avoidance of gharar is
a fundamental request according to Islamic discourse,
e.g., numerous hadiths forbidding gharar sales. Having
accepted this, it is becoming clear that, modern tools of risk
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Concept and mathematics of Islamic valuation and financial engineering
management and hedging should be applied appropriately
using all the knowledge we currently have in order to reduce
risk exposures. Quantitative evaluation of risks involved
therefore is a necessary precondition in order to enable
the counterparties to first obtain transparency about the
existing risks and, on the basis of this information, to reach
a fair agreement about the mutual distribution of risks. As
far as Islamic finance is concerned, we currently perceive
commonly loose interpretation of quantitative evaluation,
and last but not least, avoidance of mathematics as being too
sophisticated. However we’d like to remind that striving for
knowledge is any Muslim’s duty, rather than condemning
it plainly as misleading since the mathematics involved
appears too complicated for many non-experts.
One historically grown reason for the adverse attitude of
many Muslims towards financial mathematics and,
in particular, the fair value approach comes from the
conventionally common practices of discounting future cash
flows with so-called “risk-free” interest rates often derived
from forward rates of inter-bank markets and certain
government bonds. Indeed these rates are deceptive. First
because they are not really “risk-free” as their name suggests.
Secondly, their level is considerably higher than sustainable,
because they are indeed driven by and related to fixed interest
loans. Moreover, the high level of these rates also contributes
to push inflation rates up. Such objections against the common
discount curves, based on the aforementioned interest rates,
are fully legitimate. Below we will argue indeed for a more
flexible and adequate construction of discount curves. We
will also point out that the mathematical framework of
fair value—based on relative prices with respect to some
reference asset—does not require at all that the reference
asset should be given by an artificial zero-coupon bond
linked to “risk-free” discount rates, which, in turn, relate to
fixed interest loans and the conventional inter-bank markets
for forward rates. The mathematical framework of fair value
works perfectly also with a universal commodity-linked
currency or even equity as a reference asset.
2. Islamic finance for sustainable
development
The role of financial institutions is to provide the capital
for projects, in particular, for those projects that are useful
or even vital for development of societies with sustainable
infrastructures. Notwithstanding, many financial institutions
(e.g., in Europe) that had been shaken by the last financial
crisis, have been observed to refuse credits or offer credits
only with unbearably high risk premiums. This attitude
has severely threatened the existence of in particular local
small- and medium-size family businesses and enterprises,
since these suffer from additional discrimination by the
traditional rating systems favoring large international
players instead. As it is known, however, the economic
and innovative power of the society is driven very much by
locally rooted small and medium size enterprises, rather
than the big transnational players. This is known to hold
true in Europe, as well. And it is very likely to hold similarly
for the group of OIC member countries that have for long
been striving similarly for their sustainable development,
while facing similar challenges along the way.
As a consequence, it is the task of each government to take
care that the financial institutions follow up their duty of
Eds. Hatem A. El-Karanshawy et al.
providing the required capital for small and medium size
companies in order for them to remain operational and
continue making essential contributions for an innovative
and vital economy. Following the financial crisis of 2007,
new regulatory frameworks such as Basel III encourage
credit institutes to impose more severe conditions and
tougher rating conditions on entrepreneurs for credits. The
regulatory requirements now became the pretence of credit
institutes for harsher credit conditions, contradicting to
generally decreasing interest rates in Europe, particularly
in Germany. Government tried to intervene on this with
several measures to enable and to push financial institutions
to follow up their duty of providing liquidity for enterprises.
However, by the time being, Europe’s financial industry
reacts only very reluctantly. In this aspect, governments
of most OIC countries should be in a better position not
only because the influence of governments on the domestic
financial sector has traditionally been more powerful, but
much more also because the basis of understanding between
governments and financial industry is derived from the
common rules of Shariah. In this context, a very different
culture of entrepreneurship has shaped the Islamic finance
industry with the commonly accepted participatory means
of financing, fixed income products being obsolete.
In the following, we will sketch some essential mathematical
aspects of Islamic finance enabling financial engineering
similarly easy and rigorous as in the conventional interestedbased finance.
3. The myth of risk-free interest
and fixed income
In this section we will argue that the notion of a risk-free
fixed income is a myth kept up by the conventional banking
sector, rather than an economic reality. Corresponding rates
are set mainly by the agreement of an inter-bank market
among conventional creditors. To calibrate a current credit
contract to such rates may be questionable particularly
for the situations where one or more counterparties of
the contract have limited or no access to this inter-bank
market.
