Download University Of Nigeria Nsukka

1
FINANCIAL DEVELOPMENT AND ECONOMIC GROWTH IN
NIGERIA
BY
OSUJI CHRISTIAN OBINNA
REG. NO: PG/M.Sc/07/42809
DEPARTMENT OF ECONOMICS
UNIVERSITY OF NIGERIA, NSUKKA
APRIL, 2010
i
TITLE PAGE
FINANCIAL DEVELOPMENT AND ECONOMIC GROWTH IN
NIGERIA
BY
OSUJI CHRISTIAN OBINNA
REG. NO: PG/M.Sc/07/42809
AN M.SC PROJECT SUBMITTED TO DEPARTMENT OF
ECONOMICS, UNIVERSITY OF NIGERIA, NSUKKA, IN
PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR
THE AWARD OF MASTER OF SCIENCE (M.SC) DEGREE IN
ECONOMICS
APRIL, 2010
ii
CERTIFICATION
This is to certify that Osuji Obinna Christian, a post-graduate student of the Department of
Economics,
University
of
Nigeria,
Nsukka,
and
whose
registration
number
is
PG/M.SC/07/42809 has satisfactorily completed the requirement for the award of Master of
Science (M.SC) in Economics.
-------------------------------MR. J. O OLERU
SUPERVISOR
-------------------------------PROF C.C AGU
Head of Department
iii
APPROVAL PAGE
This project has been approved for the award of the Degree of Master of Science (M.SC) of the
Department of Economics, University of Nigeria Nsukka.
-------------------------------MR. J. O OLERU
SUPERVISOR
-------------------------------PROF C.C AGU
Head of Department
-------------------------------Prof (Mrs.) Patience Onokala
Dean, Faculty of Social Sciences
-------------------------------External Examiner
iv
DEDICATION
To my hardworking, industrious, resourceful and resilient mother who sacrificed all she had to
ensure that I tread on the path which she never trod.
v
ACKNOWLEDGEMENT
To Almighty God who laid the foundation of my education and enabled me to read and
understand, I give praise, lord may your name be praised both now and ever. Amen!
I appreciate the contribution of my supervisor Mr. J. O Oleru who amidst his tight schedules
made out time to scrutinize, correct and guide me throughout the duration I wrote this project.
I sincerely thank all my lecturers in the department of economics, Prof’s Agu, Onah, Ikpeze, Rev
Dr (Fr.) Ichoku, Dr Amuka, Dr Asogwa, Dr Moses O, Dr Fonta W, E. O Onyeukwu, Urama N,
for their commitment and handwork. I also appreciate all the non-academic staff in the Dept.
My immeasurable thanks go to my Uncle Mr. Osuji Ndubuisi for his unprecedented love and
providence. I am not forgetting Mr. Anyanwu Cletus, thanks for your hospitality.
A home outside the home is what I found amongst my roommates Baba, Ojini (Obi my Man),
Isaach, and Mark, your goodwill and support is highly appreciated.
Without some deep friendship my life would have been zestless; to you Chiamaka I say
“Gracias” for your memorable companionship. Worthy of remembrance are my classmates
Austin, Pricilla, Sammy, Ebinwa, Ify, CJ, Mike, Aribaba, Sir Pee, Sunny, Chris, Gabdon,
Amogu, Lizzy, Chi, Spencer my other friends Uncle Dee, Bode, Papilo, SPG, Sam, Nono, Nk
and Chigo.
Fondly remembered are my younger ones Nene, Chi baby, Misco, Senior, Eme, Nneka, Bombom
and Junior. You people are wonderful; your warmth was inspiring and your love soothing.
Finally, I thank in a special way my parents Mr. and Mrs. Linus Osuji for being the being the
pillar of my life. Thanks for solidifying my aspirations with your support; I am proud to have
you as parents.
To God be the Glory!!!
vi
TABLE OF CONTENTS
Title page ---------------------------------------------------------------------------------------------- i
Certification ------------------------------------------------------------------------------------------- ii
Approval page ---------------------------------------------------------------------------------------- iii
Dedication --------------------------------------------------------------------------------------------- iv
Acknowledgement ------------------------------------------------------------------------------------ v
Table of Contents ------------------------------------------------------------------------------------ vi
Abstract ------------------------------------------------------------------------------------------------ viii
CHAPTER ONE: Introduction
1.1 Background of the Study ------------------------------------------------------------------------- 1
1.2 Statement of the Problem -------------------------------------------------------------------------- 3
1.3 Objectives of the Study --------------------------------------------------------------------------- 7
1.4 Statement of Hypothesis --------------------------------------------------------------------------- 7
1.5 Significance of the Study -------------------------------------------------------------------------- 8
1.6 Scope of the Study -----------------------------------------------------------------------------------9
CHAPTER TWO: Literature Survey
2.1 Theoretical Studies ---------------------------------------------------------------------------------- 10
2.2 Empirical Studies -------------------------------------------------------------------------------------12
2.3 Overview of the Financial Sector in Nigeria ----------------------------------------------------- 20
2.4 Limitations of Previous Studies -------------------------------------------------------------------- 23
vii
CHAPTER THREE: Methodology
3.1 The Model ------------------------------------------------------------------------------------------- 25
3.2 Model Specification -------------------------------------------------------------------------------- 26
3.3 Justification of Model ------------------------------------------------------------------------------ 30
3.4 Estimation Procedure ------------------------------------------------------------------------------- 30
3.5 Data and Sources of Data -------------------------------------------------------------------------- 33
CHAPTER FOUR: Data Presentation, Analysis and Interpretation
4.1 Unit Roots Test Result ----------------------------------------------------------------------------- 36
4.2 Co integration Test Result ------------------------------------------------------------------------- 37
4.3 Vector Error Correction Model Result ----------------------------------------------------------- 40
4.4 Impulse Response Function ------------------------------------------------------------------------ 46
CHAPTER FIVE: Summary of Findings and Recommendation
5.1 Summary of Findings ------------------------------------------------------------------------------- 48
5.2 Policy Recommendation ---------------------------------------------------------------------------- 50
References ------------------------------------------------------------------------------------------------- 52
Appendix
viii
ABSTRACT
The study examines the causal relationship between financial development and economic growth
in Nigeria for the period 1960 to 2008 using dynamic time series model. Granger causality is
tested within multivariate co integration and vector error correction model (VECM) framework.
Four different measures of financial development are used to capture the different channels
through which finance can affect growth. The empirical findings provide evidence that there is a
stable positive long run relationship between financial development and economic growth. The
result further showed that in Nigeria the direction of causality between financial development
and economic growth is sensitive to the choice of proxy used for financial development.
Financial development cause economic growth when private sector credit is used as proxy but
when money to income ratio, bank deposits and domestic credit ratios are alternatively used as
proxies, growth is found to cause financial development.
1
CHAPTER ONE
INTRODUCTION
1.1 BACKGROUND OF THE STUDY
The pioneering contributions of Schumpeter (1911), and later McKinnon (1973) and
Shaw (1973) regarding the relationship between financial development and economic growth has
remained an important issue of debate in developing economies. The theoretical argument for
linking financial development to growth is that a well-developed financial system performs
several critical functions to enhance the efficiency of intermediation by reducing information,
transaction, and monitoring costs. A modern financial system promotes investment by
identifying and funding good business opportunities, mobilizes savings, monitors the
performance of managers, enables the trading, hedging, and diversification of risk, and facilitates
the exchange of goods and services. These functions result in a more efficient allocation of
resources, in a more rapid accumulation of physical and human capital, and in faster
technological progress, which in turn feed economic growth [Creane, et al. (2004)].
Long-term sustainable economic growth depends on the ability to raise the rates of
accumulation of physical and human capital, to use the resulting productive assets more
efficiently and ensuring the access of the whole population to these assets. Financial
intermediation supports this investment process by mobilizing household and foreign savings for
investment by firms; ensuring that these funds are allocated to the most productive use,
spreading risk and providing liquidity so that firms can operate the new capacity efficiently.
Financial development thus involves the establishment and expansion of institutions,
instruments and markets that support this investment and growth process through improvements
in quantity, quality and efficiency of these financial intermediary services.
2
However, economist still holds startling different opinions regarding the importance of the
financial system for economic growth. While many economists have underlined the importance
of financial sector development in the process of economic development others still think that its
importance is over stressed.
McKinnon and Shaw argued that government repression of financial systems through interest
rate ceilings and directed credit to preferential non-productive sectors, among other restrictive
measures, impedes financial development which they claim is essential for economic growth.
The endogenous growth literature as well stresses the significance of financial development for
long-run economic growth through the impact of financial sector services on capital
accumulation and technological innovation. These services include mobilizing savings, acquiring
information about investments and allocating resources, monitoring managers and exerting
corporate control, and facilitating risk amelioration (Greenwood and Jovanovic, 1990;
Bencivenga and Smith, 1991).
On the other hand, influential economists such as Robinson (1952), Kuznets (1955), Jung (1986)
Lucas (1988), and Ireland (1994) contend that the role of financial development is either
overstated or that financial development follows expansion of the real economy. This would
indicate, in contrast to McKinnon and Shaw and the endogenous growth theorists that causality,
if it exists, runs from economic growth to financial development. Somewhere between these two
views is also the one that claims a mutual impact of finance and growth otherwise known as “the
bidirectional causality view”. Demetriades and Hussein (1996), Greenwood and Smith (1997)
and Luintel and Khan (1999) are some of the studies that provide evidence of bi-directional
causality
3
It is in the light of these conflicting views that this study aims at explaining the
relationship between financial development and growth in Nigeria.
For Nigeria, studying the relationship between financial development and economic
growth is a vital one considering the continuing progress/reforms in its financial sector.
Considered as an integral part of macroeconomic policy the financial sector reforms are expected
to bring about significant economic benefits particularly through a more effective mobilization of
savings and a more efficient allocation of resources thus putting the economy on the path of
sustainable economic growth and development.
Although it is common to consider cross-country regression to judge the growth effects of
financial development, it is also important to study individual-country evidence at least at a
simple level to see whether higher levels of financial development are significantly and robustly
correlated with faster rates of economic growth, physical capital accumulation and economic
efficiency improvements.
1.2 STATEMENT OF THE PROBLEM
Economic development in Nigeria has been characterized by fits and starts. Several
factors have hampered economic growth; one of such factors often associated with this
unsatisfactory growth performance is shallow finance. This factor is critical in view of the
general belief that scarcity of long term finance in developing countries is the major impediment
to higher investment and output growth in these economies Nnanna etal (2004).
Thus, it is now well recognized that financial development is crucial for economic growth. This
recognition dating back to Schumpeter (1911) who argued that the services provided by the
financial intermediaries are important for innovation and development.
4
As this assertion may seem, there is also the argument that financial development contributes
significantly to the economic development of developed countries while the developing countries
have not fared well in this regard Singh (1997).
Hence, even though a growing body of work reflects the close relationship between
financial development and economic growth, it is still possible to encounter especially empirical
researches evidencing all possibilities as positive, negative, no association or negligible
relationships. While Cross country and more recently, panel data studies show evidence of a
positive impact of financial development on growth, Time series on the other hand, offer
contradictory results. Furthermore, the direction of causality between financial development and
economic growth is crucial because it has significantly different implications for development
policy; however, this causal relationship remains unclear.
Nigeria on its part has made notable efforts over the past years to reform its financial system
going from the deregulation and liberalization of the financial sector activities under SAP 1986,
banking consolidation of 2004 and the recent financial system strategy (FSS) 2007 which hopes
to make the country’s financial sector the growth catalyst that would ultimately engineer
Nigeria’s evolution into an international financial center and a natural destination for financial
products and services. Despite these great efforts Nigeria’s economic growth has been dwindling
and has still remained fragile not strong enough to significantly reduce the prevailing level of
poverty even though the various indicators used in measuring financial development has been
increasing steadily over the years. See table 1 below:
5
Table 1
YEAR
FINANCAIL
FINANCAIL
DEVELOPMENT
DEVELOPMENT
(M2/GDP)%
(CPS/GDP)%
1960
12.2
4.9
1965
15.0
7.5
1970
18.5
6.7
1975
19.7
7.8
1980
30.4
15.0
1985
38.7
20.8
1990
25.7
25.7
1995
16.5
10.9
2000
22.6
13.0
2005
19.3
13.8
2006
21.7
14.3
2007
28.1
25.5
2008
38.4
33.8
Where CPS = Credit to Private Sector; M2 = Money Supply
Source CBN Statistical Bulletin 2008
From the table above, one would ask; what is the relationship between the growth in these
financial development indicators and Nigeria’s economic growth over the years? Secondly,
considering the fact that correlation does not imply causation in any way; can changes in
Nigeria’s growth rate be attributed to improvements and innovations in its financial sector? Does
this financial growth promote economic development in Nigeria?
Referring to earlier studies done on this area especially the works of Aigbokan (1995),
Odedokun (1995) and Ndebbio (2004) as it pertains to Nigeria, their studies in testing for
causality failed to employ multivariate co integration and vector error correction model (VECM)
thus failed in making a clear distinction between long run and short run causality.
This distinction is very important since as Darrat (1999) states “most of the benefits of higher
levels of financial development could be realized in the short run while in the long run as the
economy grows and becomes mature these effects slowly disappear”.
6
Moreso, the time span of some of the studies were too brief to capture the long run relationship
between financial development and economic growth as the number of observations did not
exceed 25 years.
The work of Ndebbio (2004) was based on a cross country regression approach and generally
there has been a growing concern about cross country empirical approach and its use for causal
inference in particular. The studies based on cross country suffer from potential biases induced
by simultaneity, omitted variables and unobserved country specific effect on the finance –
growth nexus. This was acknowledged by Levine (1997) and Luintel and Khan (1999) who
further explained that aggregation blurs important events and differences across countries.
Similarly, the studies were based on financial measures that may not capture the mechanism
through which financial development cause economic growth.
Accordingly, it is appropriate and timely to empirically re-examine the financial development
and economic growth relationship in Nigeria, utilizing larger sample size, introducing alternative
indicators of financial development and testing for causality in a multivariate co integration
framework. This forms the bedrock of this research effort. In essence the study will seek to
answer the following research questions:
(i)
What is the relationship between financial development and economic growth in
Nigeria?
(ii)
What are the direction of causality between financial development and economic
growth in Nigeria?
(iii)
Finally, what are the policy implications of these relationships?
7
Answering the above research questions would help us to see whether, how and to what extent
the financial system contributes to economic growth in Nigeria.
1.3 OBJECTIVE OF THE STUDY
This study intends to investigate the link between financial development and economic
growth in Nigeria from 1960-2008. Are differences in financial development and structure
importantly associated with differences in economic growth rates?
Specifically, the study will address the following objectives:
1. To examine the relationship between financial development and economic growth in
Nigeria.
2. To determine the direction of causality between financial development and economic
growth.
1.4 STATEMENT OF HYPOTHESES
To assess the nature of the finance-growth relationship, the following hypotheses have
been suggested:
1. Financial development has no positive significant relationship with economic growth in
Nigeria.
2. There is no direction of causality between financial development and economic growth in
Nigeria.
8
1.5 SIGNIFICANCE OF THE STUDY
Whether financial development influences economic growth is not just a matter of
intellectual curiosity, it is a crucial policy issue as research that clarifies our understanding about
the role of finance in economic growth will have policy implications and shape policy oriented
research.
Thus, the importance of this study lies in the fact that it will provide an insight as to
whether financial sector development is a necessary and sufficient condition for higher growth
rates in Nigerian economy.
Information about the impact of finance on economic growth will influence the priority
that policy makers and advisers attach to reforming financial sector policies. Furthermore,
convincing evidence that the financial system influences long-run economic growth will
advertise the urgent need for research on the political, legal, regulating and policy determinants
of financial development.
In contrast, if a sufficiently abundant quantity of research indicates that the operation of
the financial sector merely responds to economic development, then this will almost certainly
mitigate the intensity of research on the determinants and evolution of financial systems.
1.6 SCOPE OF THE STUDY
To assess the current state of knowledge on the finance-growth nexus in Nigeria, a time
series data will be employed. The study will cover the period 1960 to 2008.
9
CHAPTER TWO
LITERATURE SURVEY
This section reviews various studies on both theoretical and empirical relationships between
financial development and economic growth.
2.1 THEORETICAL STUDIES
Studies undertaken to examine the relationship between financial development and economic
growth goes far back to the work of Bagehot (1873), Schumpeter (1911), and Hicks (1969).
