Recent evidence on the oil price shocks on GCC stock markets

Transcription

1 Recent evidence on the oil price shocks on GCC stock markets Suzanna ElMassah Victor Wong Abstract The recent plunge in the oil price affected many countries, especially major oil producers and exporters, such as the Gulf Corporation Council (GCC), which accounts for half the global oil reserves. This paper examines the impact of oil price changes on GCC stock markets including Bahrain, Kuwait, Oman, Qatar, KSA and UAE over a 10-year period, We examine the direction of influence and influence absorption through Granger causality and impulse response function. The results are important for portfolio management at the international level, and provide insights for government and regulatory authorities in times of oil price change. Additionally, the evidence suggests the need for more economic diversification at the country level in the GCC region to mitigate high volatility in the event of oil shocks. Keywords: GCC countries, stock markets, oil prices, stock market volatility, Granger causality, impulse response functions JEL Classification: C32, C51, G15 1. Introduction The recent oil price plunge has had a considerable impact in many oil-exporting countries. The falling oil price is mainly due to the increase in shale oil production in the US, causing significantly lower oil prices, which have dropped from about $145 per barrel (in July 2008) to $36 per barrel (in December 2015). A falling oil price reduces profitability in oil-exporting countries and slows their output/export growth (The Economist, 2015). Middle Eastern countries are sensitive to changes in oil prices, as some of them are major oil producers and exporters and are responsible for a large proportion of estimated global oil reserves with close to 810 billion barrels, representing 47.7% of the global stock (see Appendix 1). GCC countries comprise the following six Middle Eastern countries: Bahrain, Kuwait, Oman, Qatar, Kingdom of Saudi Arabia (KSA), and United Arab Emirates (UAE) which entered into an economic and trade agreement in An examination of the links between oil prices and GCC stock markets is important for the following reasons. First, GCC countries are the major producers 1

2 and exporters in the global oil market and are members of the Organization of the Petroleum Exporting Countries (OPEC). Movements in the GCC stock markets would, therefore, be expected to be strongly related to the movements in oil prices. Second, the GCC region is independent of international stock markets, but GCC countries are sensitive to regional political events and regional geopolitical characteristics are of particular interest to other oil producers and exporters. Finally, the implications of our findings would be expected to provide insights for policy makers, government regulators, institutional stakeholders and investors. The largest oil reserves are found in KSA, followed by Kuwait and UAE, respectively (see Appendix 2). KSA was the founder of OPEC, and since the organization s foundation, the kingdom has demonstrated significant leadership for several reasons. First, KSA has the world s largest proven crude oil reserves, surpassing even such regions as the Americas, Asia Pacific, Europe and Eurasia. By 2013, the total share of the world s proven crude oil reserves in KSA accounted for 21% of OPEC reserves and almost 18% globally (OPEC, 2014). Second, KSA is also OPEC s top oil exporter, and the country has and continues to account for almost one-third of extracted oil (Hamilton, 2014). Third, KSA has the largest spare production capacity worldwide and this enables the KSA to act as swing producer. That is, it can increase or decrease oil supply at minimal additional internal cost, and is therefore able to affect oil prices and balance the market; such a position provides KSA with downside protection in the short and medium terms. Oil plays an important role within the GCC economies; any change in oil prices would be expected to impact the countries stock markets. When oil affects real economic activity, it will affect earnings of companies which are directly or indirectly involved in the oil industry. Hence, a reduction in oil prices will cause reduced earnings, in turn reducing stock prices. There exists a wide range of literature evidencing linkages between crude oil prices and stock markets in developed markets. There is also some literature on the linkages in the GCC region. This study aims to add to that literature by examining the relationship between oil prices and GCC stock markets during the period The paper addresses the following two questions: 2

3 1. What is the causality relationship between oil prices and GCC stock markets? and 2. How long does it take for GCC stock markets to absorb the oil price s influence? The study also combines all the GCC stock markets by exploiting the Vector Autoregressive (VAR) model and its corresponding impulse response plots, in order to understand better how oil price movements and its impact on GCC stock markets. Our work relates to several strands of literature. First, the present study relates closely to work by Bouri (2015a) who explores the relationship between oil shock spillovers on stock markets of Lebanon and Jordan using an extension of causality-in-variance procedure. Bouri has made an important contribution to the literature, highlighting the importance of heterogeneity between countries responsiveness to oil shocks. Although the oil price volatility spillover is present in Jordan, where the stock markets react strongly to oil shocks, the same cannot be said for Lebanon, where oil price volatility does not predict stock market volatility. Bouri (2015b) extends this line of research with an in-depth focus on Lebanon. Using a VAR-GARCH framework that takes into account both the dynamics and behavior of unexplained shocks (through modelling the error terms), Bouri documents weak unidirectional return and volatility transmission from oil prices to the Lebanese stock markets. Similarly, Bouri and Demirer (2016) extends the analysis of volatility transmission between oil and stock markets to Kuwait, Saudi Arabia and UAE using causality-invariance tests. The authors also found significant spillovers from oil prices to emerging importers with the results being even more pronounced following the Great Recession in Our study contributes to this literature by focusing on oil-exporting countries, unlike Bouri (2015a) which focuses on two oil-importing countries and it also takes a more extensive approach examining the GCC countries. Like, Bouri (2015b), we use a VAR model but whereas she examines the case of Lebanon (an oil importer) we study the whole of the GCC, where the focus is on oilexporting countries. Furthermore, Bouri and Demirer (2016) consider causality in volatility (variances), while our study considers causality in stock returns (levels) within a VAR framework. Finally, our paper addresses important measurement error concerns in the literature (as pointed out 3