After we have seen that fixed interest rates by themselves are
not suitable as a basis for a risk-neutral measure, we conclude
that a new basis for such a measure must be sought. This
should be done by taking into account the position of the agent
in the market, i.e., by carefully investigating her exposure
to the various risks of the relevant markets and economies
involved. Furthermore, credit spreads should be considered
more flexible, i.e., not constant, and not necessarily always
positive. A negative credit spread for some period would
reflect the possibility that the credibility of the considered
enterprise is in fact better than that of the reference.
It is one main issue in Islamic finance that a supposed
risk-free interest rate should be close to zero. The reference
rates of some countries like Switzerland and Japan have
already come very close to this, since many years. Also in
Europe, after the financial crisis of late 2000s, interest rates
dropped drastically. This indicates that the economical
reality in interest-based financial markets might essentially
honour – sooner or later – the fact that there is no free lunch
in any market and that the assumption of a risk free return
is a myth.
183
Aydin et al.
4. Inflation of currencies, time-value
and time-less value
To the extent that inflation of any currency is inevitable,
a time-value that compensates the expected inflation rate
should be disputable. If we decide that full participation
in the inflation risk is not bearable to the investor, at least
a partial compensation of hedging against inflation risk
should be admitted. On the other hand, one might argue
that investor and entrepreneur perhaps share some common
risks, such as the risks of everyday life, and accordingly, the
inflation risk – as it is inevitable to both parties – should
also be shared among them rather than being put on one
party’s shoulders.
Real contracts usually involve several cashflows and/or
depend on asset values at different times. A fair contract
evaluation requires the ability to compare cashflows and/or
asset values at different times. The nominal value of assets
is usually measured in units of a certain currency. However,
the real economic value of this currency may changes with
time, e.g., due to inflation. For our fair valuation, we are
nevertheless interested in the intrinsic real value of the
currency, and of our assets. E.g., in the face of inflation, this
real value of assets might be measured by inflation-adjusted
prices. These are obtained by adjusting future cash flows by
discount factors and account just for the expected inflation
of the currency.
Viewed superficially, discount factors might be objected for
introducing undesirable time-value to cash flows. However,
if they are chosen just such to compensate for the timedependency of the nominal value of the currency, they in
fact may yield in fact just the desired timeless real value
standard. We would like to emphasize that the timelessness
of value as suggested by Islam has to be requested for the
real value rather than the nominal value in the first place.
Islam advocates in fact the use of time-independent measures
for value. At times of the prophet (pbuh) gold was much more
rare than today, since there was no mining industry yet. It
was a fairly good inflation-free currency. Its value was stable
and hardly to be influenced. Only exceptional political events
such as conquests and sieges could trigger sudden regional
gains and losses of huge gold treasures, which then could
indeed change locally the real value. Except such extreme
events, the real value of gold was stable. Today, however,
developments of the mining industry, demands from high
technology, and different market pressures apply almost
continuously on the value of gold in different directions.
Although gold might be still more stable than most paper
money, its real value has become much more volatile than
at times of the prophet (pbuh).
Hence the challenge is to find the reference assets, which
represent a timeless stable standard of value, similar to
gold in previous times.
5. The gold of 21st century: Back to
commodity-based currency
From the very early times of Islam, up to the 20th century,
the commodities of gold and silver have played an important
role in defining the modern currencies. The histories of
the GBP and USD in this respect and, in particular, their
184
successive decoupling from their reference commodities
simultaneously with their devaluation is described, e.g., in
El Diwany (2010).
Within the EU, after some period of fixed cross currency
rates, the Euro was introduced. Less known however is that
an early predecessor of the Euro was defined already in
the early 1930s. The universal European currency intended
as a “currency for peace” was called “l’Europa” (Le Fédériste
(1933)). It was defined as a basket of several valuable
commodities. Much later, Lietaer (2001), introduced a
global Trade Reference Currency (TRC), dubbed also as
“Terra,” comprising a basket of a dozen internationally
circulated currencies.
