The effect of Brent oil price on the AEX index
Kit Yin Liu 10466460
University of Amsterdam Economics & Business
Bachelor thesis Finance & Organization 2016/2017 – Semester 1
January 2017
Abstract
Many studies have found a significant negative relationship between oil price and the stock markets. This study investigates the relationship between Brent crude oil price and the Dutch stock market during the period January 2003 till November 2016 using monthly data. A multiple linear regression model (OLS) will be used to capture the dynamic relationship between these variables and other control variables, including the interest rate, the EUR/USD exchange rate, U.S. field production of crude oil and GDP level. Results indicate that there is a significant positive relationship between the Brent crude oil price and the AEX index.
Statement of Originality
This document is written by Kit Yin Liu who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for contents.
3 1. Introduction
In a world where industrialization depended so heavily on oil, there is no doubt that oil price shocks have implications for the global economy. Oil is so connected with the international economy that economic growth forecasts are usually marked with the footnote: ‘Provided there is no oil shock’ (Adelman, 1993). Even today, the industrialized countries continue to be heavily dependent on oil, much of which is imported. According to Nandha and Faff (2008), oil is a fundamental driver of modern economic activity and the general market perception is that oil price shocks influence the stock market. An oil price shock may influence the world economy through a variety of mechanisms: transfer of wealth from oil importing countries to oil exporting countries and impact on inflation, consumer confidence and the stock market (Nandha and Faff, 2008). Political actions, such as oil embargoes and collaborative price setting from the Organization of the Petroleum Exporting Countries (OPEC) can have serious consequences for the global economy. Even when the price mechanism of oil is regulated by supply and demand forces in the free market, huge oil price shocks can affect the global economy by triggering inflationary trends, economic slowdowns and bear market in the world stock markets (Elyasiani et al, 2011). At microeconomy level, oil price shocks affect financial performance and economic activity of firms since oil is a major factor in the production process (Huang et al, 1996). Jones et al. (2004) also share this view and conclude that oil price shocks affect stock market returns through their effect on expected earnings.
According to Mussa (2000), oil price affecting financial performance and economic activity of firms should have implications for share prices and financial markets. When a firm is not doing well then it is reflected by a decrease of the share price, provided the firm is listed on a stock exchange. At first sight, oil price seem to influence the stock market. Stock markets with relatively large share of oil industry companies should notice the impact of an oil price shock. The Dutch stock market (AEX index) is a good example. The AEX index consists of oil and gas industry companies (Shell and SBM Offshore) with a weight of 16.55%, which is the largest industry represented on the AEX index (Euronext, 2017). So without any prior knowledge, one would assume oil price has an impact on the AEX index and might even be used as a prediction variable for the AEX index return.
4
This study attempts to examine the extent of the relationship between crude oil price and stock market returns. It will focus on the Dutch stock market over the period January 2003 and November 2016. This study is connected to the research of Driesprong et al. (2008), in which they investigated the effect of oil price on different stock markets in the period of 1973 to 2003. The focus of this study will be the AEX index and crude oil since Driesprong et al. (2008) also studied the effect on the AEX index in their study. This study strives to answer the following research question: what is the effect of Brent crude oil price change on the AEX index in the period January 2003 till November 2016? An ordinary least squares regression (OLS) will be used on the monthly data of Brent crude oil return, AEX index return and other control variables.
The remainder of the paper is structured as follows. The following section offers some background information with regards to the crude oil market and the AEX index. Section 3 discusses the relevant literature. Section 4 outlines the regression model and variables used in the analysis and describes the nature and sources of the data. Section 5 then presents the empirical results from the analysis. Section 6 concludes this paper with a discussion and conclusion of the main findings as well as outlining the study’s limitations and suggestions for further research.
5 2. Background information
2.1 Crude oil market
The crude oil market is one of the most important and largest commodity market, with a total world consumption of nearly 80 million barrels a day (Driesprong et al, 2008). There are many different types of crude oil but only three of them serve as benchmark oil; Brent crude oil, West Texas Intermediate (WTI) crude oil and Dubai crude oil. The value of crude oil is based on its density, sulphur content and location. Crude oils that have high density and high sulphur content are more expensive to process than crude oils with low density (also known as light) and low sulphur content (also known as sweet). For example, WTI crude oil is often more expensive than Brent crude oil as it is sweeter and lighter (Driesprong et al, 2008). However, Brent crude oil is still considered a sweet and light crude oil from the North Sea (OPEC, 2016). According to a report by OPEC (2016), Brent crude oil accounts for 62.5% of the total world oil consumption with 50 million barrels a day. WTI crude oil and Dubai crude both represent 18.75% of the total consumption with nearly 15 million barrels a day. The demand of oil for The Netherlands amounts to nearly 1 million barrel a day according to a report by the International Energy Agency (2015). The industry and transformation sectors are the main users with 80% of all oil used in The Netherlands (IEA, 2015).