4 in Hamilton, 2014). We do this by considering weekly data which is plausibly less noisy and prone to attenuation biases relative to daily data used in much of the literature. The rest of the paper is organized as follows. Section 2 discusses the background of GCC markets. Section 3 provides the literature review. Sections 4 and 5 describe the methodology and data. Section 6 presents the main results of the paper and includes the discussion of Granger causality and impulse response analysis. Section 7 concludes the paper. 2. Background of GCC markets The GCC is a political and economic union of six Arab countries sharing the Arabian Gulf, namely Bahrain, Kuwait, Oman, Qatar, KSA, and UAE. The six countries are among world s major oil producers and are OPEC members; they entered into economic and trade agreement in The most common feature of GCC countries is their extensive dependence on non-renewable natural resources (National Bank of Kuwait, 2011), though with different intensity levels (Arouri and Rault, 2012). Monetary policies in the GCC countries are connected to the US because their exchange rates are tied to the US dollar (Karam, 2001) 1. Although stock trading began in some GCCs in the 1930s, the formal market system was only established by Kuwait in Since then, GCC markets have grown markedly to reach 658 listed companies in seven stock markets 2 by the end of From an international standpoint, GCC market capitalization in 2012 accounted for 1.32% and 3.75% of world and US market capitalization, respectively. GCC stock markets share common institutional characteristics, such as a high concentration of financial and banking sectors that serve as mediators of regional and global volatility to the GCC markets during financial crises (Jamaani and Roca, 2015). Additionally, the GCC markets experience low levels of foreign participation due to government restrictions on foreign ownership and asymmetric information resulting from weak financial market development (Jamaani and Roca, 2015). Compared to developed markets, GCC markets regularly experience thinner trading and are 4

5 more easily manipulated by a few large dealers (Bley, 2011). Moreover, the capital markets of GCC countries are known to be small in size with high short-term volatility and low liquidity (Bahlous, 2013). Table 1 presents the volume of trades in the six GCC stock markets and shows the relatively low level compared with FTSE350 and S&P500, for example. Table 1: Volume of trades in GCC stock markets Stock Exchange 50-day average volume % of GCC 50-day average volume Abu Dhabi Stock Exchange, UAE 124,073, Bahrain Stock Exchange, Bahrain 2,232, Dubai Financial Market, UAE 294,553, Kuwait Stock Exchange, Kuwait 138,592, Muscat Securities Market, Oman 14,853, Qatar Exchange, Qatar 6,237, Saudi Stock Exchange, KSA 202,324, GCC Region 782,867, Source: (data obtained on 20 September 2015) Note: For comparison purposes, the 50- day average volume as at 20 September 2015 for the United Kingdom FTSE 350 index and the United States S&P500 composite index was 1,022,881 and 669,361, respectively. Table 1: Volume of trades in GCC stock markets Stock Exchange 50-day average volume % of GCC 50-day average volume Abu Dhabi Stock Exchange, UAE 113,084, Bahrain Stock Exchange, Bahrain 2,675, Dubai Financial Market, UAE 275,372, Kuwait Stock Exchange, Kuwait 69,025, Muscat Securities Market, Oman 13,786, Qatar Exchange, Qatar 11,442, Saudi Stock Exchange, KSA 157,521, GCC Region 642,908, Source: (data obtained on 30 July 2017) Note: For comparison purposes, the 50-day average volume as at 28 July 2017 for the United Kingdom FTSE 350 index and the United States S&P500 composite index was 1,223, and 2,045,102.44, respectively. Comment [EM1]: Is this right is less than the countries in the table. Maybe the units used are different please check Comment [SEM2]: Have checked our dataset and this is correct based on 20 Sept 2015 for the last 50 days average volume. The lower volume is unique to that period in the last 50 days for that period. Alternatively, we would like to update this table to shows the latest numbers, as follows. 3. Literature review on oil price changes and GCC stock markets Theoretically, oil prices will have a direct impact on stock prices, as stock prices at any given point in time reflect the discounted future value of expected cash flows (Huang, Masulis, and Stoll, 1996). A falling oil price may be beneficial for end consumers; however, it would also mean reduced profits for GCC oil exporters. Hence, a lower oil price would increase aggregate stock 5 Comment [EM3]: Check this is right Comment [SEM4]: That is correct. A lower oil price will lead to lower cost of productions. The lower retail price would possibly encourage consumer spending and higher consumer spending would possibly lead to better company performance, hence better performance on the stock market.