One advantage of the general concept of a basket of
commodities underlying to a reference currency is the
increased stability. With its currency linked to a basket
of commodities and its monetary authority backed 100
percent by a sufficiently large reserve of these commodities,
artificial depreciation (appreciation) of a country’s local
currency (domestic prices) would come to an end, together
with an improvement in its immunity to the associated
monetary and fiscal challenges facing its economy.
An inevitable effect of any commodity-based currency
is the increase of efforts for production of or mining for
the underlying commodities. Taking into account the
challenges of 21st century and beyond, previous choices
of baskets did not yet account sufficiently for the aspect
of sustainability and desirability of the production of the
commodities chosen for the basket.
According to Islam, at times of the Prophet Muhammad
(pbuh), exploration and production of gold was still
desirable as the most precious commodity known at that
time. It was not yet challenged by an excessive mining
industry with all its social and ecological problems. From our
current modern point of view, in our opinion, the negative
effects related to the intensive industrial production, such
as ecological damages and exploitation of workers, cannot
be ignored. In essence, what is true for gold and other
metals is also true for oil and gas as well as agricultural
commodities. For any commodity, consideration should
me made whether and to what extent its production is
desirable. So from an Islamic perspective, we would like to
argue in favor of a historically conscious reading that the
“gold” mentioned in the Holy Qur’an should be read just as
synonymous for the “most precious commodity.” If we ask
ourselves – from the perspective of the global challenges
ahead of us – what are these “most precious commodities”,
we might identify such commodities as cleanly produced
renewable energy, clean drinking water, and agricultural
products produced according to ecological standards.
6. Euro lesson: A multi-dimensional
currency system reflecting diversity
and community appropriately
The Euro currency was introduced well-intended, with
the idea to enhance the political and economical solidarity
between the countries involved and the ultimate goal to make
them together more competitive in the world markets arena.
However, the recent Euro crisis has shown that as long as
the economic foundations of the constituents are basically
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Concept and mathematics of Islamic valuation and financial engineering
too different, the hard entry of one or more countries into
a currency union may be a problem. If productivity of the
countries highly differs, the stronger countries have to aid
the weaker ones in order to avoid a potential collapse in the
common currency. As a result, the political disputes going
along with this process not only put the common currency
union at risk, but even worse may damage the political union,
as such concerns have been central to the Euro crisis.
For example, the EC currently is split about whether to
introduce Euro bonds or not. The introduction of Euro
bonds would be consequent since there exists already a
common currency. The opponents however argue in view
of the different economic realities within the member
countries. The dilemma is that, any currency has to reflect
the economic realities, but the EUR member countries now
discover that they will not be able to sufficiently converge
on their economic realities.
In our opinion, the choices for countries that belong to
a certain political entity (e.g., the EU) to either share a
common currency 100 percent or withdraw from the
monetary partnership, are insufficient in number in the face
of the current economic and political realities. The currently
practiced “0 or 1” hard entries into a currency system
apparently ignore the risk that a similar hard exit from it
may endanger the whole system. Therefore, we propose a
structured system of world currencies, starting from a global
reference currency Cw. It is defined by a basket of commodities
S1 … Sn whose productions are globally desirable, i.e.:
n
C w = ∑ w i Si (1)
i=1
Similarly, we may consider a basket Cr of commodities, the
production of which is desirable just within a certain region,
e.g., the political entity of the OIC member countries. Cr
acts as the currency of this region. ∆Cr: = Cr − Cw accounts
for structural differences of the region from the rest of the
world. It may also give additional (non-negative) weight to
some of the commodities in Cw, e.g., clean drinking water,
which is already within the global world basket. It may also
contain new commodities specific for the region. Hence:
n
ΔC r = ∑ w(i r )Si +
i=1
m
∑w
i=n+1
S (2)
(r )
i
i
Finally, we admit an extra basket for individual countries
in order to take into account their particular situation,
differing from that of their region as well as the world at
large:
m
ΔC c = ∑ w (i c )Si +
i=1
l
∑w
i=m+1
S (3)
(c)
i
i
The currency Cc of any country will be composed from a
worldwide defined fraction βw of the world currency, the
regionally-defined fraction βr of its regional commodity
basket, and the remaining local country fraction (1-βw-βr)
of its local commodity basket, i.e.:
C c = β w C w + β r ΔC r + (1 - β w - β r )ΔC c (4)
Eds. Hatem A. El-Karanshawy et al.