In his study for instance, Schumpeter (1911) discusses the finance growth relationship as a
supply-leading one, in which the financial sector leads economic growth by successfully
identifying and funding high yielding projects. This is based on the view that a financial system
that is functioning well would encourage technological innovation by selecting and financing
businesses that are expected to be successful.
Bagehot (1873) and Hicks (1969) on the other hand, argued that financial development was an
important channel in the industrialization of England, by helping the movement of large amounts
of funds for “immense” works.
Later works include that of Greenwood and Jovanovic (1990), Levine (1991), Bencivenga and
Smith (1991) and Saint-Paul (1992), which involved theoretical models, wherein an efficient
financial market raises the quality of investments, thus leading to economic growth. Specifically,
Greenwood and Jovanovic (1990) built in their model a financial sector whose main objective is
to direct funds to high-yielding investments with the assistance of information. This then would
lead to economic growth, which would in turn enable the implementation of costly financial
structures.
10
In his model, Levine (1991) explains how stock markets influence growth by improving firm
efficiency. Furthermore, Bencivenga and Smith (1991) explain in their study, that a wellfunctioning financial system would improve the level of investment towards non-liquid objects,
which will be beneficial to the economy. Saint-Paul (1992) on the other hand, explains the role
of the financial sector in helping business enterprises in specialization by allowing investors to
hedge by holding a diversified portfolio. This in turn would lead to productivity growth.
Supporting this, Atje and Jovanovic (1993) explain how the financial system can help investors
disperse risk and provide funding, thereby guiding them to the best investments which are
profitable to the economy.
More studies include that of Maurice Obstfled (1994) who argued that financial openness and
access to international financial markets bring benefits to businesses as well as the economy.
Bencivenga, Smith and Starr (1995) argued that industries, which require a longer period to
implement new technologies benefit more relatively, from developments in the financial market.
Rajan and Zingales (1998) concluded that as the market develops, firms that are less-firmly
established and have difficulty with self-funding projects, would benefit better from external
funding methods, and therefore expand relatively faster.
Blackburn and Hung (1996) found that in a developed financial system, the task of monitoring
projects can be undertaken by financial intermediaries, lowering transaction costs and channeling
greater savings towards new investments, thus boosting economic growth. Moreover, the authors
explain how a country can be trapped in a situation of low economic growth and low financial
development.
More recently, Levine and Zervos (1998) in their study argued that higher returns and improved
risk could encourage a lower savings rate, which would lower economic growth with more liquid
11
and internationally integrated financial markets. In line with this, Tsuru (2000) explained how
the development of the financial sector is able to affect the saving rate, thus affecting the rate of
economic growth.
2.2 EMPIRICAL LITERATURE
An extensive amount of empirical investigations have been conducted aimed at testing the
conflicting theoretical developments in the finance- growth nexus using different techniques.
While some studies looks at the relationship between the two, a large number of studies are
concerned with the question first asked by Patrick (1966) as to what direction does causality
between finance and growth run, others in a similar vein deal specifically with stock market
indicators.
Using a broad cross county analysis Goldsmith (1969) reports a rough correlation
between financial development and economic activity for a sample of 35 countries covering the
period 1860-1963.
King and Levine (1993 a, b) using IMF data and various financial indicators for roughly 80
countries over the 1960-1989 period concluded that there is a positive relationship between
financial indicators and growth and that financial development is robustly correlated with
subsequent rates of growth, capital accumulation and economic efficiency. They emphasized
those policies that alter efficiency of financial intermediation exact a first order influence on
growth.
In another cross-section study, DeGregorio and Guidotti (1995) examining a sample of
about 100 countries during 1960-1985 found out that financial development is associated with
improved growth performance. The authors clearly stated that the impact of financial
development increases from high to low income countries.
12
They subsequently analyzed panel data of 12 latin American countries, using six year averages
for 1950-85. Particularly for the 1970’s and 1980’s unregulated financial liberalization and
expectations of government bailouts explain a reversed relationship between financial
development and economic growth. This result leads to the policy that financial liberalization
requires appropriate regulation framework in order to avoid financial crisis.
Ram (1999) finds a positive association between financial factors and economic development
only for high growth countries applying panel techniques to a sample of four Latin American and
South East Asian countries over the period 1965-1985.
Levine etal (2000) using a sample of 74 development and less developed countries over the
period 1960-1995, go beyond previous studies recognizing the potential biases induced by
simultaneity, omitted variables and unobserved county – specific effect on the finance – growth
nexus. They found that the strong positive relationship between financial development and
output growth can be partly explained by the impact of the exogenous components like finance
development on economic growth. They interpreted these results as supportive of the growth
enhancing hypothesis of financial development.
Deidda and Fattouh (2002) state that financial depth and growth are only associated positively
for a high income sub sample. They offer the explanation that in developing countries the fixed
resources cost associated with the provision of financial services inhibits growth they therefore
conclude that there is non-linearity between finance and growth.
On the question of causality, Fritz (1984) uses the Philippines as a sample economy to conduct
causality test for time series of quarterly date from 1969 till 1981. He finds evidence for the
hypothesis that in the initial stages of the development process causality is reversed with the real
economy demanding an increase in financial services.
13
Other studies include that by Jung (1986) which used a vector auto regressive (VAR) approach
to test the causality between financial development and economic growth for a sample of 56
countries (both developed and developing). He uses two alterative measure of financial
development, one is a currency ratio (currency over M1.) and the other one a monetization
variable (M2 over GDP). For the LDCS he finds a supply leading more often than a demand
following indicating the importance of financial development for developing countries. The
opposite is true for developed countries even though only when the currency is used. As far as
the temporal causality pattern is concerned the finding moderately supports Patrick hypothesis.
Murinde and Eng (1994) in their study found the causality between financial development and
economic growth running in both directions in the case of Singapore.
Aigbokan (1995) applying granger causality model to investigate the causal relationship between
real and financial sector growth in Nigeria finds evidence that largely supports the supply leading
hypothesis that financial development induces real growth.
Odedokun (1996) analyses a sample of 71 developing countries over varying periods that
generally span the 1960’s and 1980’s the findings are strongly in favour of the finance causes
growth hypothesis. Using time series regression analysis, the author comes to the conclusion that
financial intermediation promotes economic growth in roughly 85 percent of the countries.
He also observed the growth-promoting effects of financial intermediation primarily in lowincome LDC’s.
Rousseau and Wachtel (1998) found one way causality in the relationship between financial
development and economic growth in the case of 5 percent OECD countries during the period of
fast industrialization (1871-1979)
14
Beck, Levine and loayza (2002) found that banks have a strong causal effect on economic
growth using panel data analysis.
Some studies however come to a different conclusion with respect to the causality issue.
Demetriades and Hussein (1996) also conducts causality test between financial development and
real GDP for 16 less developed countries using time series techniques. Their findings provide
little support for the hypothesis that financial factors play a leading role in the process of
economic development. There is more evidence for the opposite pattern of growth causing
finance and for bi-directional causality. Another important finding of their study is that causality
patterns vary across countries indicating a need for case studies and careful time series analysis.
In a case study of India Demetriade and Luintel (1996) show empirical links between banking
sector controls and the process of financial deepening and between financial deepening and
growth.
Arestis and Demetriade (1997) use time series analysis and Johnson Co integration analysis for
the US and Germany. For Germany, they find an effect of banking development on growth.
In the US there is insufficient evidence to claim a growth effect of financial development and the
data point to direction that GDP contributes to both banking system and stock market
development.
Luintel and Khan (1999) apply a multivariate vector auto regression framework to a sample of
ten mostly developing countries over a period of 36- 41 yrs and find evidence of bi – directional
causality for all countries.
15
On specific aspects of financial development, Roubini and Sala i Martin (1992) using a crossSection of 98 countries for the period 1960-1985, showed that various measures of financial
repression affect growth negatively. More so inflation rates and reserve ratios are negatively
correlated with growth after controlling for other factors.
Dealing specifically with stock market indicators, Atje and Jovanovich (1993) examined the role
of stock market on development using two proxies of financial development, one measuring
bank intermediation and one approximating stock market activity. They concluded that stock
market activities had positive effect on growth. They could not however establish or significant
relationship between bank liabilities and growth. In contrast to that Harris (1997) comes to the
conclusion that stock market activity has at best weak explanatory power for long-run growth in
per capita output. He uses data on 49 of the 60 countries teat had official stock markets in 1991
covering the period 1980-1991. Splitting the sample into developed and developing countries he
found that for the developing countries stock markets do not seem to promote long-run growth
where as for the developed countries they have some explanatory power.
Levine and Zervos (1998) found evidence that stock market liquidity and banking development
have a positive relationship with economic growth.
Arestis, Demetriades and Luintel (2001) conducted time series analysis using data on five
industrialized countries covering the period 1968-1998.
They report that stock market have made significant contributions to growth in Germany, Japan
and France. The effect of stock market, however, is weaker than the impact of banking. For USA
and UK the link between finance and growth is not very robust and rather seems to run from
growth to finance.
16
More recently, Shan Jordan (2003) using VAR techniques of innovation accounting for china
found that financial development come as a second force (after the contribution from labour
input) in leading economic growth in china.
The empirical evidence supports the view that financial development and economic growth
exhibit a two way causality and hence in contrast with the finance lead growth hypothesis.
Christopoulos and Tsionas (2004) using panel unit root test and panel co integration analysis to
examine the relationship between financial development and Economic growth in ten developing
countries find strong evidence in favour of the hypothesis that long-run causality runs from
financial development to growth and there is no evidence of bi-directional causality.
Furthermore, they find a unique co integrating vector between growth and financial development
and emphasize the long-run nature of the relationship between finance and growth.
Ndebbio (2004) using two financial deepening variables namely the degree of financial
intermediation measured by M2 as ratio to GDP and the growth rate in per capita real money
balances; conducted within an ordinary least square (OLS) multiple regression method reports
that a developed financial sector spurs overall high but sustainable growth of an economy.
Khan etal (2005) using an autoregressive distributed lag (ARDL) approach found a stable longrun relationship between economic growth and financial development for Pakistan over the
period 1971-2004, concluding that economic growth is an outcome of financial development.
Suleiman and Aamer (2006) on the other hand report a weak support for long-run relationship
between financial development and growth and for the hypothesis that finance leads to growth.
Employing four different measures of financial development in examining the causal link
between financial development in five Middle Eastern and North African countries for the
periods 1960-2004 within a trivariate VAR framework they found that in cases where co
17
integration was detected granger causality was either bidirectional or it ran from output to
financial development.
Eita and Andre (2007) in their study found a stable long run relationship between financial
development and economic growth with causality running from financial development to
economic growth in the case of Botswana.
Acaravci (2007) in the case of Turkey for the periods of 1986:1-2006:4 found a one way shortrun causal relationship between financial development and economic growth.
Guryay etal (2007) using the OLS method to examine the relationship between financial
development and economic growth found that there is a negligible positive effect of financial
development on economic growth in Northern Cyprus . Furthermore the granger causality test
showed evidence of economic growth causing financial development in Northern Cyprus.
Chakraborty (2008) employing five different measures of financial development found out that
in an overall sense economic growth has caused financial development in India for the period
1996-2005.
Majid (2008) using a quarterly data from 1998-2008 and ARDL model documents a long run
relationship between financial development and economic growth in the case of Malaysia. The
granger causality test based on the vector error correction model (VECM) revealed that there is a
unidirectional causality running from finance to growth thus supporting the finance led growth
hypothesis.
Abdul and Ying Ma (2008) applying the ARDL approach to explore finance-growth
relationships for China and Pakistan over the period 1960-2005 reports different results for both
countries. Using deposit liability ratio (DLR) and credit to Private sector (CPS) as proxies for
financial development they found that both DLR and CPS had significant impacts on the
18
economic growth of Pakistan which is line with the study of khan etal (2005). In the case of
China, the study finds a positive and significant relationship for DLR and a positive yet
insignificant relationship with CPS.
Agu and Chukwu (2008) employing the augmented granger causality test approach developed by
Toda and Yamamotto (1995) to ascertain the direction of causality between bank based financial
deepening variables and economic growth in Nigeria between 1970-2005 found that financial
deepening and economic growth are positively co integrated with one co integrating vector
indicating a stable and sustainable long run relationship. On the issue of causality, their findings
suggested that the choice of bank based financial deepening variables influences the causality
outcome. While variables like private sector credit, broad money supported the demand
following hypothesis variables like loan deposit ratio and bank deposit liabilities were in favour
of the supply leading hypothesis.
19
2.3 OVERVIEW OF THE FINANCIAL SECTOR IN NIGERIA
The financial sector is a segment of the economy that provides the market environment
for savers and investors to interact and trade in financial assets. The facilitators in the sector are
banks, non-bank financial institutions, and regulatory agencies. The evolution of this sector in
Nigeria dates back to the pre-independence era when it was virtually dominated and controlled
by foreigners. The establishment of the Central Bank of Nigeria (CBN) in 1958 gave impetus to
internally propelled growth of the sector. The CBN became the first regulatory organ in the
sector saddled with the responsibility of developing domestic money market, among other things.
The Nigerian Deposit insurance Corporation (NDIC) was established in 1988 to assist the CBN
in its regulatory functions. NDIC operations include the provision of insurance cover for bank
deposits, liquidation of banks whose licenses have been revoked by the CBN, and payment of
compensation to depositors of such banks.
The Capital Issues Committee (1961), which later transformed into Securities and
Exchange Commission (1977), is the second regulatory body in the sector charged with the
responsibility of developing domestic capital market.
The concurrent establishment of the
Nigerian Stock Exchange (then Lagos Stock Exchange) in 1961 was intended to assist the
commission in developing the capital market by promoting the liquidity of its securities.
The
third and most recent regulatory agency in the sector is the National Insurance Commission
(1977), which was established to enhance the effective control of the financial system where
insurance companies have become prominent.
Since 1970, the sector has grown quite significantly due to the monetization of export
earnings and the consequent increase in economic activities (Ojo, 1993), coupled with the
various policies adopted by the government during the period under review. Between 1970 and
20
1985, growth of the section was a gradual one because it was predominantly under government
regulation. The period 1986-2000 however witnessed a phenomenal growth of the sector as a
result of the financial deregulation policy of government (1986).
That policy led to the
emergence of appreciable number of banks and other financial institutions, with some of them
poorly equipped to properly function in the sector. The problem became glaring by 1991 as the
number of distressed banks rose astronomically, leading to a reduction in the size of the sector.
This development prompted government to set up the failed Banks Tribunal (1994) in order to
sanitize the sector. In addition, the capital base for banks and other financial institutions was
significantly increased to enhance stability in the sector. These measures, among others, assisted
in the repositioning and restoration of growth to the sector by year 2000.
In order to further ensure stability in the banking system the central bank of Nigeria
CBN, on 6th July 2004, announced a required minimum capital base of N25 billion for banks
effective 31st December 2005. The purpose of the recapitalization process was to enable the
Nigerian banks give priority to and make available cheap and competitive credit to the real sector
so that the economy can grow.
The first phase of the banking sector reforms and consolidation which was concluded at
the end of December 2005 resulted in 25 banks emerging from seventy five (75) out of eighty
Nine (89) banks that had existed prior to the end of December 2005. This has brought a number
a positive developments to the banking sector in particular and the economy in general. The
banking consolidation has produced relatively well capitalized banks which has engendered
greater public confidence in the system. It has brought greater public awareness and a deepening
of the capital market. The resulting liquidity in the system also induced a significant fall in
interest rate. Banks now have greater potential to finance big ticket transactions with higher
21
oblige limit. Ownership of banks has been diluted and this in no small way, tamed the monster
of insider and corporate governance abuse. The banks now enjoy economies of scale where the
benefits in the form of reduced bank charges are passed to customers. With virtually all the
banks now public quoted, there is wider regulatory oversight with the securities and exchange
commission (SEC) and Nigerian stock Exchange (NSE) joining the regulatory team. These
regulators now oversee and focus on fewer banks thereby fostering effectiveness and efficiency
in supervision.
The growth in the Nigerian financial sector continued in 2008 as indicated by the various
performance measures in all segments. The depth of the financial sector increased as broad
money supply to nominal GDP ratio rose to 38.4% from 28.1% at the end of 2007. The banking
sector also showed stronger capacity to finance real sector activities with substantial credit flow
to the core private sector as CP/GDP ratio increased from 24.5% in 2007 to 33.8% in 2008.
In addition, the increased use of the various electronic money products reflected the shift away
from cash transactions and thus an improvement in the efficiency of funds intermediation.
Consequently, the ratio of currency outside banks to broad money supply fell further to 9.7 per
cent in 2008 from 12.7 percent at the end of 2007.
Overall, the financial system at the end of December 2007 composed the central Bank of Nigeria
(CBN).