6 prices (Hamilton, 1983), either directly affecting future cash flows or indirectly through the effect of interest rates used to discount the cash flows (Basher et al., 2012). The relationships between macroeconomic events are important to GCC oil exporters, particularly the interaction between oil prices and stock market prices, as well as an understanding how long it takes GCC economies to fully absorb the movements in oil prices these are the objectives of our study. Nevertheless, the exact magnitude of these effects remains an empirical question. Many studies have evaluated the effect of oil prices on capital market dynamics. This strand of literature initially focused on stock markets in developed economies, which are mostly net oilimporters, (e.g. Kling 1985; Jones and Kaul 1996). They have generally confirmed a significant negative impact of oil price changes on stock markets (Charles and Kaul, 1996; Huang et al., 1996; Papapetrou, 2001; Park and Ratti, 2008; Kaneko and Lee 1995 and many others). However, the more recent literature has seen an increasing number of studies focusing their analyses on developing markets. Narayan and Narayan (2010) investigated the dynamic relationship between oil prices and Vietnamese stock market from 2000 to 2008 using daily data and found that oil prices have a significant positive impact on Vietnamese stock market in the long-run. Williams and Saadi-Sedik (2011) found that GCC stock markets are strongly affected by the Global Financial Crisis as well as falling oil prices. Sahu et al. (2014) examined the relationship between oil prices and Indian stock market, and found a unidirectional volatility spillovers effect from oil prices to Indian stock market. Similarly, Bouri (2015b) documented weak unidirectional relationship from oil prices to the Lebanese stock market. Another study by Bouri (2015a), also mentioned earlier, extended the sample to four oil-importing countries such as, Lebanon, Jordan, Tunisia and Morocco, and found bidirectional effect from oil prices and Jordan stock market, as well as a unidirectional effect from oil prices and Tunisia stock market. The later study by Bouri and Demirer (2016) found unidirectional effect from oil prices to net oil-exporting countries, such as Kuwait, Saudi Arabia and UAE. 6

7 Few studies investigated this relation in the net oil-exporting countries, emerging markets, and GCC in particular, and their results are inconsistent. The focus on oil-importing countries, though important, limits our understanding of capital market dynamics of countries especially in the Gulf Region, which are large oil-exporting economies. The present study focuses on the relationship between oil prices on the stock markets of GCC countries. The following two sections review the literature relating to oil prices and GCC stock markets categorizing studies with findings of positive influence in Section 3.1 and studies that found weak or no relationship between oil prices and GCC stock markets in Section Studies suggesting a positive relationship between oil and GCC stock markets A substantial empirical literature supports significant volatility spill-over from the oil market to each individual GCC stock markets, and the spill-over from one GCC stock market to another GCC stock market as well. Using VAR models and cointegration tests, Hammoudeh and Aleisa (2004) found KSA stock returns have a bidirectional relationship with oil price changes. However, no direct relation was noticed in other GCC stock markets which appeared to be rather more influenced by their own domestic concerns. Zarour (2006) found that only KSA and Oman have the power to affect oil prices, whereas in all other GCC excluding Qatar stock markets returns were significantly influenced by positive oil price shocks. Hammoudeh and Choi (2006) found evidence of a short run, indirect effect of oil prices in KSA and Oman but not in other GCC countries. On the contrary, Jouini (2013) and Sbeiti and Haddad (2013) 3 found a significant and positive long-run impact of oil price changes on stock markets for all GCC countries. Arouri et. al. (2011) confirmed the existence of considerable short and long run return and volatility spillovers between oil prices and GCC stock markets, with greater response to negative oil price shocks than to positive oil shocks. Similarly, Maghyereh and Al-Kandari (2007) concluded that oil prices affect GCC stock markets 4 in a nonlinear fashion. These findings, however, were not supported by Malik and Hammoudeh (2007), who depicted unidirectional 7