The Brent crude oil price has been quite volatile since 2003 as figure 1 shows. The figure shows an all-time high price of around $135 per barrel in July 2008 before it decreases to nearly $40 per barrel by February 2009. By January 2011, the Brent crude oil increased back to above $100 and mostly maintained that level until the middle of 2014 due to declining demand from emerging markets and increased oil production from OPEC, Russia and the United States (OPEC, 2016).
6 Figure 1
Historic price of Brent crude oil
This figure shows the historical price of Brent crude oil in the period January 2003 till November 2016.The x-axis represents the year and y-axis represents the price of Brent crude oil in US dollar per barrel. The prices are not corrected for inflation.
Source: DataStream
2.2 AEX index
The AEX index is a Dutch stock market index that trade on the Euronext Amsterdam. It is a large cap index that consists of the 25 largest (in terms of market capitalization) and most actively traded Dutch companies. Figure 2 shows the historical price of the AEX index over the period January 2003 till November 2016. The weights of the AEX index can be found in the appendix.
Figure 2
Historic price of AEX index
This figure shows the historical price of the AEX index the period January 2003 till November 2016. The x-axis represents the year and y-axis represents the value of the AEX index. Source: DataStream
7 3. Literature review
There is no doubt that oil price shocks influence the economic activity of firms and the macroeconomy. According to Nandha and Faff (2008), oil price shocks have an adverse effect on the economic activity and the real output of firms. An increase in oil price leads to higher production costs and lower corporate earnings of firms where oil is mainly used as a production factor. Because stock prices are the present discounted value of expected future cash flow of a firm, such oil price shocks will be absorbed quickly in the stock market if the stock market is efficient (Jones et al, 2004). Moreover, oil price shocks can influence the interest rate for cash flow through the expected rate of inflation and the expected real interest rate (Miller & Ratti, 2009). Rising oil price is often an indication of inflationary pressure, which the central bank can control by increasing the interest rate. Higher interest rate will lead to bonds being cheaper and more attractive over stocks, which will in turn decrease stock prices (Basher & Sadorsky, 2006). According to Nandha and Faff (2008), oil price shocks also influence consumer confidence by creating uncertainty, leading to a temporarily decrease in spending on large consumption and investment goods such as cars, housing and other appliances. This temporarily uncertainty will often result in a decrease of the stock market.
The relationship between oil price shocks on stock market returns has been investigated by a number of researchers. In a comprehensive study, Jones and Kaul (1996) attempt to identify whether the effect of oil price on international stock markets (Canada, Japan, United States and United Kingdom) can be justified by current and future changes in real cash flows and changes in expected return. Using quarterly data from 1947 to 1991 and a standard cash-flow dividend valuation model, Jones and Kaul (1996) find a negative relationship between oil prices and stock market returns. The reaction of stock prices to oil price shocks can be completely accounted for by the impact of these shocks on real cash flows. The results for Japan and the UK are, however, not so significant. In general, they find that international stock markets do react to oil price shocks and present evidence that favors a negative relationship.
Another study by Sadorsky (1999) also concludes the negative relationship of oil price and stock market. Sadorsky (1999) uses a vector autoregression model with monthly data from January
8
1947 till April 1996 to examine the relationship between oil price change and the S&P 500. In his analysis, he finds that oil price changes and oil price volatility are important in explaining movements in the S&P 500 and that this effect is significantly negative. Another important finding by Sadorsky (1999) is that there is evidence that oil price shocks may have asymmetric effects on the economy. An increase in oil price may have a greater impact on stock returns and economic activity than a decrease in oil price. He also finds that oil price movements explain a larger portion of the forecast error variance in the S&P 500 than interest rates do.
Driesprong et al. (2008) find that a rise in oil price lowers future stock returns considerably. Their study also provide evidence towards a negative relationship between oil price and the stock market. They find that one-standard deviation increase in oil price (a rise of around 10%) in one month lowers the expected return in the next month to just below zero (- 0.1%). This means that for every one-standard deviation decrease in the oil price, the world stock market return will increase 1% per annum (Driesprong et al, 2008).
Other researchers have come to the same conclusion. Papapetrou (2001) shows that an oil price shock has a negative impact on stock prices in Greece. In addition, he concludes that oil price negatively affect industrial production and employment growth in Greece. O’Neill et al. (2008) find that an oil price increase lead to lower stock returns in France, the United States and the United Kingdom. Park and Ratti (2008) report that the relationship between oil price shocks and stock returns is significant negative in the United States and 12 European oil importing countries.