Nigerian Deposit Insurance Corporation (NDIC), the securities and Exchange
commission (SEC), the National Insurance Commission (NAICOM), the National Insurance
Commission (PENCOM), 24 deposit money banks, 5 discount houses, 709 microfinance banks,
112 finance companies, 703 bureaux-de change, I stock Exchange, I commodity Exchange, 93
primary mortgage institutions, 5 development finance institutions and 117 insurance companies.
22
2.4 LIMITATIONS OF PREVIOUS STUDIES
There are number of concerns with previous empirical work done by Nigerian researchers
on this area:
(1)
Earlier studies by Aigbokan (1995), Odedokun (1996), Ndebbio (2004) except for Agu and
Chukwu (2008) failed to employ multivariate co integration and vector error correction
mechanism (VECM) in testing for causality.
(2)
Most of the studies were based on financial measures that may not capture the
mechanism through which financial development can cause economic growth. According to
Luintel & Khan (1999) the standard measure of financial development used in literature is the
ratio of broad money – usually M2 to the level of nominal GDP. However, strictly speaking this
ratio measures the extent of monetization rather than that of financial depth. In developing
countries such as ours a large component of the broad money stock is currency held outside the
banking system and a such monetization can be increasing without financial development
occurring. Thus it is not an entirely satisfactory indicator of financial depth.
(3)
Finally, the above studies test for granger causality between finance and growth within
the sample period and did not attempt to evaluate the strength of their findings beyond the
sample period by applying variance decompositions or impulse response functions.
In the present study we try to overcome the limitations of the aforementioned studies by
using four alternative proxies of financial development namely the ratio of broad money (M2) to
income (M2Y), the ratio of total bank liabilities to nominal GDP (BDY), the ratio of private
sector credit to GDP (CPSY) and the ratio of Bank credit to GDP (DCY).
23
To overcome the misspecification bias we apply a multivariate vector autoregressive (VAR)
system that includes a complete conditioning information set consisting of variables borrowed
from the relevant literature. The share of investment in GDP is included as an additional
variable as this allows us to test whether financial development affects economic growth through
increasing productivity or through accumulation of resources. Interest rate is also included as a
real positive interest increases financial depth through increased savings mobilization and
promotes growth through increasing the volume of productivity of capital.
We also test the direction of causality between financial development and economic
growth using Unit root and co integration test within a multivariate VAR model.
Furthermore, using variance decompositions we estimate the relative importance of
financial development in explaining changes in the growth of per capita GDP beyond and the
sample period.
24
CHAPTER THREE
METHODOLOGY
This chapter deals with model specifications, data definitions, estimation procedures,
evaluation techniques, data and sources of data.
3.1 THE MODEL
Given the nature of the objectives of this study, the research employs time series
econometric methodology using the vector autoregressive model (VAR) which is transformed
into the vector error correction model (VECM).
In structural equation models some variables are treated as endogenous and some as exogenous
before estimating such models, the equations in the system has to be identified. However this
decision is often subjective and has been severely criticized by Christopher Sims. According to
Sims, if there is true simultaneity among a set of variables, they should all be treated on an equal
footing; there should not be any a priori distinction between endogenous and exogenous
variables. It is in this spirit that the VAR model was developed.
Thus, in VAR models all variables are treated as endogenous, each variable is explained by its
own lagged values as well as the lagged values of all other endogenous variables in the model
Sims vector autoregressive model offers an easy solution in explaining and predicting the
values of a set of economic variables at any given point in time. VAR is a straight forward,
powerful statistical forecasting technique that can be applied to any set of historical data. Like
the structural model, the VAR system also generates system of equations that can project the
future paths of economic variables extrapolating from their past historical values. However, the
main difference between the VAR system and the structural models is that the VAR system is
based entirely on empirical regularities embedded in the data.
25
The structural model is tied closely to the economic theory and has to follow the assumption and
the a priori restrictions imposed there in. VAR, on the other hand does not have to resort to the
theory per say, in fact, the data determines the final system (Chisiti, et al 1992).
3.2 MODEL SPECIFICATION
The research will be guided by the model specified below; first in its functional form then
transformed into a VAR model.
RGDP =
F (FD, INV, r)
(1)
Where
RGDP =
Real Gross Domestic Product Per capita (proxy for economic growth in Nigeria)
FD
Financial development indicators which is proxied by the ratio of broad money
=
stock (M2) to GDP (M2Y), the ratio of total bank liabilities to nominal GDP
(BDY), the ratio of private sector credit to GDP (CPSY) and the ratio of Bank
credit to GDP (DCY).
Total bank liabilities are obtained by taking the ratio of broad money stock (M2)
minus currency in circulation to GDP.
INV
=
Output share of Investment (ie Gross capital formation). This variable is
considered as having a robust correlation to economic growth.
r
=
Interest rate (maximum lending rates of commercial banks)
INV and r are information set of control variables commonly used in literature to avoid
problems posed by bivarate VAR. INV used in (christopoulos and Tsionas 2004) and r used in
(Luintel and Khan 1999).
26
Putting equation (1) in a VAR model we have:
n
n
n
n
i 1
i 1
i 1
i 1
RGDPt   1  11  RGDPt i  12  FD t -i  13  INVt i  14  rt i  U 1
n
n
n
n
i 1
i 1
i 1
i 1
n
n
n
n
i 1
i 1
i 1
i 1
FDt   2   21  RGDPt i   22  FDt i   23  INVt i   24  rt i  U 2
INVt   3   31  RGDPt i   32  FDt i   33  INVt i   34  rt i  U 3
n
n
n
n
i 1
i 1
i 1
i 1
rt   4   41  RGDPt i   42  FDt i   43 INVt i   44  rt i  U 4
(2)
(3)
(4)
(5)
Where
i is the lag length, ’s is the constant terms, U’s are the stochastic error terms which in the
language of VAR is referred to as Impulses or Innovations and RGDP, FD, INV, r, are as defined
earlier.
The model above can be stated more compactly as below:
n
n
i 1
i 1
Yit = i + i  y t -i  i  x it -i  Vi
(6)
Where
Yit = vector of endogenous variables (such that yit = RGDPt….rt)
i = vector of constant terms
i = Coefficient of the autoregressive terms
i = Coefficients of the explanatory variables (vector of coefficients)
vi = Vector of innovations.
If the variables are non-stationary say I (1), it may be helpful to take the first difference of the
variables to make them I (0) and then use the differenced variables in the VAR.
27
However, if the I (1) variables are co integrated, by differencing the variables, there will be loss
of important information about the long-run relationships.
Omitting the co integrating combination is a specification error in a VAR (in first differencing)
and in addition such a VAR provides no information about the long-run which is often
considerable interest to economist (Patterson 2000).
More so, the co integration technique pioneered by Engel and Granger (1987) makes a
significant contribution towards testing causality. According to this technique once a number of
variables (say RGDP and FD) are found to be co integrated, there always exist a corresponding
error correction representation which implies that changes in the dependent variable are a
function of the level of disequilibrium in the co integration relationship (captured by the error
correction term) as well as changes in other explanatory variables. A consequence of co
integration is that either RGDPt or FDt or both must be caused by the lagged error- correction
term which itself is a function of RGDPt-i, FDt-i.
A vector error correction model (VECM) is a restricted VAR designed to use with non
stationary variables that are known to be co integrated. It restricts long-un behaviour of the
endogenous variables to converge to their co-integrating relationships while allowing for shortrun adjustment dynamics.
The VECM corresponding to our situation is
n
n
n
n
i 1
i 1
i 1
i 1
RGDPt  1  11  RGDPt i  12  FDt  i  13  INVt  i  14  rt  i  15 ECM t  i  U1
n
n
n
n
i 1
i 1
i 1
i 1
FDt   2   21  RGDPt i   22  FDt i   23  INVt i   24  rt i   25ECM t i  U 2
n
n
n
i 1
i 1
i 1
INVt   3   31  RGDPt i   32  FDt i   33  INVt i   34 rt i   35 ECM t i  U 3
n
n
n
n
i 1
i 1
i 1
i 1
rt   4   41  RGDPt i   42  FDt i   43  INVt i   44  rt i   45 ECM t i  U 4
(7)
(8)
(9)
(10)
28
Where
 is the difference operator, ECM is the error correction term, i is the lag length which translates
to a lag of i-1 in the VECM.
Equation (7), (8) (9) and (10) can be summarized in the form below:
n
n
i 1
i 1
Yit   i   i  Yt i  i  X it i   i ECM t i  U i
3.3 JUSTIFICATION OF MODELS
The choice of a VAR model to be transformed into a vector error correction model
(VECM) is made because it is one of the models that are superior and are not highly vulnerable
to simultaneity bias. It has the ability to test for weak exogeneity and parameter restrictions.
It also assumes there is no priory direction of causality among variable. It is a theoretical i.e does
not require any explicit economic theory to estimate the model (Gujarati 2003).
VECM offers a way of analyzing the dynamic relationships between variables such as financial
Development and economic growth.
The transformation of the VAR into VECM is to account for the speed of adjustment in
the long-run and short-run dynamic of the model and it has a co integration restriction embedded
in the specification.
3.4 ESTIMATION PROCEDURE
The empirical investigation on the relationship between financial development and
economic growth will be performed in two steps. First we define the order of integration in series
and explore the long- run relationship between the variables by using unit root test and co
integration test respectively.
29
Second, we test the long run or short- run causal relationship between financial development and
Economic growth which is carried out in VECM or VAR framework.
Because the order of integration of a time series is of great importance for the analysis, we use
the Augmented-Dickey fuller (ADF) unit root test to examine the stationary of the variables.
If the variables are integrated of order 1 ie I (1) (becoming stationary after first difference) then
we search for the co integrating relationship between these variable.
If there is no co integrating relationship we make the variables stationary by first differencing
and test for non- causality in a VAR context.
For non- stationary variables and a co integrated relationship we estimate a vector error
correction model and again test for granger non-causally in this context.
Finally the impulse response function (IRF) is applied. The IRF gives the proportions of the
movement in one variable which are because of its own shocks and the shocks in the other
variables. Thus, in our context it allows to explore the relative importance of financial
development in accounting for innovations in economic growth.
OPTIMAL LAG-LEMGHT
This is used to ascertain the actual number of the lag to be introduced so as to avoid too
much lag or few lag. Too much lag will consume degree of freedom and multi nonlinearly will
set in while few lag will lead to specification error.
Therefore based on the Akaike information criterion (AIC) and Schwarz information criterion
(SIC), we shall choose the lag level that has the minimum information criterion for the VAR and
VECM estimation.
30
THE UNIT ROOT TEST
The variable of this study shall be subjected to stationary test and made stationary using
the Augmented Dickey fuller (ADF) test because it eliminates the problem of auto correlation by
including enough terms so that the error term is serially uncorrelated. The regression form of the
ADF Unit root test is stated below
n
Yt = a0 + a1Yt-1 + a2t +   i Yt-i + Ui
i 1
Where a0 = is the intercept
t = is linear trend
n = no of lagged differences
Ui = is the error term
 = is the first difference operator.
The null hypothesis is that a unit root problem exists and the alternative hypothesis is that there
exists no unit root problem.
COINTGERATION TEST
This test will be used to establish the existence of long-run relationship between financial
development and economic growth in Nigeria. The Johansen (1988) procedure is utilized. The
approach is chosen because it does not suffer normalization problem (Gujarati 2003).
The Johansen (1988) co integration test is done in order to obtain the random innovation which
will be used in estimating the VECM.
This is specified as follows:
n
Yt = 1 +   i Xt-i + Vt
i 1
31
Where
Yt = Dependent variables
Xti = Explanatory variables
i = lag length
vt = residuals (error term)
 i = parameter coefficient
The random innovation is defined below
Vt = Yt - 1 -
n

1 1
 i Xti.
3.5 DATA AND SOURCES OF DATA
Annual data are used and the study covers the period of 1960 to 2008. The study uses the ratio of
broad money supply (M2) to GDP (M2Y), the ratio of total bank deposit liabilities to nominal
GDP (BDY), the ratio of private sector credit to GDP (CPSY) and the ratio of Bank credit to
GDP (DCY) as proxies of financial development.
In the literature, the most commonly used measure of financial development is the ratio of some
broad measure of the stock usually M2 to the level of nominal GDP (king and Levine 1993a).
This simple indicator measures the degree of monetization in the economy (ie in which money
provides valuable payment and saving services).
However in developing countries a large part M2 stock consists of currency held outside banks.
As such, an increase in the M2 may reflect extensive use of currency rather than increase in bank
deposits and for this reason this measure is less indicative of the degree of financial
intermediation by banking institutions.
32
In order to obtain a more representative measure of financial development, Demetriades and
Hussien (1996); Luintel and Khan (1999) proposed to subtract currency in circulation from the
broad money stock giving rise to the ratio of bank deposit liabilities to income.
The study also uses the of private sector credit to GDP as a second proxy for financial
development. This indicator is frequently used to assess the allocation of financial assets as it is
directly related to investment and growth and such a more direct measure of financial
intermediation. Furthermore, it is assumed that credit provided to the private sector generates
increase in investment and productivity to a much larger extent to credits to public sector.
It is also argued that loans to the private sector are given more stringently and that the improved
quality of investment emanating from financial intermediaries evaluation of project viability is
more significant for private sector credits. Thus an increase in the ratio of credit extended to the
private sector to GDP is also interpreted as financial deepening (CBN 2008).
The ratio of Bank credit to GDP is also used as another proxy for financial development. This
represents the domestic assets of the financial sector.
For economic growth real GDP per capita (RGDP) is used. Investment (INV) and interest rate (r)
are included to avoid problem of simultaneity bias in VAR.
Finally the data are sourced from the central Bank of Nigeria statistical bulletin (various Issues)
ECONOMETRIC SOFTWARE
The data series will be subjected to the appropriate time series tests using EVIEWS 5.0
econometric software.
33
CHAPTER FOUR
DATA PRESENTATION, ANALYSIS AND INTERPRETATION
The chapter reviews and analyzes the various data used for this study.
4.1. UNIT ROOTS TEST RESULT
In this study the augmented dickey-fuller (ADF) unit roots test is employed to test for the
stationarity of the variables. The null hypothesis is that the variable under investigation has a unit
root against the alternative that it does not. The lag length was set based on the Akaike and
Schwartz information criteria. The decision rule is to reject the null and accept the alternative if
the ADF statistics value exceeds the critical value at chosen level of significance in absolute
terms and vice versa. These results are presented in table 2.
Table 2 Unit Root Test
Variable
ADF statistics
Critical values
ADF statistics
(1st
(level form)
RGDP
M2Y
BDY
CPSY
DCY
INV
R
-1.49
-1.60
-1.01
-1.35
-2.48
-1.43
-1.12
1%
-3.62
5%
10%
1%
-3.57
5%
-2.92
10%
-2.66
1%
-3.57
5%
Critical values
Diff)
-5.37
1%
-3.62
-2.94
5%
-2.94
-2.61
10
-2.61
1%
-3.58
5%
-2.93
10%
-2.60
1%
-3.58
-2.93
5%
-2.93
10%
-2.66
10%
-2.60
1%
-3.57
1%
-3.58
5%
-2.92
5%
-2.93
10%
-2.66
10%
-2.60
1%
-3.58
1%
-3.58
5%
-2.93
5%
-2.93
10%
-2.60
10%
-2.60
1%
-3.58
1%
-3.58
5%
-2.93
5%
-2.93
10%
-2.60
10%
-2.60
1%
-3.58
1%
-3.58
5%
-2.93
5%
-2.93
-6.00
-5.30
-5.90
-5.86
-6.63
-7.95
34
10%
-2.60
10%
-2.60
The results of table 2 show that all the variables are non-stationary i.e. has unit roots in levels
since their ADF values are less than the critical values at 1,5, and 10 percent significance level.
However they became stationary in first-difference and at 1%, 5%, and 10% the null hypothesis
of no unit root was accepted for all series. Thus, we conclude that the variables under
investigation are integrated of order one I (1). Since the variables are I (1) i.e. becoming
stationary at first difference we examined their co integrating relationship using Johansen’s full
information maximum likelihood.
4.2. CO INTEGRATION TEST RESULT
A necessary but insufficient condition for co-integration is that each of the variables are
integrated of the same order I (1) Granger (1986).
