• No results found

The relationship between oil prices and stock market indices in eight European countries for the recent 15 years: 2002-2016

N/A
N/A
Protected

Academic year: 2021

Share "The relationship between oil prices and stock market indices in eight European countries for the recent 15 years: 2002-2016"

Copied!
24
0
0

Bezig met laden.... (Bekijk nu de volledige tekst)

Hele tekst

(1)

1

The relationship between oil prices and stock

market indices in eight European countries for

the recent 15 years: 2002-2016

Bachelor Thesis

Abstract

The aim of this thesis is to investigate the relationship between oil price and the stock market indices, focusing on the correlation between WTI crude oil prices and stock market indices in eight European countries over the period of 2002-2016. The empirical results show that there are a positive and significant relationships between the returns of oil price and the returns of stock market indices in France and Germany. The rest of the countries do not clearly respond to the changes in oil prices over the 15 years sample period.

Student: Guseok Song

Student number: 10918531

Supervisor: Leva Sakalauskaite Bachelor program: Economics and Business Specialization: Finance and Organization

(2)

2

Statement of originality

This document is written by Guseok Song 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 the contents.

(3)

3

Content

1. Introduction 4

2. Literature review 6

2.1 Relationship between oil prices and stock markets 6

2.2 Financial crisis 7

3. Methodology 9

3.1 Multiple regression model 9

3.2 Hypotheses 10

4. Empirical analysis 11

4.1 Data 11

4.2 Data analysis 13

4.3 Results 14

4.3.1 Multiple regression model 14

4.3.2 Summary of results 16

5. Robustness check 17

5.1 Different oil price benchmark 17

6. Conclusion 18

7. Limitations and Further research 19

8. References 20

(4)

4

1 Introduction

During the 1970s, the dramatic oil shocks had severely influenced the world economy. Accordingly, the attentions on the movements of oil prices had been considerably increased from economists. For instance, Hamilton (2011) stated that most post-World War II recessions in the U.S. had been occurred by an increase in oil prices. Moreover, empirical studies find that the changes in oil prices may help to predict and explain real economic activity. According to Cunado and Gracia (2014), the movements of oil prices not only trigger relevant macroeconomic factors, but it may also affect the factors in financial markets such as exchanges rates and returns on stock markets.

Although the changes in oil prices may play a crucial role in the global stock markets, the results of empirical studies about the relationship between oil price movements and returns on stock markets are largely dissimilar. For example, Sadorsky (1999) who examined the relationship between oil prices and stock market returns in the U.S. and Cunado and Gracia (2014) that investigated the same relationship in Europe find that the relationships between oil prices and stock markets are negative, reflecting stock market returns decrease when oil prices increase, while Chen et al. (1986) and Huang et al. (1996) do not show a significant relationship between oil prices and stock market returns. Moreover, the changes in oil prices help to predict returns on stock markets (Diesprong et al., 2008). Arouri (2011) who examined the relationship between oil prices and stock market returns based on sector locations instead of countries finds that there are different sensitivities of sectors to oil prices changes.

Therefore, the purpose of this study is to investigate the relationship between oil prices and stock market indices returns in some European countries by using multiple regression model and the model will examine the period from 02/01/2002 to 31/12/2016. The collection of the European countries consists of eight countries: Austria, Belgium, France, Ireland, Netherlands, Portugal, Spain and Germany.

This paper contributes to the literature on the relationship between oil prices and stock market returns in the following ways. First, the paper covers the recent 15 years and the period includes the responses of stock markets in Europe both before and after the global financial crisis in 2008, which is the period that the crude oil price was at peak of US$145 per barrel in July 2008. According to the existing literature, the results are slightly outdated and the results of this paper are more recent in case of European stock market investigation. Second, there

(5)

5

have been relatively less studies analyzing the relationship between oil prices and stock market returns (Chen, 2010). Considering the relative sensitivities of stock market returns to oil price changes would be beneficial for risk management purposes for investors and businesses. Lastly, this paper proposes an alternative way to investigate the relationship between oil prices and stock market returns based on multiple regression model while other researchers use another model such as Vector Autoregressive (VAR) (i.e., Apergis and Miller, 2009; Cunado and Gracia, 2014; Park and Ratti, 2008; Sadorsky, 1999; Zhang, 2017). Furthermore, while the relationship between oil prices and stock market returns in Europe are examined based on sector investigation by Aruori (2011), this paper will investigate the relationship between oil prices and stock market returns in Europe based on country-specific. The research question to be answered is as follows:

Do the stock market indices in eight European countries respond to the changes in oil prices during the period from 02/01/2002 to 31/12/2016?