There are also various studies on industry sectors as opposed to on a stock market wide level. Narayan and Sharma (2011) investigate the effect of oil price on the returns of 560 U.S. firms listed on the NYSE. They find evidence that oil price affects industry returns differently with a positive relationship between oil price and oil and gas industry. Similarly, Faff and Brailsford (1999) investigate the relationship between the Australian industry returns and oil price shocks. Using monthly data over the period 1983 to 1996, they find a positive and significant impact of oil price on the oil and gas index and a negative and significant impact on the transportation industry. Nandha and Faff (2008) also find a detrimental effect of oil price on equity returns except for mining, and oil and gas industries. McSweeney and Worthington (2008) examine the impact
9
of oil price on the Australian monthly industry stock returns using data from January 1980 to August 2006. They find that oil price has a negative effect on banking, transportation and retail industry. They also find a positive effect on oil and gas industry and no effect on financial, insurance, materials, media and property trust industry. In addition, Sadorsky (2001) uses a multifactor market model which takes into consideration the presence of several risk premiums, interest rate, exchange rate along with the oil price as the main variable of the Canadian oil and gas industry returns. Using monthly data between 1983 and 1999, Sadorsky (1999) reports a significant positive relationship between the oil prices and stock returns coming from oil and gas firms. A 1% increase in oil prices will lead to a 0.305% increase in the oil and gas index. Sadorsky’s (2001) findings also reveal that an increase in oil prices will lead to an increase in the oil and gas index and that there is a significant positive relationship between the oil and gas index and the stock market as a whole. Huang et al. (1996) also investigate the relationship of oil price and U.S. stock returns on an industry level. Using a vector autoregression approach and daily data from 1979 to 1990, they find that oil price returns do lead the oil and gas index but they find no evidence of a broad-based market index impact like the S&P 500.
El-Sharif et al. (2005) use a multi-factor model to investigate the relationship between oil price and oil and gas firms returns in the UK. Their results are very similar to Sadorsky’s (2001) findings. According to El-Sharif et al. (2005), oil and gas industry is always positively associated with the oil price. In addition, they also concludes that the relationship between oil price and non-oil and gas sectors is weak. According to El-Sharif et al. (2005), this weak relationship is surprising, given the commonly asserted view that oil price change have impact on general macro-economic conditions.
In contrary to the studies mentioned above, Chen et al. (1986) investigate the impact of macroeconomic variables on stock price returns. They conclude that interest rate, inflation rate, bond yield spread and industrial production are priced in the stock market. However, Chen et al. (1986) did not find any evidence that oil price influences the stock market and therefore it is not priced in the stock market. Hamao (1988) also did not find any evidence that oil price influences the stock market on a sample of Japanese equity data. Aspergis and Miller (2009) also conclude that stock markets tend not to react to oil price shocks.
10
Whilst the literature above suggest that increasing oil price is generally bad for the economic growth, the true impact of an oil price increase on stock market returns depends on various factors such as whether oil is an input or output for an industry (Nandha & Faff, 2008). Furthermore, some industries might have monopoly power and might be able to pass on the higher oil price cost to their customers, thus minimizing the negative effect of an oil price increase on their profitability. This ‘pass-through’ effect depends on the degree of price elasticity and market power within an industry (Nandha & Faff, 2008). In addition to the direct effects of oil price shocks, there are also an indirect effects according to Nandha and Faff (2008). They argue that higher oil prices might influence stock markets through monetary policy, consumer confidence and employment rate, a view also shared by Bjørnland (2009).
11 4. Data and methodology
