Download Oil Shocks and Economic Growth in OPEC countries

Business School
WORKING PAPER SERIES
Working Paper
2014-064
Oil Shocks and Economic Growth
in OPEC countries
Zied Ftiti
Khaled Guesmi
Frédéric Teulon
http://www.ipag.fr/fr/accueil/la-recherche/publications-WP.html
IPAG Business School
184, Boulevard Saint-Germain
75006 Paris
France
IPAG working papers are circulated for discussion and comments only. They have not been
peer-reviewed and may not be reproduced without permission of the authors.
Oil Shocks and Economic Growth
in OPEC countries
Zied Ftiti* Department of Finance – IPAG Business School (France). Khaled Guesmi Department of Finance – IPAG Business School (France). Frédéric Teulon Department of Finance – IPAG Business School (France). * Corresponding author. Abstract
This paper assesses the impact of oil prices on economic growth of the four major OPEC countries (United Arab
Emirates, Kuwait, Saudi Arabia and Venezuela) over the period spanning from 03/09/2000 to 03/12/2010. We aim at
complementing the results from existing analyses (mainly focused on oil-importing countries) by using the
evolutionary co-spectral analysis as defined by Priestley and Tong (1973). We find that co-movements between oil
and economic growth have different patterns depending of the studied horizons. This interdependence is a mediumlived phenomenon, revealed on a three years and one quarter horizon, being weak in the short-run (ten months). We
show that oil price shocks in periods of world turmoil or during fluctuations of the global business cycle (downturn
or growth, as for instance the 2008 financial crisis) have a significant impact on the relationship between oil and
economic growth in oil-exporting countries.
Keywords: oil prices shocks, stock markets, evolutionary co-spectral analysis, OPEC
JEL Classifications: C14, C22, G12, G15, Q43
1 1.
Introduction
The world has witnessed a sharp rise in oil prices during the decade of the 2000s. This increase has led oil prices to record levels in 2008 (US 140 $ per barrel) before the trend turns. The impact of oil prices on economic growth has mainly been studied from the point of view of importing countries (the United States and more broadly the OECD countries). The situation of producers and exporters has been the subject of less attention. Among the few studies that have focused on this issue, the papers of Mehrara and Oskoui (2007) and
that of Berument and Ceylan (2010) may be mentioned. A priori, for the OPEC (Organization of Petroleum Exporting Countries), the soaring price of crude oil has a stimulating effect on the production of wealth, provided it does not lead a global recession and a reduction in energy demand. Oil is one of the most important energy, and known as ‘the blood of industry’, thus the macro-­‐
economic situation is necessary affected by the world crude oil price. This relation has captured increasing attention of academic researchers such as : Sadorsky (2000), Lynch (2002), Merino and Ortiz (2005), Kaufmann et al. (2004), Fan et al. (2007), Hamilton (2009a, 2009b), He et al. (2010), Kilian and Vigfusson (2011) and others. There are a great deal of researches on the relationship between oil price shocks and economic activity after the oil crisis of 1970s and the subsequent recession. The seminal work is completed by Hamilton (1983), who firstly analyse in depth the relationship between oil prices and recessions. Hamilton has made a series of contribution (see Hamilton, 1983, 2003, 2009a, and 2009b). Our study tackles the issues of oil and economic growth interdependence in oil-­‐exporting and importing countries by measuring the interaction between oil price indices and growth rate according the evolutionary co-­‐spectral analysis as defined by Priestley and Tong (1973). Our analysis leads to two main findings. Firstly, the exporting countries are more sensible to oil price shocks than importing countries. Secondly, we show that oil price shocks in periods of world turmoil or during fluctuations of the global business cycle (downturn or expansion) exhibit a significant impact on the relationship between oil and real economic activity in OPEC countries ; aggregate demand-­‐side oil price shocks (housing market boom, Chinese economic growth, and the latest global financial crisis) cause a significant higher correlation between real activity and oil prices, and important precautionary demand-­‐side oil price shocks (i.e. second war in Iraq, terrorist attacks) tend to cause lower coherence. This paper is organized as follows. Section 2 presents a review of the literature. Section 3 presents the empirical methodology and the data. Section 4 is concerned to results and discussion. Section 5 concludes. 2.