This paper is structured in the following order. The second chapter will provide an overview of the existing literature about the relationship between oil prices and stock markets in various settings. The third chapter will present the methodology used and also the hypothesis will be stated. In the fourth chapter, the data used in the paper will be described and analyzed. In addition, the empirical results of the analysis will be discussed. The fifth chapter is devoted to robustness checks for the results. The sixth chapter will provide a main conclusion of the paper. The last chapter will address limitations of the paper and suggestions for further research.

(6)

6

2 Literature review

In this chapter, the existing literature on the relationship between oil prices and stock market returns are reviewed. Also, the reasons why the 2008 global financial crisis matter for this paper are discussed.

2.1 Relationship between oil prices and stock markets

Some papers investigate the relationship between oil prices and stock markets based on sector- or industry-level (Arouri, 2011; Elyasiani et al., 2011). For example, Aruori (2011) examines the short-term links between the changes in oil prices and sector stock market returns in Europe. The author uses weekly stock market indices over the sample period from 1998 to 2010. His paper shows a strong and significant relationship between the changes in oil prices and sector markets for most European countries. The results of his paper also suggest that the reactions of stock market returns to oil price changes differ substantially sector by sector.

Likewise, Elyasiani et al. (2011) examines the impact of changes in the oil returns and its volatility on excess returns and return volatilities of 13 U.S. industries during the period from 1998 to 2008. Their results bring a strong evidence to support the view that the changes in oil prices contain a systematic asset price risks at the industry level. Most industries examined in their paper show a significant relationships between oil-futures return and industry excess return.

This paper will attempt to investigate the relationship between oil prices and stock markets focusing on country-specific instead of analyzing the stock market returns based on industries or sectors. There are also other papers that examine the relationship based on country-specific and national indices (i.e., Sadorsky, 1999; Park and Ratti, 2008; Apergis and Miller, 2009; Cunado and Gracia, 2014; Narayan and Gupta, 2015; Balcilar et al., 2017; Zhang, 2017). Sadorsky (1999), using monthly data for the period from 01/1947 to 04/1996, suggests that oil price changes affect economic activity but, changes in economic activity have little impact on oil prices. It was also mentioned by his research that movements of oil prices are important in explaining movements of stock returns in the U.S. market. In another study, Park and Ratti (2008), using monthly data for the U.S. and in 13 European stock markets over the period from 01/1986 to 12/2005, find that oil price shocks contain a statistically significant impact on stock market returns. They also show that there is not enough evidence of asymmetric impacts on stock market returns of oil price shocks for oil importing European

(7)

7

countries. Apergis and Miller (2009), using monthly stock market data for Australia, Canada, France, Germany, Italy, Japan, the United Kingdom, and the United States over the period from 1981 to 2007, show that international stock market returns do not largely respond to oil price shocks. Cunado and Gracia (2014), using monthly data for 12 European stock markets during the period from 02/1973 to 12/2011, find that the responses of the European real stock returns to oil price shocks may differ substantially depending on the underlying causes of the oil price change. Narayan and Gupta (2015), using monthly data for the U.S. during the period from 1859 to 2013, show that both positive and negative oil price changes are important predictors of the U.S. stock market returns. Balcilar et al. (2017), using the stock market data for the U.S. over the period from 09/1859 to 07/2015, oil prices and the stock markets have a common stochastic trend over the sample period, but a common cycle only exists during the post-World War II period. Lastly, Zhang (2017), using monthly data for Germany, Japan, Shanhai, Singapore, the United Kingdom, and the United States during the period from 01/2000 to 03/2016, finds that oil shocks may be important to a single financial market, but have no serious impacts on the six major financial markets in general. Also it was mentioned in the paper that the 2008 global financial crisis significantly influenced the world economy and thus it is necessary to be taken into account in order to capture a more systematic responses of stock markets to oil price changes after the crisis.