The main objective of this research is to find a relationship between the Brent crude oil price and the Dutch stock market index. A multiple linear regression model (OLS) will be used to capture the dynamic relations between these variables and other control variables, including the interest rate, the EUR/USD exchange rate, GDP and U.S. field production of crude oil. All data except U.S. field production of crude oil come from DataStream and cover the period January 2003 till November 2016 for a total of 166 monthly observations. U.S. field production of crude oil come from the database of the U.S. Energy Information Administration and cover the same period for a total of 166 monthly observations. Monthly frequency has been chosen because daily frequency is too noisy. In addition, most relevant studies use monthly data. The reason for the period January 2003 till November 2016, is that this study is related to the study by Driesprong et al. (2008). Their data ends in 2003 and this study starts where Driesprong et al. (2008) ended their research. However, this study has a much smaller focus and scale. Instead of investigating the stock markets of 48 countries and their industry sectors, this study is limited to only the stock market of The Netherlands. Another important difference is that this study will not take the lagged effect of oil price into account. In addition, Driesprong et al. (2008) compared trading strategy of oil price change with a simple buy-and-hold strategy. They compared both trading strategies and concluded that oil price strategy outperformed the buy-and-hold strategy, indicating economic significance results. This study will not conduct any trading strategies to investigate the economic significance of the findings. Lastly, this study will use more control variables whereas Driesprong et al. (2008) used none. The following model will be used to estimate the effect of Brent crude oil price and AEX index return:
𝑅𝑡 = 𝑐 + 𝛽1𝑂𝐼𝐿𝑡+ 𝛽2𝐸𝑋𝐶𝐻𝑡+ 𝛽3𝐺𝐷𝑃𝑡+ 𝛽4𝐼𝑁𝑇𝑡+ 𝛽5𝑃𝑅𝑂𝐷𝑡−1+ 𝜀𝑡 (1)
Where 𝑐 is the constant term while 𝑅𝑡 represents the dependent variable AEX index return in
month t. The MSCI reinvestment index for the AEX index has been used since it assumes the dividend paid being reinvested. Dividend payout affect AEX index return and therefore it will be captured in this model by using the MSCI reinvestment index for the AEX index. The main
12
independent variable 𝑂𝐼𝐿𝑡 is the Brent crude oil return in month t. Variable 𝐸𝑋𝐶𝐻𝑡 is the
EUR/USD exchange rate return in month t, variable 𝐺𝐷𝑃𝑡 is the GDP return in month t and variable
𝐼𝑁𝑇𝑡 is the 3-month Euribor interest rate in month t. Variable 𝑃𝑅𝑂𝐷𝑡−1 is the U.S. field production
of crude oil return in the previous month. This production variable will be lagged one month to allow the supply and demand mechanism of oil to be incorporated properly in the financial markets and economy. The error term is represented by 𝜀𝑡. All variables have been transposed
to return variables for easier interpretation. GDP data is only available on a quarterly basis and therefore interpolation is needed. The quarterly GDP value has been divided by three for a monthly GDP value.
Exchange rate control variable has been added since a depreciation (appreciation) leads to foreign investors decreasing (increasing) their investment in the AEX index. Interest rates also affect AEX index return in the way that an increase in the interest rates leads to bonds being cheaper and more attractive to hold over equities for investors. In addition, interest rate is the price paid for money and therefore an increase in interest rate leads to lower earnings for firms with a high debt-to-equity ratio (Sadorsky, 1999). Changes in GDP should also affect the AEX index since it influences the economic activity of those listed firms. Lastly, the U.S. field production of crude oil has been added as control variable to account for exogenous forces such as oil supply side shock.
Equation 1 will test whether the coefficient 𝛽1 is significantly different from zero. When 𝛽1 is significant, the null hypothesis of no oil effect is rejected. In addition, a robustness check will be performed to see whether weekly data yield different returns than monthly data since investors have different investment horizons. The following model will be used to estimate the effect of Brent crude oil price and AEX index return using weekly data:
𝑅𝑡 = 𝑐 + 𝛽1𝑂𝐼𝐿𝑡+ 𝛽2𝐸𝑋𝐶𝐻𝑡+ 𝛽3𝐼𝑁𝑇𝑡+ 𝛽4𝑃𝑅𝑂𝐷𝑡−1+ 𝜀𝑡 (2)
All the variables are defined as in equation 1 with the only difference being t as the return in week
t. The GDP variable has been excluded since it is quarterly data. To interpolate for weekly returns
13
findings are based on equation 1. Table 1 reports summary statistics for the monthly data used in the study and table 2 reports the correlation between the variables.
Table 1
Descriptive statistics of the data
Variable N Mean S.D. Minimum Maximum Kurtosis Skewness
𝑅𝑡 166 0.00398 0.05108 -0.19715 0.13716 5.31372 -0.83436 𝑂𝐼𝐿𝑡 166 0.00765 0.09744 -0.36594 0.30078 4.15939 -0.23924 𝐸𝑋𝐶𝐻𝑡 166 0.00052 0.03007 -0.08715 0.10772 4.20930 0.44790 𝐺𝐷𝑃𝑡 166 0.00105 0.00239 -0.01074 0.00495 11.6748 -2.07267 𝐼𝑁𝑇𝑡 166 0.01650 0.01523 -0.00309 0.05113 2.45433 0.68306 𝑃𝑅𝑂𝐷𝑡−1 166 0.00316 0.03294 -0.20515 0.19075 23.9442 -1.40840 The table provides information regarding the mean, standard deviation, minimum, maximum, kurtosis and skewness of the variables used in the regression analysis. 𝑅𝑡 represents the dependent variable AEX index return in month t. Variable 𝑂𝐼𝐿𝑡 is the Brent crude oil return in month t, variable 𝐸𝑋𝐶𝐻𝑡 is the EUR/USD exchange rate return in month t, variable 𝐺𝐷𝑃𝑡 is the GDP return in month t, variable 𝐼𝑁𝑇𝑡 is the Euribor 3-months interest rate in month t and variable 𝑃𝑅𝑂𝐷𝑡−1 is the U.S. field production of crude oil in the previous month.