8 volatility transmission from oil to stock markets, while bidirectional volatility transmission was only noticed in KSA. Using weekly data, Arouri and Rault (2012) showed, the existence of a strong causal short run transmission from oil prices to GCC stock markets. However, there was an asymmetric effect of negative and positive shocks; stock markets appeared to react more to negative oil shocks than to positive oil shocks (while no long run relations were reported). While most studies have found that KSA is heavily influenced by oil price changes, Daly and Fayyad (2011) found that Qatar and UAE were more sensitive to oil shocks and that oil price transmission to their stock markets was higher during positive shocks and global financial crises. In the event of crisis periods, Basher and Sadorsky (2016) suggested that oil is the best asset to hedge emerging market stock prices. 3.2 Studies suggesting a negative e relationship between oil and GCC stock markets Another stream of research concludes there are weak or no effects, or different results amongst individual GCC in regard to the relation between stock markets and oil price changes. Arouri and Fouquau (2009) found significant transmission of oil prices to only three markets (Qatar, UAE, and Oman) with a greater effect from negative oil shocks. However, no volatility appeared in Bahrain, Kuwait, or KSA s stock markets. Similarly, Arouri, Dinh, and Nguyen (2010) found evidence of a short-run effect of oil prices on stock markets in KSA, UAE, Qatar, and Oman, however, no effect was found in Bahrain and Kuwait. Lescaroux and Mignon (2008) examined thirty-six countries and found that OPEC countries showed weak evidence in favor of cointegration between oil and share prices. Rizvi and Masih (2014) found inconsistent results among GCC countries: three stock markets (KSA, Oman and Kuwait) were not influenced by oil price neither in the short nor the long run; another two stock markets (Qatar and Bahrain) were affected by oil prices in both short and long 8

9 runs, while the UAE stock market was affected only in the long run. In contrast, Alhayki (2014) concluded that the oil price had a negative impact on the stock markets of Bahrain, KSA and UAE. However, the oil price had a positive relationship in Kuwait, Qatar and Oman stock markets. Furthermore, bidirectional causal relationship between oil prices and stock market returns was noticed, except for the case of Bahrain. As can be seen from the literature above, there are no consensus on the magnitude or the direction of oil price shocks on the stock market. The mixed findings within the literature can also be attributed to the short study periods, noisy daily data and the assumptions imposed on estimation methods. It is noticeable that there is no consistent classification of GCC countries used in these past studies. Our study extends the understanding of the relationship between oil prices and stock markets in GCC countries from 2005 to 2015 at weekly frequency by testing the causality relationship and gauging how long it takes for GCC stock markets to absorb oil price influence in levels using VAR and impulse response functions. 4. Methodology We conduct a number of diagnostic tests before performing the vector autoregressive (VAR) analysis. In order to determine the stationarity of the data, unit root tests based on the Augmented Dickey-Fuller and Phillips-Perron tests are performed first. The optimal lags to be used in the model are also tested first, based on the Akaike information criterion, which suggest two lags to be specified in the VAR model. As VAR tests are already well-known in the literature, discussions of these tests are briefly provided below Vector autoregressive model The VAR model is among the most flexible, well established, and proven models in multivariate time series analysis. Ideally, the model provides a natural extension of the univariate autoregressive model, giving rise to multivariate models including deterministic, endogenous, and exogenous variables. Over the years, the VAR model has proven to be rather useful in describing the dynamic 9

10 characteristic of economic and financial time series analysis; as well as in making predictions and forecasting. Being more sophisticated than univariate models, it provides superior forecasts compared with both their univariate cousins and other theory-based simultaneous equation models. The VAR has a multivariate advantage, as the forecasts developed can be made conditional on potential future trends in the other given variables used in the model. The study investigates the relationship between oil prices and GCC stock markets using VAR models. Suppose that represents a time series variable vector then an autoregressive model of the basic vector, having order, VAR(p) becomes: Π Π Π (1) Where and and Π are coefficient matrices, is a ( 1) constant vector and a vector process for white noise with an unobservable zero mean, with covariance matrix. Consider a special case, for example, where is a two-dimensional vector, the VAR (1) comprises two equations and can be written in matrix form (bivariate VAR) as: ( ) ( ) ( ) ( ) ( ) ( ) ( ) (2) Where ( ) for. The lag operators can also be used to represent VAR(p) where it may look like a univariate AR process as follows: Π, (3) Where Π Π Π. Imposing stationarity on gives unconditional unexpected value as: Π Π. (4) In many cases, stochastic exogenous variables and/or other deterministic terms are part of the VAR specification. Therefore, the VAR(p) can have a general form as: 10

11 Π Π Π Θ (5) Where denotes a deterministic or exogenous variable matrix of, and Θ denotes a parameters matrix. It is important to note that is a vector and the system can be represented in matrix form as shown by the two by two cases earlier. In this study, there will be both endogenous and exogenous variables to include the effect of oil prices on stock market indices. There are a number of aspects involved in the VAR methodology and these are discussed next Granger causality test A key part of employing the VAR model is its use in forecasting. Its structure gives information on the ability of variables or variable groups to forecast other variables. Granger (1969) introduced this intuitive notion of a variable s ability to forecast. Should a variable or variable groups be instrumental in another variable or variable group s prediction, then Granger-causes. In the opposite case, does not Granger-cause if for all the mean squared error of a forecast of based on ( ) is the same as the mean squared error of a forecast of, based on ( ) and ( ). It is worth noting that Granger s causality notion only suggests the ability to forecast. A VAR(p) bivariate model for ( ), sees the failure of to Granger-cause, given that all p VAR matrices of coefficients are lower triangular. The Wald statistic can test p linear restrictions on coefficients. The coefficient matrices of VAR are diagonal in the event that both a d fail to Granger-cause each other. It is important to note that Granger causality is rather useful in finance and continues to be extensively used because it shows bidirectional as well as unidirectional causality of the time series data. In essence, how and in what manner the other variables contribute to the prediction process of a given variable that is, which variables or which information in terms of the variables is crucial for the prediction process contribute significantly to the forecasting of a given variable. 11