Instead of using monthly data for the investigation of the relationship between oil prices and stock market returns, this paper will use daily data for eight European stock market returns and oil prices over the recent 15 years. It is because of that daily returns can deal with holidays and their lag relationships, and that daily returns are superior for short or medium tactical forecasting as days of the week may have different patterns. Moreover, this paper will take the 2008 global financial crisis into account to produce a meaningful outcome about the relationship between oil prices and stock market returns as suggested by Zhang (2017).

2.2 Financial crisis

Generally, it is quite vague to clearly define the start and end of financial crisis, but it is necessary to take into account in order for this paper to investigate any impacts of the crisis on the eight European countries.

This paper investigates the relationship between oil prices and stock markets and examines the period from 2002 to 2016, which includes the 2008 global financial crisis. Since

(8)

8

the 2008 global financial crisis, the world economic and financial system had changed significantly. Markets are more integrated and thus different sectors and countries react in a more systematic way (Zhang, 2017). He also suggests that it is beneficial to be aware of the increasing connectedness in the economic and financial system worldwide for policy makers and financial practitioners.

While this paper mainly investigates the impacts of oil prices on stock markets, it can additionally observe the before and after the crisis impacts on the stock markets. In order to see the impacts of the crisis, it is required to define the start date of the 2008 global financial crisis to separate both before and after crisis period. The start date of the crisis for this paper is considered as 15th of September 2008, which is the day when the Lehman brothers that was the fourth-largest investment bank in the U.S. had collapsed.

2.3 Summary of literature review

Previous studies show a various results on the links between oil prices and stock markets and also the importance of considering the financial crisis. Therefore, this paper will investigate the relationship between oil prices and stock market returns in Europe with looking at the impacts of the 2008 global financial crisis. If a significant relationship between oil prices and stock markets exists in the eight European countries, then the characteristics of the relationship will also be identified in the later chapter.

(9)

9

3 Methodology

This chapter introduces the methodology conducted in order to analyze the correlations between the returns of oil prices and the returns on stock markets in Austria, Belgium, France, Ireland, Netherlands, Portugal, Spain and Germany during the period from 02/01/2002 to 31/12/2016. First, multiple regression model used is described. Next, the hypothesis relating to the correlation between the returns on stock market indices in Europe and West Texas Intermediate (WTI) crude oil returns are presented.

3.1 Multiple regression model

The multiple regression model used for the paper estimates the returns on stock market indices in the eight European countries based on the three variables introduced in equation (2).

(2) 𝑅𝐶𝑜𝑢𝑛𝑡𝑟𝑦,𝑡 = 𝐵0+ 𝐵1𝑅𝑂𝑖𝑙,𝑡+ 𝐵2𝑟𝐸𝑈𝑅𝐽,𝑡+ 𝐵3𝐷𝑡+ 𝜀𝑡 Where:

𝑅𝐶𝑜𝑢𝑛𝑡𝑟𝑦,𝑡 The dependent variable: the natural logarithmic daily return on stock market index for each country

𝑅𝑂𝑖𝑙,𝑡 The explanatory variable: the natural logarithmic daily return on the WTI crude oil prices

𝐷𝑡 The dummy variable being equals to 1 after the 2008 global financial crisis

𝑟𝐸𝑈𝑅𝐽,𝑡 The control variable: the European market index returns filtered by the WTI

crude oil price return.

