THE JOURNAL OF ENERGY AND DEVELOPMENT

Mohamed El Hdi Arouri, The Impact of Oil Price Shocks on the European

Stock Market (1998-2008): A Sector Analysis,

Volume 35, Number 2

Copyright 2011

www.iceed.org

THE IMPACT OF OIL PRICE SHOCKS ON THE EUROPEAN STOCK MARKET (1998-2008): A SECTOR ANALYSIS
di Mohamed El He Arouri*

Introduction nderstanding the dynamics of stock prices is an issue of ongoing research in financial markets literature. In particular, identifying the factors that affect stock market returns is of the utmost relevance and importance to investors and policy makers. Although an abundance of theoretical and empirical works focused on asset pricing, there is no consensus about the nature and number of factors that affect stock prices. Furthermore, as the oil price has changed with sequences of very large increases and decreases over recent years, it is now quite opportune to augment the existing research on its impacts on stock market returns. Various transmission channels exist through which oil price fluctuations may affect stock prices. Indeed, the value of stock in theory equals the discounted sum of expected future cash flows. These discounted cash flows reflect economic conditions (inflation, interest rates, production costs, income, economic growth, investor and

*Mohamed Arouri is currently a full Professor of Finance at the Universite dAuvergne, France, and an associate researcher at EDHEC Financial Analysis and Accounting Research Centre. He holds a masters degree and a Ph.D. in economics from the Universite de Paris X Nanterre and a Ph.D. in management science from the Universite dOrleans. His principal research areas concern energy economics, emerging markets finance, market efficiency, volatility modeling, and risk management in international capital markets. The authors most recent articles have been published in such refereed journals as Macroeconomic Dynamics, Energy Economics, Journal of International Money and Finance, Energy Studies Review, Economic Modelling, Energy Policy, International Journal of Finance and Economics, Managerial Finance, Manchester School, and Applied Financial Economics. The Journal of Energy and Development, Vol. 35, Nos. 1 and 2 Copyright 2011 by the International Research Center for Energy and Economic Development (ICEED). All rights reserved.

165

166

THE JOURNAL OF ENERGY AND DEVELOPMENT

consumer confidence, etc.) and then they are affected by macroeconomic events that possibly can be influenced by oil shocks. Thus, oil price changes may impact stock prices. In the literature, there has been a large volume of works on the linkages between oil prices and economic variables. Most of these studies have shown significant effects of oil price fluctuations on economic activity for several developed and emerging countries.1 In contrast, there have been relatively few attempts to study the relationship between oil price variations and stock markets. The pioneering paper by C. Jones and G. Kaul tests the reaction of stock returns in developed markets (Canada, Japan, the United Kingdom, and the United States) to oil price fluctuations on the basis of the standard cash flow dividend valuation model.2 They find that for the United States and Canada this reaction can be accounted for entirely by the impact of the oil shocks on cash flows. The results for Japan and the United Kingdom, however, were inconclusive. Using an unrestricted vector autoregressive (VAR), R. Huang et al. find no evidence of a relationship between oil prices and the S&P (Standard & Poors) 500 market index.3 P. Sadorsky, by contrast, applies an unrestricted VAR with GARCH effects to U.S. monthly data and shows a significant relationship between oil price changes and aggregate stock returns in the U.S. market.4 J. Park and R. Ratti show that oil prices have a negative impact on stock returns in the United States and in 12 European nations; stock markets in Norway, an oil-exporting country, on the other hand, respond positively to rises in the price of oil.5 More recently, N. Apergis and A. Miller have examined whether structural oil-market shocks affect stock prices in eight developed countries.6 They find that international stock market returns do not respond significantly to oil price shocks. Very few studies have looked at the impact of oil price changes on the stocks of individual sectors. In addition, most of these studies are country specific and thus do not provide a global perspective. For instance, P. Sadorsky and M. Boyer and D. Filion show that oil price increases affect the stock returns of Canadian oil and gas companies positively.7 I. El-Sharif et al. reach the same conclusion for oil and gas returns in the United Kingdom.8 However, the authors show that non-oil and gas sectors are weakly linked to oil price changes. More recently, M. Nandha and R. Faff study the short-term link between oil prices and 35 DataStream global industries and show that oil price rises have a negative impact on all but the oil and gas industries.9 Finally, M. Nandha and R. Brooks look into the reaction of the transport sector to oil prices in 38 countries and find that, in developed economies, oil prices have some influence on the returns to the sector.10 However, there appears to be no evidence of a significant role for oil in Asian and Latin American economies. Taken together, the results from the work on the relationships between oil prices and sector stock returns differ from country to country and from sector to sector. The current article extends the understanding of the relationship between oil prices and stock returns at the disaggregated sector level in Europe by investigating