12 4.3. Impulse response analysis An impulse response method measures the duration and effect of oil and gas prices in one index to another by tracing the effects of a shock to one endogenous variable on the other variables in the VAR model. The dynamic relationships between variables are captured by the impulse response function. Processes of vector moving averages can be used to represent VAR(p) as follows:, (6) in which the moving average matrices are of matrix type. The coefficient matrix s elements then represent the shock effects on. Based on this, an impulse response of the following form can be determined: where the coefficient sets, (7) represent the impulse response functions and where. 5. Data description This study investigates international integration among oil prices and GCC stock markets focusing particularly on Bahrain, Kuwait, Oman, Qatar, KSA, UAE. The stock market data used in the study is weekly data from the S&P Broad Market Index. All indices are expressed in US dollars and the data used are in the form of returns on the price indices calculated by the formula: R t = ln(price t /price t-1 ) Subsequently, the indices are weighted based on weekly stock market values to control for market sizes. The study also uses oil price data from Brent Crude Oil traded on the Intercontinental Exchange. All data are collected from Thomson Reuters DataStream database. The sample period for the study is from 7 January 2005 to 25 December 2015, with a total observation of 573 weeks. 12

13 Stock Market Index Oil Price Figure 1 presents the relative growth of oil prices and GCC stock market indices from January 2005 to December Figure 1: Movements of stock market indices and oil prices, Bahrain Kuwait Oman Qatar Saudi Arabia UAE Oil During the sample period, the oil price fluctuated from $145 per barrel (in July 2008) to $36 per barrel (in December 2015). As seen in Figure 1, the stock market indices are volatile. The KSA index peaked in 2005 and fell greatly in 2008, which could be explained by the Global Financial Crisis in Similar behaviors are shown by other GCC stock markets. 6. Empirical results 6.1. Data preliminaries Table 2 shows that weekly returns were highest for KSA followed by Qatar, whereas Kuwait, Bahrain, and oil index had negative returns over the sample period. The standard deviation indicates high market volatility in the oil index and KSA. All countries stock markets are negatively skewed implying a long left tail distribution. Kurtoses for all markets are above three, suggesting the data is leptokurtic or peaked relative to the normal distribution. The Jarque-Bera test confirms the markets are non-normally distributed. Augmented Dickey-Fuller (ADF) and Phillips-Perron (PP) tests consistently reject the null hypothesis (i.e. H 0 : = 0) that the series has a unit root and thereby 13

14 confirm that all returns series are stationary. Correlation between each GCC stock market index and the oil index are positive and generally low, the highest being (Oman) and and (Qatar), respectively. Table 2: Descriptive statistics for the S&P Broad Market Index Bahrain Kuwait Oman Qatar KSA UAE OIL Mean Std. Dev Skewness Kurtosis Jarque-Bera Probability ADF PP Correlation with OIL Notes: * The Jarque-Bera, Augmented Dickey-Fuller (ADF) and Phillip-Perron (PP) tests are significant and the results show the data is non-normally distributed and stationary Significant linkages between oil prices and GCC stock markets We first perform the multivariate Granger causality analysis in a VAR model to determine whether oil prices are affecting the GCC stock markets. In performing this analysis, we control for the effect of the size of each market. The VAR model 6 is estimated based on the weighted returns of each market, where the weight for each market is the proportion of the capitalization of each market in relation to the total capitalization of the GCC stock markets derived from S&P Broad Market Index. The results from the Granger causality are presented in Table 3 and graphically represented in Figure 2. It can be seen from the table that all markets have a significant coefficient at 1%, 5% and 10% levels of significance either as a market influencing another market or as a market that is being influenced. Table 3: Granger causality Dependent variables Independent variables Bahrain Kuwait Oman Qatar KSA UAE OIL Bahrain *** ** ** Kuwait *** * * Oman *** * Qatar KSA * * ** UAE * *** 14