The fraction βw should be negotiated on a world conference,
common for all regions. It should be as high as possible
in order to meet global challenges. However, the case
where βw < 1 is desirable in order to admit the differences
according to the particular situation of different regions of
the world.
Similarly, the fraction βr should be negotiated on a regional
conference, common for all countries. It should be as high
as possible in order to meet common regional challenges.
However, βr < 1 is desirable in order to admit in this case
the differences according to the particular situation of the
different countries of the region.
By construction, the higher the value of βw, the more
strongly any two world currencies are correlated. The
same is true for the currencies of the countries within a
region: the higher the value of βw + βr, the stronger the
correlation between the currencies of any two countries
within a region will be. In particular, the negotiable weight
βr may give to a region, e.g., Europe, the flexibility to
regulate a partial entry into a currency union with some
weight βr, realistically reflecting the degree of economical
unification.
7. Fighting usury: The risk-neutral
fair value
The value of an asset can equivalently be quoted in the
form of a price or in the form of a rate, which is relative
to a reference price. Hence, a quoted rate per se does not
yet imply ribā, while, vice versa, quoted prices may conceal
ribā. The form of the quotation, whether it is given price St,
or rate, Rt, is irrelevant to the question of usury.
The equivalence relation is:
S(t)
= 1 + R(t)⋅ t (5)
S(0)
When St is unknown and subject to risk, it will be stochastic
and one has to consider its expected value EQ[St] under a
certain risk measure Q. The conventional present value (at
time 0) of the stochastic variable S is then given by:
S(0) =
1
E Q[ S(t)] (6)
1 + tR(t)
The spot rate R(t) may equivalently be converted into a
discount factor, corresponding to the initial value P(0, t) of
a zero-coupon bond with a nominal value of P(t, t) = 1 and
time-to-maturity t, i.e.:
P(0, t) =
1
(7)
1 + tR(t)
The relation (6) above may then be written as:
E [ S(t)]
 S(t) 
S(0)
= Q
= EQ 
 (8)
P(0, t)
P(t , t)
 P(t , t) 
185
Aydin et al.
This is simply the martingale expression for the relative
price of S with respect to the zero-coupon bond P(·,t),
whose price is inversely proportional to the spot rate R(t).
In order to determine whether a certain quoted price or
rate implies usury, it is rather important to have at hand an
independent method to determine a fair price with a riskneutral position. The expected price should be computed
under a measure Q, which in particular compensates
any risk that is out of the responsibility of the considered
counterparty. The computed value is supposed to match as
close as possible the realistic economic expectation.
Conventional finance usually assumes depreciation of
cash flows, e.g., according to S(t), in a certain currency
during time. It further assumes that the appropriate way
to account for this is to discount with a particular spot rate
curve R(t) derived from forward rates Fi (t) = F(t; Ti-1 , Ti )
corresponding to quoted interest rates from certain interbank markets (e.g., LIBOR).
From an Islamic finance perspective, the questionable
point here is the market from which the quotes are taken.
The method of calibration of the expected spot value to the
forward curve by itself agrees with the principle of fair value.
Even within the still very conventional setting of R(t)
and P(·,t) and relations (6) and (8), there is already the
flexibility to adjust the rate R(t) (and equivalently the
zero-coupon bond) to the any expected time-dependent
performance. It is not at all required that it is linked to the
conventional inter-bank markets for deposits or swaps.
If it is linked to performance of equity, commodity, or (as
we will consider below) to currency, it may be consistent
with principles of Islamic finance. Also compensation of
inflation of a currency may be considered as legitimate
under circumstances that the counterparty cannot bear to
be exposed to inflation risk, e.g. because it is too small and
does not have the possibility to protect itself. In this case
R(t) could be linked to inflation rate.
Concerning relation (2) above, remarkably, in mathematical
finance the meaningful quantities are mostly relative.
In particular, price dynamics is investigated mainly with using
the relative price Sˆ := S/S0 with respect to the price S0 of a
reference asset. Several desirable properties for the reference
asset are usually postulated. Best stability of its value is one of
them. The more the price is stable, the lower its volatility, i.e.,
the less risk of change is connected to the asset price. In the
past, the commodity gold was considered as the most stable
physical asset with the least volatility of its price.