OIL PRICE SHOCKS & EUROPEAN STOCK MARKETS

167

short linkages using different econometric techniques over the last turbulent decade (1998-2008). Studying the short effects of oil price fluctuations sector by sector is important for several reasons. First, any marketwide consequence may mask the performance, not necessarily uniform, of various sectors. Sector sensitivities to changes in the price of oil can be asymmetric; some sectors may be more severely affected by these changes than others. The sector sensitivities to oil depend on whether oil is an input or an output for the industry, on the indirect effect of oil prices on the industry, on the degree of competition and concentration in the industry, and on the capacity of the industry to transfer oil price shocks to its consumers and thus to minimize the impact of these shocks on its profitability. Second, the industrial base varies from one European market to another. Large, mature markets such as France and Germany are more diversified, whereas small markets such as Switzerland usually concentrate on a few industries. Thus, the results of studies based on a national stock market index, such as J. Park and R. Ratti and N. Apergis and S. Miller, should be considered with a precaution.11 Indeed, it is interesting to know whether sector indices rather than national indices are sensitive to oil price fluctuations. Finally, from the point of view of portfolio management, identifying the heterogeneity of sector sensitivities to oil implies that there are sectors that can still provide a channel for international diversification during large swings in oil prices. The rest of this paper is organized as follows. We first present the data and some preliminary analysis. This is followed by a discussion of the empirical results. The summary conclusions and policy implications are provided in the last section. Data and Preliminary Analysis The aim of this paper is to examine the existence of relationships between oil prices and sector stock returns in Europe. We use weekly stock market indices over the period January 1, 1998 to November 13, 2008. Over this sample period, the relationship between oil prices and economic activity was ambiguous. Indeed, increases in oil prices were indicative, at the same time, of higher production costs and inflation pressure and synonyms of higher expected economic growth and higher levels of consumers and investors confidence. We think that weekly data may adequately capture the interaction of oil and stock prices. We study the DJ Stoxx 600 and 12 European sector indices: (a) automobile and parts, (b) financials, (c) food and beverages, (d) oil and gas, (e) health care, (f) industrials, (g) basic materials, (h) personal and household goods, (i) consumer services, (j) technology, (k) telecommunications, and (l) utilities. Stock market data are obtained from the DataStream database. For oil, we use the weekly Brent oil crude price obtained from the Energy Information Administration (EIA). We express Brent oil prices in euros using exchange rates from DataStream.

168

THE JOURNAL OF ENERGY AND DEVELOPMENT

In order to determine the order of integration of our series, we apply three standard unit root tests: augmented Dickey-Fuller (ADF), Phillips-Perron (PP), and Kwiatkowski et al. (KPSS) tests. The ADF and PP tests are based on the null hypothesis of a unit root, while the KPSS test considers the null of no unit root. Results are reported in table 1. All the series appear to be integrated of order one, which is a standard result in the literature for such series. Descriptive statistics for return series (first logarithmic differences) are summarized in table 2. Oil prices on average have increased more than European stock prices over our sample period. Technology stocks have the highest volatility, followed by oil and automobile stocks. Skewness is negative in most cases and the Jarque-Bera test statistic (JB) strongly rejects the hypothesis of normality. Correlations between oil price changes and European sector returns are weak on average and surprisingly they are positive except for three sectors: automobile and parts, food and beverages, and health care. This suggests that oil price increases over the last decade were seen as indicative of higher expected economic Table 1
UNIT ROOT TEST Levels ADF OilBrent DJ Stoxx Automobile & parts Financials Food & beverages Oil & gas Health care Industrials Basic materials Personal & household goods Consumer services Technology Telecommunications Utilities 0.60 b -0.38 b 0.03 b -0.78 b 0.43 b 0.09 b 0.02 b -0.48 b 0.34 b 0.30 b -0.9 b -0.66 b -0.30 b 0.75
b

First Difference KPSS 0.69** c 0.37* c 0.67** c 0.26 c 1.55* c 1.14* c 0.22 c 0.495** c 2.03* c 1.55* c 1.00* c 1.25* c 1.09* c 1.45*
c

PP -2.68 b -0.26 c -2.14 d -0.66 c -2.11 c -2.01 c -2.07 b -0.25 b 0.322 b 1.87 d 1.34 c 2.00 d -2.10 b 0.76
c

ADF -11.48* b -9.16* b -27.48* b -8.29* b -21.83* b -15.54* b -25.85* b -5.83* b -22.29* b -23.19* b -6.21* b -6.49* b -7.79* b 17.4*
b

PP -18.35* b -23.6* b -27.46* d -23.91* b 21.83* b -23.38* b -25.87* b -22.12* b -22.47* b -23.23* b -22.76* b -24.07* b -23.89* b 22.78*
b

KPSS 0.08 c 0.27 c 0.058 c 0.35* c 0.08 c 0.139 c 0.12 c 0.199 c 0.139 c 0.178 c 0.23 c 0.206 c 0.218 c 0.19
c

a All variables are in natural logs. ADF is the augmented Dickey-Fuller test, PP the PhillipsPerron test, and KPSS the Kwiatkowski-Phillips-Schmidt-Shin test; DJ Stoxx is the European market index; * = indicates significance of coefficients at the 1-percent level; ** = indicates significance of coefficients at the 5-percent level; and *** = indicates significance of coefficients at the 10-percent level. b Model without constant or deterministic trend. c Model with constant without deterministic trend. d Model with constant and deterministic trend.