(3) 𝑅𝐸𝑈 = 𝑎 + 𝐵1𝑅𝑜𝑖𝑙+ 𝑟𝐸𝑈𝑅𝐽

The multiple regression model contains the three variables that have a different characteristics. First, the natural logarithmic daily return on the WTI crude oil prices is used as the main explanatory variable. Second, the dummy variable is applied in order to capture and check if there exists any systematic risks or impacts on the returns on stock market indices before and after the 2008 global financial crisis. Lastly, the control variable (𝑟𝐸𝑈𝑅𝐽) is obtained from equation (3). More precisely, 𝑅𝐸𝑈 is the natural logarithmic daily return on Euronext100 (a proxy of European stock market index), and 𝑟𝐸𝑈𝑅𝐽 is merely the residual of the Ordinary Least Squares (OLS) regression of the Euronext100 returns on the WTI oil price returns (𝑅𝑂𝑖𝑙).

The reason of using the control variable is to filter out the crude oil price impacts on the European market index. This method is implemented in order to see whether the index of each country has any sensitivities to the European market index, Euronext100.

(10)

10

3.2 Hypotheses

Given the multiple regression model described in the previous section, the hypotheses with regard to the relationship between WTI crude oil prices and stock market indices in Europe is formulated in statement (1). According to the existing literature, it is still controversial whether the relationship is positive or negative. By the means of being positive or negative, if it is positive, the stock market returns in Europe decrease when oil prices go up.

(1) H0: the returns of WTI crude oil prices do not have a significant influence on the returns of stock market index in each country in Europe

H1: the returns of WTI crude oil prices do have a significant influence on the returns of stock market index in each country in Europe

In other words, the hypotheses expect that the stock market indices in Europe do not respond to the changes in WTI crude oil prices. It is meaningful to have the expectation because of that it is believed in this research that the relationship is not significant in European stock markets.

In the case of that the null hypothesis is rejected, it will be concluded that the stock market in the country that is being rejected respond to the changes in the oil prices. This paper will therefore investigate the relationship between oil prices and stock market indices based on the possible hypotheses.

(11)

11

4 Empirical analysis

In this chapter, the data used will be described and an analysis of the data will be presented. Also, the results of the analysis will be discussed.

4.1 Data

This paper uses the data with a daily basis during the period from 02/01/2002 to 31/12/2016. The data for stock market indices for the eight European countries, except for Belgium, and the European market index that is Euronext100 was obtained from Thomson Reuters Datastream. The index for Belgium is separately obtained from Euronext.

The stock market indices are AEX, ATX, BSI20, CAC40, DAX30, IBEX35, ISEQ, and PSI20 for Netherlands, Austria, Belgium, France, Germany, Spain, Ireland, and Portugal respectively. Those indices are used as a proxy for a country’s stock market and its movements are presented in the Appendix. For the 2008 global financial crisis dummy variable, the starting date of the crisis, 15/09/2008, is taken in order to divide the periods before and after the crisis. In order to make this paper more representative, the description of the multiple stock market indices that have been used is listed in Table 4.1.

Table 4.1: Data description for stock market indices

Country Proxy index

Austria ATX Belgium BSI20 France CAC40 Germany DAX30 Ireland ISEQ Netherlands AEX Portugal PSI20 Spain IBEX35 Europe Euronext 100

The data for West Texas Intermediate (WTI) crude oil spot prices is obtained from Energy Information Administration (EIA) in the United States. The WTI crude oil is the underlying commodity of the New York Mercantile Exchange’s oil futures and is widely used as an oil price benchmark worldwide. According to EIA, WTI crude oil is simply petroleum

(12)

12

that are used as fuels for transportation, heating, paving roads, and generating electricity. Balcilar et al. (2015) that examine the relationship between WTI crude oil prices and S&P500 stock market state that the WTI crude oil price remained relatively stable from World War II until the emergence of Organization of Petroleum Exporting Countries (OPEC) in 1973. It is also mentioned that the volatility of WTI crude oil had become larger since 1973.

Thus, it is appropriate to choose the WTI crude oil as an oil price benchmark as this paper deals the data for the recent years. There are other researchers that also used the WTI crude oil as a benchmark (i.e., Basher el al., 2012; Balcilar et al., 2015; Narayan and Gupta, 2015; Balcilar et al., 2017; Joo and Park, 2017; Zhang, 2017). The movements of WTI crude oil prices during the sample period is presented in Figure 4.1

Figure 4.1: The historical price of WTI crude oil

Notes: This figure shows the historical price of WTI crude oil during the period from January 2002 to December 2016. The x-axis represents the dates and y-axis represents the price of WTI crude oil in U.S. dollar per barrel.