Note, however, that considering “risk-free” zero bonds or
rates is unrealistic.3 Therefore, we propose, instead of the
zero-bond linked to a deterministic rate R(t) to consider
another reference asset instead, reflecting the stochastic
risk. We propose the commodity-linked currency (4) as a
natural reference asset for this purpose. With this choice,
relation (8) will be replaced by:
186
 S(t) 
S(0)
= EQ 
 (9)
C country (0)
 C country (t) 
Note that this relation holds, independent of the auxiliary
currency units within which S and Ccountry are evaluated.
This invariance is an advantage of relative prices. In general,
the relative price in (8) will be stochastic. However, the
relative price of Ccountry to itself becomes trivially constant
equal to 1.
8. Conventional versus Islamic valuation
Let us give first a simple description of a forward contract
and its conventional valuation. A forward contract agrees to
a fixed price: the fair forward price F to be paid at delivery
time T in exchange for an asset S (say a commodity) having
present value S(0) and unknown value S(T) at delivery.
A fair value forward contract on some traded asset S should
always have value 0 at time 0, the time of contraction.
Using relative prices w.r.t. zero bonds having price P(t, T),
conventional financial engineering demands:
0=
S(0)
F
S(0)
, or equivalently F =
. (10)
P(0, T )
P(0, T ) P(T , T )
Hence P(0, T) acts as a discount factor. Note that (10) implies
a nominal value concept with respect to some conventional
currency but tries to compensate this by taking zero bonds
as numéraire. Above, (10) is postulated directly, with
deterministic P(0, T). Note that, for independent unknown
stochastic prices S(t) of the asset and B(t, T) of the zero
bond, it holds
 S(t) 
E P[ S(t)]
E [ S(t)]
EP 
= P
, (11)
=
 B(t , T )  E P[ B(t , T )] P(t , T )
with deterministic prices P(t, T) = EP[B(t, T)], from
which (10) follows immediately, assuming F = EP[S(T)].
The independency assumption may hold for equity or
commodity assets, when it appears plausible that such
assets develop essentially independent from inter bank
markets determining B(t, T). When asset prices are nontrivially correlated with the stochastic discount factors, the
above derivation of (10) has to be replaced by:
 S(T ) 
 S(t) 
S(0)
, (12)
F = E P[ S(T )] = E P 
 = EP 
=
 B(T , T ) 
 B(t , T )  P(0, T )
whence again the fair price of the future contract is simply
S(0)/P(0, T).
Let us now consider the valuation of a forward contract
within an Islamic financial engineering framework relative
to a reference currency C. According to (9) above, we
replace (12) by:
 S(T ) 
 S(t)  S(0)
EQ 
 = E Q  C(t)  = C(0) . (13)
C
(
T
)




Unlike (12), where B(T, T) = 1 by definition of the zero
bond, here C(T) is in general still unknown, stochastic.
Furthermore the stochastic forward price S(t) and the
stochastic currency C(t) are unlikely to become either
independent or fully dependent at maturity. Then only
Islamic banking and finance – Essays on corporate finance, efficiency and product development
Concept and mathematics of Islamic valuation and financial engineering
value that can be estimated is their relative value. The
fair relative value w.r.t. to the given currency is constant.
For the future contract hence the fair price should be
specified as relative to the reference currency, simply as
S(0)/C(0). Remarkably, the fair price is quite analogous to
the conventional future price, which also can be viewed as
the relative price w.r.t. a zero bond.
Important from the perspective of Islamic principles is that,
in the conventional case the zero bond numeraire is a pure
interest based asset, which is linked essentially to inter
bank markets, while in the Islamic case the numeraire is
a currency, which we propose to be linked to a commodity
basket supposed to reflect a more appropriate and relevant
economic reality.
Now let us consider the special case of full dependence,
S(t) = a · C(t), i.e., the traded asset is proportional to the
reference currency (with constant factor a), and hence
proportional to the commodity basket defining C. Then
(13) yields:
E [S(T)]
F
 S(T )  S(0)
EQ 
=a= Q
=
, (14)
=
E Q[C(T)] E Q[C(T)]
 C(T )  C(0)
and therefore F = a · EP[C(T)].