OIL PRICE SHOCKS & EUROPEAN STOCK MARKETS Table 2

DESCRIPTIVE STATISTICS OF RETURN SERIES
a

169

Standard SkewMean Errors ness Kurtosis Oil Brent DJ Stoxx Automobile & parts Financials Food & beverages Oil & gas Health care Industrials Basic materials Personal & household goods Consumer services Technology Telecommunications Utilities
a

JB 91.36 291.89 5930.1 512.7 217.0 91.31 118.5 341.6 459.5

Correlation with Oil 0.10 1.00 -0.02 0.04 -0.10 0.33 -0.07 0.10 0.12

Correlation with DJ Stoxx 1.00 0.10 0.67 0.92 0.57 0.63 0.63 0.91 0.76

0.21 -0.02 0.02 -0.09 0.05 0.01 0.01 -0.02 0.06

4.61 2.56 4.32 3.29 2.18 3.18 2.54 2.90 3.34

-0.47 -0.68 0.54 -0.53 -0.54 -0.52 -0.03 -0.71 -0.90

4.72 6.21 18.80 7.53 5.82 4.66 5.23 6.52 7.02

0.04 -0.09 -0.11 -0.01 0.07

2.57 2.65 5.62 3.63 2.25

-0.76 -0.48 -0.24 -0.07 -1.57

6.70 5.25 4.70 4.01 13.08

378.6 142.9 73.78 24.70 2635.3

0.03 0.04 0.08 0.04 0.08

0.97 0.88 0.81 0.73 0.72

The test for Kurtosis coefficient has been normalized to zero; JB is the Jarque-Bera test for normality based on excess skewness and kurtosis; DJ Stoxx is the European stock market index.

growth and earnings. The oil and gas sector has the highest correction with oil (33 percent), followed by basic materials (12 percent). Correlations between the European market index (DJ Stoxx 600) and sector returns are high on average. The personal and household goods sector has the highest correlation (97 percent) and food and beverages the lowest (57 percent). Empirical Analysis We investigate the relationships between oil price variations and stock market returns in Europe over the last turbulent decade of 1998-2008. We begin our analysis with the estimation of a multifactor asset-pricing model to investigate the sensitivities of the sector stock returns to oil price and European market changes, then perform the Granger causality test to examine their causal linkages, and, finally, study cyclical comovements.
Sector Returns, Oil Price Changes, and Market Sensitivities: The multifactor model we employ to examine whether the European sector stock returns are sensitive to oil price and global European market changes takes the following form:12

170

THE JOURNAL OF ENERGY AND DEVELOPMENT 1

rit = a + b 3 roilu + c 3 rdji0 + d 3 ri;t1 + eit t t eit ! N 0; hit h2 it =a+

q X k=1

bk e2 i;t1

p X l=1

gl h2 i;t1

where rit is the weekly stock return in country i; roilu is an unexpected change in the oil price measured as the difference bet tween the observed oil price and the expected value of oil price using an AR(1) process; rdj0 represents the world market returns filtered by the oil price returns. More t precisely, rdj0 is the residuals of the ordinary least squares (OLS) regression of the t world market returns (Rdjt) on the oil price changes (roilt): rdjt = a + broilt + rdj0 . t The return innovations are assumed to be normally distributed with zero mean and conditional variance governed by a standard GARCH(q,p) process. The model is estimated using the quasi-maximum likelihood (QML) method; the results are presented in table 3.13 Table 3 contains parameter estimates. The coefficients relating the sector return series to the European filtered returns (coefficients c) are highly significant for all sectors. They vary from 0.52 (defensive sector) for food and beverages to 1.48 (offensive sector) for technology. More interestingly, the coefficients relating the return series to oil price changes (coefficients b) are significant in nine cases, indicating significant short-term effects of oil price fluctuations on European sector stock prices. Oil price increases negatively affect sector returns in two cases (food and beverages and health care), and positively in seven cases, which is consistent with the correlations reported in table 2. In fact, while for some industries higher oil prices imply higher production costs and lower earnings, increases in oil prices in the last decade were synonymous with increases in world demand; thus, higher oil prices are generally observed in periods of high economic growth and then high customer and investor confidence levels. Furthermore, the sign of the oilstock price relationships depends also on the capacity of the industry to transfer oil price shocks to its consumers and thereby to minimize the impact of these shocks on its profitability. Finally, our results show that there is no relationship between oil price changes and stock returns for two European sectors (personal and household goods and telecommunications), while for the automobile and parts industry a weak link is obtained. For these sectors, table 2 reports the weakest correlations. The model we estimated seems to satisfactorily fit the data. The ARCH and GARCH coefficients are significant. We further observe that, in most cases, conditional volatility does not change very sharply as the ARCH coefficients are relatively small in size. By contrast, it tends to fluctuate gradually over time because

OIL PRICE SHOCKS & EUROPEAN STOCK MARKETS Table 3

a b c a b1 g1
a

171

ESTIMATES RESULTSMODEL (1)