Following objective one of our study, we examined the long-run relationship between financial
development and economic growth (RGDP) using four different indicators namely ratio of
money supply to income (M2Y), bank liabilities (BDY), private sector credit (PSCY) and
domestic/bank credit (DCY). To overcome simultaneity bias in VAR we included investment
(INV) and interest rate (R) as ancillary variables. The Johansen co integration test includes both
the trace and the maximum Eigen value statistics. The first row in each of the tables test the
hypothesis of no co integration, the second row test the hypothesis of one co integrating relation
and so on, all against the alternative of full rank of co integration. The results are presented in
tables 3- 6 below:
35
Table 3: Co integration Test Result between RGDP and M2Y
Hypothesized
Trace
0.05
No. of CE(s)
Eigen value
Statistic
Critical Value
Prob.**
None *
0.431905
48.15180
47.85613
0.0469
At most 1
0.209953
22.14033
29.79707
0.2908
At most 2
0.147865
11.29983
15.49471
0.1937
At most 3 *
0.082074
3.939367
3.841466
0.0472
Trace test indicates 1 co integrating eqn(s) at the 0.05 level
Hypothesized
Max-Eigen
0.05
No. of CE(s)
Eigen value
Statistic
Critical Value
Prob.**
None*
0.431905
39.01147
27.58434
0.0384
At most 1
0.209953
10.84049
21.13162
0.6632
At most 2
0.147865
7.360467
14.26460
0.4474
At most 3 *
0.082074
3.939367
3.841466
0.0472
Max-Eigen value test indicates 1 co integrating eqn(s) at the 0.05 level
* denotes rejection of the hypothesis at the 0.05 level
Table 4: Co integration Test Result between RGDP and BDY
Hypothesized
Trace
0.05
No. of CE(s)
Eigen value
Statistic
Critical Value
Prob.**
None *
0.425469
50.28226
47.85613
0.0290
At most 1
0.216329
24.78901
29.79707
0.1691
At most 2
0.158850
13.57577
15.49471
0.0953
At most 3 *
0.114976
5.618464
3.841466
0.0178
Trace test indicates 1 co integrating eqn(s) at the 0.05 level
Hypothesized
Max-Eigen
0.05
No. of CE(s)
Eigen value
Statistic
Critical Value
Prob.**
None*
0.425469
35.49325
27.58434
0.0404
At most 1
0.216329
11.21324
21.13162
0.6260
At most 2
0.158850
7.957306
14.26460
0.3830
At most 3 *
0.114976
5.618464
3.841466
0.0178
Max-Eigen value test indicates 1 co integrating eqn(s) at the 0.05 level
* denotes rejection of the hypothesis at the 0.05 level
36
Table 5: Co integration Test Result between RGDP and
CPSY
Hypothesized
No. of CE(s)
Trace
0.05
Eigen value
Statistic
Critical Value
Prob.**
None *
0.794927
102.5839
47.85613
0.0000
At most 1 *
0.332094
31.28647
29.79707
0.0334
At most 2
0.181745
13.12415
15.49471
0.1103
At most 3 *
0.087043
4.098010
3.841466
0.0429
Trace test indicates 2 co integrating eqn(s) at the 0.05 level
Hypothesized
Max-Eigen
0.05
No. of CE(s)
Eigen value
Statistic
Critical Value
Prob.**
None *
0.794927
71.29742
27.58434
0.0000
At most 1
0.332094
18.16232
21.13162
0.1239
At most 2
0.181745
9.026136
14.26460
0.2840
At most 3 *
0.087043
4.098010
3.841466
0.0429
Max-Eigen value test indicates 1 co integrating eqn(s) at the 0.05 level
* denotes rejection of the hypothesis at the 0.05 level
Table 6: Co integration Test Result between RGDP and DCY
Hypothesized
Trace
0.05
No. of CE(s)
Eigen value
Statistic
Critical Value
Prob.**
None *
0.496994
59.33647
47.85613
0.0029
At most 1
0.277643
27.04032
29.79707
0.1006
At most 2
0.128834
11.75421
15.49471
0.1691
At most 3 *
0.106105
5.271856
3.841466
0.0217
Trace test indicates 1 co integrating eqn(s) at the 0.05 level
Hypothesized
Max-Eigen
0.05
No. of CE(s)
Eigen value
Statistic
Critical Value
Prob.**
None *
0.496994
32.29616
27.58434
0.0115
At most 1
0.277643
15.28611
21.13162
0.2693
At most 2
0.128834
6.482353
14.26460
0.5521
At most 3 *
0.106105
5.271856
3.841466
0.0217
Max-Eigen value test indicates 1 co integrating eqn(s) at the 0.05 level
* denotes rejection of the hypothesis at the 0.05 level
37
In tables 3-6, the trace and the maximum Eigen value statistics indicates the presence of
one co integrating equation at 5 percent significance level which implies that economic growth
(RGDP) and the measures of financial development (M2Y, BDY, CPSY and DCY) are co
integrated. In addition, the result also revealed that economic growth (RGDP), investment (INV)
and interest rates (INV) are also co integrated. The Johansen co integration result above confirms
the rejection of the null hypothesis of no co integration and the acceptance of the alternative of
co integration i.e. a long run relationship exist.
Hence, the test suggests the existence of a stable long-run relationship between economic growth
proxied by real GDP per capita (RGDP) and financial development variables proxied by M2Y,
BDY, CPSY and DCY.
The long-run relationship between economic growth and financial development was found to be
positive in each of the co integrating vectors suggesting causality in at least one direction.
4.3. VECTOR ERROR CORRECTION MODEL RESULTS
Since there is co integration, the direction of causality is tested using the vector error
correction model (VECM). This is done using Johansen co integrating vectors. The results are
presented in table 7 below:
38
TABLE 7: VECM RESULTS
7a. Variables included in the VAR: RGDP and M2Y, INV, R
Variables
β
ECM
RGDP
1.0000
-0.064657
(-4.81437)
M2Y
-19.63721
(-9.24983)
INV
-20.10934
(-5.3005)
R
137648.9
(4.43941)
C
-2596956
7b. Variables included in the VAR: RGDP and BDY, INV, R
Variables
β
ECM
RGDP
1.0000
-0.060944
(-4.23222
BDY
-17.8198
(-6.7649)
INV
-23.7368
(-5.30513)
R
169114.7
(4.51923)
C
-2497008
39
7c. Variables included in the VAR: RGDP and CPSY, INV, R
Variables
β
ECM
RGDP
1.0000
-0.309626
(-2.53376)
CPSY
-29.936
(-3.36881)
INV
-5.46607
(-5.13752)
R
66334.87
(6.73364)
C
-287736
7d. Variables in the VAR: RGDP and DCY, INV, R
Variables
β
ECM
RGDP
1.0000
-0.013196
(-9.0384)
DCY
-36135.7
(-1.7507)*
INV
-32.7070
(-7.2775)
R
181821.2
(4.0291)
C
-1357239
Notes * indicate significance at the 10%, level
The co integrating vector is normalized on RGDP.
The t-statistics are in parentheses
The signs of the coefficients (β) changes since they still have to be taken to the right hand side of the equation. The
positive sign turns to negative while the negative turns to positive.
40
From table 7, we can formally state the normalized long-run co integrating relationship involving
financial development and Economic growth in the equation form below where the sign changes:
RGDP = 2596956 + 19.63721M2Y + 20.10934INV – 1376489R
(1)
RGDP = 2497008+ 17.8198BDY + 23.7368INV – 169114.7R
(2)
RGDP = 287736 + 29.936 CPSY + 5.4660 INV – 66334.87R
(3)
RGDP = 1357239 + 36135.7DCY + 32.7070INV-181821.2R
(4)
From equations 1 and in table 7(a) the VECM result shows that there is a significant positive
long-run relationship between financial development proxied by M2Y and economic growth
(RGDP) implying that an increase in M2Y will lead to an increase in RGDP.
From the result a unit increase in M2Y will lead to a 19.64 units increase in RGDP. This finding
is consistent with a prior economic expectation. The t-statistics was also significant.
Investment (INV) had a significant positive impact on RGDP with an estimated coefficient of
20.11 suggesting that an increase in INV will lead to a corresponding increase in RGDP. This
conforms to economic theory
Interest rate (R) on the other hand has an indirect relationship with RGDP with an estimated
coefficient of -1376489. This means that a rise in interest rate would lead to fall in economic
growth which is in line economic theory.
Similar results were obtained in 7(b), (c) and (d) when BDY, CPSY and DCY were used as
financial development indicators respectively.
Each of the financial development proxies exerted a significant positive impact on RGDP. This
also is consistent with a priori economic expectation.
41
Precisely, a unit increase in BDY, CPSY and DCY led to a 17.82, 29.94 and 36135.7 units
increase in RGDP respectively.
Investment (INV) and interest rate (R) as in 7(a) had a significant positive and negative influence
on real GDP respectively.
For the error correction term (ECM) the VECM result distinguishes between short-run
and long-run granger causality. The coefficient of the lagged error correction term (ECM) in 7(a)
is negative and significant.
This significance shows that there is a long-run causal relationship between economic growth
and measures of financial development. It also indicates that M2Y and RGDP are adjusting to
their long-run equilibrium relationships. The negative coefficient and the magnitude of the ECM
indicate the speed of adjustment to the long-run equilibrium relationship which in 7(a) is 6.4
percent.
The coefficients of the lagged error terms in 7(b), (c) (d) are also negative and significant
implying that each measure of financial development and economic growth are adjusting to their
long-run equilibrium relationships, the speeds of adjustment are 6.1, 30.9 and 1.32 percent
respectively.
The results of the granger causality test based on the VECM representations of the co
integrated variables are presented in table 8 below.
The null hypothesis of no direction of causality is tested against the alternative that there exists a
direction of causality amongst the variables in question.
42
In our case there are four possibilities
1. Unidirectional causality from financial development (FD) to Economic Growth (EG)
when the FD coefficient is statistically significant.
2. Unidirectional causality from EG to FD, when the EG coefficient is statistically
significant
3. Feedback or bilateral causality when the sets of FD and EG coefficients are statistically
significant.
4. Finally, independence, when FD & EG coefficients are not statistically significant.
Table 8 Granger causality test result
Cause and effect (Ho)
Chi-sq
P-value
Conclusion
RGDP does not granger cause M2Y
18.912
0.0008
Reject Ho
M2Y does not granger cause RGDP
2.373
0.3052
Do not reject Ho
RGDP does not granger cause BDY
11.925
0.0027
Reject Ho
BDY does not granger cause RGDP
3.643
0.1618
Do not reject Ho
RGDP does not granger cause CPSY
0.4557
0.9285
Do not reject Ho
CPSY does not granger cause RGDP
9.798
0.0499
Reject Ho
RGDP does not granger cause DCY
4.785
0.0287
Reject Ho
DCY does not granger cause RGDP
0.192
0.6611
Do not reject Ho
From table 8, the causality test revealed that economic growth proxied by real GDP per capita
causes M2Y, BDY and DCY without a feedback.
This implies a unidirectional causality running from economic growth to financial development,
the conclusion was arrived based on the fact that their chi-sqs were statistically significant at less
than 5% indicated by their p-values.
43
These three outcomes support the demand following view otherwise known as the growth-led
finance hypothesis which says that as the real side of the economy expands, its demand for
certain financial instruments, arrangements and markets increases, leading to the growth of these
services.
Finally, the causality test also showed that financial development caused economic growth when
proxied by the ratio of private sector credit to nominal GDP; its p-value was less than 5%. This
finding of unidirectional causality from finance to growth supports the supply leading view of
finance led growth hypothesis which posits that there is a robust effect that runs from financial
intermediation to economic growth exercised either by raising the efficiency of capital
accumulation, savings rate and thus investment rate
4.4 IMPULSE RESPONSE FUNCTION
The impulse response functions (IRF) traces out the response of the dependent variable in
the VAR system to shocks in the error terms for over a 10 year period.
The impulse response graph (see appendix 3) represents the various responses of economic
growth proxied by real GDP per capita (RGDP) to a one standard deviation (0.25 percentage
point) positive shock on the financial development variables proxied by M2Y, BDY, CPSY and
DCY.
The empirical evidence shows that an increase in M2Y by 0.25% points will leave the RGDP
unaffected in the first period. However, in the second period, the effect of shock from M2Y taps
off and persisted throughout the sample period.
44
The response of RGDP to the shocks from BDY is very much similar to its response to M2Y
shocks. Shocks resulting from BDY had permanent positive effect on RGDP throughout the
entire period. However shocks from BDY indicate much stronger effect on RGDP than M2Y.
The graph further showed that DCY shocks stimulate RGDP more in the short run than M2Y and
BDY. In the long run (towards the end of the period) shocks to M2y and BDY are slightly higher
on GDP.
The behaviour of GDP due to shocks from CPSY is quite different from M2Y, BDY, and DCY.
Shocks to CPSY exerted no impact or effect on RGDP from first through the 4 th period. At the
5th period, the effect of CPSY shocks on GDP taps off rather slowly till the end of the sample
period with little fluctuations in some periods.
From the analysis above we conclude that all the financial development indicators; M2Y, BDY,
DCY and CPSY have a measurable positive effect on economic growth in Nigeria.
45
CHAPTER FIVE
SUMMARY OF FINDINGS AND RECOMMENDATIONS
5.1 SUMMARY OF FINDINGS
The focus of this research project was to investigate the link between financial
development and economic growth in Nigeria from 1960-2008. Specifically the study examined
the relationship between financial development and economic growth and its direction of
causality.
Given the nature of the objectives, the vector autoregressive model (VAR) which was
later transformed into a vector error correction model (VECM) was utilized. In other to see the
impact of various aspects of financial development, the study employed four different measures
of financial development namely the ratio of broad money supply (M2) to income (M2Y), the
ratio of total bank deposit liabilities to nominal GDP (BDY), the ratio of private sector credit to
GDP (CPSY) and the ratio of Domestic/Bank credit to GDP (DCY). To avoid misspecification
and simultaneity bias in VAR, Investment (INV) and Interest rate (R) were included as ancillary
variables.
Using each of the four financial development indicators we ran four different VAR
models. The summary of the major empirical findings are stated below: The Johansen co integration test result showed a one co integrating equation/vector
between Real GDP per capita and measures of financial development. The implication of
the above is that there exist a stable long-run equilibrium relationship between economic
growth proxied by RGDP and financial development variables proxied by M2Y, BDY,
CPSY, and DCY.
46
 The long-run relationship between economic growth and financial development was also
found to be positive in each co integrating vector.
 The estimated co integrating vectors within the context of VECM suggest that real
economic activity is affected by changes in financial development. The magnitude of the
estimated coefficients shows that financial development contributes significantly to
determining the magnitude of real economic growth in the long-run.
 The estimated Co integrating relationship further revealed that investment exerted a
positive influence on real GDP implying that an increase in stock of capital is associated
with a corresponding increase real GDP.
 Interest rate on the other hand had a negative impact on real GDP suggesting that
increase in interest rate affects GDP negatively while lower interest rates increases
growth.
 The result from the impulse response functions/graphs also confirms the positive impact
of the various financial development indicators on growth beyond the sample period.
 Error correction mechanism that measures the speed of adjustment to equilibrium has the
expected negative relationship with the dependent variable in all the models. This means
that any presence of error in the model will be corrected
Given that correlation does not imply causation in any way, the study carried out a granger
causality test within co integration and vector error correction mechanisms to determine the
direction of causality between financial development and economic growth in Nigeria.
47
The findings showed that:
 There is unidirectional causality between financial development and economic growth
with economic growth causing financial development when M2Y, BDY and DCY were
used as proxies.
 On the other hand financial development caused economic growth when credit to private
sector (CPSY) was used as proxy thus favouring the supply leading view.
In general the empirical results show that the direction of causality between financial
development and economic growth is sensitive to the choice of proxy used.
5.2 POLICY RECOMMENDATIONS
This result obtained in the study suggests that financial development has a positive
impact on Nigeria’s economic growth in the long-run. Thus
1. If policy makers want to promote sustainable growth then attention should be focused on longrun policies for example the creation of modern, efficient and strong financial institutions that
can drive Nigeria’s economic growth.
2. To achieve the above desired benefits efforts should be devoted to
 Deepening the financial sector especially the microfinance system in Nigeria.
 Legal reforms to fast track markets and institutions for efficient credit system
 Strengthening of supervisory and regulatory bodies in the financial system.
 Enhancing competition, investing in human resources and the legal environment on the
one hand and to improving the quality of the institutions on the other hand.
3. Since high but sustainable economic growth leads financial development in Nigeria, there is
need to address the decay in the critical infrastructures - power, transport, water etc as this will
reduce the cost of funds for banks and also deepen and sustain the momentum for growth.
48
REFERENCES
Abdul Jalil and Ying Ma (2008), “Financial Development and Economic Growth: Time Series
Evidence from Pakistan and China”. Journal of Economic Co-Operation 29(2): 29-68.