Related Literature
The breakdown of Bretton Woods regime in 1971 and the occurrence of the primary oil shocks in 1973 has been challenged some classical macroeconomic schema, such as the transmission mechanism. Indeed, the abrupt rise in oil prices due the OPEC oil embargo is followed by a global recession. At this period 2 some pioneer researches (Rasche and Tatom, 1977, 1981; Darby 1982; Hamilton, 1983; Burbidge and Harrison, 1984 ; Santini, 1985; Gisser and Goodwin, 1986) has been studied the relationship between oil price shocks and economic activity in order to check whether the observed recession (1970s) was be attributable the oil shock of 1973. Hamilton (1983) is the most influential paper in this field; he argued that oil-­‐price increases were at least partially responsible for every post-­‐second WORLD WAR U.S. recession except the one in 1960. In addition, to Hamilton (1983), the above-­‐cited pioneer researches are interested to the US economy. These studies have identified a weak effect on the US economic growth and oil price movement. After these studies, a large number of studies were carried out in various branches. Principally two methodologies were followed: 1/ The first strand of literature evaluates the relationship between oil price shocks and macro-­‐
economy through a theoretical modelling by studying the transmission mechanisms channels. Indeed, Hooker (1996) finds that oil prices did not affect many of U.S macroeconomic indicator variables after 1973 period (in sense of Granger causality). Rotemberg and Woodford (1996) prove that, through using a model involving implicit collusion between oligopolists, an increase in oil prices leads to decrease in both output and real wages. However, Bernanke et al. (1997) show that an important part of the effect of oil price shocks on the economy results not from the change in oil prices but from the resulting tightening of monetary policy. Rogoff (2005) argues that, compared with decades ago, higher energy efficiency, greater concentration of oil consumption in final demand, better anchored monetary policy, deeper financial markets and more flexible labor markets have weakened the effects of oil shocks. Barsky and Kilian (2001, 2004) show that the disturbances in the oil market are likely to matter less for the US macroeconomic performance than has commonly been thought. 2/ The second strand of literature has focused mainly on the empirical investigation of the relationship between oil-­‐price change and national aggregate economic activity. Most of these researches used times series analysis and reached to mixed results. Lee et al. (1996), by taking into account both the unanticipated component of real oil price movement and the time-­‐varying conditional variance of oil price change forecasts by using a GARCH model, show a negative effect of real oil price changes on the industrial production. Papapetrou (2001) reaches to the same results by examining the case of Greece through a VAR model. Miguel et al. (2003) focus on the impact of the oil-­‐price change on economic activity under the case of Spain. They used a VAR model and they show a negative impact. Limin et al. (2010) investigate the relationship between the world oil price and China’s macro-­‐economy through a VAR model. They show that the world oil price affects the economic growth and inflation of China significantly, and the impact is non-­‐
linear. Contrary to these findings, Cunado and Perez de Gracia (2003) obtained different results on studying this relationship on 15 European countries. Moreover, they could not find any cointegrating long-­‐run relationship between oil prices and economic activity except for the United Kingdom and Ireland. Therefore, they suggest that the impact of oil shocks on economic activity is limited to the short-­‐run. They show that results are depending on whether they use a world oil price index or a national real price index. In addition, 3 Levin and Loungani (1996) also reported significant differences in GDP response to oil price shocks for G-­‐7 countries. Limin et al. (2010) investigate the relationship between the world oil price and China’s macro-­‐
economy through a VAR model. They show that the world oil price affects the economic growth and inflation of China significantly, and the impact is non-­‐linear. The oil price crisis observed at the beginning of 1970 due to the OPEC oil embargo had been followed by the global recession. Consequently, many works (Rasche and Tatom, 1977, 1981; Darby 1982; Hamilton 1983; Burbidge and Harrison, 1984 ; Santini, 1985 ; Gisser and Goodwin, 1986) are interested to check if that the recessions were attributable to the oil price shocks. However, those researches pointed out to the negative correlation between oil prices and real output. The most of studies examining the relationship between oil price movement and economic growth are interested to the US economy. These studies have identified a weak effect on the US economic growth and oil price movement. Between 1999 and 2008, the oil prices are rising continually; this is the fourth rally in the last three decades. A further research on the effects of the oil price movements, especially pertaining to the developing country situation, is needed. Such a study would not only fill the gap that the oil macroeconomics literature lacks but would also serve to the needs of the policy makers. As summarized above concisely, the literature on oil is mainly dominated by the studies aiming to explore the oil price changes-­‐GDP growth or the mechanisms, in which such effects work for the US economy. A proportionally lesser number of studies in the relevant literature comprise the oil importing developed countries, especially that of the OECD countries. In other hand, a limited number of studies related to the issue of interest have been made on oil producing economy like Saudi Arabia. In such a setting, where the oil prices are on the rise; and relevant studies on the matter regarding the oil price transmission over the individual economies are limited there is an urgent need of closing this gap. There are few studies that have interested on the relationship between energy and economic growth for the case of exporting countries. For example, Mehrara and Oskoui (2007) have been interested on the sources of economic fluctuations in the case of exporting countries (Iran, Saudi Arabia, Kuwait, and Indonesia). Through a structural VAR approach, four structural shocks are identified: nominal demand, real demand, supply, and oil price shocks. We show that Oil price shocks are the main source of output fluctuations in Saudi Arabia and Iran, but not in Kuwait and Indonesia. Other works have analysed the relationship between energy and economic growth. However, this relationship is made out through the energy consumption rather than oil prices movement (Mehrara, 2007; Ozturk et al., 2010; Sadr Seyed, 2012; Damette and Seghir, 2013; Behmiri and Pires Manso, 2013). We aim at complementing the existing studies to uncover whether the results of the previous literature are robust to model specification, in particular in the dynamic dimension of the oil-­‐economic growth relationship when importing and exporting countries are accounted for. 4 3.