R2

ARCH

DW

Automobile & parts -0.025 0.037*** 1.242* (0.082) (0.021) (0.036) Financials -0.101** 0.028** 1.120* (0.043) (0.008) (0.015) Food & beverages 0.102 -0.036* 0.515* (0.071) (0.011) (0.019) Oil & gas -0.055 0.197* 0.808* (0.080) (0.018) (0.033) Health care 0.014 -0.043* 0.608* (0.062) (0.014) (0.024) Industrials 0.028 0.061* 1.073* (0.041) (0.009) (0.017) Basic materials 0.105 0.083** 1.026* (0.071) (0.014) (0.026) Personal & household goods 0.044 0.011 0.889* (0.042) (0.008) (0.015) Consumer services -0.085** 0.028** 0.9334* (0.042) (0.011) (0.019) Technology -0.127 0.071* 1.4823* (0.095) (0.024) (0.035) Telecommunications -0.061 0.005 0.978* (0.075) (0.017) (0.034) Utilities 0.119** 0.024** 0.605* (0.057) (0.012) (0.022)

0.223*** (0.134) 0.869* (0.039) 0.595** (0.028) 0.091** (0.047) 0.045* (0.019) 0.015** (0.007) 0.273* (0.034) 0.028*** (0.017) 0.021*** (0.012) 0.151* (0.052) 0.025** (0.012) 0.401** (0.202)

0.144* (0.037) 0.539* (0.078) 0.157* (0.045) 0.078* (0.026) 0.068* (0.018) 0.094* (0.027) 0.430* (0.110) 0.095* (0.024) 0.083* (0.015) 0.079* (0.024) 0.090* (0.026) 0.122* (0.058)

0.841* (0.044) -

0.454

0.621

2.077

0.851

0.601

1.960

0.335

0.041

1.923

0.907* (0.030) 0.922* (0.019) 0.902* (0.026) -

0.474

0.381

2.089

0.423

0.271

1.957

0.828

0.260

1.998

0.588

0.987

2.097

0.890* (0.028) 0.908* (0.018) 0.907* (0.024) 0.915* (0.021) 0.707* (0.152)

0.763

0.891

2.084

0.790

1.212

1.936

0.664

0.219

2.063

0.536

0.999

1.943

0.527

1.173

2.028

a * = indicates significance of coefficients at the 1-percent level; ** = indicates significance of coefficients at the 5-percent level; *** = indicates significance of coefficients at the 10-percent level; and robust standard errors are in parentheses. We also tested for GACH in mean effect; coefficients are not significant. Models are estimated with AR(1) term when the latter is significant.

172

THE JOURNAL OF ENERGY AND DEVELOPMENT

of the large GARCH coefficients. Note also that the estimates coefficients g and b satisfy the stationary conditions. Finally, we test for the absence of autocorrelation and ARCH effects in the residual series. The results indicate that the model specification we use is flexible enough to capture the dynamics of returns. To sum up, our analysis shows strong linkages between oil price changes and most European sector returns over the period under consideration. The sign and intensity of these linkages differ from one sector to another. In the following section, we test for asymmetries in the responses of European sector returns to oil price shocks.

Asymmetric Reaction to Oil Shocks: Some recent papers have shown that the link between oil and economic activity is not entirely linear and that negative oil price shocks (price increases) tend to have larger impacts on growth than positive shocks do (J. Hamilton, D. Zhang, and A. Cologni and M. Manera).14 Thus, we should expect that oil prices equally affect stock markets in an asymmetric fashion. In order to test for asymmetry in the reaction of European sector returns to oil price shocks, we estimate the following asymmetric multifactor model:

rit = a + b + Dt 3 roilu + b 1Dt 3 roilu + c 3 rmsci0 + d 3 ri;t1 + eit t t t eit ! N 0; hit h2 it =a+
q X k=1

bk e2 i;t1

p X l=1

gl h2 i;t1

where Dt is a dummy variable taking one if unexpected change in oil price is positive and zero if it is negative. Thus, b+ and b are the coefficients corresponding to increases and decreases in oil price expectations, respectively. There is no asymmetry if b+ and b are not statistically different from each other. Then, we test for this null hypothesis: b+ = b. Moreover, we test for the hypothesis of non-asymmetry and null sensitivities to oil price increases and decreases: b+ = b = 0. Our main empirical results are summarized in table 4. Wald tests show that the hypothesis b+ = b = 0 is rejected mostly at the 1-percent level in 10 cases, which confirms the significance of oil price shocks as a factor affecting sector returns in Europe. Oil price changes do not significantly affect stock returns for automobile and parts and telecommunications industries, which is consistent with our findings reported in table 3. The only exception is for personal and household goods, for which we found no significant reaction to oil shocks using a symmetric asset-pricing model. Indeed, asymmetric results in table 4 show that stock returns in this sector react negatively to expected oil price increases and negatively to expected oil price decreases. Wald tests confirm this result and show that the hypothesis b+ = b is rejected for personal and household

OIL PRICE SHOCKS & EUROPEAN STOCK MARKETS Table 4

b
+

173

ESTIMATES RESULTSMODEL (2) b

b =b =0

b =b

R2

ARCH

DW

Automobile & parts 0.002 0.067 (0.037) (0.042) Financials 0.022 0.033* (0.016) (0.012) Food & beverages -0.089* 0.001 (0.024) (0.018) Oil & gas 0.255* 0.224* (0.043) (0.029) Health care 0.014 -0.094* (0.032) (0.026) Industrials 0.058* 0.074* (0.021) (0.019) Basic materials 0.048 0.128* (0.030) (0.028) Personal & household goods -0.030*** 0.051* (0.017) (0.018) Consumer services 0.029 0.028 (0.022) (0.023) Technology 0.107** 0.032 (0.045) (0.040) Telecommunications 0.031 0.007 (0.041) (0.037) Utilities 0.046** 0.602* (0.022) (0.022)