Source: Energy Information Administration (EIA)

In Figure 4.1, the oil price was at peak of US$145 per barrel in July 2008. It interestingly coincides with the period of the 2008 global financial crisis. After the peak, it started to drop considerably to US$30 per barrel in December 2008 and dramatically fluctuated after 2008. At a glance, the movements of oil price are quite volatile for the last 15 years and thus this paper expects that there may be a significant impacts on the stock markets in Europe.

0 20 40 60 80 100 120 140 160

(13)

13

4.2 Data analysis

Instead of investigating the correlations between the stock market indices and oil prices, this paper examines the correlation coefficients between the returns of the indices and the returns of oil prices.

In Table 4.2 below, the descriptive statistics show that the returns of crude oil outperform the returns of stock market indices in Europe. However, the returns of stock market indices have a lower variation in daily returns compared to oil price returns. In addition, the maximum and minimum value of oil price returns is higher than the values of stock market indices returns. In case of the indices, the returns of ATX (Austria), DAX30 (Germany) and IBEX35 (Spain) are higher than other indices while the lowest return is BSI20 (Belgium). The variation of indices returns are relatively higher in ATX (Austria), DAX30 (Germany), IBEX35 (Spain) and CAC40 (France), while PSI20 (Portugal) and BSI20 (Belgium) have a relatively lower variation in daily returns.

Table 4.2: Descriptive Statistics of Variables for Daily Returns, 2002-2016

N Mean Std. Dev. Min Max

AEX 3768 .000144 .014988 -.137208 .112884 ATX 3768 .000341 .015213 -.140443 .120205 BSI20 3768 .000069 .013037 -.083193 .093340 CAC40 3768 .000166 .015082 -.131628 .105944 DAX30 3768 .000231 .015314 -.117500 .107975 IBEX35 3768 .000235 .0152 -.131859 .134831 ISEQ 3768 .000137 .014679 -.139637 .097331 PSI20 3768 .000022 .012435 -.103797 .113110 Euronext100 3768 .000187 .013834 -.124401 .103225 Crude oil 3768 .000264 .024423 -.151909 .164137

Additionally, the Jarque-Bera statistic that tests for skewness, kurtosis and normality is implemented to test if the variables are normally distributed, and its descriptive statistics are described in Table 4.3 below. The probability values of the JB test are zero for all cases reported, which indicates that the null hypothesis of normality of the variables is rejected. Likewise, the distributions of the returns on BSI20 (Belgium), CAC40 (France), IBEX35 (Spain) and Crude

(14)

14

oil are positively skewed while other indices returns show a value of zero. For all the variables, the hypotheses that daily returns on the European stock market indices and oil prices are normally distributed are clearly rejected by the statistics. However, while the t-test is invalid for small samples that are not-normally distributed, it is valid for a large number of samples with a non-normal distribution. Since the samples consists of 3768 observations, the t-test will be valid.

Table 4.3: Descriptive Statistics of the Variables for Jarque-Bera test, 2002-2016

N Skew Kurt JB AEX 3768 .0000 .0000 .0000 ATX 3768 .0000 .0000 .0000 BSI20 3768 .1776 .0000 .0000 CAC40 3768 .0076 .0000 .0000 DAX30 3768 .0130 .0000 .0000 IBEX35 3768 .0256 .0000 .0000 ISEQ 3768 .0000 .0000 .0000 PSI20 3768 .0000 .0000 .0000 Euronext100 3768 .0000 .0000 .0000 Crude oil 3768 .9866 .0000 .0000

4.3 Results

This section will analyze the results obtained from the multiple regression model and also a summary of the results will be discussed.

4.3.1 Multiple regression model

The output of the multiple regression model (2) is presented in Table 4.4 below. According to the results, the correlation coefficients relating to the returns on indices on the returns on Euronext100 are highly significant for the all eight European countries as the coefficients are significantly different from 0. They vary from 0.09 (a less reactive country) for BSI20 (Belgium) to 1.07 (a more reactive country) for CAC40 (France). Next, the coefficients relating to the crisis dummy variable are significantly negative for ATX (Austria) and IBEX35 (Spain) with -0.00058 and -0.00042 respectively, reflecting the negative impacts of the 2008 global financial crisis on stock returns in Austria and Spain.