Agu C. Cletus and Chukwu Jude (2008), “Toda and Yamamoto Causality Test between Bank
Based Financial Deepening and Economic Growth in Nigeria”. European Journal of
Social Sciences Vol 7 No 2.
Acaravci, A; I. Ozturk and S. K (2007), “Finance-Growth Nexus: Evidence from Turkey”.
International Research Journal of Finance and Economics ISSN 1450-2887 Issue 11
Aigbokhan, K. Ben (1995), “Financial Development and Economic Growth: A Test of
Hypothesis on Supply Leading and Demand Following with Evidence on Nigeria”. The
Nigerian Economic and Financial Review NEFR Vol 1, No 2, Dec 1995.
Arestis, P. and Demetriades, P. (1997), “Financial Development and Economic Growth:
Assessing the Evidence”. Economic Journal 107: 783-799.
Arestis, P., Demetriades P., O, and K. B. Luintel (2001), “Financial Development and Economic
Growth: The Role of Stock Markets”. Journal of Money, Credit, and Banking, 33: 16-41.
Atje, R., and Jovanovic B. (1993), "Stock Markets and Development”. European Economic
Review, 37: 632 640.
Bagehot, Walter (1873), Lombard Street. Homewood, IL: Richard D. Irwin, 1962 edition.
Beck, T., Levine, R., Loyaza, N., (2000), “Finance and the sources of growth”. Journal of
Financial Economics 58: 261– 300.
Bencivenga, V. and Smith B. (1991), “Financial Intermediation and Endogenous Growth”.
Review of Economic Studies 58 (2):195-209.
Bencivenga, V. R., Smith, B. D., and Starr R., M. (1995), “Transactions Costs, Technological
Choice and Endogenous Growth”. Journal of Economic Theory, 67: 53-177.
Blackburn, K. and Hung, V., T. (1998), “A Theory of Growth, Financial Development, and
Trade”. Economica, 65: 107-24.
49
Chakraborty Indrani (2008), “Does Financial Development Cause Economic Growth? The case
of India”. South Asia Economic Journal 9(1):109-139.
Chisti, Hasan, U.S.A and Mahmud S.F (1992), “Macroeconometric modeling and Pakistan’s
Economy: A vector autoregression approach”.Journal of Development Economics, 38:
353-370
Christopoulos, D. K. and Tsionas E. G. (2004), “Financial Development and Economic Growth:
Evidence from Panel Unit Root and Cointegration Tests”. Journal of Development
Economics, 73: 55-74.
Creane, S., Goyal, R., Mushfiq M, and Randa, S., (2004), “Financial Sector Development in the
Middle East and North Africa”. IMF Working Paper 04/201, Washington D.C
Darrat, A., (1999), “Are financial deepening and economic growth causally related: Another
look at the evidence”. International Economic Journal, 13:19–35
De Gregorio, J., and Guidotti, P. (1995), “Financial Development and Economic Growth”. World
Development; 23: 433-448
Deidda, L. and Fattouh B. (2002), “Non-Linearity between Finance and Growth”. Economics
Letters 74: 339-345.
Demetriades, P. and Hussein K. (1996), “Does Financial Development Cause Economic
Growth? Time Series Evidence from 16 Countries”. Journal of Development Economics,
51: 387-411.
Demetriades, Panicos O., and Luintel, K., (1996), “Financial Development, Economic Growth
and Banking Sector Controls: Evidence from India”. The Economic Journal 106: 359374.
Eita, J.H and Andre C. J (2007), “A Causality Analysis between Financial Development and
Economic Growth for Botswana”. Working Paper 22, Dept of Economics, University of
Pretoria South Africa
Engle, R.F, and Granger, C.W.J. (1987), “Co-integration and Error Correction: Representation,
Estimation and Testing”, Econometrica, 55:251-276
50
Fritz, Richard G. (1984), “Time Series Evidence on the Causal Relationship between Financial
Deepening and Economic Development”. Journal of Economic Development, July: 91112
Goldsmith, R. W. (1969), Financial Structure and Development. New Haven, CT: Yale
University Press.
Greenwood, J. and Jovanovic, B. (1990), “Financial Development, Growth, and the Distribution
of Income”. Journal of Political Economy 98 (5):1076-1107.
Greenwood, J., and Smith, B. (1997), “Financial Markets in Development and the Development
of Financial Market”. Journal of Economic Dynamic and Control 21(1): 145-181.
Gujarati N. D (2003) Basic Econometrics. New York: McGraw, Inc, 4th Edition
Guryay E.; O.V Safakli and B. Tuzel (2007), “Financial Development and Economic Growth:
Evidence from Northern Cyprus”. International Research Journal of Finance and
Economics ISSN 1450-2887 Issue 8
Harris, R. D.F. (1997). “Stock Markets and Development: A Re-Assessment”. European
Economic Review 41:139-146.
Hicks, J. (1969), A Theory of Economic History. Oxford: Clarendon Press.
Ireland, P. N. (1994), “Money and Growth: An Alternative Approach”. American Economic
Review 84 (1): 47–65.
Johansen, S. (1988), “Statistical Analysis of Cointegration Vectors”, Journal of Ecoomic
Dynamics and Control, 12: 231-254
Jung, W. S. (1986), “Financial Development and Economic Growth: International Evidence”.
Economic Development and Cultural Change, 34: 333-346.
Khan, A. M; Abdul Q; Saeed A. (2005), “Financial Development and Economic Growth: The
Case of Pakistan”. The Pakistan Development Review, (4)819-837.
King, R. G. and Levine R. (1993a), “Finance and Growth: Schumpeter Might Be Right”.
Quarterly Journal of Economics 108: 717-738.
51
King, R. G. and Levine R. (1993b), “Finance, Entrepreneurship, and Growth: Theory and
Evidence”. Journal of Monetary Economics 32: 513-542.
Kuznets, S. (1955), “Economic Growth and Income Inequality”. The American Economic
Review 45 (1):1-28.
Levine, R. (1991), “Stock Markets, Growth, and Tax Policy”. The Journal of Finance, 46:144565.
Levine, R. (1997), “Financial Development and Economic Growth: Views and Agenda”. Journal
of Economic Literature, 35: 688-726.
Levine, R. and Zervos S. (1998), “Stock Markets, Banks, and Economic Growth”. American
Economic Review, 88: 537-558.
Levine, R., Loayza N. and Beck T. (2000), “Financial Intermediation and Growth: Causality and
Causes”. Journal of Monetary Economics, 46: 31-77.
Levine, R., Zervos, S., (1996), “Stock market development and long-run growth”. World Bank
Economic Review 10: 323–339.
Lucas, R. E. (1988), “On the Mechanics of Economic Development”. Journal of Monetary
Economics (22): 3-42.
Luintel, B. K., and Khan, M. (1999), “A Quantitative Re-assessment of the Finance-Growth
Nexus: Evidence from a Multivariate VAR”. Journal of Development Economics 60:
381–405.
Mackinnon, R. I. (1973) Money and Capital in Economic Development. Washington, D. C:
Brooking Institution
Majid, Shabi A (2008), Does Financial Development Matter for Economic Growth in Malaysia?
An ARDL Bound Testing Approach”. Journal of Economic Co-Operation 29(1):61-82.
Murinde, V. and Eng, F.S.H. (1994), “Financial Development and Economic Growth in
Singapore: Demand-Following or Supply-Leading”. Applied Financial Economics 4(6):
391-404.
Ndebbio, J.E.U (2004), “Financial Deepening, Economic Growth and Development: Evidence
from Selected SSA Countries”. AERC Research Paper 142
52
Obstfeld, M. (1994), "Risk-Taking, Global Diversification and Growth”. American Economic
Review, 84: 1310-1329.
Odedokun, M. O. (1996), “Alternative Econometric Approaches for Analyzing the Role of the
Financial Sector in Economic Growth: Time Series Evidence from LDCs”. Journal of
Development Economics, 50: 119-146.
Patrick, H. T. (1966), “Financial Development and Economic Growth in Underdeveloped
Countries”. Economic Development and Cultural Change 14: 174-189.
Patterson, K (2000), An introduction to Applied Econometrics: A Time Series Approach.
Palgrave, 599-707
Rajan, R. G. and Zingales L. (1998), “Financial Dependence and Growth”. American Economic
Review, 88: 559-586.
Ram, Rati (1999). “Financial Development and Economic Growth”. The Journal of Development
Studies, 35: 164-174.
Robinson, J. (1952), “The Generalization of the General Theory”, In: the Rate of Interest and
Other Essays, London: MacMillan.
Roubini, N. and Sala-i-Martin X. (1992), “Financial Repression and Economic Growth”. Journal
of Development Economics, 39: 5-30.
Rousseau, P., and Wachtel P. (1998), “Financial Intermediation and Economic Performance:
Historical Evidence from Five Industrialized Countries”. Journal of Money, Credit, and
Banking, 30: 657-678.
Saint-Paul, G. (1992), “Technological Choice, Financial Markets and Economic Development”.
European Economic Review, 36: 763-781.
Schumpeter, J. A, (1911) The Theory of Economic Development. Cambridge, MA: Harvard
University Press.
Shan, J. (2003), “Financial Development and Economic Growth: The Empirical Evidence from
China”. Proceedings of the 15th Annual Conference of the Association for Chinese
Economics Studies Australia
53
Shaw E. S, (1973), Financial Deepening in Economic Development New York: Oxford
University Press.
Singh, A. (1997), “Financial liberalization, stock markets and economic development”.
Economic Journal (107): 771– 782.
Suleiman, A. and Aamer, A. (2006), “Financial Development and Economic Growth Nexus:
Time Series Evidence from Middle Eastern and North African Countries”. MPRA Paper
No 972. Http: mpra.ub.uni-muenchen.de/972/
Tsuru, K. (2000). “Finance and Growth: Some Theoretical Considerations and A Review of the
Empirical Literature”. Economics Department Working Papers, OECD, January, No.
228.
54
APPENDIX 1
UNIT ROOTS TEST RESULTS
UNIT ROOTS TEST: RGDP (LEVEL FORM)
Null Hypothesis: RGDP has a unit root
Exogenous: Constant
Lag Length: 0 (Automatic based on AIC, MAXLAG=10)
t-Statistic
Prob.*
Augmented Dickey-Fuller test statistic
-1.408894
0.7988
Test critical values:
1% level
-3.574446
5% level
-2.923780
10% level
-2.599925
*MacKinnon (1996) one-sided p-values.
Augmented Dickey-Fuller Test Equation
Dependent Variable: D(RGDP)
Method: Least Squares
Date: 04/22/10 Time: 18:31
Sample (adjusted): 1961 2008
Included observations: 48 after adjustments
Variable
Coefficient
Std. Error
t-Statistic
Prob.
RGDP(-1)
-0.012961
0.009200
-1.408894
0.1656
C
0.042973
0.109976
0.390747
0.6978
R-squared
0.041367
Mean dependent var
0.193247
Adjusted R-squared
0.020527
S.D. dependent var
0.187600
S.E. of regression
0.185665
Akaike info criterion
4.488974
Sum squared resid
1.585686
Schwarz criterion
4.411007
Log likelihood
13.73537
F-statistic
1.984983
Durbin-Watson stat
1.625771
Prob(F-statistic)
0.165591
UNIT ROOTS TEST: RGDP (1ST DIFF)
Null Hypothesis: D(RGDP) has a unit root
Exogenous: Constant
Lag Length: 0 (Automatic based on AIC, MAXLAG=10)
t-Statistic
Prob.*
Augmented Dickey-Fuller test statistic
-5.366580
0.0000
Test critical values:
1% level
-3.577723
5% level
-2.925169
10% level
-2.600658
55
*MacKinnon (1996) one-sided p-values.
Augmented Dickey-Fuller Test Equation
Dependent Variable: D(LOG_RGDP,2)
Method: Least Squares
Date: 04/22/10 Time: 18:35
Sample (adjusted): 1962 2008
Included observations: 47 after adjustments
Variable
Coefficient
Std. Error
t-Statistic
Prob.
D(RGDP(-1))
-0.775387
0.144484
-5.366580
0.0000
C
0.152527
0.039010
3.909988
0.0003
R-squared
0.390245
Mean dependent var
0.001863
Adjusted R-squared
0.376695
S.D. dependent var
0.235190
S.E. of regression
0.185682
Akaike info criterion
4.487941
Sum squared resid
1.551501
Schwarz criterion
4.409212
Log likelihood
13.46662
F-statistic
28.80019
Durbin-Watson stat
1.995262
Prob(F-statistic)
0.000003
UNIT ROOTS TEST: M2Y (LEVEL FORM)
Null Hypothesis: M2Y has a unit root
Exogenous: Constant
Lag Length: 0 (Automatic based on AIC, MAXLAG=10)
t-Statistic
Prob.*
Augmented Dickey-Fuller test statistic
-1.604451
0.4726
Test critical values:
1% level
-3.574446
5% level
-2.923780
10% level
-2.599925
*MacKinnon (1996) one-sided p-values.
Augmented Dickey-Fuller Test Equation
Dependent Variable: D(M2Y)
Method: Least Squares
Date: 04/19/10 Time: 16:32
Sample (adjusted): 1961 2008
Included observations: 48 after adjustments
Variable
Coefficient
Std. Error
t-Statistic
Prob.
M2Y(-1)
-0.122340
0.076250
-1.604451
0.1155
C
3.347413
1.845735
1.813594
0.0763
56
R-squared
0.052996
Mean dependent var
0.545833
Adjusted R-squared
0.032409
S.D. dependent var
4.212846
S.E. of regression
4.144015
Akaike info criterion
5.721981
789.9517
Schwarz criterion
5.799948
F-statistic
2.574263
Prob(F-statistic)
0.115459
Sum squared resid
Log likelihood
-135.3275
Durbin-Watson stat
1.698468
UNIT ROOTS TEST: M2Y (1ST DIFF)
Null Hypothesis: D(M2Y) has a unit root
Exogenous: Constant
Lag Length: 0 (Automatic based on AIC, MAXLAG=10)
t-Statistic
Prob.*
Augmented Dickey-Fuller test statistic
-6.004550
0.0000
Test critical values:
1% level
-3.577723
5% level
-2.925169
10% level
-2.600658
*MacKinnon (1996) one-sided p-values.
Augmented Dickey-Fuller Test Equation
Dependent Variable: D(M2Y,2)
Method: Least Squares
Date: 04/19/10 Time: 16:38
Sample (adjusted): 1962 2008
Included observations: 47 after adjustments
Variable
Coefficient
Std. Error
t-Statistic
Prob.
D(M2Y(-1))
-0.951215
0.158416
-6.004550
0.0000
C
0.536688
0.629592
0.852437
0.3985
R-squared
0.444819
Mean dependent var
0.214894
Adjusted R-squared
0.432481
S.D. dependent var
5.708726
S.E. of regression
4.300602
Akaike info criterion
5.797008
Sum squared resid
832.2830
Schwarz criterion
5.875738
F-statistic
36.05462
Prob(F-statistic)
0.000000
Log likelihood
Durbin-Watson stat
-134.2297
1.896214
57
UNIT ROOTS TEST: BDY (LEVEL FORM)
Null Hypothesis: BDY has a unit root
Exogenous: Constant
Lag Length: 0 (Automatic based on AIC, MAXLAG=10)
t-Statistic
Prob.*
Augmented Dickey-Fuller test statistic
-1.011906
0.7418
Test critical values:
1% level
-3.574446
5% level
-2.923780
10% level
-2.599925
*MacKinnon (1996) one-sided p-values.
Augmented Dickey-Fuller Test Equation
Dependent Variable: D(BDY)
Method: Least Squares
Date: 04/19/10 Time: 16:40
Sample (adjusted): 1961 2008
Included observations: 48 after adjustments
Variable
Coefficient
Std. Error
t-Statistic
Prob.
BDY(-1)
-0.068431
0.067626
-1.011906
0.3169
C
1.743543
1.233150
1.413894
0.1641
R-squared
0.021775
Mean dependent var
0.595000
Adjusted R-squared
0.000509
S.D. dependent var
3.340558
S.E. of regression
3.339707
Akaike info criterion
5.290417
Sum squared resid
513.0676
Schwarz criterion
5.368384
F-statistic
1.023954
Prob(F-statistic)
0.316876
Log likelihood
Durbin-Watson stat
-124.9700
1.543876
UNIT ROOTS TEST: BDY (1ST DIFF)
Null Hypothesis: D(BDY) has a unit root
Exogenous: Constant
Lag Length: 0 (Automatic based on AIC, MAXLAG=10)
t-Statistic
Prob.*
Augmented Dickey-Fuller test statistic
-5.296887
0.0001
Test critical values:
1% level
-3.577723
5% level
-2.925169
10% level
-2.600658
*MacKinnon (1996) one-sided p-values.