1.574 (0.207) 6.885 (0.001) 6.374 (0.018) 100.7 (0.000) 7.405 (0.001) 20.97 (0.000) 17.42 (0.000) 4.321 (0.013) 3.185 (0.042) 4.498 (0.000) 0.482 (0.619) 2.404 (0.091)

0.910 (0.339) 0.196 (0.658) 4.669 (0.031) 0.229 (0.631) 4.492 (0.034) 0.204 (0.651) 2.597 (0.108) 7.430 (0.007) 0.002 (0.964) 1.069 (0.301) 0.123 (0.726) 1.732 (0.188)

0.453

1.692

2.078

0.850

0.599

1.961

0.354

0.061

1.912

0.475

0.286

2.096

0.425

0.400

1.961

0.825

0.157

2.002

0.585

0.022

2.025

0.763

0.817

2.093

0.789

1.119

1.937

0.663

0.103

2.063

0.537

0.558

1.923

0.528

1.168

2.034

a * = indicates significance of coefficients at the 1-percent level; ** = indicates significance of coefficients at the 5-percent level; *** = indicates significance of coefficients at the 10-percent level; and robust standard errors are in parentheses. We also tested for GACH in mean effect; coefficients are not significant. Models are estimated with AR(1) term when the latter is significant.

174

THE JOURNAL OF ENERGY AND DEVELOPMENT

goods and for two other industries (food and beverages and health care). For these industries, reactions to oil price changes are asymmetric.
Causality Tests: In order to further examine the relationships between oil price changes and sector stock returns in Europe, we proceed to Granger causality tests on the return series. Results are reported in table 5. Since some variables as well as their bilateral effects are very sensitive to the selected number of lags in the analysis, we decide to implement this test for different lags. The results show that there is a bidirectional causality between oil price changes and DJ Stoxx returns. Indeed, DJ Stoxx returns Granger-cause changes in oil prices at the 10-percent level, and oil price shocks Granger-cause changes in DJ Stoxx returns. Similar results are obtained for automobile and parts, food and beverages, oil and gas, industrials, personal and household goods, consumer services, and the utilities industries. There are unidirectional Granger causalities from oil to stock returns for financials, health care, and technology industries and from stock returns to oil for the basic materials industry. Finally, there was no significant causality between oil price changes and telecommunications stock returns. Taken together, our causality test results corroborate our previous findings and suggest significant interactions between oil prices and stock prices, except for telecommunications stocks. Our causality test results are interesting for at least two reasons. First, they imply some predictability in oil and European stock price dynamics. Second, in contrast to several works in the literature, which assume exogeneity of oil prices with respect to macroeconomic and financial variables, our findings suggest that a reverse relationship may exist. Indeed, changes in some European sector returns do significantly affect world oil prices (B. Ewing and M. Thompson and F. Lescaroux and V. Mignon).15 Cyclical Correlations between Oil Prices and Stock Markets: Finally, we examine

cyclical correlations as a measure of short-run linkages between oil price changes and sector stock returns in Europe. To this end, we follow the methodology introduced by A. Serletis and A. Shahmoradi and applied by several papers to study the links between energy prices and economic activity.16 First, the HodrickPrescott (HP) filter is employed to decompose each time-series variable in our study into long-run and business cycle components. Next, we compute the crosscorrelations between the cyclical component of oil price (coilt) and that of stock market indices (cstockt). We denote these correlations by r(j) and they are computed for j = 0, 1, 2, 3, 4, 5, and 6. By doing so, the cyclical correlations assess the linkages that may exist between oil price and stock markets over the business cycle. They enable the investigation of the dynamics of the short-run component comovements by providing information about both their strength and their synchronization. Following A. Serletis and A. Shahmoradi and B. Ewing and M. Thompson, we consider that the two cyclical components are strongly

OIL PRICE SHOCKS & EUROPEAN STOCK MARKETS Table 5

RESULTS OF THE GRANGER CAUSALITY TESTS Lags 1 2 3 4 6 8
a

175

10

12

DJ Stoxx Europe S!O 0.073 O!S 0.253 Automobile & parts S!O 0.016 O!S 0.133 Financials S!O 0.345 O!S 0.137 Food & beverages S!O 0.055 O!S 0.102 Oil & gas S!O 0.068 O!S 0.635 Health care S!O 0.632 O!S 0.128 Industrials S!O 0.030 O!S 0.807 Basic materials S!O 0.005 O!S 0.625 Personal & household S!O 0.070 O!S 0.069 Consumer services S!O 0.056 O!S 0.131 Technology S!O 0.529 O!S 0.324 Telecommunications S!O 0.582 O!S 0.607 Utilities S!O 0.046 O!S 0.673
a

0.075 0.035 0.019 0.007 0.137 0.076 0.023 0.064 0.060 0.040 0.441 0.055 0.022 0.057 0.004 0.264 goods 0.042 0.097 0.038 0.069 0.803 0.084 0.811 0.116 0.091 0.087