15 Oil *** *** *** ** ** Note: *** Significant at 1% level. ** Significant at 5 % level. * Significant at 10 % level. Generally, the results show that GCC markets are interconnected in the short run and, thus, there is significant interdependence among all markets. The Granger causality results show that oil-price shocks changes in all GCC stock markets at 1% and 5% significance except Oman, which has a relatively small oil market. This finding is in line with other studies (Hammoudeh and Aleisa, 2004; Zarour, 2006; Hammoudeh and Choi, 2006; Malik and Hammoudeh, 2007). Figure 2: Granger causality graphical representation Bahrain OIL Kuwait UAE Oman KSA Qatar 1% significant 5% significant 10% significant Granger causality results indicate that Kuwait is most influenced by other markets, followed by Bahrain. The relationship and/or volatility between GCC stock markets is affected by the degree of market openness in each country, as well as foreign ownership restrictions. Figure 2 shows the strong and heavy bidirectional volatility and shock spillover among four GCC stock markets, namely Bahrain, Kuwait, Oman, and UAE, which can probably be explained by their openness and lack of restrictions on foreign ownership. Qatar and KSA are less open to foreign investors; a GCC citizen may own up to 49% of company shares in Qatar, a percentage that decreases to 25% in KSA. Closed market policies in the GCC countries prevent information flows from developed countries to GCC markets, while fewer restrictions on GCC citizens ownership allow for more volatility among the GCC markets (Eissa and Elgammal, 2015). 15

16 6.3. Duration and speed of interaction between oil and stock markets This section, investigates the degree and manner of interaction among the markets that are significantly linked, using impulse response analysis. Impulse responses provide evidence on how much and how quickly the movement of one market is transmitted to the others. The shorter and the faster the interaction is, the more integrated the markets are. Conversely, the longer and the slower the interaction, the less integrated the markets. The impulse responses of each GCC market to price movements in other markets are plotted in Figure 3. Figure 3 reveals that all responses were immediate that is, each market immediately responded to information from other markets during the first week. Only Bahrain, Kuwait and KSA responded with a slight positive response in the second week and correction takes place after the third week when the responses start to reverse it could imply that shocks are absorbed initially with a positive response, then an adjustment or correction happens slowly. All other GCC (except Bahrain, Kuwait and KSA) are affected by the oil shocks and these majority of the responses are absorbed within two weeks and these responses are fully absorbed within a period of five to six weeks after which responses slowly fade to zero. KSA responded with the highest magnitude to oil shocks followed by UAE and Qatar. A possible explanation for this magnitude is due to the oil dependency and/or exports by KSA, UAE and Qatar. In line with Williams and Saadi-Sedik (2011), our results suggest that GCC stock markets are affected by the oil prices and GCC economics have a strong dependence on oil and a low level of sectoral diversity, and therefore their economies are intrinsically more volatile than other economies at the same level of development. In summary, the impulse response analysis points to increasing integration and volatility among GCC stock markets after oil price shocks, which are consistent findings with Daly and Fayyad (2011) and Bouri (2015a; 2015b). Figure 3: Responses of GCC stock markets to oil price shocks 16

17 BAHRAIN KUWAIT OMAN QATAR KSA UAE 7. Conclusion and Policy Implications Our paper investigates the impact of oil price changes on the GCC stock markets, including Bahrain, Kuwait, Oman, Qatar, KSA, and UAE during using weekly data from the S&P Broad Market Index. The GCC countries represent major oil producers and exporters, such that movements in oil prices would influence their stock markets. The results from the Granger causality analysis show that KSA, UAE, and Kuwait are significantly affected by oil price changes. The analysis also revealed that oil price changes are affected by Bahrain, Kuwait and Qatar, as well as, GCC markets are generally interconnected and, thus, there is significant interdependence among all markets. Oil dependency and cross-listing in GCC stock markets may play a role in such interconnections. The results from impulse response analysis revealed that all responses were immediate from other markets during the first week, though shocks are absorbed slowly and completed within a period of five to six weeks. Such results point toward increased integration and volatility between GCC stock markets during the recent oil price shocks. These findings are consistent with Williams and Saadi- Sedik (2011), Daly and Fayyad (2011) and Bouri (2015a; 2015b) and confirm that GCC stock markets are not immune from global financial shocks, including those related to oil price shocks. 17 Comment [EM5]: Any comment on direction of influence arising from this work? Earlier you highlighted difference in earlier studies but don t seem to cover this either in findings or conclusions Comment [SEM6]: Have added more discussion on the direction in the paragraph above. Figure 2 shows the details on the directions.