Empirical analysis
Previous studies, cited above, reach different conclusions and focus mainly on importing countries. In this paper, we aim at complementing the existing studies uncovering the oil-­‐economic growth relationship under the case of the exporting countries. In addition, from an empirical point of view, we propose a new perspective in this literature. We employ a frequency approach through the evolutionary co-­‐spectral proposed by Priestley and Tong (1973). 3.1. Methodology
The interaction between the series is measured according to a frequency approach based on the theory of
evolutionary co-spectral analysis of Priestley and Tong (1973). This methodology has been adopted by Ftiti (2010),
Allégret and Essadi (2010) and Bouichouicha and Ftiti (2012).1
While existing studies applies either VAR model, cointegration or volatility analysis, we choose evolutionary
co-spectral analysis because it presents several advantages. First, this kind of analysis does not impose any restrictions or pre-­‐treatment of the data (as it is the case of volatility analysis, for instance, which requires the series to be stationary or cointegration techniques which can be only applied to time series data integrated of order one). Second, it does not have an “end-­‐point problem”: no future information is used, implied or required as in band-­‐pass or trend projection methods. The most important advantage of frequency analysis consists of the decomposition of series on two dimensions, that is frequency and time occurrence of the dependence. This allows to study time series according to different horizons, for instance short and medium-­‐
term. This additional information allows understanding which cycles and periods are more synchronized than others. In addition, the evolutionary co-­‐spectral analysis gives a robust frequency representation of non-­‐
stationary series. We measure co-movement between series by the coherence function. In our analysis, we are interested in
time series as discrete process. We analyse two pairs of series (oil price and economic growth) for each country.
Therefore, we detail the procedure to estimate the evolutionary cross-spectral density function.
Let a non-stationary discrete bivariate process
{ X(t),Y (t)} have
the Gramer representation for each
−π  w  π :
Xt = ∫
π
−π
At , X ( w) e iwt dZ X ( w)
and
Yt = ∫
π
−π
At ,Y ( w) e iwt dZ Y ( w)
(1)
with
1
For more details on the evolutionary co-spectral analysis estimation, see Ftiti (2010).
5 E[dZ x ( w1 ) dZ x* ( w2 )] = E[dZ y ( w1 ) dZ *y ( w2 )]
= E[dZ x ( w1 ) dZ *y ( w2 )] = 0
2
for w1 = w2
2
E[ dZ x ( w1 ) ] = dµ xx( w1 ) , E[ dZ y ( w1 ) ] = dµ yy ( w1 ), and
E[dZ X ( w1 ) dZ *y ( w1 )] = dµ xy ( w)
By virtue of the Cauchy–Schwarz inequality, we can write that:
2
dH t , XY ( w) ≤ dH t, xx (w) dH t,YY (w),
And
dH t , XY ( w) = ht , XY ( w)dw
where
ht , XY ( w)dw
for each
t and w
may then be termed the evolutionary cross-spectral density function.
The estimation of the evolutionary cross-spectral density function needs two filters. For the discrete univariate
process, Priestley (1966) gives two relevant windows. These are relevant filters and they are tested by several works,
such as Ahamada and Boutahar (2002) or Ftiti and Essaadi (2008).