Comparing the conventional with the Islamic approach, a
practically important difference is that the zero bond prices
are computed conventionally purely deterministically from
available quotes of money market, future, and swap prices;
while for the reference currency and the commodity basket
their forward prices are currently not easily available from
quotes. Their estimation is however possible via calibrated
stochastic processes. When the asset S(T) is non-trivially
correlated with the reference C(T), then stochastic simulat­
ion of their relative value may be necessary in order to
determine the expected value in (13). This shows the need of
an analogous stochastic forward rate model for commodities
in analogy to the conventional LIBOR market model.
We finally emphasize that, the numeraires in (12) and (14)
are not equivalent numeraires, i.e., they are not resulting
from transformation of each other by the change of
numeraire theorem.4 They are rather the different choices
for the primary numeraire, defining the martingale measure
P, respectively Q, i.e., defining what is meant by fair value.
This fair value according to (14) becomes more than
another arbitrary definition exactly when the reference
currency C reflects a real value linked to the real economy.
Only in this way the Islamic choice (14) becomes really
superior to the conventional choice (13). Otherwise,
“value” would remain just a conventional concept. This
is the case in the conventional case, since the interbank
interest markets behind zerobond curves for conventional
currencies do in most cases not really reflect the economic
reality to which the counterparties are exposed. This
becomes evident particularly in situations, when the finance
industry meets some of their homemade crises. Then, the
conventional zero curves tend to exhibit artefacts from the
conventional banking sector. In particular the bias towards
short maturities introduces basis spreads into the forward
rates traded and quoted in the conventional sector. Since
Eds. Hatem A. El-Karanshawy et al.
the 2007 financial crisis, it has been realized now within
the conventional banking sector that, because of the basis
spread included in quoted forward prices, the discount
curve which is supposed to yield fair arbitrage-free prices,
can no longer be equal to the (inter bank) market forward
curve, rather a 2 curve approach becomes necessary.5 The
interpretation is that the conventional inter bank markets
are biased, rather than risk-neutral.
9. Risk profiles and equivalent
martingale measures
Let us now consider the dynamics of the relative price,
Sˆ(t) = S(t)/S0 (t). Since a future price S(t), t > 0, is subject
to the risk of change, its expected value E Q Sˆ(t) is of
particular importance, since it is the most objective value
one can assign to them. The expected value EQ depends
on the measure Q related to the risk of change. Under the
assumptions of no arbitrage and a complete market, there
exists a unique measure Q with a risk-neutral expectation,
i.e.:
E Q Sˆ(t) = Sˆ(0). (15)
This is called the martingale measure.
However, if the market is not complete, e.g., due to
illiquidity, extreme events, or other reasons, the martingale
measure Q and the derived arbitrage-free value are no
longer unique. In the case of an incomplete market,
it makes even more sense to consider also alternative
reference assets. In particular, the reference curve R(t) may
be chosen according to a particular tailor-made dynamical
risk-aversion profile, agreed by the counterparties of a
contract. A particular risk aversion profile corresponds to a
certain utility function, which in turn may be used to select
a particular martingale measure among several equivalent
ones. One might even consider time periods within which
R(t) becomes negative.
The important issue from the Islamic perspective is the
mutual agreement about the risk profile implied by the
choice of R(t).
Determining R(t) rather freely according to a mutually
agreed risk-profile, rather than determining it from the
corresponding reference currency, might be interesting
particularly for direct contracts between investor and
entrepreneurs, where both of them wave developed a
common understanding of the risks involved in the project,
and correspondingly have consciously both agreed on a
particular non-standard form of the R(t) curve. Because
for the project under consideration, this choice is more
suitable than the plain comparison with the value of the
currency.
The last mentioned alternative approach of direct
contractual agreement on a risk-profile and corresponding
R(t) requires outmost transparency about the risks involved
in the contract. Therefore, in cases where there is no detailed
understanding and clarity about risks and the counter party
attitudes appear to be rather indifferent, then it might be
preferable to stick to a standard choice for the reference
asset, namely our proposed commodity-linked currency.
187
Aydin et al.
10. Consequences for Islamic derivatives
as hedging instruments
We exemplify the consequences of our modified evaluation
framework for the design of Islamic derivatives. It is
commonly agreed that such derivatives should be permitted
only for the purpose of hedging.
For example, in Jobst 2010 three examples of synthetic
instruments from asset-based investment finance have
been given. The valuation nevertheless has been based on
conventional measures of value using risk-free discount
factors. As a consequence, the call-put parity could be used
to synthesize a conventional loan.