58
Augmented Dickey-Fuller Test Equation
Dependent Variable: D(BDY,2)
Method: Least Squares
Date: 04/19/10 Time: 16:40
Sample (adjusted): 1962 2008
Included observations: 47 after adjustments
Variable
Coefficient
Std. Error
t-Statistic
Prob.
D(BDY(-1))
-0.864138
0.163141
-5.296887
0.0000
C
0.545719
0.498285
1.095193
0.2793
R-squared
0.384043
Mean dependent var
0.207234
Adjusted R-squared
0.370355
S.D. dependent var
4.269511
S.E. of regression
3.387865
Akaike info criterion
5.319898
516.4933
Schwarz criterion
5.398628
F-statistic
28.05701
Prob(F-statistic)
0.000003
Sum squared resid
Log likelihood
-123.0176
Durbin-Watson stat
1.881855
UNIT ROOTS TEST: CPSY (LEVEL FORM)
Null Hypothesis: CPSY has a unit root
Exogenous: Constant
Lag Length: 0 (Automatic based on AIC, MAXLAG=10)
t-Statistic
Prob.*
Augmented Dickey-Fuller test statistic
-1.348899
0.5992
Test critical values:
1% level
-3.574446
5% level
-2.923780
10% level
-2.599925
*MacKinnon (1996) one-sided p-values.
Augmented Dickey-Fuller Test Equation
Dependent Variable: D(CPSY)
Method: Least Squares
Date: 04/19/10 Time: 16:43
Sample (adjusted): 1961 2008
Included observations: 48 after adjustments
Variable
Coefficient
Std. Error
t-Statistic
Prob.
CPSY(-1)
-0.127564
0.094569
-1.348899
0.1840
C
2.323394
1.405396
1.653195
0.1051
R-squared
0.038050
Mean dependent var
0.602083
Adjusted R-squared
0.017138
S.D. dependent var
4.115073
59
S.E. of regression
Sum squared resid
Log likelihood
4.079659
Akaike info criterion
5.690677
765.6063
Schwarz criterion
5.768644
F-statistic
1.819527
Prob(F-statistic)
0.183973
-134.5763
Durbin-Watson stat
1.709853
UNIT ROOTS TEST: CPSY (1ST DIFF)
Null Hypothesis: D(CPSY) has a unit root
Exogenous: Constant
Lag Length: 1 (Automatic based on AIC, MAXLAG=10)
t-Statistic
Prob.*
Augmented Dickey-Fuller test statistic
-5.901828
0.0000
Test critical values:
1% level
-3.581152
5% level
-2.926622
10% level
-2.601424
*MacKinnon (1996) one-sided p-values.
Augmented Dickey-Fuller Test Equation
Dependent Variable: D(CPSY,2)
Method: Least Squares
Date: 04/19/10 Time: 16:43
Sample (adjusted): 1963 2008
Included observations: 46 after adjustments
Variable
Coefficient
Std. Error
t-Statistic
Prob.
D(CPSY(-1))
-1.403120
0.237743
-5.901828
0.0000
D(CPSY(-1),2)
0.391469
0.166329
2.353585
0.0232
C
0.687294
0.601315
1.142986
0.2594
R-squared
0.521645
Mean dependent var
0.186957
Adjusted R-squared
0.499396
S.D. dependent var
5.719969
S.E. of regression
4.047071
Akaike info criterion
5.696857
704.2876
Schwarz criterion
5.816116
F-statistic
23.44572
Prob(F-statistic)
0.000000
Sum squared resid
Log likelihood
Durbin-Watson stat
\
-128.0277
1.919191
60
UNIT ROOTS TEST: DCY (LEVEL FORM)
Null Hypothesis: DCY has a unit root
Exogenous: Constant
Lag Length: 1 (Automatic based on AIC, MAXLAG=10)
t-Statistic
Prob.*
Augmented Dickey-Fuller test statistic
-2.483480
0.1259
Test critical values:
1% level
-3.577723
5% level
-2.925169
10% level
-2.600658
*MacKinnon (1996) one-sided p-values.
Augmented Dickey-Fuller Test Equation
Dependent Variable: D(DCY)
Method: Least Squares
Date: 04/19/10 Time: 16:49
Sample (adjusted): 1962 2008
Included observations: 47 after adjustments
Variable
Coefficient
Std. Error
t-Statistic
Prob.
DCY(-1)
-0.201293
0.081053
-2.483480
0.0169
D(DCY(-1))
0.217699
0.146402
1.486998
0.1441
C
4.769768
2.118166
2.251839
0.0294
R-squared
0.136189
Mean dependent var
0.312340
Adjusted R-squared
0.096925
S.D. dependent var
8.017347
S.E. of regression
7.618905
Akaike info criterion
6.960844
Sum squared resid
2554.099
Schwarz criterion
7.078938
F-statistic
3.468544
Prob(F-statistic)
0.039922
Log likelihood
Durbin-Watson stat
-160.5798
2.026942
UNIT ROOTS TEST: DCY (1ST DIFF)
Null Hypothesis: D(DCY) has a unit root
Exogenous: Constant
Lag Length: 0 (Automatic based on AIC, MAXLAG=10)
t-Statistic
Prob.*
Augmented Dickey-Fuller test statistic
-5.864103
0.0000
Test critical values:
1% level
-3.577723
5% level
-2.925169
10% level
-2.600658
61
*MacKinnon (1996) one-sided p-values.
Augmented Dickey-Fuller Test Equation
Dependent Variable: D(DCY,2)
Method: Least Squares
Date: 04/19/10 Time: 16:50
Sample (adjusted): 1962 2008
Included observations: 47 after adjustments
Variable
Coefficient
Std. Error
t-Statistic
Prob.
D(DCY(-1))
-0.875911
0.149368
-5.864103
0.0000
C
0.291905
1.173667
0.248712
0.8047
R-squared
0.433162
Mean dependent var
0.147660
Adjusted R-squared
0.420565
S.D. dependent var
10.56808
S.E. of regression
8.044485
Akaike info criterion
7.049472
Sum squared resid
2912.118
Schwarz criterion
7.128202
F-statistic
34.38770
Prob(F-statistic)
0.000000
Log likelihood
-163.6626
Durbin-Watson stat
1.973452
UNIT ROOTS TEST: INV (LEVEL FORM)
Null Hypothesis: INV has a unit root
Exogenous: Constant
Lag Length: 2 (Automatic based on AIC, MAXLAG=10)
t-Statistic
Prob.*
Augmented Dickey-Fuller test statistic
-1.426917
0.5610
Test critical values:
1% level
-3.581152
5% level
-2.926622
10% level
-2.601424
*MacKinnon (1996) one-sided p-values.
Augmented Dickey-Fuller Test Equation
Dependent Variable: D(INV)
Method: Least Squares
Date: 04/22/10 Time: 18:20
Sample (adjusted): 1963 2008
Included observations: 46 after adjustments
Variable
Coefficient
Std. Error
t-Statistic
Prob.
INV(-1)
-0.176320
0.123567
-1.426917
0.1610
D(INV(-1))
-0.310495
0.173855
-1.785945
0.0813
D(INV(-2))
-0.238751
0.163437
-1.460812
0.1515
C
1.593039
1.137438
1.400550
0.1687
62
R-squared
0.196503
Mean dependent var
0.074313
Adjusted R-squared
0.139110
S.D. dependent var
2.399217
S.E. of regression
2.226093
Akaike info criterion
4.521315
208.1306
Schwarz criterion
4.680327
F-statistic
3.423826
Prob(F-statistic)
0.025614
Sum squared resid
Log likelihood
-99.99024
Durbin-Watson stat
1.898411
UNIT ROOTS TEST: INV (1ST DIFF)
Null Hypothesis: D(INV) has a unit root
Exogenous: Constant
Lag Length: 1 (Automatic based on AIC, MAXLAG=10)
t-Statistic
Prob.*
Augmented Dickey-Fuller test statistic
-6.628196
0.0000
Test critical values:
1% level
-3.581152
5% level
-2.926622
10% level
-2.601424
*MacKinnon (1996) one-sided p-values.
Augmented Dickey-Fuller Test Equation
Dependent Variable: D(INV,2)
Method: Least Squares
Date: 04/22/10 Time: 18:21
Sample (adjusted): 1963 2008
Included observations: 46 after adjustments
Variable
Coefficient
Std. Error
t-Statistic
Prob.
D(INV(-1))
-1.717120
0.259063
-6.628196
0.0000
D(INV(-1),2)
0.301414
0.159312
1.891976
0.0652
C
0.039153
0.332391
0.117792
0.9068
R-squared
0.650397
Mean dependent var
0.124530
Adjusted R-squared
0.634137
S.D. dependent var
3.724381
S.E. of regression
2.252753
Akaike info criterion
4.525176
218.2205
Schwarz criterion
4.644436
F-statistic
39.99837
Prob(F-statistic)
0.000000
Sum squared resid
Log likelihood
Durbin-Watson stat
-101.0791
1.950090
63
UNIT ROOTS TEST: R (LEVEL FORM)
Null Hypothesis: R has a unit root
Exogenous: Constant
Lag Length: 2 (Automatic based on AIC, MAXLAG=10)
t-Statistic
Prob.*
Augmented Dickey-Fuller test statistic
-1.128923
0.6965
Test critical values:
1% level
-3.581152
5% level
-2.926622
10% level
-2.601424
*MacKinnon (1996) one-sided p-values.
Augmented Dickey-Fuller Test Equation
Dependent Variable: D(R)
Method: Least Squares
Date: 04/19/10 Time: 16:59
Sample (adjusted): 1963 2008
Included observations: 46 after adjustments
Variable
Coefficient
Std. Error
t-Statistic
Prob.
R(-1)
-0.087410
0.077427
-1.128923
0.2653
D(R(-1))
-0.371164
0.147494
-2.516470
0.0158
D(R(-2))
-0.359451
0.143234
-2.509547
0.0160
C
1.827367
1.324924
1.379224
0.1751
R-squared
0.250734
Mean dependent var
0.286522
Adjusted R-squared
0.197215
S.D. dependent var
4.207061
S.E. of regression
3.769455
Akaike info criterion
5.574679
Sum squared resid
596.7693
Schwarz criterion
5.733692
F-statistic
4.684944
Prob(F-statistic)
0.006534
Log likelihood
Durbin-Watson stat
-124.2176
1.867452
UNIT ROOTS TEST: R (1ST DIFF)
Null Hypothesis: D(R) has a unit root
Exogenous: Constant
Lag Length: 1 (Automatic based on AIC, MAXLAG=10)
t-Statistic
Prob.*
Augmented Dickey-Fuller test statistic
-7.948964
0.0000
Test critical values:
1% level
-3.581152
5% level
-2.926622
10% level
-2.601424
64
*MacKinnon (1996) one-sided p-values.
Augmented Dickey-Fuller Test Equation
Dependent Variable: D(R,2)
Method: Least Squares
Date: 04/19/10 Time: 17:01
Sample (adjusted): 1963 2008
Included observations: 46 after adjustments
Variable
Coefficient
Std. Error
t-Statistic
Prob.
D(R(-1))
-1.810510
0.227767
-7.948964
0.0000
D(R(-1),2)
0.389436
0.141198
2.758084
0.0085
C
0.470843
0.559952
0.840862
0.4051
R-squared
0.702279
Mean dependent var
0.063913
Adjusted R-squared
0.688431
S.D. dependent var
6.774593
S.E. of regression
3.781466
Akaike info criterion
5.561094
Sum squared resid
614.8779
Schwarz criterion
5.680353
F-statistic
50.71519
Prob(F-statistic)
0.000000
Log likelihood
Durbin-Watson stat
-124.9052
1.876170
65
APPENDIX 2
CO INTEGRATION & VECTOR ERROR CORRECTION MODEL (VECM) RESULTS
FD = M2Y
Date: 04/20/10 Time: 12:29
Sample (adjusted): 1963 2008
Included observations: 46 after adjustments
Trend assumption: Linear deterministic trend
Series: RGDP M2Y INVT R
Lags interval (in first differences): 1 to 2
Unrestricted Cointegration Rank Test (Trace)
Hypothesized
Trace
0.05
No. of CE(s)
Eigenvalue
Statistic
Critical Value
Prob.**
None *
0.431905
48.15180
47.85613
0.0469
At most 1
0.209953
22.14033
29.79707
0.2908
At most 2
0.147865
11.29983
15.49471
0.1937
At most 3 *
0.082074
3.939367
3.841466
0.0472
Trace test indicates 1 cointegrating eqn(s) at the 0.05 level
* denotes rejection of the hypothesis at the 0.05 level
**MacKinnon-Haug-Michelis (1999) p-values
Unrestricted Cointegration Rank Test (Maximum Eigenvalue)
Hypothesized
Max-Eigen
0.05
No. of CE(s)
Eigenvalue
Statistic
Critical Value
Prob.**
None*
0.431905
39.01147
27.58434
0.0384
At most 1
0.209953
10.84049
21.13162
0.6632
At most 2
0.147865
7.360467
14.26460
0.4474
At most 3 *
0.082074
3.939367
3.841466
0.0472
Max-eigenvalue test indicates 1 cointegrating eqn(s) at the 0.05 level
* denotes rejection of the hypothesis at the 0.05 level
**MacKinnon-Haug-Michelis (1999) p-values
Unrestricted Cointegrating Coefficients (normalized by b'*S11*b=I):
RGDP
M2Y
INVT
R
1.35E-06
-0.002650
-2.71E-05
0.185767
5.01E-08
-0.132712
-1.47E-05
0.044824
1.03E-06
0.058329
-1.07E-05
-0.111149
1.17E-06
-0.039948
6.48E-06
-0.035658
66
Unrestricted Adjustment Coefficients (alpha):
D(RGDP)
47909.06
-83219.10
-136344.2
29813.47
D(M2Y)
-0.463550
1.293927
0.407335
0.431165
D(INVT)
48258.38
12457.20
8109.020
856.3570
D(R)
0.057434
-0.978301
0.723240
0.451706
Log likelihood
-863.950
1 Cointegrating Equation(s):
Normalized cointegrating coefficients (standard error in parentheses)
RGDP
M2Y
INVT
R
1.000000
-19.63721
-20.10934
137648.9
(2.12998)
(3.79384)
(31006.1)
Adjustment coefficients (standard error in parentheses)
D(RGDP)
- 0.064657
(0.01343)
D(M2Y)
-6.26E-07
D(INVT)
- 0.065128
D(R)
-7.75E-08
(7.7E-07)
(0.01467)
(7.3E-07)
2 Cointegrating Equation(s):
Log likelihood
-858.529
Normalized cointegrating coefficients (standard error in parentheses)
RGDP
M2Y
INVT
1.000000
0.000000
-19.90718
137087.3
(3.63774)
(29069.2)
0.000103
-0.285959
(6.3E-05)
(0.49983)
0.000000
1.000000
R
Adjustment coefficients (standard error in parentheses)
D(RGDP)
D(M2Y)
D(INVT)
D(R)
0.060484
10917.23
(0.09160)
(9003.60)
-5.61E-07
-0.170492
(7.1E-07)
(0.06985)
0.065753
-1781.118
(0.01441)
(1416.71)
2.85E-08
0.129680
(7.0E-07)
(0.06865)
3 Cointegrating Equation(s):
Log likelihood
-854.849
Normalized cointegrating coefficients (standard error in parentheses)
RGDP
M2Y
INVT
R
1.000000
0.000000
0.000000
-1125678.
67
(386067.)
0.000000
1.000000
0.000000
6.244106
(2.19730)
0.000000
0.000000
1.000000
-63432.62
(19755.4)
Adjustment coefficients (standard error in parentheses)
D(RGDP)
D(M2Y)
-0.079345
2964.408
1.379453
(0.10838)
(9266.22)
(2.08669)
-1.43E-07
-0.146732
-1.07E-05
(8.8E-07)
(0.07566)
(1.7E-05)
0.074069
-1308.127
-1.579121
(0.01795)
(1535.00)
(0.34567)
7.70E-07
0.171866
5.04E-06
(8.5E-07)
(0.07292)
(1.6E-05)
D(INVT)
D(R)
Vector Error Correction Estimates
Date: 04/20/10 Time: 12:41
Sample (adjusted): 1963 2008
Included observations: 46 after adjustments
Standard errors in ( ) & t-statistics in [ ]
Co integrating Eq:
CointEq1
RGDP(-1)
1.000000
M2Y(-1)
-19.63721
(2.1299.8)
[-9.24983]
INVT(-1)
-20.10934
(3.79384)
[-5.30053]
R(-1)
137648.9
(31006.1)
[ 4.43941]
C
-2596956.