(15)

15

More interestingly, the coefficients with regard to the returns on indices to the returns on crude oil prices are significantly different from 0 and show a positive relationship in CAC40 (France) and DAX30 (Germany) at 1% level of statistical significance with 0.01436 and 0.01524 respectively. For the rest of the countries shows no significant relationship between the returns on oil prices and the returns on stock market indices. Moreover, the explanatory power of the multiple regression model, which is shown by the R2, is quite large for the seven countries as it varies from 0.5145 to 0.9732, except for Belgium case, 0.0084.

Table 4.4: The estimation results for the multiple regression model (2), 2002-2016

Index Crude oil Euronext100 Crisis dummy R2 N

AEX .00346 (.00285) 1.03859* (.00503) .00016 (.00014) .9189 3768 ATX .00419 (.00705) .79141* (.01245) -.00058* (.00035) .5178 3768 BSI20 .00058 (.00867) .08633* (.01530) -.00003 (.00043) .0084 3768 CAC40 .01436* (.00165) 1.07541* (.00291) -.00009 (.00008) .9732 3768 DAX30 .01524* (.00441) .99836* (.00779) .00003 (.00022) .8138 3768 IBEX35 .00743 (.00498) .95737* (.00879) -.00042*** (.00025) .7590 3768 ISEQ -.00256 (.00682) .76135* (.01204) .00038 (.00034) .5152 3768 PSI20 -.00120 (.00578) .64487* (.01021) -.00033 (.00028) .5145 3768

Notes: Numbers in parentheses are robust standard errors.

(16)

16

4.3.2 Summary of the results

The hypotheses presented in statement (1) are rejected for the two cases of the eight cases. The relationships between oil prices and stock market indices in France and Germany are significantly positive and thus oil prices positively influence the stock markets over the sample period, reflecting that the stock market indices for France and Germany increase when oil prices go up. For the rest of the countries, the movements of the stock markets do not largely respond to the changes in oil prices.

Interestingly, by looking at the correlation coefficients relating to the oil prices, CAC40 (France) and DAX30 (Germany) that are rejected by the statistics show the top two highest correlation coefficients among the eight indices. The reasons maybe as follows.

Firstly, Germany and France have a relatively much higher GDP through the sample period than the rest of the countries that examined in the paper. This implies that countries with a more developed economy are more affected by the changes in oil prices. This is in line with Elyasiani et al. (2011) that show that a more industrialized countries are more responsive to oil price changes.

Additionally, it is important to consider the magnitude of imports and exports of oil by the eight countries, as presented in the Appendix. According to Energy Information Administration (EIA), both exports and imports of crude oil in Netherlands are the highest, while it is the lowest in Austria. Interestingly, Germany and France that are rejected by the statistics show that its exports and imports of crude oil are smaller than Netherlands. If the magnitude of exports and imports in a country is important factor affecting the relationship between oil prices and stock markets, the stock market in Netherlands, AEX, should be more affected by oil price changes. However, the relationship between oil prices and the stock market in Netherlands is not significant by the analysis.

Consequently, it is not very clear in this research whether the impacts of oil price changes on stock markets in European countries are more influenced by the magnitude of exports and imports of oil.

(17)

17

5 Robustness check

In this chapter, a robustness check is done by choosing another oil price benchmark and also the results of the robustness check are briefly discussed.

5.1 Different oil price benchmark

As a robustness check, Brent oil is taken as a proxy of oil prices instead of WTI crude oil. The indices of the eight European countries remain the same so that the different oil price impacts on the stock markets can be tested upon robustness. Additionally, the same methodology as described in the section 3.1 and the same hypotheses as stated in the section 3.2 have been applied. The estimation results for the multiple regression model with Brent oil is in the Appendix.