18 Our findings support Bouri and Demirer (2016), which call into question the shortsightedness of continuation of the current policy of dependence on oil as sole vehicle of output growth. Therefore, it is important for GCC countries look for potential opportunities to diversify into different economic sectors. One possible avenue is to invest in energy-efficient technologies (such as solar and wind power in the manufacturing and/or services sector), where the investment funding could be raised from issuing Green Sovereign Wealth Fund. Thus, by developing appropriate policies that support this path, the GCC countries should be able to devise more effective economic diversification and less dependent on adverse oil price shocks. Further studies can be conducted to determine the feasibility and implementation of these energy-efficient technologies in the GCC economies. Aknowledgements We would like to acknowledge the helpful comments and suggestions received during the anonymous refereeing process and also thank the editor for her constructive suggestions. References Akoum, Ibrahim, Michael Graham, Jarno Kivihaho, Jussi Nikkinen, and Mohammed Omran Co-Movement of Oil and Stock Prices in the GCC region: A Wavelet Analysis. The Quarterly Review of Economics and Finance 52 (4): Alhayki, Zainab Jaafar The Dynamic Co-movements between Oil and Stock Market Returns In: The Case of GCC Countries. Journal of Applied Finance and Banking, 4 (3): Arouri, Mohamed El Hedi, Amine Lahiani, and Duc Khuong Nguyen Return and Volatility Transmission between World Oil Prices and Stock Markets of the GCC Countries. Economic Modelling 28 (4): Arouri, Mohamed El Hedi, and Christophe Rault Oil Prices and Stock Markets: What Drives What in the Gulf Corporation Council? CESifo Working Paper Series No Arouri, Mohamed El Hedi, and Christophe Rault Oil Prices and Stock Markets In GCC Countries: Empirical Evidence From Panel Analysis. International Journal of Finance and Economics 17 (3):

19 Arouri, Mohamed El Hedi, and Julien Fouquau On the Short-Term Influence of Oil Price Changes on Stock Markets in GCC Countries: Linear and Nonlinear Analyses. Economics Bulletin, 29 (2): Arouri, Mohamed El Hedi, Jamel Jouini, and Duc Khuong Nguyen On the Relationship Between World Oil Prices and GCC Stock Markets. Journal of Quantitative Economics 10 (1): Arouri, Mohamed El Hedi, Thanh Huong Dinh, and Duc Khuong Nguyen Time-Varying Predictability in Crude Oil Markets: The Case of GCC Countries. Energy Policy 38 (8): Bahlous, Mejda Does Cross Listing Benefit the Shareholders? Evidence from Companies in the GCC Countries. Asia-Pacific Financial Markets 20 (4): Basher, Syed Abul, Alfred A. Haug, and Perry Sadorsky Oil Prices, Exchange Rates and Emerging Stock Markets. Energy Economics 34 (January 2012): Basher, Syed Abul, and Perry Sadorsky Hedging Emerging Market Stock Prices with Oil, Gold, VIX, and Bonds: A Comparison between DCC, ADCC and GO-GARCH. Energy Economics 54 (February 2016): Bley, Jorg Are Stock Markets Predictable? Emerging Markets Review 12 (3): Bouri, Elie, and Riza Demirer On the Volatility Transmission between Global Oil and Stock Markets: A Comparison of Emerging Importers and Exporters. Economia Politica 33 (1): Bouri, Elie. 2015a. Oil Volatility Shocks and the Stock Markets Of Oil-Importing MENA Economies: A Tale from the Financial Crisis. Energy Economics 51 (September 2015): Bouri, Elie. 2015b. Return and Volatility Linkages between Oil Prices and the Lebanese Stock Market in Crisis Periods. Energy 89 (2015): Brailsford, Tim, Richard Heaney, and Chris Bilson. 2004, Investments: Concepts and Applications, 2nd ed.,thomson, South Melbourne. Campbell, John Y., and John H. Cochrane By Force of Habit: A Consumption-Based Explanation of Aggregate Stock Market Behavior. Journal of Political Economy 107 (2): Daly, Kevin James, and Abdallah Fayyad Can Oil Prices Predict Stock Market Returns? Modern Applied Science 5 (6): Eissa, Mohamed Abdelaziz, and Mohammed M Elgammal Do Developed Markets affect Emerging Markets? Evidence from the Gulf Cooperation Council. Journal of Business and Policy Research 10 (1):

20 Granger, Clive W.J Investigating Causal Relations by Econometric Models and Cross- Spectral Methods. Econometrica 37 (3): Hamilton, James. D Oil and the Macroeconomy since World War II. Journal of Political Economy 91 (2): Hamilton, James. D The Changing Face of World Oil Markets. NBER Working Paper (No ). Hammoudeh, Shawkat, and Eisa Aleisa Dynamic Relationships among GCC Stock Markets and NYMEX Oil Futures. Contemporary Economic Policy 22 (2): Hammoudeh, Shawkat, and Kyongwook Choi Behaviour of GCC Stock Markets and Impacts of US Oil and Financial Markets. Research in International Business and Finance 20 (1): Harrathi, Nizar, and Ahmed Almohaimeed, Interdependence between GCC Stock Market and Oil Prices and Portfolio Management Strategies under Structural Breaks. African Journal of Business Management 9 (5): Huang, Roger D., Ronald W. Masulis, and Hans R. Stoll Energy Shocks and Financial Markets. Journal of Future Markets 16 (1): Jamaani, Fouad, and Eduardo Roca Are the Regional Gulf Stock Markets Weak-Form Efficient as Single Stock Markets and as a Regional Stock Market? Research in International Business and Finance 33 (January 2015): Jones, Charles M., and Gautam Kaul Oil and the Stock Markets. Journal of Finance 51 (2): Jones, Charles M., and Gautam Kaul Oil and the Stock Markets. Journal of Finance 51 (2): Jouini, Jamel Stock Market In GCC Countries and Global Factors: A Further Investigation. Economic Modelling 31 (March 2013): Kaneko, Takashi, and Bong-Soo Lee Relative Importance of Economic Factors in the U.S. and Japanese Stock Markets. Journal of the Japanese and International Economies 9 (3): Karam, Philippe. D Exchange Rate Policies in Arab Countries: Assessment and Recommendation. Abu Dhabi, Arab Monetary Fund, Kling, John. L Oil Price Shocks and Stock Market Behavior. Journal of Portfolio Management 12 (1): Le, Thai-Ha, and Youngho Chang Effects of Oil Price Shocks on the Stock Market Performance: Do Nature of Shocks and Economies Matter? Energy Economics 51 (September 2015):