3.2. Coherence Function
According to Priestley and Tong (1973), the evolutionary cross-spectral density function may be written as:
ht , XY ( w) = C t , XY ( w) − i Qt , XY ( w)
(2)
C t , XY ( w) = R{h XY ( w j , t )}
(3)
Qt , XY ( w) = Im{h XY ( w, t )}
The real-valued functions
C t , XY ( w) and Qt , XY ( w) termed the evolutionary co-spectrum and the evolutionary
quadrature spectrum, respectively. If the measures
µ XX ( w) and µ YY ( w)
are absolutely continuous, Priestley and
Tong (1973) similarly define the evolutionary auto-spectral density functions,
h XX ( w j , t ) , hYY ( w j , t ) . The
coherency function is defined by the following expression:
C t , XY ( w) =
ht , XY ( w )
{h
=
( w) ht ,YY ( w}
1/ 2
t , XX
{E dZ (w)
x
2
E dZ Y ( w)
Priestley and Tong (1973) interpret
(4)
E[dZ Y ( w) dZ * ( w)]
}
2 1/ 2
C t , XY ( w) as the modulus of the correlation coefficient between
6 dZ X ( w), dZ Y ( w)
or, more generally, as a measure of the linear relationship between corresponding components
at frequency w in the processes
{Y (t )} and {X (t )} .
The estimation of the coherency function is based on the estimation of the cross-spectral density function
between two processes
{Y (t )} and {X (t )} ,
and the estimation of the auto-spectral density function of each
process. So, the estimation coherency can be written as follows:
Cˆ t , XY ( w) =
hˆt , XY ( w )
{hˆ
t , XX
}
1/ 2
( w) hˆt ,YY ( w
(5)
3.3. Data description
In this study, we use monthly data for oil prices and Gross domestic product. The sample consists of oil-­‐
importing (US) and exporting countries (Saudi Arabia, United Arab Emirates, Kuwait and Venezuela). The selected sample is based on the criteria of a rank in the top 10 oil-­‐importers and exporters countries. The Brent crude oil index is used as it accounts for the 65% of the world oil daily production (IMF, 2010). We use gross domestic product (GDP) to determine the economic growth rate (𝑌! ) for the selected coutries through the following formula: 𝑌! = 100 ∗ ln(
!"!!
!"#!!!
) (6)
The data range from 03/09/2000 to 03/12/2010 and have been extracted from Federal Reserve Bank of Saint Louis and Datastream Database. The time horizon depends on data availability and includes, in addition to the major economic crisis and political events such as the different monetary and financial crises in Asian and Latin American and Middle East region, the first and the Gulf war, the Russian economic crisis and the terrorist attack in US. Note that for the evolutionary spectral estimation necessity, we lose ten observations at the beginning and at the end. So, the estimated evolutionary spectral density functions of inflation series for our countries sample are started from 03/08/2001 to 03/02/2010. 4. Empirical results and discussion
4.1. Oil price movement
Figure 1 presents the Brent crude oil prices, in dollars, from September 2000 to October 2010 in levels.2 2
The series appears to have a non-stationary behavior in the sense that it does not converge towards its long term means while the series in first
difference fluctuates around zero and seems to be stationary.
7 Figure 1. Brent crude oil price ($), 2000-2010
Oil price movements show some important peaks and troughs during the period of the study. The main peaks are observed between 2007 and 2008. Another peak is observed in June 2009, where prices increased by more than 60% since the January 2009 price levels. All these changes are linked to aggregate demand-­‐side oil price shocks (table 1). The first one occurred during the Asian economic crisis, the second took place in 2000, where interest rates decreased significantly creating a bust in the housing market and construction industries. The third one took place in the period 2006–2007, a result from the rising demand of oil from China and the fourth demand-­‐side oil price shock took place in the global financial crisis of 2008. 4.2. Time-varying coherence functions
The analysis resulting from the time-­‐varying coherence functions as computed from equation (5) between each stock market index and the Crude oil prices is shown in Figure 2, for oil-­‐importing and 3 to 6, for oil-­‐
exporting countries respectively. Figures 2a to 6b : Coherence functions between the growth rate and the oil price Fig 2a : The Medium-­‐run coherence function between the growth rate of USA and the oil price Fig 2b : The short-­‐run coherence function between the growth rate of USA and the oil price 8 Fig 3a : The Medium-­‐run coherence function between the growth rate of Venezuela and the oil price Fig 3b : The short-­‐run coherence function between the growth rate of Venezuela and the oil price Fig 4a : The Medium-­‐run coherence function between growth rate of Saudi Arabia and the oil price Fig 4b : The short-­‐run coherence function between growth rate of Saudi Arabia and the oil price Fig 5a : The Medium-­‐run coherence function between the growth rate of the United Arab Emirates and the oil price Fig 5b : The short-­‐run coherence function between the growth rate of the United Arab Emirates and the oil price 9 Fig 6a : The Medium-­‐run coherence function between the growth rate of Kuwait and the oil price Fig 6b : The short-­‐run coherence function between the growth rate of Kuwait and the oil price The above graphics show a divergence between the medium-­‐term interdependence of the oil price and economic growth and the short-­‐term one for the case of exporting countries. However, for the case of the USA (importing country) there is a similar pattern between the short-­‐run and the long-­‐run interdependance. 