Table 2
Correlation matrix of the variables
Variable 𝑅𝑡 𝑂𝐼𝐿𝑡 𝐸𝑋𝐶𝐻𝑡 𝐺𝐷𝑃𝑡 𝐼𝑁𝑇𝑡 𝑃𝑅𝑂𝐷𝑡−1 𝑅𝑡 1.00 - - - - - 𝑂𝐼𝐿𝑡 0.28 1.00 - - - - 𝐸𝑋𝐶𝐻𝑡 -0.16 -0.29 1.00 - - - 𝐺𝐷𝑃𝑡 0.14 0.06 -0.06 1.00 - - 𝐼𝑁𝑇𝑡 0.21 -0.01 -0.05 0.01 1.00 - 𝑃𝑅𝑂𝐷𝑡−1 0.16 0.03 -0.18 -0.11 -0.05 1.00
This table provides information regarding the degree of correlation between the variables used in the regression analysis. 𝑅𝑡 represents the dependent variable AEX index return in month t. Variable 𝑂𝐼𝐿𝑡 is the Brent crude oil return in month t, variable 𝐸𝑋𝐶𝐻𝑡 is the EUR/USD exchange rate return in month t, variable 𝐺𝐷𝑃𝑡 is the GDP return in month t, variable 𝐼𝑁𝑇𝑡 is the Euribor 3-months interest rate in month t and variable 𝑃𝑅𝑂𝐷𝑡−1 is the U.S. field production of crude oil in the previous month.
Results from table 1 show that the average return of the AEX index is 0.40% per month in the period January 2003 till November 2016. In the same period, the Brent crude oil return per month averaged at 0.77%. The GDP growth rate of the Netherlands amounts to 0.11% per month at an average. Interest rate and exchange rate return averaged at respectively 1.65% and 0.05% per month. Lastly, the U.S. field production of crude oil average return is 0.32% per month.
More interesting are the minimum and maximum observations. For the AEX index return, the minimum observation is -19.72% and maximum observation is 13.72%. The maximum of 13.72%
14
increase was in April 2003 and is the fifth largest increase per month in the history of the AEX index. The decrease of -19.72% occurred in September 2008. This is most probably due to the collapse of Lehman Brothers on September 15 and the start of the subprime mortgage crisis in the U.S. which led to the financial crisis and worldwide recession. This worldwide recession caused the largest decrease of GDP in The Netherlands with a minimum of -1.07% in January 2009. The maximum increase of GDP amounts to 0.49% in March 2006. However, these GDP value are interpolated values based on quarterly GDP data so careful interpretation is required. The reported minimum and maximum values are the same for two consecutive months after since the quarterly GDP value is simply divided by three months. The minimum and maximum observation of both the exchange rate return and Brent crude oil return also occurred around the beginning of the financial crisis. The maximum of 10.77% increase in the exchange rate was in October 2008 and caused the euro to depreciate against the US dollar. The minimum of -8.72% was 2 months later, in December 2008. According to a study by the European Central Bank (2009), a repatriation of capital to the U.S. and flight to safe havens by investors may have caused the need for US dollar liquidity and the appreciation of the US dollar during the financial crisis. In addition, the wild fluctuations of risk aversion level after the collapse of Lehman Brothers may have been the cause of both extremes in the EUR/USD exchange rate (ECB, 2009). The Brent crude oil price has dropped -36.59% in October 2008 from $93.84 to $59.50 per barrel due to reduced economic activity during the financial crisis. However, the highest increase of Brent crude oil price is 30.08% in May 2009, rising from $49.87 to $64.87 per barrel. The highest observed interest rate is 5.11% in October 2008. The collapse of Lehman Brothers and the possibility of bad mortgages banks had on their balance sheet caused banks to be reluctant to lend out money which may have increased the Euribor to a maximum of 5.11%. The minimum observation is in November 2016 with a negative interest rate of -0.31% due to ECB’s monetary policy and in particular the asset purchase programme (APP). Lastly the minimum and maximum observation of -20.52% and 19.08% for the U.S. field production of crude oil happened in respectively September 2008 and October 2008, at the start of the U.S. subprime mortgage crisis.
Based on the results from table 1, the kurtosis of all return series except interest rate is greater than 3, indicating leptokurtic distributions across the variables. This implies that there is a
15
prevalence of extreme values in the data studied. However, the high kurtosis should not be a problem since the data set is large enough. The Brent crude oil, AEX index, GDP and U.S. field production of crude oil exhibit (slight) negative skewness implying that these variables had higher probabilities of negative returns during the period January 2003 and November 2016.
Table 2 indicates that there is a positive correlation between the AEX index return and Brent crude oil return, GDP return, interest rate, U.S. field production of crude oil. A negative correlation is found between the AEX index return and exchange rate. There seems to be little correlation between the independent variables. Only the exchange rate is negatively correlated with Brent crude oil and U.S. field production of crude oil.