21 Lescaroux, François, and Valérie Mignon On the Influence of Oil Prices On Economic Activity and Other Macroeconomic and Financial Variables. OPEC Energy Review 32 (4): Liu, Li, Feng Ma, and Yudong Wang Forecasting Excess Stock Returns with Crude Oil Market Data. Energy Economics 48 (March 2015): Maghyereh, Aktham, and Ahmad Al-Kandari Oil Prices and Stock Markets in GCC Countries: New Evidence from Nonlinear Cointegration Analysis. Managerial Finance 33 (7): Malik, Farooq, and Shawkat Hammoudeh Shock and Volatility Transmission in the Oil, US and Gulf Equity Markets. International Review of Economic and Finance 16 (3), Mohanty, Sunil K., Mohan Nandha, Abdullah Q. Turkistani, and Muhammed Y. Alaitani Oil Price Movements and Stock Market Returns: Evidence from Gulf Cooperation Council (GCC) Countries. Global Finance Journal 22 (1): Narayan, Paresh Kumar, and Seema Narayan Modelling The Impact of Oil Prices on Vietnam s Stock Prices. Applied Energy 87 (1): National Bank of Kuwait The GCC Hydrocarbon Sector: Big and Getting Bigger. Kuwait: National Bank of Kuwait. OPEC OPEC Annual Statistical Bulletin Web. 15 June < 4.pdf> Papapetrou, Evangelia Oil Price Shocks, Stock Markets, Economic Activity and Employment in Greece. Energy Economics 23 (5): Park, Jungwook, and Ronald A. Ratti Oil Price Shocks and Stock Markets in the U.S. and 13 European Countries. Energy Economics 30 (5): Rizvi, Aun, and Mansur Masih Oil Price Shocks and GCC Capital Markets: Who Drives Whom? MPRA Paper (56993). Sadorsky, Perry Oil Price Shocks and Stock Market Activity. Energy Economics 21 (5): Sahu, Tarak Nath, Kalpataru Bandopadhyay, and Debasish Mondal An Empirical Study on the Dynamic Relationship between Oil Prices and Indian Stock Market. Managerial Finance 40 (2): Salisu, Afees A., and Tirimisiyu F. Oloko Modeling Oil Price-US Stock Nexus: A VARMA- BEKK-AGARCH approach. Energy Economics 50 (July 2015): Sbeiti, Wafaa, and Ayman E. Haddad Stock Markets Dynamics in Oil-Dependent Economies: Evidence from the GCC Countries. International Journal Research of Applied Finance 2 (3):

22 The Economist The Great Bear Market. 21 July Web. 15 June < Williams, Oral H., and Tahsin Saadi-Sedik Global and Regional Spillovers to GCC Equity Markets. 29 June Web. 15 June < Zarour, Bashar Abu Wild Oil Prices, But Brave Stock Markets! The Case of the GCC Markets. Operational Research 6 (2): Appendices Appendix 1: Global total proven oil reserves (at end of 2014). Region Barrels (in billions) % share of total Middle East % South & Central America % North America % Europe & Eurasia % Africa % Asia Pacific % World 1, % Source: BP (2014). Appendix 2: GCC proven oil reserves (2015). GCC Countries Barrels (in billion) % share of GCC countries KSA Kuwait United Arab Emirates Qatar Oman Bahrain Total Source: EIA (2015) Endnotes: 1 As exchange rates are benchmarked against the US dollar, the data used in this analysis will also be standardised in US dollars for comparison purposes. 2 UAE has 2 financial markets (Dubai and Abu Dhabi). 3 The authors examined Kuwait, KSA, Bahrain and Oman. 4 The authors examined Bahrain, Kuwait, Oman and KSA. 5 The returns on price indices are calculated using continuous returns formula. The formula is well-known to provide more accurate measurement of returns compared to discrete returns formula (see Brailsford, et. al., 2004, pp. 9). 6 Lag length test is carried out and Akaike information criterion suggested VAR model with 2 lags. 22