0.111 0.055 0.037 0.007 0.216 0.117 0.005 0.150 0.058 0.136 0.250 0.149 0.043 0.063 0.009 0.426 0.074 0.144 0.063 0.090 0.903 0.112 0.902 0.222 0.138 0.212

0.099 0.097 0.062 0.006 0.191 0.155 0.002 0.086 0.069 0.144 0.324 0.272 0.025 0.106 0.007 0.195 0.085 0.259 0.116 0.154 0.400 0.208 0.606 0.233 0.240 0.322

0.216 0.159 0.049 0.005 0.363 0.277 0.004 0.119 0.056 0.306 0.456 0.510 0.060 0.154 0.003 0.338 0.163 0.343 0.308 0.184 0.454 0.112 0.791 0.196 0.393 0.597

0.112 0.172 0.041 0.011 0.234 0.077 0.003 0.142 0.082 0.341 0.158 0.499 0.032 0.087 0.001 0.251 0.078 0.566 0.155 0.247 0.469 0.016 0.807 0.213 0.387 0.768

0.104 0.185 0.035 0.023 0.140 0.150 0.004 0.178 0.063 0.470 0.161 0.377 0.039 0.072 0.001 0.280 0.060 0.308 0.127 0.141 0.532 0.005 0.771 0.149 0.313 0.871

0.362 0.372 0.138 0.048 0.436 0.272 0.015 0.315 0.160 0.487 0.468 0.445 0.180 0.166 0.008 0.404 0.246 0.451 0.303 0.285 0.852 0.020 0.909 0.247 0.531 0.943

This table provides the P-values of rejection of the null hypothesis we consider; S!O is the null hypothesis of no causality from stock market returns to oil price changes; O!S is the null hypothesis of no causality from oil price changes to stock market returns; and DJ Stoxx Europe is the European market index.

176

THE JOURNAL OF ENERGY AND DEVELOPMENT

correlated, weakly correlated, or uncorrelated for a shift j based on 0.23 jr(j)j < 1, 0.10 jr(j)j < 0.23, and 0 jr(j)j < 0.10, respectively. If jr(j)j is high for a positive, zero, or negative value of j, then the cycle of oil prices is leading the cycle of stock markets by j periods, is synchronous, or is lagging the cycle of stock markets by j periods, respectively. The results for leads and lags from 1 to 6 are shown in table 6. They globally confirm previous results. Positive weak-cyclical correlations are observed for the DJ Stoxx 600 Index for negative values of j, suggesting that oil prices are procyclical and lagging the whole stock market generally by a few weeks ahead in Europe over the 1998-2008 time frame. As we expect, oil prices and stock markets are strongly and contemporaneously correlated for oil and gas and basic materials industries. Furthermore, for two sectors (automobile and parts and food and beverages) negative high cyclical correlations are observed for positive values of j, indicating that oil prices are counter-cyclical and leading the stock markets generally by a few weeks ahead in these sectors. Weak correlations are obtained for four sectors: health care, industrials, personal goods, and consumer services. Surprisingly, we find no significant cyclical correlations between oil prices and stock market returns in technology and utilities industries, but, as we expect, no correlation also is found for telecommunications stocks. To sum up, our analysis shows significant relationships between most sector returns in Europe and oil price changes. However, there is some asymmetry in the reaction of stock returns in some sectors to oil price shocks, and in all cases the value and sign of the sensitivities of stock returns to oil price changes vary significantly across sectors. In what follows, we make some robustness checks.
Robustness Checks: Given previous evidence in the literature on the significance of world market risk and exchange rate risk in international asset-pricing relationships, especially in developed stock markets such as European markets, we conduct the following robustness checks.17 Effects of Exchange Rate: World oil prices are denominated in U.S. dollars and thus the latter plays an important role in the energy industry since the dollar exchange rate affects the price perceived by oil and oil-related-products consumers and producers. Therefore, it is interesting to take into account the links between the dollar exchange rates and the oil prices in our empirical investigation. In literature, the link between the dollar and oil prices has been examined at a theoretical level by, among others, P. Krugman and K. Rogoffat an empirical leveland S. Zhou and S. Dibooglu.18 The findings of these studies suggest the existence of a positive relationship between the two variables: an increase in oil prices is linked to a dollar appreciation. Moreover, recent papers show that exchange rate risk is internationally priced (De Santis et al. and references therein).19 Thus, the statistical significance of oil price changes in models (1) and (2) could be due to the failure to include the euro-