4.3. Discussion In the short-­‐term, the average interdependence does not exceed 20%, while in the medium-­‐term, on average, it reaches more than 60%. Hence economic environement reacts weakly to transitory fluctuations of oil price (short-­‐term interdependence). However, economic activities for all countries, instead, react to persistent fluctuations of oil price (medium-­‐term interdependence). These findings are in line with Barsky and Kilian (2001, 2004). From the graphics 2b to 6b, we observe low interdependencies betewen oil price and real activity in the short-­‐run particulary in a stable period. These interdependences rise slightly in crisis period. We show that 10 the interaction between real activity and oil price index are higher in 2001 and in 2008 than other period. In other words, economic environnement is slowely sensible to some exogenous shocks (see the appendix). Our results are in line with previous studies that showing a partial relationship between oil prices mouvement and economic growth (Hamilton, 1983; Papapetrou, 2001; Miguel et al.,2003; Limin et al., 2010…) Graphics 2a to 6a show that the dynamic interaction pattern in the medium-­‐term is different form the short-­‐run one. Firstly, the medium-­‐term intercation is very higher than in the short-­‐run for the case of exporting countries. In the case of this kind of country (exporting countries) the medium-­‐term interaction is, on average, more than 60-­‐70%. Moreover, OPEC countries face higher and multiple peaks in the dynamic coherence patterns, which coincide with very important events (such as the 2008 oil price crisis). We conclude that real activity in exporting countries is highly interdependent to oil price. This outcome is explained by the fact that oil sector is the mainly important one in the case of exporting countries specially for the GCC countries. Our result is contrary on previous studies that showing that is no long-­‐run relationship between oil-­‐price and economic growth and (Cunado and Perez de Gracia, 2003). More explicitly, for the case of all exporting countries, we also observe a peak in the medium-­‐term coherence pattern observed around the year 2001 for all countries (80%). This high level of coherence between oil and real activity is due to the rapid increase in the housing market and construction industry, a result of decreasing interest rates worldwide in 2000. In addition, the 2001 attack can explain the higher coherence level observed in this period. In 2003 the coherence pattern is smoothed. This result can be explained by the war in Iraq in Mars 2003 and PdVSA Strike in Venezuela. We observe a breakdown, for all exporting countries, in coherence pattern in 2006. We explain this decrease in interdependence between oil price and stock market index by the military attack in Nigeria which caused the shutting down of more than 600 000 billion barrel per day. Sin the end of 2006 to the end of 2008, we aboserve an continually increase of interaction between oil price and real activity. This increase is explained mainly by the high of oil prices due to the rinsing of China demand. This aggregate demand-­‐side oil price shock has a positive effect on real activity, as it signals an increase in world trade. These findings are in line with Hamilton (2009b) and Kilian and Park (2009), who suggest that aggregate demand-­‐
side oil price shocks, originated by world economic growth, have a positive impact on macroeconomic performance. For the case of importing countries (USA), Figures 2a and 2b show a simillar pattern than the case of exporting-­‐countries. However, the diffrence between the two kind of countries is the magnitude of the dynamic interaction. There are only three periods of noteworthy higher or lower coherence between oil prices and real activity for exporting countries. These are the early 2000 until 2001; 2006-­‐2009 (aggregate demand-­‐side oil price shocks — higher coherence), and 2002-­‐2006 low coherence (precautionary demand-­‐side oil price shocks — low coherence). 