If the present value would be computed using our proposed
reference asset-based currency, the value of a certain
amount of this currency would be constant without any
interest. Call- put parity can no longer be used to synthesize
a conventional interest-based loan. Hence motivation for
such constructions will become void.
11. Concluding remarks
We have shown that the mathematical framework of riskneutral valuation and arbitrage free contract prices may
be applied for Islamic finance, similarly as in conventional
finance, provided some modification on some conventional
inputs. In particular, the conventional discount rates drawn
from inter-bank markets forward interest rates have to be
replaced.
A rather mild modification for this goal consists in
replacing them by performance rates of underlying equity
or commodity. Shifting the reference is from fixed income
interest to equity or commodity, compatibility with Islamic
principles might be achieved.
As a more challenging method, we propose to replace
the conventional interest-based zero rate curve by an
alternative reference asset. We propose in particular a
commodity-linked multi-dimensional reference currency
system. Globally negotiated fraction βw of the world part,
and regionally agreed fraction βr may provide easy political
instruments to regulate the degree of required homogeneity
of currencies globally or within a defined region such as
Europe respectively.
The most challenging approach is nevertheless the one that
leaves also the most contractual freedom and responsibility
at the same time to the contracting counterparties. Provided
all counterparties have both, the necessary capability for
risk analysis and commitment for risk transparency then
they might agree on a particular zero rate curve freely,
according to their commonly-agreed risk aversion profile.
The zero rates in this case may be linked closely to their
common judgement of risk for the project underlying to the
contract. The zero rates in this case may be even negative,
meaning that a later cash flow is strategically preferable to
an earlier one.
188
Notes
1.From 1831 to 2001 the Greek currency was the
Drachma.
2. From 1869 to 1914 the Drachma was effectively coupled
100% to the union monétaire latine.
3. Even a conventional fixed income market in fact is not
risk-free, but suffers in particular also from the risk of
changing market interest rates.
4. See e.g., Brigo & Mercurio 2007.
5. See e.g., Bianchetti 2009 for the theoretical background,
and Ametrano & Bianchetti 2009 for practial
implementation of the 2 curve approach.
References
Ametrano FM, Bianchetti M. (2009) Bootstrapping the
Illiquidity: Multiple Yield Curves Construction for Market
Coherent Forward Rate Estimation. SSRN Working Paper.
Anonymous. (1933) L’Europa – M nnaie de la Paix. Le Fédériste.
1 January.
Al-Bashir M, Al-Amine M. (2005) Commodity Derivatives:
An Islamic analysis. In: Iqbal M, Khan T. (2005) (Eds.).
Financial Engineering and Islamic Contracts. Palgrave
MacMillan. New York. 58–98.
Brigo D, Mercurio F. (2007) Interest Rate Models–Theory
and Practice, 2nd Edition. Springer Verlag, Heidelberg.
Bianchetti M. (2009) Two Curves, One Price – Pricing and
Hedging Interest Rate Derivatives Using Different Yield
Curves for Discounting and Forwarding. SSRN Working
Paper.
El Diwany T. (2010) The Problem with Interest, 3rd Edition.
Kreatoc, Ltd. London.
Hamzawy A. (2003) Hatar al-’Aulama: Die Arabische
Globalisierungsdebatte – Eine Neuauflage der
Kontroverse über Moderne und Authentizität? Welt des
Islams. 43(2):173–213.
IMF. (2010) Data and Statistics. World Economic Outlook
Data Base. July 2010. Available at: http://www.imf.
org/external/pubs/ft/weo/2010
Jobst A. (2010) Derivatives in Islamic Finance. Paper
presented at SSRN. Preprint.
Latif AM, Janahi AR. (2005) Shariah Alternatives to
Government Bonds. In: Iqbal M, Khan T (Eds.). Financial
Engineering and Islamic Contracts. Palgrave MacMillan.
New York. pp. 99–122.
Lietaer B. (2001) The Future of Money. Random House.
London.
Qutb S. (1949) Al-Adala al-Ijtima’iyya fi’l-Islam. Translation
by Hardie JB (1980). Social Justice in Islam. Octagon
Press. New York.
Rogoff K. (2011) Summary of Comments. Banque de France
Conference. March 4.
Schacht J. (1982) An Introduction to Islamic Law. Clarendon
Press. Oxford.
Islamic banking and finance – Essays on corporate finance, efficiency and product development