Table 6
a

CYCLICAL CORRELATIONS OF OIL PRICES WITH STOCK MARKET INDICES -4 0.087 0.046 -0.044 -0.045 0.330 -0.203 0.089 0.455 -0.001 -0.100 0.037 -0.024 0.011 0.107 0.008 -0.035 -0.074 0.356 -0.294 0.079 0.465 -0.025 -0.108 0.035 -0.030 0.016 0.119 0.023 -0.025 -0.093 0.382 -0.191 0.069 0.452 -0.048 -0.118 0.036 -0.031 0.024 0.118 -0.064 -0.031 -0.127 0.352 -0.194 0.043 0.423 -0.086 -0.138 0.036 -0.031 0.025 0.116 -0.000 -0.187 -0.158 0.385 -0.183 0.016 0.422 -0.120 -0.164 0.033 -0.034 0.011 0.070 -0.170 -0.202 -0.166 0.320 -0.153 -0.014 0.360 -0.155 -0.182 0.026 -0.039 0.002 0.023 -0.200 -0.206 -0.164 0.271 -0.128 -0.041 0.304 -0.151 -0.195 0.018 -0.046 -0.024 -0.011 -0.204 -0.196 -0.169 0.231 -0.101 -0.054 0.263 -0.157 -0.200 0.012 -0.050 -0.033 -0.042 -0.243 -0.183 -0.256 0.196 -0.073 -0.069 0.222 -0.162 -0.207 0.005 -0.052 -0.040 -3 -2 -1 0 1 2 3 4 5

-6

-5

6 -0.080 -0.257 -0.181 -0.161 0.151 -0.055 -0.090 0.180 -0.173 -0.216 0.000 -0.052 -0.065 -0.907 -0.282 -0.169 -0.167 0.105 -0.036 -0.108 0.131 -0.178 -0.217 0.005 -0.049 -0.095

DJ Stoxx Europe Automobile & parts Financials Food & beverages Oil & gas Health care Industrials Basic materials Personal goods Consumer services Technology Telecommunications Utilities

0.024 0.020 -0.073 0.023 0.281 -0.193 0.096 0.422 0.032 -0.086 0.038 -0.020 0.002

0.051 0.026 -0.066 -0.012 0.301 -0.202 0.088 0.430 0.014 -0.097 0.034 -0.021 0.001

OIL PRICE SHOCKS & EUROPEAN STOCK MARKETS

This table provides the cyclical correlations between oil price and stock market prices measured by r(j) = r(coilt, cstockt+j); bold type indicates high absolute value correlations; and DJ Stoxx Europe is the European market index.

177

178

THE JOURNAL OF ENERGY AND DEVELOPMENT

dollar exchange rate changes. To shed light on this issue, we re-estimate the models (1) and (2) with the euro-dollar exchange rate changes as a risk factor. Overall, we find that exchange-rate risk changes are significant for most European sectors. More interestingly, the relationships between oil price shocks and European sector returns are mainly unchanged. Effects of World Factors: European markets are integrated into the world market and thus European stock returns are subject to global factor effects. Hence, as a robustness check, we test for the significance of oil price changes within the framework of a model that allows for global risk factors. The inclusion of global factors is motivated by the fact that, if oil prices and global factors are correlated, tests based on models (1) and (2) may result in a spurious significance of the oil price changes because we are failing to account for global factors. More precisely, we re-estimate augmented versions of models (1) and (2), which include the MSCI world returns unexplained by the DJ Stoxx returns and Brent oil price changes. We show that global factors orthogonalized to European and oil factors are significant for several European sectors. More importantly, we find that the significance of the oil price changes seems to be largely unaffected by the use of global factors in addition to European factors. Real versus Nominal Prices: Finally, in order to take into account the effects of inflation, we re-conduct our empirical analysis using real oil and stock prices instead of nominal prices. Overall, our results do not change significantly and our main conclusion remains valid: oil price changes significantly affect sector stock returns in Europe. In summary, the evidence reported in this section suggests that oil price changes significantly affect several European sectors, even after accounting for European, global, and exchange-rate risk factors, and for the effects of inflation.

Conclusion and Policy Implications In this article, we investigated the linkages between oil and stock prices. Unlike other empirical investigations, which have focused largely on broad market indices (national and/or regional indices), we add to the knowledge of the relationship between oil prices and the stock markets in Europe by testing for shortterm links in the aggregate as well as sector by sector. Our results show strong significant linkages between oil price changes and stock markets for most European sectors. However, the nature and sensitivity of the reaction of stock returns to oil price shocks change considerably across sectors. Our findings should be of interest to researchers, regulators, and market participants. First, traders who are interested in investing in oil-sensitive stocks in Europe may, when oil prices are expected to remain high, select stocks from