11 The explanation of such findings can depend on the boom that the housing market experienced in 2000 creating a positive environment for world markets and at the same time a high demand for oil, driving an increase in the real activity. The 9/11 terrorist attack created significant uncertainty in all economies, causing similar recession environement and thus similar coherence with oil prices. In addition, the Chinese growth and its impact in the world trade caused euphoria in all stock markets regardless the country of origin. Similarly, the last world financial crisis influenced all stock market similarly and thus their co-­‐movements. Through our result, we suggest, on the one hand, that news about global growth predicts much of the surge in the real price of oil from mid 2003 until mid 2008 and much of its subsequent decline, and in the other hand, that the surge in the real price of oil is explained primarily by rising global demand for industrial commodities driven by unexpected economic growth. From the transmission channel between business cycle and oil price, our results have some important explanation of these transmission mechanisms. Indeed, there are different issues why an oil shock should affect macroeconomic variables. For example, the non-­‐linear specification of the oil price–macroeconomy relationship. For example, the oil shock can lead to lower aggregate demand since the price rise redistributes income between the net oil import and export countries. In addition, the oil price increase reduces aggregate supply since higher energy prices mean that firms purchase less energy. As a consequence, the productivity of any given amount of capital and labor declines and potential output falls. The decline in factor productivity implies that real wages will be lower. If some labor supply is withdrawn voluntarily as a result, potential output will be lower than it would otherwise be, thus compounding the direct impact of lower productivity. Our analysis shows two main findings. Firstly, the exporting countries are more sensible to oil price shocks than importing countries. Secondly, we show that oil price shocks in periods of world turmoil or during fluctuations of the global business cycle (downturn or expansion) exhibit a significant impact on the relationship between oil and real economic activity in OPEC countries ; aggregate demand-­‐side oil price shocks (housing market boom, Chinese economic growth, and the latest global financial crisis) cause a significant higher correlation between real activity and oil prices, and important precautionary demand-­‐side oil price shocks (i.e. second war in Iraq, terrorist attacks) tend to cause lower coherence. 5. Conclusion and policy implications This study aims to contribute to the literature on the macroeconomic consequences of oil shocks. Firstly, we complement the existing studies, highlighting the oil-­‐economic growth relationship under the case of the exporting countries. Secondly, we propose an empirical methodology (the evolutionary co-­‐spectral analysis) that distinguishes between the short-­‐run dynamic co-­‐movement and the long-­‐run one. Our results show that oil price shocks have both the medium-­‐term and short-­‐term effects on economic growth. However, the impact of the medium-­‐term one is more greater than those of short-­‐run. These results find out about the 12 transmission mechanism channel of the relationship between oil price and business cycle. Firstly these transmission mechanisms are active in short-­‐run and the medium-­‐run. More explicitly, the dependence between oil price and business cycle can be explained through the impact of oil shock on aggregate demand. Indeed, an increase of an oil price reduces aggregate supply since higher energy prices mean that firms purchase less energy. As a consequence, the productivity of any given amount of capital and labor declines and potential output falls. The decline in factor productivity implies that real wages will be lower. The major contribution of this paper is to distinguish these two types of effects. The medium-­‐term effect is due to the aggregate demand-­‐side oil price shocks (housing market boom, Chinese economic growth, and the latest global financial crisis) cause a significant higher correlation. While the short-­‐term effect is caused by important precautionary demand-­‐side oil price shocks (i.e. second war in Iraq, terrorist attacks) tend to cause lower coherence. From our analysis there are three main lessons in terms of policy prescriptions. First, a long oil price run up contributes to declines in the GDP components as during the period of the subprime crisis (figure 1). This in turn generates persistent declines in oil prices. This run up traduces the dominance of the demand, rather than supply which mean that barring any other intervening events, the oil price slump may last only as long as the recession. This is the consequence of many reasons such as the lack of any large disrupted capacity that would create additional surplus capacity once resolved, and the potentially large demand from emerging economies, such as India, China. The second lesson is that exogenous events affecting the oil market, whether producing supply shortfalls or not, tend to generate short-­‐ to medium-­‐term oil price effects. This