Error Correction:
D(RGDP)
D(M2Y)
D(INVT)
D(R)
CointEq1
- 0.064657
-6.26E-07
0.065128
7.75E-08
D(RGDP(-1))
(0.01343)
(7.7E-07)
(0.01467)
(7.3E-07)
[ -4.81437]
[-0.81507]
[ 4.43837]
[ 0.10591]
0.510440
1.34E-06
-0.029206
-3.86E-06
(0.24177)
(2.0E-06)
(0.03797)
(1.9E-06)
[ 2.11124]
[ 0.67645]
[-0.76918]
[-2.03891]
68
D(RGDP(-2))
D(M2Y(-1))
D(M2Y(-2))
D(INVT(-1))
D(INVT(-2))
D(R(-1))
D(R(-2))
C
R-squared
0.249435
2.45E-06
-0.135356
2.85E-06
(0.22742)
(1.9E-06)
(0.03572)
(1.8E-06)
[ 1.09678]
[ 1.31334]
[-3.78970]
[ 1.59857]
8371.129
0.152442
5003.222
-0.074976
(22564.3)
(0.18536)
(3543.70)
(0.17674)
[ 0.37099]
[ 0.82242]
[ 1.41186]
[-0.42422]
19731.18
0.118125
-5626.124
0.015149
(20808.1)
(0.17093)
(3267.89)
(0.16298)
[ 0.94825]
[ 0.69106]
[-1.72164]
[ 0.09295]
2.463704
-1.16E-05
-0.168948
1.28E-06
(1.67826)
(1.4E-05)
(0.26357)
(1.3E-05)
[ 1.46801]
[-0.84093]
[-0.64100]
[ 0.09770]
0.375959
-9.78E-06
-0.247380
5.81E-06
(1.28568)
(1.1E-05)
(0.20191)
(1.0E-05)
[ 0.29242]
[-0.92607]
[-1.22517]
[ 0.57699]
-20641.92
0.629199
-11801.20
-0.430795
(20656.7)
(0.16969)
(3244.11)
(0.16180)
[-0.99929]
[ 3.70798]
[-3.63773]
[-2.66257]
-13489.54
0.073054
-14191.14
-0.448381
(22768.6)
(0.18704)
(3575.79)
(0.17834)
[-0.59246]
[ 0.39058]
[-3.96867]
[-2.51420]
150709.2
-1.053837
92163.63
1.053534
(134140.)
(1.10192)
(21066.6)
(1.05068)
[ 1.12352]
[-0.95636]
[ 4.37486]
[ 1.00272]
0.834310
0.357620
0.823375
0.388585
Adj. R-squared
0.792888
0.197025
0.779219
0.235731
Sum sq. resids
7.94E+12
535.6338
1.96E+11
486.9747
S.E. equation
469561.1
3.857294
73744.18
3.677917
F-statistic
20.14153
2.226845
18.64686
2.542200
-660.3728
-121.7318
-575.2178
-119.5413
Log likelihood
Akaike AIC
29.14664
5.727469
25.44425
5.632230
Schwarz SC
29.54418
6.125000
25.84178
6.029761
Mean dependent
518250.6
0.563043
15610.14
0.286522
S.D. dependent
1031786.
4.304590
156944.8
4.207061
Determinant resid covariance (dof adj.)
1.38E+23
Determinant resid covariance
5.16E+22
Log likelihood
-863.950
69
Akaike information criterion
35.56303
Schwarz criterion
37.31216
VEC Granger Causality/Block Exogeneity Wald Tests
Date: 04/20/10 Time: 13:23
Sample: 1960 2008
Included observations: 46
Dependent variable: D(RGDP)
Excluded
D(M2Y)
Chi-sq
df
Prob.
18.91163
2
0.0008
D(INVT)
5.760485
2
0.0561
D(R)
1.073699
2
0.6415
All
25.74588
6
0.2248
Dependent variable: D(M2Y)
Excluded
Chi-sq
df
Prob.
D(RGDP)
2.373327
2
0.3052
D(INVT)
0.865281
2
0.6488
D(R)
14.56128
2
0.0007
All
16.71762
6
0.0104
Dependent variable: D(INVT)
Excluded
Chi-sq
df
Prob.
D(RGDP)
15.65119
2
0.0004
D(M2Y)
3.999479
2
0.1354
D(R)
21.76162
2
0.0000
All
26.87845
6
0.0002
Dependent variable: D(R)
Excluded
Chi-sq
df
Prob.
D(RGDP)
6.147163
2
0.0463
D(M2Y)
0.180118
2
0.9139
D(INVT)
0.941389
2
0.6246
All
7.143326
6
0.3078
70
FD = BDY
Date: 04/20/10 Time: 13:44
Sample (adjusted): 1963 2008
Included observations: 46 after adjustments
Trend assumption: Linear deterministic trend
Series: RGDP BDY INVT R
Lags interval (in first differences): 1 to 2
Unrestricted Cointegration Rank Test (Trace)
Hypothesized
Trace
0.05
No. of CE(s)
Eigenvalue
Statistic
Critical Value
Prob.**
None *
0.425469
50.28226
47.85613
0.0290
At most 1
0.216329
24.78901
29.79707
0.1691
At most 2
0.158850
13.57577
15.49471
0.0953
At most 3 *
0.114976
5.618464
3.841466
0.0178
Trace test indicates 1 cointegrating eqn(s) at the 0.05 level
* denotes rejection of the hypothesis at the 0.05 level
**MacKinnon-Haug-Michelis (1999) p-values
Unrestricted Cointegration Rank Test (Maximum Eigenvalue)
Hypothesized
Max-Eigen
0.05
No. of CE(s)
Eigenvalue
Statistic
Critical Value
Prob.**
None*
0.425469
35.49325
27.58434
0.0404
At most 1
0.216329
11.21324
21.13162
0.6260
At most 2
0.158850
7.957306
14.26460
0.3830
At most 3 *
0.114976
5.618464
3.841466
0.0178
Max-eigenvalue test indicates 1 cointegrating eqn(s) at the 0.05 level
* denotes rejection of the hypothesis at the 0.05 level
**MacKinnon-Haug-Michelis (1999) p-values
Unrestricted Cointegrating Coefficients (normalized by b'*S11*b=I):
RGDP
BDY
INVT
R
1.21E-06
-0.021560
-2.87E-05
0.204607
-0.051852
4.09E-07
0.139192
1.21E-05
-1.34E-06
-0.019736
1.12E-05
0.108333
-7.86E-07
0.082852
-7.64E-06
-0.000838
Unrestricted Adjustment Coefficients (alpha):
D(RGDP)
50372.32
80037.81
128114.8
-66810.30
D(BDY)
-0.359303
-0.768108
-0.641173
-0.400606
71
D(INVT)
48817.49
-6567.337
-12319.62
678.5075
D(R)
-0.205608
1.243398
-0.712840
-0.048344
Log likelihood
-953.024
1 Cointegrating Equation(s):
Normalized cointegrating coefficients (standard error in parentheses)
RGDP
BDY
INVT
R
1.000000
-17.81975
-23.73682
169114.7
(2.6341.5)
(4.47431)
(37421.1)
Adjustment coefficients (standard error in parentheses)
D(RGDP)
-0.060944
D(BDY)
-4.35E-07
D(INVT)
0.059063
(0.01440)
(5.3E-07)
(0.01334)
D(R)
-2.49E-07
(6.5E-07)
2 Cointegrating Equation(s):
Log likelihood
-947.418
Normalized cointegrating coefficients (standard error in parentheses)
RGDP
BDY
INVT
R
1.000000
0.000000
-21.08980
154391.9
0.000000
1.000000
(4.18780)
(34287.8)
0.000149
-0.826204
(5.9E-05)
(0.48313)
Adjustment coefficients (standard error in parentheses)
D(RGDP)
0.093681
10054.59
(0.08745)
(9644.35)
D(BDY)
-7.49E-07
-0.099168
(5.3E-07)
(0.05890)
D(INVT)
0.056377
-1966.606
(0.01401)
(1545.41)
D(R)
2.60E-07
0.177504
(6.3E-07)
(0.06994)
3 Cointegrating Equation(s):
Log likelihood
-943.439
Normalized cointegrating coefficients (standard error in parentheses)
RGDP
BDY
INVT
R
1.000000
0.000000
0.000000
-290308.9
0.000000
1.000000
0.000000
2.306007
0.000000
0.000000
1.000000
-21086.06
(98439.0)
(0.88480)
72
(5135.26)
Adjustment coefficients (standard error in parentheses)
D(RGDP)
-0.078305
7526.095
0.950979
(0.12054)
(9252.94)
(2.15289)
D(BDY)
1.12E-07
-0.086514
-6.11E-06
(7.5E-07)
(0.05750)
(1.3E-05)
D(INVT)
0.072915
-1723.463
-1.618915
(0.01997)
(1532.94)
(0.35667)
1.22E-06
0.191572
1.29E-05
(8.9E-07)
(0.06857)
(1.6E-05)
D(R)
Vector Error Correction Estimates
Date: 04/20/10 Time: 13:50
Sample (adjusted): 1963 2008
Included observations: 46 after adjustments
Standard errors in ( ) & t-statistics in [ ]
Cointegrating Eq:
CointEq1
RGDP(-1)
1.000000
BDY(-1)
-17.81975
(2.6341.5)
[-6.76490]
INVT(-1)
-23.73682
(4.47431)
[-5.30513]
R(-1)
169114.7
(37421.1)
[ 4.51923]
C
-2497008.
Error Correction:
D(RGDP)
D(BDY)
CointEq1
-0.060944
(0.01440)
[ -4.23222]
D(RGDP(-1))
D(RGDP(-2))
D(INVT)
D(R)
-4.35E-07
0.059063
-2.49E-07
(5.3E-07)
(0.01334)
(6.5E-07)
[-0.82157]
[ 4.42734]
[-0.38212]
-3.48E-06
0.519862
8.80E-07
-0.017295
(0.22300)
(1.4E-06)
(0.03525)
(1.7E-06)
[ 2.33124]
[ 0.62930]
[-0.49067]
[-2.02044]
0.268772
2.03E-06
-0.103891
3.51E-06
(0.21064)
(1.3E-06)
(0.03329)
(1.6E-06)
73
D(BDY(-1))
D(BDY(-2))
D(INVT(-1))
D(INVT(-2))
D(R(-1))
D(R(-2))
C
[ 1.27599]
[ 1.53524]
[-3.12042]
[ 2.15809]
9499.229
0.147386
7738.790
-0.192814
(29525.2)
(0.18510)
(4666.86)
(0.22774)
[ 0.32173]
[ 0.79625]
[ 1.65825]
[-0.84665]
23068.49
0.204335
-2806.895
0.037080
(27838.5)
(0.17453)
(4400.25)
(0.21473)
[ 0.82865]
[ 1.17081]
[-0.63789]
[ 0.17268]
2.554454
-9.28E-06
-0.092358
-4.77E-06
(1.75916)
(1.1E-05)
(0.27806)
(1.4E-05)
[ 1.45208]
[-0.84121]
[-0.33215]
[-0.35146]
0.407711
-7.30E-06
-0.209353
2.50E-06
(1.29872)
(8.1E-06)
(0.20528)
(1.0E-05)
[ 0.31393]
[-0.89669]
[-1.01984]
[ 0.24931]
-21237.42
0.448895
-11421.67
-0.397102
(20437.0)
(0.12812)
(3230.35)
(0.15764)
[-1.03916]
[ 3.50359]
[-3.53574]
[-2.51907]
-14254.49
0.035453
-13125.52
-0.405969
(22244.6)
(0.13946)
(3516.06)
(0.17158)
[-0.64081]
[ 0.25422]
[-3.73302]
[-2.36606]
134675.9
-0.658774
70747.68
0.743571
(111786.)
(0.70081)
(17669.3)
(0.86225)
[ 1.20477]
[-0.94002]
[ 4.00399]
[ 0.86236]
R-squared
0.831783
0.395890
0.818357
0.398023
Adj. R-squared
0.789729
0.244862
0.772946
0.247528
Sum sq. resids
8.06E+12
316.7294
2.01E+11
479.4577
S.E. equation
473128.8
2.966149
74784.45
3.649420
F-statistic
19.77882
2.621310
18.02120
2.644769
-660.7210
-109.6475
-575.8622
-119.1835
Akaike AIC
29.16178
5.202066
25.47227
5.616674
Log likelihood
Schwarz SC
29.55931
5.599597
25.86980
6.014205
Mean dependent
518250.6
0.607609
15610.14
0.286522
S.D. dependent
1031786.
3.413343
156944.8
4.207061
Determinant resid covariance (dof adj.)
8.56E+22
Determinant resid covariance
3.21E+22
Log likelihood
-953.024
Akaike information criterion
45.08801
Schwarz criterion
46.83715
74
VEC Granger Causality/Block Exogeneity Wald Tests
Date: 04/20/10 Time: 14:10
Sample: 1960 2008
Included observations: 46
Dependent variable: D(RGDP)
Excluded
Chi-sq
df
Prob.
D(BDY)
11.924919
2
0.0027
D(INVT)
5.470083
2
0.0649
D(R)
1.156680
2
0.2608
All
18.55168
6
0.3284
Dependent variable: D(BDY)
Excluded
Chi-sq
df
Prob.
D(RGDP)
3.642765
2
0.1618
D(INVT)
0.822841
2
0.6627
D(R)
13.52137
2
0.0012
All
18.19067
6
0.0058
Dependent variable: D(INVT)
Excluded
Chi-sq
df
Prob.
D(RGDP)
11.92579
2
0.0026
D(BDY)
2.857852
2
0.2396
D(R)
19.33147
2
0.0001
All
23.32212
6
0.0007
Dependent variable: D(R)
Excluded
Chi-sq
df
Prob.