16 5. Empirical results
According to the prevalent literature, the effect of oil price on market index return is significantly negative (e.g. Jones and Kaul, 1996; Sadorsky, 1999; Driesprong et al, 2008; Hamilton, 1996). Most of these studies are based on the stock market indices of Australia, Canada, United States and United Kingdom. However, Driesprong et al. (2008) studied the relationship between stock market index and oil price and found that there is a significant negative relationship between the AEX index and oil price in the period 1973 till 2003. This study found a significant positive relationship between the AEX index and oil price in the period 2003 till 2016. The regression estimates results are shown in table 3.
Table 3
Regression results
Variable Coefficient Estimate t-statistic p-value
𝑐 0.0106 (0.0057) 1.84 0.068 𝑂𝐼𝐿𝑡 0.1317 (0.0396) 3.33 0.001 𝐸𝑋𝐶𝐻𝑡 -0.1074 (0.1311) -0.82 0.414 𝐺𝐷𝑃𝑡 2.8973 (1.5583) 1.86 0.065 𝐼𝑁𝑇𝑡 -0.6834 (0.2431) -2.91 0.004 𝑃𝑅𝑂𝐷𝑡−1 0.2217 (0.1151) 1.93 0.056
This table reports the estimation results of regression equation 1 in the period January 2003 till November 2016 with 166 monthly observations. Standard deviations values are in parentheses, figures in bold are significant at the 95% confidence level. 𝑅𝑡 represents the dependent variable AEX index return in month t. Variable 𝑂𝐼𝐿𝑡 is the Brent crude oil return in month t, variable 𝐸𝑋𝐶𝐻𝑡 is the EUR/USD exchange rate return in month t, variable 𝐺𝐷𝑃𝑡 is the GDP return in month t, variable 𝐼𝑁𝑇𝑡 is the Euribor 3-months interest rate in month t and variable 𝑃𝑅𝑂𝐷𝑡−1 is the U.S. field production of crude oil in the previous month.
The regression coefficient of 0.1317 for the oil price return is significant at a 5% significance level and indicates a 0.13% increase in the AEX index for every 1% increase in the Brent crude oil price per month. This results is not surprising given the relatively heavy weight of oil and gas industry on the AEX index. Previous related studies (e.g., Sadorsky, 2011; Faff and Brailsford, 1999; Nandha and Faff, 2008) found that the relationship between Brent crude oil and oil and gas industry is always positive and very significant. The result of this study is similar to the findings of Sadorsky
17
(2001). According to Sadorsky (2001), increases in oil prices will lead to increases in the oil and gas index and there is a significant positive relationship between the oil and gas index and the stock market as a whole. Based on the latest weighing of Euronext (2017), oil and gas industry companies (Royal Dutch Shell and SBM Offshore) add up to 16.55% of the market capitalization, which is the largest industry represented on the AEX index. Therefore it is not surprising that there is a positive relationship between the Brent crude oil and AEX index.
The joint F-test value constraining all slope coefficients of the control variables equal to zero is 4.12 and is therefore rejected at the 5% significance level. From table 3, the regression coefficient of GDP return is 2.8973 but not significant at a 5% significance level. This indicates that the AEX index will increase with 2.90% for every 1% increase in GDP per month. Levine and Zervos (1998) also found this positive correlation between GDP and stock market index. It is not surprising that GDP growth affects the AEX index positively since GDP increases lead to better economic activity and generally better financial performances of firms that will be reflected on the AEX index. Based on table 3, the regression coefficients of the interest rate shows a negative effect on the AEX index. The regression coefficient of -0.6834 for the interest rate is significant at a 5% significant level and indicates a -0.68% decrease in the AEX index for every 1% increase in the interest rate per month. This negative relationship between interest rate and stock market is also found in other related studies (e.g. Pearce and Roley, 1983; Sadorsky, 1999). According to Sadorsky (1999), changes in interest rate affects the stock market in different ways. First, interest rate influences the level of corporate profits since it is the price charged for borrowing money. Second, movements in interest rates influence the degree in which investors hold competing financial assets, such as bonds or equities. Third, changes in the interest rate influence the desire or ability of investors to speculate with stocks purchased on margin. Therefore, increases in interest rates lower stock returns. The results in table 3 further show that the regression coefficient of the exchange rate return is -0.1074 and is not significant at a 5% significance level. This indicates that a 1% increase in exchange rate will yield a -0.11% decrease in the AEX index per month. An increase in the EUR/USD exchange rate implies a depreciation of the Euro and investors switching to US dollar assets. The coefficient being not significant might indicate that there are not that many US investors on the AEX index. The AEX is relatively small compared to other international
18
stock indices and might not attract enough US investors for the EUR/USD exchange rate to make a statistical difference. Lastly, the coefficient of U.S. field production of crude oil is 0.2217, indicating a 0.22% increase in the AEX index for every 1% increase in the U.S. field production of crude oil. This result is however, just slightly not significant at a 5% significance level.