is shown in the Iraq invasion that did not lead to a significant supply shortfall, but produced both short-­‐ and medium-­‐run oil price increases. Thus, the multitude of exogenous events during the 2000s (table 2, appendix), though coincident with rising demand, by themselves played an important role in the oil price increases of that period. The approach in Kilian (2007a, b) emphasize the context of individual oil price shocks by separating the ‘‘precautionary demand’’ following such events from actual supply losses. However, because these ‘‘precautionary demands’’ are direct consequences of the underlying events, it seems appropriate to also count the price, and other implications, as part of the overall effects of those events. The third lesson is that the OPEC countries are still too dependent on oil exports, which is for them a source of instability. Countries such as Kuwait, United Arab Emirates, Saudi Arabia and Qatar have shown the way by seeking new growth drivers and actively preparing for the post-­‐oil stage. Public investment policies of some of the oil surpluses are needed. These policies have several advantages: 1 / diversification of income sources, 2 / reduce the impact of oil price shocks on growth 3 / defense of national interests in areas where deficiencies are evident (eg agriculture for countries of the Gulf). Being long remained the "rentier states" -­‐ prone to develop policies based on the allocation of income by the State and not the creation of new wealth through the production (Mahdavy 1970; Luciani, 1987) -­‐ is a handicap to overcome. This paper has a number of potential extensions. Among these is development of capabilities to study through of frequency approach the causality between oil price and growth rate. This kind of study can have an 13 impact to more understand the transmission mechanisms channels between oil price and real sector. In addition, it is useful to study the relationship between oil prices and other economic variables, such as inflation, interest rates, and employment could also be analysed using the same methodology. These and other extensions are left to future efforts. References
Ahamada I., Boutahar M. (2002), Tests for covariance stationarity and white noise, with application to Euro/US Dollar exchange rate. Economics Letters 77, 177-­‐186. Barsky R., Kilian L. (2001), Do we really know that oil caused the great stagflation ? A monetary alternative, NBER Working Papers 8389. Barsky R., Kilian L. (2004), Oil and the Macroeconomy Since the 1970s. Journal of Economic Perspectives 18 (4), 115-­‐
134. Behmiri N., Pires Manso J. (2013), Hox Crude Oil impacts on Economic Growth : An empirical Investigation in Multiple Regions, Working Paper, available at SSRN. Bernanke B., Gertler M., Watson M. (1997), Systematic Monetary Policy and the Effects of Oil Price Shocks. Brookings Papers on Economic Activity 1, 91-­‐157. Berument H., Ceylan N.B. (2010), The impact of oil price shocks on the economic growth of selected MENA countries. The Energy Journal 31 (1), 149-­‐176. Bouchouicha R., Ftiti Z. (2012), Real estate markets and the macroeconomy: a dynamic coherence framework. Economic Modelling 29, 1820–1829. Burbidge J., Harrison A. (1984), Testing for the Effects of Oil-­‐Price Rises Using Vector Autoregressions. International Economic Review, vol. 25 (2), 459-­‐84. Cunado J., Perez de Gracia F. (2003), Do oil price shocks matter ? Evidence from some European countries, Energy Economics 25 (2), 137-­‐154. Damette O., Seghir M. (2013), Energy as a Driver of Growth in oil exporting countries ? Energy Economics 37, Issue C, 193-­‐199. Darby M. (1982), The Price of oil and World inflation and Recession. American Economic Review 72 (4), 738-­‐751. Fan Y., Jiao J.-­‐L., Liang Q.-­‐M., Han Z.-­‐Y., Wei Y.-­‐M. (2007), The impact of rising international crude oil price on China's economy: an empirical analysis with CGE model. International Journal of Global Energy Issues 27 (4), 404-­‐424. Ftiti Z. (2010), The macroeconomic performance of the inflation targeting policy: An approach based on the evolutionary co-­‐spectral analysis (extension for the case of a multivariate process). Economic Modelling 27 (1), 468-­‐476. Ftiti Z., Essaadi E. (2008), The inflation Targeting effect on the inflation series: A new Analysis Approach of evolutionary spectral analysis. GATE working paper No. 08/32. Gisser M., Goodwin T. (1986), Crude Oil and the Macroeconomy: Tests of Some Popular Notions. Journal of Money, Credit, and Banking 18, 95-­‐103. Gudarzi Y., Sadr Seyed M. (2012), Causality between oil consumption and economic growth in Iran: an ARDL testing approach. Asian Economic and Financial Review 2 (6), 678-­‐686. Hamilton J. (1983), Oil and the Macroeconomy Since World War II. Journal of Political Economy 91, 228-­‐248. Hamilton J. (2003), What is an Oil Shock? Journal of Econometrics 113 (2), 363-­‐398. Hamilton J. (2009a), Understanding crude oil prices. The Energy Journal 30 (2), 179-­‐206. Hamilton J. (2009b), Causes and consequences of the oil shock of 2007-­‐08. Brooking papers on Economic Activity, spring, 215-­‐261. He Y., Wang S., Lai K. (2010), Global economic activity and crude oil prices : A cointegration analysis. Energy Economics 32 (4), 868-­‐876. Hooker M.A. (1996), What Happened to the Oil Price-­‐Macroeconomy Relationship? Journal of Monetary Economics 38, 195-­‐213. 