D(RGDP)
6.751302
2
0.0342
D(BDY)
0.716956
2
0.6987
D(INVT)
1.293326
2
0.5238
All
7.578875
6
0.2706
75
FD= CPSY
Date: 04/20/10 Time: 14:59
Sample (adjusted): 1964 2008
Included observations: 45 after adjustments
Trend assumption: Linear deterministic trend
Series: RGDP CPSY INVT R
Lags interval (in first differences): 1 to 3
Unrestricted Cointegration Rank Test (Trace)
Hypothesized
Trace
0.05
No. of CE(s)
Eigenvalue
Statistic
Critical Value
Prob.**
None *
0.794927
102.5839
47.85613
0.0000
At most 1 *
0.332094
31.28647
29.79707
0.0334
At most 2
0.181745
13.12415
15.49471
0.1103
At most 3 *
0.087043
4.098010
3.841466
0.0429
Trace test indicates 2 cointegrating eqn(s) at the 0.05 level
* denotes rejection of the hypothesis at the 0.05 level
**MacKinnon-Haug-Michelis (1999) p-values
Unrestricted Cointegration Rank Test (Maximum Eigenvalue)
Hypothesized
Max-Eigen
0.05
No. of CE(s)
Eigenvalue
Statistic
Critical Value
Prob.**
None *
0.794927
71.29742
27.58434
0.0000
At most 1
0.332094
18.16232
21.13162
0.1239
At most 2
0.181745
9.026136
14.26460
0.2840
At most 3 *
0.087043
4.098010
3.841466
0.0429
Max-eigenvalue test indicates 1 cointegrating eqn(s) at the 0.05 level
* denotes rejection of the hypothesis at the 0.05 level
**MacKinnon-Haug-Michelis (1999) p-values
Unrestricted Cointegrating Coefficients (normalized by b'*S11*b=I):
RGDP
CPSY
INVT
R
2.74E-06
-0.008216
-1.50E-05
0.182064
-2.86E-07
0.020281
2.99E-05
-0.222319
-2.72E-06
0.058977
3.25E-05
-0.049556
-8.97E-07
-0.259352
-3.12E-06
0.151117
Unrestricted Adjustment Coefficients (alpha):
D(RGDP)
112811.7
117877.8
104008.0
38890.75
76
D(CPSY)
-0.412844
0.739325
-1.212507
-0.194152
D(INVT)
28050.66
-19314.76
-13521.51
-3521.658
D(R)
-0.090130
-0.015099
0.353893
-0.784214
Log likelihood
-1170.906
1 Cointegrating Equation(s):
Normalized cointegrating coefficients (standard error in parentheses)
RGDP
CPSY
INVT
R
1.000000
-29.93610
-5.466015
66334.87
(8.862.51)
(1.08153)
(10014.1)
Adjustment coefficients (standard error in parentheses)
D(RGDP)
- 0.309626
(0.12220)
D(CPSY)
-1.13E-06
(1.6E-06)
D(INVT)
0.076989
(0.02493)
D(R)
-2.47E-07
(1.4E-06)
2 Cointegrating Equation(s):
Log likelihood
-1161.825
Normalized cointegrating coefficients (standard error in parentheses)
RGDP
CPSY
INVT
R
1.000000
0.000000
-1.105957
34996.48
(1.52371)
(12276.4)
0.001456
-10.46843
(0.00038)
(3.07457)
0.000000
1.000000
Adjustment coefficients (standard error in parentheses)
D(RGDP)
0.275901
1463.735
(0.16567)
(1313.68)
D(CPSY)
-1.34E-06
0.018386
(1.6E-06)
(0.01242)
0.082515
-622.1893
(0.02310)
(183.152)
D(INVT)
D(R)
-2.43E-07
0.000434
(1.4E-06)
(0.01113)
3 Cointegrating Equation(s):
Log likelihood
-1157.312
Normalized cointegrating coefficients (standard error in parentheses)
RGDP
CPSY
INVT
R
1.000000
0.000000
0.000000
22002.63
0.000000
1.000000
0.000000
(11711.1)
6.643407
(2.03681)
77
0.000000
0.000000
1.000000
-11748.96
(2558.03)
Adjustment coefficients (standard error in parentheses)
D(RGDP)
-0.006726
7597.821
5.204639
(0.22057)
(3582.65)
(2.65374)
D(CPSY)
1.95E-06
-0.053124
-1.11E-05
(2.0E-06)
(0.03288)
(2.4E-05)
D(INVT)
0.119257
-1419.648
-1.436591
(0.03097)
(503.102)
(0.37266)
D(R)
-1.20E-06
0.021306
1.24E-05
(2.0E-06)
(0.03173)
(2.4E-05)
Vector Error Correction Estimates
Date: 04/20/10 Time: 15:10
Sample (adjusted): 1964 2008
Included observations: 45 after adjustments
Standard errors in ( ) & t-statistics in [ ]
Co integrating Eq:
CointEq1
RGDP(-1)
1.000000
CPSY(-1)
-29.93610
(8.862.51)
[-3.36881]
INVT(-1)
-5.466015
(1.06394)
[-5.13752]
R(-1)
66334.87
(9851.26)
[ 6.73364]
C
-2877366.
Error Correction:
D(RGDP)
D(CPSY)
D(INVT)
D(R)
CointEq1
- 0.309626
-1.13E-06
0.076989
-2.47E-07
D(RGDP(-1))
D(RGDP(-2))
(0.12220)
(1.6E-06)
(0.02452)
(1.4E-06)
[ -2.53376]
[-0.71930]
[ 3.13945]
[-0.18018]
0.056865
3.16E-06
-0.167859
-3.89E-06
(0.42274)
(3.9E-06)
(0.06020)
(3.4E-06)
[ 0.13452]
[ 0.81722]
[-2.78814]
[-1.15416]
-0.043945
2.37E-06
-0.130854
2.79E-06
(0.35367)
(3.2E-06)
(0.05037)
(2.8E-06)
78
D(RGDP(-3))
D(CPSY(-1))
D(CPSY(-2))
D(CPSY(-3))
D(INVT(-1))
D(INVT(-2))
D(INVT(-3))
D(R(-1))
D(R(-2))
D(R(-3))
C
R-squared
[-0.12425]
[ 0.73275]
[-2.59799]
[ 0.98914]
-0.562722
1.59E-06
0.048269
2.39E-06
(0.33177)
(3.0E-06)
(0.04725)
(2.6E-06)
[-1.69610]
[ 0.52507]
[ 1.02159]
[ 0.90501]
12790.69
0.021986
-2091.042
-0.078031
(22215.6)
(0.20323)
(3163.80)
(0.17713)
[ 0.57575]
[ 0.10818]
[-0.66093]
[-0.44053]
1837.604
-0.442094
-2994.248
0.350145
(21403.7)
(0.19581)
(3048.18)
(0.17066)
[ 0.08585]
[-2.25780]
[-0.98231]
[ 2.05175]
-3005.033
-0.156466
-1503.185
0.420733
(24664.2)
(0.22564)
(3512.52)
(0.19665)
[-0.12184]
[-0.69345]
[-0.42795]
[ 2.13946]
2.404523
-1.29E-05
-1.118334
4.92E-06
(0.90703)
(8.3E-06)
(0.12917)
(7.2E-06)
[ 2.65098]
[-1.55039]
[-8.65760]
[ 0.68054]
0.052982
-6.22E-06
-1.264281
1.13E-05
(1.17400)
(1.1E-05)
(0.16719)
(9.4E-06)
[ 0.04513]
[-0.57937]
[-7.56179]
[ 1.20407]
-2.253979
1.46E-05
-1.483038
2.06E-06
(1.66944)
(1.5E-05)
(0.23775)
(1.3E-05)
[-1.35014]
[ 0.95453]
[-6.23778]
[ 0.15444]
-7164.856
0.375659
-10650.07
-0.468817
(25861.0)
(0.23658)
(3682.95)
(0.20620)
[-0.27705]
[ 1.58785]
[-2.89172]
[-2.27365]
-12395.69
-0.017619
-11061.95
-0.232980
(25734.7)
(0.23543)
(3664.97)
(0.20519)
[-0.48167]
[-0.07484]
[-3.01829]
[-1.13544]
33161.20
0.002295
-8191.826
0.192400
(28475.6)
(0.26050)
(4055.32)
(0.22704)
[ 1.16455]
[ 0.00881]
[-2.02002]
[ 0.84742]
698045.5
-2.349252
188607.6
-0.003474
(389509.)
(3.56334)
(55471.4)
(3.10565)
[ 1.79212]
[-0.65928]
[ 3.40009]
[-0.00112]
0.884720
0.422397
0.899505
0.561676
79
Adj. R-squared
0.836377
0.180176
0.857362
0.377863
Sum sq. resids
5.49E+12
459.5481
1.11E+11
349.0757
S.E. equation
420866.3
3.850212
59937.13
3.355668
F-statistic
18.30085
1.743851
21.34410
3.055689
-638.2202
-116.1328
-550.5144
-109.9463
Akaike AIC
28.98757
5.783680
25.08953
5.508725
Log likelihood
Schwarz SC
29.54964
6.345753
25.65160
6.070798
Mean dependent
529763.8
0.613333
15955.37
0.292889
S.D. dependent
1040452.
4.252304
158700.5
4.254376
Determinant resid covariance (dof adj.)
1.51E+22
Determinant resid covariance
3.40E+21
Log likelihood
-1170.906
Akaike information criterion
53.59584
Schwarz criterion
56.00473
VEC Granger Causality/Block Exogeneity Wald Tests
Date: 04/20/10 Time: 15:55
Sample: 1960 2008
Included observations: 45
Dependent variable: D(RGDP)
Excluded
Chi-sq
df
Prob.
D(CPSY)
0.455695
3
0.9285
D(INVT)
15.53681
3
0.0014
D(R)
3.854121
3
0.2777
All
28.41365
9
0.0008
Dependent variable: D(CPSY)
Excluded
Chi-sq
df
Prob.
D(RGDP)
9.798188
3
0.0499
D(INVT)
3.548964
3
0.3145
D(R)
4.355336
3
0.2256
All
17.70249
9
0.5900
Dependent variable: D(INVT)
Excluded
Chi-sq
df
Prob.
D(RGDP)
17.16512
3
0.0007
D(CPSY)
1.626494
3
0.6534
D(R)
11.11858
3
0.0111
80
All
26.39642
9
0.0018
Dependent variable: D(R)
Excluded
Chi-sq
df
Prob.
D(RGDP)
7.076236
3
0.0695
D(CPSY)
11.45595
3
0.0095
D(INVT)
2.293519
3
0.5138
All
20.97190
9
0.0128
FD = DCY
Date: 04/20/10 Time: 16:07
Sample (adjusted): 1962 2008
Included observations: 47 after adjustments
Trend assumption: Linear deterministic trend
Series: RGDP DCY INVT R
Lags interval (in first differences): 1 to 1
Unrestricted Cointegration Rank Test (Trace)
Hypothesized
Trace
0.05
No. of CE(s)
Eigenvalue
Statistic
Critical Value
Prob.**
None *
0.496994
59.33647
47.85613
0.0029
At most 1
0.277643
27.04032
29.79707
0.1006
At most 2
0.128834
11.75421
15.49471
0.1691
At most 3 *
0.106105
5.271856
3.841466
0.0217
Trace test indicates 1 cointegrating eqn(s) at the 0.05 level
* denotes rejection of the hypothesis at the 0.05 level
**MacKinnon-Haug-Michelis (1999) p-values
Unrestricted Cointegration Rank Test (Maximum Eigenvalue)
Hypothesized
Max-Eigen
0.05
No. of CE(s)
Eigenvalue
Statistic
Critical Value
Prob.**
None *
0.496994
32.29616
27.58434
0.0115
At most 1
0.277643
15.28611
21.13162
0.2693
At most 2
0.128834
6.482353
14.26460
0.5521
At most 3 *
0.106105
5.271856
3.841466
0.0217
81
Max-eigenvalue test indicates 1 co integrating eqn(s) at the 0.05 level
* denotes rejection of the hypothesis at the 0.05 level
**MacKinnon-Haug-Michelis (1999) p-values
Unrestricted Co integrating Coefficients (normalized by b'*S11*b=I):
RGDP
DCY
INVT
R
6.24E-07
-0.022550
-2.04E-05
0.113464
3.31E-07
-0.016329
8.17E-06
-0.009400
2.18E-07
0.073255
1.69E-06
0.000762
-2.08E-07
-0.024153
1.77E-06
0.146705
Unrestricted Adjustment Coefficients (alpha):
D(RGDP)
21145.57
138487.7
82492.45
78857.73
D(DCY)
1.407676
1.656652
-2.185673
-0.555787
D(INVT)
71960.52
-5038.659
4814.818
-8580.530
D(R)
-0.509208
1.023539
0.543497
-0.917774
Log likelihood
-1017.404
1 Co integrating Equation(s):
Normalized cointegrating coefficients (standard error in parentheses)
RGDP
DCY
INVT
R
1.000000
-36135.77
-32.70701
181821.2
(20640.5)
(4.49424)
(45127.4)
Adjustment coefficients (standard error in parentheses)
D(RGDP)
-0.013196
D(DCY)
-8.78E-07
D(INVT)
0.044906
(0.00146)
(7.0E-07)
(0.00768)
D(R)
-3.18E-07
(3.7E-07)
2 Co integrating Equation(s):
Log likelihood
-1019.761
Normalized cointegrating coefficients (standard error in parentheses)
RGDP
DCY
INVT
R
1.000000
0.000000
-189.8161
757354.1
(38.9769)
(374211.)
0.000000
1.000000
-0.004348
15.92696
(0.00104)
(9.94079)
Adjustment coefficients (standard error in parentheses)
D(RGDP)
D(DCY)
0.059032
-2738.174
(0.04437)
(1748.87)
1.43E-06
-0.058795
(7.7E-07)
(0.03048)
82
D(INVT)
0.043239
D(R)
-1540.449
(0.00867)
(341.840)
2.10E-08
-0.005230
(4.0E-07)
(0.01581)
3 Co integrating Equation(s):
Log likelihood
-1026.519
Normalized cointegrating coefficients (standard error in parentheses)
RGDP
DCY
INVT
1.000000
0.000000
0.000000
R
58556.89
(66147.9)
0.000000
1.000000
0.000000
-0.079017
0.000000
0.000000
1.000000
-3681.444
(0.76365)
(2060.68)
Adjustment coefficients (standard error in parentheses)
D(RGDP)
0.077024
3304.847
0.839039
(0.04545)
(4818.08)
(1.35559)
D(DCY)
9.50E-07
-0.218907
-1.89E-05
(7.7E-07)
(0.08152)
(2.3E-05)
D(INVT)
0.044289
-1187.737
-1.501772
(0.00906)
(960.403)
(0.27021)
D(R)
1.40E-07
0.034584
1.97E-05
(4.2E-07)
(0.04400)
(1.2E-05)
Vector Error Correction Estimates
Date: 04/19/10 Time: 20:05
Sample (adjusted): 1962 2008
Included observations: 47 after adjustments
Standard errors in ( ) & t-statistics in [ ]
Co integrating Eq:
CointEq1
RGDP(-1)
1.000000
DCY(-1)
-36135.77
(20640.5)
[-1.75072]
INVT(-1)
-32.70701
(4.49424)
[-7.27753]
R(-1)
181821.2
(45127.4)
[ 4.02907]
C
-1357239.
83
Error Correction:
D(RGDP)
D(DCY)
D(INVT)
D(R)
CointEq1
-0.013196
-8.78E-07
0.044906
-3.18E-07
D(RGDP(-1))
D(DCY(-1))
D(INVT(-1))
D(R(-1))
C
R-squared
(0.00146)
(7.0E-07)
(0.00768)
(3.7E-07)
[ -9.0384]
[ -1.25137]
[ 5.84887]
[-0.86322]
0.914311
-7.01E-07
-0.020572
-4.76E-07
(0.09438)
(1.6E-06)
(0.01748)
(8.4E-07)
[ 9.68767]
[-0.43839]
[-1.17702]
[-0.56842]
19619.94
0.207441
-777.1675
-0.054061
(8969.34)
(0.15187)
(1661.07)
(0.07964)
[ 2.18745]
[ 1.36589]
[-0.46787]
[-0.67880]
-2.13E-06
2.196554
5.87E-06
0.048782
(0.78784)
(1.3E-05)
(0.14590)
(7.0E-06)
[ 2.78809]
[ 0.43998]
[ 0.33434]
[-0.30421]
-16457.65
0.558962
-6315.568
-0.303717
(16753.6)
(0.28368)
(3102.66)
(0.14876)
[-0.98234]
[ 1.97040]
[-2.03553]
[-2.04166]
73364.17
0.377686
25101.78
0.596235
(76488.2)
(1.29513)
(14165.1)
(0.67916)
[ 0.95916]
[ 0.29162]
[ 1.77208]
[ 0.87790]
0.823434
0.175304
0.736897
0.158188
Adj. R-squared
0.801901
0.074731
0.704811
0.055528
Sum sq. resids
8.50E+12
2438.447
2.92E+11
670.5469
S.E. equation
455454.5
7.711959
84347.33
4.044107
F-statistic
Log likelihood
38.24148
1.743055
22.96648
1.540891
-675.8460
-159.4909
-596.5874
-129.1518
Akaike AIC
29.01472
7.042165
25.64202
5.751142
Schwarz SC
29.25091
7.278354
25.87821
5.987331
Mean dependent
507229.0
0.312340
15276.55
0.280426
S.D. dependent
1023303.
8.017347
155246.3
4.161291
Determinant resid covariance (dof adj.)
1.23E+24
Determinant resid covariance
7.11E+23
Log likelihood
-1017.404
Akaike information criterion
47.46398
Schwarz criterion
48.56620
84
VEC Granger Causality/Block Exogeneity Wald Tests
Date: 04/20/10 Time: 16:23
Sample: 1960 2008
Included observations: 47
Dependent variable: D(RGDP)
Excluded
Chi-sq
df
Prob.
D(DCY)
4.784919
1
0.0287
D(INVT)
7.773425
1
0.0053
D(R)
0.964984
1
0.3259
All
15.06185
3
0.0018
Dependent variable: D(DCY)
Excluded
Chi-sq
df
Prob.
D(RGDP)
0.192188
1
0.6611
D(INVT)
0.193583
1
0.6600
D(R)
3.882477
1
0.0488
All
4.656276
3
0.1988
Dependent variable: D(INVT)
Excluded
Chi-sq
df
Prob.
D(RGDP)
1.385384
1
0.2392
D(DCY)
0.218905
1
0.6399
D(R)
4.143396
1
0.0418
All
4.767695
3
0.1896
Dependent variable: D(R)
Excluded
Chi-sq
df
Prob.
D(RGDP)
0.323101
1
0.5698
D(DCY)
0.460770
1
0.4973
D(INVT)
0.092541
1
0.7610
All
1.146229
3
0.7659
85
Appendix 3
RESPONSE TO CHOLESKY ONE S.D INNOVATIONS