OIL PRICE SHOCKS & EUROPEAN STOCK MARKETS

179

sectors, such as oil and gas, with high positive sensitivity to oil prices and, alternatively, when they expect decreases in oil price, they may select sectors with negative sensitivity such as food and beverages. They also can use oil-related derivatives. Thus, our results can be used to build profitable investment strategies. Second, that sectors in Europe have different sensitivities to oil price changes means great risk diversification possibilities across industries in Europe. Selecting portfolios across rather than within sectors would be more efficient. Finally, investors and portfolio managers should rebalance their portfolios in keeping with their views of the sign of coming changes in oil prices (rises or falls), and our findings suggest that diversification can be achieved across sectors in all cases of oil price changes.
NOTES J. D. Hamilton, What is an Oil Shock? Journal of Econometrics, vol. 113, no. 2 (April 2003), pp. 363-98; J. Cunado and F. Perez de Gracia, Oil Prices, Economic Activity and Inflation: Evidence for Some Asian Countries, The Quarterly Review of Economics and Finance, vol. 45, no. 1 (2005), pp. 65-83; P. Balaz and A. Londarev, Oil and its Position in the Process of Globalization of the World Economy, Politicka Ekonomie, vol. 54, no. 4 (2006), pp. 508-28; M. Gronwald, Large Oil Shocks and the US Economy: Infrequent Incidents with Large Effects, Energy Journal, vol. 29, no. 1 (2008), pp. 151-71; A. Cologni and M. Manera, Oil Prices, Inflation and Interest Rates in a Structural Cointegrated VAR Model for the G-7 Countries, Energy Economics, vol. 30, no. 3 (2008), pp. 856-88; and L. Kilian, Exogenous Oil Supply Shocks: How Big Are They and How Much Do They Matter for the US Economy? Review of Economics and Statistics, vol. 90, no, 2 (2008), pp. 216-40. C. M. Jones and G. Kaul, Oil and the Stock Markets, Journal of Finance, vol. 51, no. 2 (1996), pp. 463-91. R. D. Huang, R. W. Masulis, and H. R. Stoll, Energy Shocks and Financial Markets, Journal of Futures Markets, vol. 16, no. 1 (1996), pp. 1-27.
4 P. Sadorsky, Oil Price Shocks and Stock Market Activity, Energy Economics, vol. 21, no. 5 (1999), pp. 449-69. 3 2 1

J. Park and R. A. Ratti, Oil Price Shocks and Stock Markets in the US and 13 European Countries, Energy Economics, vol. 30, no. 5 (2008), pp. 2587-608. N. Apergis and S. Miller, Do Structural Oil-Market Shocks Affect Stock Prices? Energy Economics, vol. 31, no. 4 (2009), pp. 569-75. P. Sadorsky, Risk Factors in Stock Returns of Canadian Oil and Gas Companies, Energy Economics, vol. 23, no. 1 (2001), pp. 17-28, and M. M. Boyer and D. Filion, Common and Fundamental Factors in Stock Returns of Canadian Oil and Gas Companies, Energy Economics, vol. 29, no. 3 (2007), pp. 428-53. I. El-Sharif, D. Brown, B. Burton, B. Nixon, and A. Russel, Evidence on the Nature and Extent of the Relationship between Oil and Equity Value in UK, Energy Economics, vol. 27, no. 6 (2005), pp. 819-30.
8 7 6

180

THE JOURNAL OF ENERGY AND DEVELOPMENT

9 M. Nandha and R. Faff, Does Oil Move Equity Prices? A Global View, Energy Economics, vol. 30, no. 3 (2008), pp. 986-97.

M. Nandha and R. Brooks, Oil Prices and Transport Sector Returns: An International Analysis, Review of Quantitative Finance and Accounting, vol. 33, no. 4 (2009), pp. 393-409.
11 12

10

J. Park and R. Ratti, op. cit., and N. Apergis and S. Miller, op. cit.

The suitability of two factor market and oil pricing models similar to the model we use in this paper was investigated in several papers; see R. Faff and T. Brailsford, A Test of a Two Factor Market and Oil Pricing Model, Pacific Accounting Review, vol. 12, no. 1 (2000), pp. 61-77.
13

We have employed other multifactor models and obtained similar results.

14 J. D. Hamilton, op. cit.; D. Zhang, Oil Shock and Economic Growth in Japan: A Nonlinear Approach, Energy Economics, vol. 30, no. 5 (2008), pp. 2374-390; and A. Cologni and M. Manera, The Asymmetric Effects of Oil Shocks on Output Growth: A Markov-Switching Analysis for the G-7 Countries, Economic Modelling, vol. 26, no. 1 (2009), pp. 1-29. 15 B. T. Ewing and M. A. Thompson, Dynamic Cyclical Components of Oil Prices with Industrial Production, Consumer Prices, Unemployment and Stock Prices, Energy Policy, vol. 35, no. 11 (2007), pp. 5535-540, and F. Lescaroux and V. Mignon, On the Influence of Oil Prices on Economic Activity and Other Macroeconomic and Financial Variables, OPEC Energy Review, vol. 32, no. 4 (2008), pp. 343-80. 16 A. Serletis and A. Shahmoradi, Business Cycles and Natural Gas Prices, OPEC Review, vol. 29, no. 1 (2005), pp.75-84; B. Ewing and M. Thompson, op. cit.; and F. Lescaroux and V. Mignon, op. cit.

In order to preserve space, the results of robustness tests are not reported here but are available from the author upon request. P. Krugman, Oil and the Dollar, National Bureau of Economic Research (NBER) Working Paper Series no. 554, Cambridge, Massachusetts, 1980; K. Rogoff, Oil, Productivity, Government Spending and the Real Yen-Dollar Exchange Rate, Federal Reserve Bank of San Francisco Working Paper no. 91-06, 1991; S. Zhou, The Response of Real Exchange Rates to Various Economic Shocks, Southern Journal of Economics, vol. 61, no. 4 (1995), pp. 936-54; and S. Dibooglu, Real Disturbances, Relative Prices, and Purchasing Power Parity, Journal of Macroeconomics, vol. 18, no. 1 (1996), pp. 69-87. G. De Santis, B. Gerard, and P. Hillion, The Relevance of Currency Risk in the EMU, Journal of Economics and Business, vol. 55, no. 5-6 (2003), pp. 427-62.
19 18

17