A robustness check has been done to see whether using weekly data will have a different outcome than monthly returns. Table 3 reports the outcome of the regression using weekly data. As shown in the table, the coefficient of the Brent crude oil variable is 0.16%, slightly higher than using monthly data. This shows that the results are fairly robust for Brent crude oil and AEX index.
Table 3
Regression results using weekly data
Variable Coefficient Estimate t-statistic p-value
𝑐 0.0031 (0.0016) 1.96 0.050 𝑂𝐼𝐿𝑡 0.1639 (0.0241) 6.79 0.000 𝐸𝑋𝐶𝐻𝑡 0.0678 (0.0816) 0.83 0.406 𝐼𝑁𝑇𝑡 -0.6152 (0.2820) -2.18 0.029 𝑃𝑅𝑂𝐷𝑡−1 0.0618 (0.0485) 1.27 0.203
This table reports the estimation results of regression equation 2 in the period January 2003 till November 2016 with 726 weekly observations. Standard deviations values are in parentheses, figures in bold are significant at the 95% confidence level. 𝑅𝑡 represents the dependent variable AEX index return in week t. Variable 𝑂𝐼𝐿𝑡 is the Brent crude oil return in week t, variable 𝐸𝑋𝐶𝐻𝑡 is the EUR/USD exchange rate return in week t, variable 𝐼𝑁𝑇𝑡 is the Euribor 3-months interest rate in week t and variable 𝑃𝑅𝑂𝐷𝑡−1 is the U.S. field production of crude oil in the previous week.
19 6. Discussion and concluding remarks
A vast number of studies report a negative relationship between crude oil prices and the stock market (e.g. Jones and Kaul, 1996; Driesprong et al, 2008; Sadorsky, 1999). Driesprong et al. (2008) found that in the period 1973 till 2003, the effect of crude oil prices on the Dutch stock market, among other stock market indices, is negative and significant. This study is based on the work of Driesprong et al. (2008) on a much smaller scale and focus. The main focus of this study is the relationship between the AEX index and the Brent crude oil in the period January 2003 till November 2016. In this study, a multiple linear regression (OLS) model is developed and estimated in order to investigate the empirical relationship between these variables and other control variables, including the interest rate, the EUR/USD exchange rate, GDP and U.S. field production of crude oil. The result of this regression analysis is different than what the prevalent literature suggest. This study found a significant positive relationship between the AEX index and the Brent crude oil with an estimated 0.13% increase in the AEX index for every 1% increase in the Brent crude oil price. The reason for this might be the relatively high weight of oil and gas industry companies on the AEX index compared to other international stock market indices. With a weight of 16.55%, the oil and gas industry is the largest industry represented on the AEX index (Euronext, 2017). These findings are similar to other related findings (e.g. Sadorsky 2001; Nandha and Faff, 2008; Faff and Brailsford, 1999), where a significant positive relationship between crude oil prices and oil and gas industry companies is found. In addition, the AEX index consists of mostly firms that do not consider oil as a major production factor. Therefore, an increase in oil price does not lower the corporate earnings of these firms as much as oil dependent firms. Oil dependent industries such as the construction & materials, industrial goods & services and chemical industry only account for less than 10% of the AEX index, as shown in the appendix. Because stock prices are the present discounted value of expected future cash flow of a firm, a rise in oil price will not lead to a high decrease in the stock price for the majority of the firms on the AEX index.
The study has a number of obvious limitations. There are many variables that will impact the stock market index but is not incorporated in the regression model (e.g. investor’s confidence, inflation). In addition, this study is based on the assumption that the relationship between Brent crude oil and AEX index is linear which it is not always the case as Ciner (2001) indicates. The GDP
20
data is also an issue since the regression model is based on monthly data whereas GDP data is publicized quarterly. Interpolation was needed for the GDP data which resulted in noisy monthly data. Lastly, Driesprong et al. (2008) found a lagged response in crude oil price changes on the stock market despite being public information. This study has not taken the lagged effect of oil price change into account.
Improvements for future research, which could eliminate some of the limitations of this study, could also investigate more stock market indices in the same period for comparison. Future research could also incorporate turbulent times such as crises and war events in the model to take the structural breaks into account. In addition, nonlinearities could also be tested with structural breaks as Ciner (2001) and others argue that the relationship between crude oil prices and stock market might be nonlinear in the long-run. Extreme values of crude oil price increases or decreases might affect the stock market differently.
21 References
Adelman, M. A. (2003). The Economics of Petroleum Supply. Cambridge, MA: The MIT Press.
Aspergis, N., & Miller, S. M. (2009). Do structural oil-market shocks affect stock prices? Energy
Economics, 31(4), 569 – 575.
Basher, S. A., & Sadorsky, P. (2006). Oil price risk and emerging stock markets. Global Finance
Journal, 17(2), 224 – 251.