14 IMF (2010), World Economic Outlook. Kaufmann R., Dees S., Karadeloglou P., Sanchez M. (2004), Does OPEC Matter ? An Econometric Analysis of Oil Prices. The Energy Journal 1 (4), 67-­‐90. Kilian L., Park C. (2009), The impact of oil price shocks on the U.S market. International Economic Review 50 (4), 1267-­‐1287. Kilian L., Vigfusson R. (2011), Nonlinearities in the oil price-­‐output relationship. Macroeconomic Dynamics 15 (3), 337-­‐363. Levin A., Loungani P. (1996), Oil, money and economic activity in the G7 countries. Paper presented at the DOE Conference, Washington D.C., october. Limin D., He Y., Chu W., (2010), The relationship between oil price shocks and China's macro-­‐economy: An empirical analysis. Energy Policy 38 (8), 4142-­‐4151. Luciani G. (1987), Allocation vs Production States : A Theoretical Framework, in H. Balhami & G. Luciani, The Rentier State. Nation, State and Integration in the Arab World, New York, Croom Helm, 63-­‐82. Lynch M. (2002), Forecasting oil supply: theory and practice. The Quarterly Review of Economics and Finance 42 (2), 373-­‐389. Mahdavi H. (1970), The Patterns and Problems of Economic Developments in Rentier States : the case of Iran, in M.A. Cook (ed.), Studies in the Economic History of the Middle-­‐East, Oxford, Oxford University Press. Mehrara M. (2007), Energy consumption and Economic Growth: The case of oil exporting countries. Energy Policy 35 (5), 2939-­‐2945. Mehrara M., Oskoui K. (2007), The source of macroeconomic fluctuations in oil exporting countries: A comparative study. Economic Modelling 24 (3), 365-­‐379. Miguel C., Manzano B., Martin-­‐Moreno J. (2003), Oil price shocks and aggregate fluctuations. The Energy Journal 24 (2), 47-­‐61. Merino A., Ortiz A. (2005), Explaining the so-­‐called “price premium" in oil markets. OPEC Review 29, 133-­‐152. Ozturk I., Aslan A., Kalyoncu H. (2010), Energy Consumption and Economic Growth relationship evidence from panel data for low and middle income countries. Energy Policy 38 (8), 4422-­‐4428. Papapetrou E. (2001), Oil prices shocks, stock market, economic activity and employment in Greece. Energy Economics 23 (5), 511-­‐532. Priestley M.B. (1966), Design relations for non stationary processes. Journal of Royal Statistic Society. ser. B, 28, 228-­‐240. Priestley M.B., Tong H. (1973), On the analysis of bivariate non-­‐stationary processes. Journal of Royal Statistic Society. ser. B, 35, 135-­‐166. Rasche R., Tatom J. (1977), The effects of the new energy regime on economic capacity, production, and prices. Review, Federal Reserve Bank of St. Louis, issue may, 2-­‐12. Rasche R., Tatom J. (1981), Energy price shocks, aggregate supply and monetary policy: the theory and the international evidence. Carnegie-­‐Rochester Conference Series on Public Policy 14 (1), 9-­‐93. Rogoff K. (2005), Oil and the global economy. International Energy Forum meeting, Riyadh, november 19. Rotemberg J., Woodford M. (1996), Imperfect Competition and the Effects of Energy Price Increases on Economic Activity. Journal of Money, Credit and Banking 28 (4), 550-­‐577. Santini D. (1985), The Energy-­‐Squeeze Model: Energy Price Dynamics in U.S. Business Cycles. International Journal of Energy Systems 5, 18-­‐25. Sadorsky P. (2000), The Empirical Relationship between Energy Futures Prices and Exchange Rates. Energy Economics 22 (2), 253-­‐266. 15 Appendix Table 1-­‐ Oil price shock origin and their main events Events Year Oil price shock origin Housing market boom 2000 Aggregate demand side 09/11 Attacks 2001 Precautionary demand PdVSA worker’s strike 2002 Supply side Second war in Iraq 2003 Precautionary demand Chinese economic growth 2006-­‐2007 Aggregate demand side Global financial crisis 2008 Aggregate demand side Golbal debt crisis 2010 Aggregate demand side Sources: Kilian’s (2009) and Hamilton (2009a,b) findings. 16 Table 2 -­‐ Oil price chronology from 2000 to 2010: the main events Months 2000 2001 2002 2003 2004 2005 2006 2008 2009 2010 01 OPEC
decides to
cut quotas
Rising
demand, low
spare capacity
02 OPEC
decides to
cut quotas
at various
meeting
03 War in
Irak
04 05 OPEC oil agree
to increase the
oil production
Breakdown
of more
600,000bbl/
d of oil
production
due to
Nigeria
attacks
U.S president
sign into law
a bill that
temporary
halts adding
oil to the
strategic
petroleum
reserve
06 07 08 Hurricane
Katrina,
Dennis, and
Rita Strike
09 09/11
Attacks
Hurricane
Ivane
Striles
10 11 12 PdVA
Strike in
Venezuela
Source: US Energy Information Administration. OPEC
decides
to cut
quotas at
various
Hurricane
Gustav strikes
OPEC
decides to cut
quotas
17