According to the results, the correlation coefficients relating to Euronext100 are very similar to the correlation coefficients of the model with WTI crude oil as the coefficients are significantly positive for all cases. This definitely should be the case since the stock market indices remained the same in the robustness checks. In case of the crisis dummy variable, the results show that ATX (Austria) and IBEX35 (Spain) are negatively and significantly influenced by the 2008 global financial crisis with -0.00064 and -0.00042 respectively, which are the same results as the results with WTI crude oil.

On the other hand, the results with Brent oil show a slightly different correlation coefficients relating to the oil price changes. For example, while CAC40 (France) and DAX30 (Germany) are positively and significantly responsive to WTI crude oil price changes, AEX (Netherlands), BSI20 (Belgium), CAC40 (France) and ISEQ (Ireland) are positively and significantly affected by the changes in Brent oil with 0.00991, 0.01906, 0.00781 and 0.02239 respectively.

Nevertheless, the results indicate that there are still positive relationships between oil prices and stock markets in some European countries. Consequently, the results for the robustness checks on applying Brent oil are largely similar to the results with WTI crude oil and therefore seem fairly robust.

(18)

18

6 Conclusion

Despite the movements of oil prices may play a significant role in the international financial market, the results of existing studies on the relationship between oil prices and stock market prices are mixed and remain quite ambiguous. This research therefore attempts to identify the relationships between WTI crude oil prices and stock market indices in eight European countries: Austria, Belgium, France, Germany, Ireland, Netherlands, Portugal and Spain for the recent 15 years from 2002 to 2016.

Empirical results after the use of the multiple regression model find three main results that can be discussed. Firstly, the stock market returns in France and Germany positively and significantly do respond to the returns on WTI crude oil prices, while the rest of the countries do not. Specifically, the relationship between oil prices and stock markets in France and Germany are statistically significant at the 1% level. This implies that stock market indices in France and Germany increase when WTI crude oil prices rise. Secondly, there are a negative and significant impacts of the 2008 global financial crisis on the stock markets in Austria and Spain. However, it does not mean that the rest of the countries had no influence from the crisis because of that the improvements in economic situations in the rest of the countries might be one reason why the impacts of the crisis are not significant on their stock market indices. Lastly, the European market index, Euronext100, is positively and significantly related to stock market indices for all eight countries, but the degree of sensitivities of the stock market indices to Euronext100 varies from country to country. For example, the indices for Netherlands and France are largely sensitive to the whole European stock market changes, while the index for Belgium is not very sensitive compared to other indices.

From the findings of the multiple regression model, it can be concluded that the impacts of the changes in oil prices differ greatly across countries in Europe for the recent 15 years. However, since only the two cases for France and Germany are significant at the 1% level from the multiple regression model, this paper cannot conclude that there are clear relationships between oil prices and stock markets in Europe. Likewise, the correlation coefficients relating to the crude oil seem very small decimals from the multiple regression and thus the relationships between oil prices and stock markets are still weak and ambiguous. Therefore, it is required for further research to gain a more in-depth knowledge about macroeconomic factors affecting the relationship between oil prices and stock market prices.

(19)

19

7 Limitations and further research

This chapter provides some limitations of this research and suggestions for further research. First, this paper deals with a time-series data and a relatively large sample that is 3780 observations. Therefore, it is also another possibility to use a panel data analysis. Although collecting panel data may be more costly than collecting time-series data, implementing panel data analysis can obtain more accurate outcomes. For example, pooling the stock market indices data rather than using the data individually can be helpful to generate more accurate predictions because of that it allows to observe an individual stock market’s behavior by observing the behavior of other stock markets.

Second, the time-series of the sample is quite limited. Despite the sample period in this research is the recent 15 years from 2002 to 2016, a longer period would be more appealing for a long-term perspective. For example, Balcilar et al. (2017) examine the relationship between oil prices and stock prices over 150 years and conclude that the oil and stock markets might commove both in the short- and long-run. With a longer-sample period, it would be able to create more reliable results. Due to the data accessibility, it was not very feasible in this paper.