Bjørnland, H. C. (2009). Oil price shocks and stock market booms in an exporting country.
Journal of Political Economy, 56(2), 232 – 254.
Chen, N., Roll, R., & Ross, S. A. (1986). Economic Forces and the Stock Market. The Journal of
Business, 59(3), 383 – 403.
Ciner, C. (2001). Energy Shocks and Financial Markets: Nonlinear Linkages. Studies in Nonlinear
Dynamics & Econometrics, 5(3), 203 – 212.
Driesprong, G., Jacobsen, B., & Maat, B. (2008). Striking oil: Another puzzle? The Journal of
Financial Economics, 89(2), 307 – 327
El-Sharif, I., Brown, D., Burton, B., Nixon, B., & Russel, A. (2005). Evidence on the nature and extent of the relationship between oil prices and equity values in the UK. Energy Economics,
27(6), 819 – 830.
Elyasiani, E., Mansur, I., & Odusami, B. (2011). Oil price shocks and industry stock returns.
Energy Economics, 33(5), 966 – 974.
Euronext (2017). AEX index [Fact sheet]. Retrieved from:
https://www.euronext.com/sites/www.euronext.com/files/aex_20161230.pdf
European Central Bank (2009). What explains global exchange rate movements during the
financial crisis? (Working paper series no 1060). Retrieved from: https://www.ecb.europa.eu/pub/pdf/scpwps/ecbwp1060.pdf
22
Faff, R. W., & Brailsford, T. J. (1999). Oil price risk and the Australian stock market. Journal of
Energy Finance & Development, 4(1), 69 – 87.
Hamao, Y. (1988). An empirical examination of the Arbitrage Pricing Theory: Using Japanese Data. Japan and the World Economy, 1(1), 45 – 61.
Hamilton, J. D. (1996). This is what happened to the oil price-macroeconomy relationship.
Journal of Monetary Economics, 38(2), 215 – 220.
Huang, R. D., Masulis, R. W., & Stoll, H. R. (1996). Energy shocks and financial markets. Journal
of Futures Market, 16(1), 1 – 27
International Energy Agency (2015). Annual Statistical Supplement. Retrieved from: https://www.iea.org/media/omrreports/2015/AnnualStatisticalSupplement_2015.pdf
Jones, M. C., & Kaul, G. (1996). Oil and stock markets. The Journal of Finance, 51(2), 463 – 491.
Jones, W. D., Leiby, P. M., & Paik, I. K. (2004). Oil price shocks and the macroeconomy: what has been learned since 1996. Energy Journal, 25(2), 1 – 32.
Levine, R., & Zervos, S. (1998). Stock Markets, Banks, and Economic Growth. The American
Economic Review, 88(3), 537 – 558.
McSweeney, E. J., & Worthington, A. C. (2008). A comparative analysis of oil as a risk factor in Australian industry stock returns, 1980-2006. Studies in Economics and Finance, 25(2), 131 – 145.
Miller, J. I., & Ratti, R. A. (2009). Crude oil and stock markets: Stability, instability, and bubbles.
Energy Economics, 31(4), 559 – 568.
Mussa, M. (2000). The impact of higher oil prices on the global Economy (International Monetary Fund). Retrieved from: http://www.imf.org/external/pubs/ft/oil/2000/.
Nandha, M., & Faff, R. (2008). Does oil move equity prices? A global view. Energy Economics,
23
Narayan, P. K., & Sharma, S. S. (2011). New evidence on oil price and firm returns. Journal of
Banking & Finance, 35(12), 3253 – 3262.
O’Neill, T. J., Penm, J., & Terrell, R. D. (2008). The role of higher oil prices: A case of major developed countries. Research in Finance.
Organization of the Petroleum Exporting Countries (2016). World Oil Outlook. Retrieved from: http://www.opec.org/opec_web/static_files_project/media/downloads/publications/WOO%20 2016.pdf
Papapetrou, E. (2001). Oil price shocks, stock market, economic activity and employment in Greece. Energy Economics, 23(5), 511 – 532.
Park, J., & Ratti, R. A. (2008). Oil price shocks and stock markets in the U.S. and 13 European countries. Energy Economics, 30(5), 2857 – 2608.
Pearce, D. K., & Roley, V. V. (1983). The Reaction of Stock Prices to Unanticipated Changes in Money: A Note. The Journal of Finance, 38(4), 1323 – 1333.
Sadorsky, P. (1999). Oil price shocks and stock market activity. Energy Economics, 21(5), 449 – 469.
Sadorsky, P. (2001). Risk factors in stock returns of Canadian oil and gas companies. Energy
24 Appendix
Figure 3
Aex index weights breakdown by company
This figure shows the breakdown of the AEX index by companies in percentage of the total market capitalization.
25
Figure 4
Aex index weights breakdown by industry