Third, this paper uses the data for Euronext100 as a proxy of the whole European market index. However, the members of Euronext100 are only France, Netherlands, Belgium and Portugal so that the index is not directly related to Austria, Germany, Ireland and Spain. Although Euronext100 can still indirectly influence the stock markets in those four countries that are not members, it would be more desirable to find a more representative market index.

A number of suggestions can be made for further research. Firstly, focusing on a single stock market rather than multiple stock markets will be more efficient and effective. For instance, adding more relevant variables such as a country’s specific events to the multiple regression model or other models may be helpful to deepen the analysis about the relationship between oil prices and stock market. Additionally, it is obvious that the stock market indices in Europe are more affected by quite many other variables rather than only oil prices. Since the increased globalization, it is important to understand how macroeconomic forces affect stock market returns (Elyasiani et al., 2011). Thus, the investigation of macroeconomic factors that may affect stock market prices would be also valuable. Lastly, it is worth for further research to investigate the method how the movements of oil prices can be used for risk diversification, which could yield more practical outcomes for investors and businesses.

(20)

20

8 References

Apergis, N., & Miller, S.M. (2009). Do structural oil-market shocks affect stock prices? Energy Economics, 31, pp. 569-575.

Arouri, M.E.H. (2011). Does crude oil move stock markets in Europe? A sector investigation. Ecnonomic Modelling, 28, pp. 1716-1725.

Balcilar, M., Gupta, R., & Wohar, M.E. (2017). Commom cycles and common trends in the stock and oil markets: Evidence from more than 150 years of data. Energy Economics, 61, pp. 72-86.

Balcilar, M., Gupta, R., & Miller, S.M. (2015). Regime switching model of US crude oil and stock market prices: 1859 to 2013. Energy Economics, 49, pp. 317-327.

Basher, S.A., Haug, A.A., & Sadorsky, P. (2012). Oil prices, exchange rates and emerging stock markets. Energy Economics, 34, pp. 227-240.

Chen, N.F., Roll, R., & Ross, S.A. (1986). Economic forces and the stock market. Journal of Business, 59, 383-403.

Chen, S.S. (2010). Do higher oil prices push the stock market into bear territory? Energy Economics, 32, pp. 490-495.

Cunado, J., & Gracia, F.P. (2014). Oil price shocks and stock market returns: Evidence for some European counries. Energy Economics, 42, pp. 365-377.

Driesprong, Gerben, Jacobsen, Ben, Maat, & Benjamin. (2008). Striking oil: another puzzle? Jounal of Financial Economics, 89(2), pp. 307-227.

Elyasiani, E., Mansur, I., & Odusami, B. (2011). Oil price shocks and industry stock returns. Energy Economics, 33, pp. 966-974.

Hamilton, J. (2011). Nonlinearities and the macroeconomic effects of oil prices. Macroeconomic Dynamics, 15, pp.364-378.

Huang, R.D., Masulis, R.W., & Stoll, H.R. (1996). Energy shocks and financial markets. Journal of Futures Markets, 16, pp. 1-27.

Joo, Y.C., & Park S.Y. (2017). Oil prices and stock markets: Does the effect of uncertainty change over time? Energy Economics, 61, pp. 42-51.

Narayan, P.K., & Gupta, R. (2015). Has oil price predicted stock returns for over a century? Energy Economics, 48, pp. 18-23.

(21)

21

Park, J., & Ratti, R.A. (2008). Oil price shocks and stock markets in the U.S. and 13 European countries. Energy Economics, 30, pp. 2587-2608.

Sadorsky, P. (1999). Oil price shocks and stock market activity. Energy Economics, 21, pp. 449-469.

Zhang, D. (2017). Oil shocks and stock markets revisited: Measuring connectedness from a global perspective. Energy Economics, 62, pp. 323-333

(22)

22

9 Appendix

Figure A.1: The stock market indices in the eight European countries, 2002-2016

0 5000 10000 15000

DAX30: Germany

0 5000 10000

ATX: Asutria

0 10000 20000 30000

IBEX35: Spain

0 1000 2000

AEX: Netherlands

0 10000 20000 30000

PSI20: Portugal

0 5000 10000 15000

CAC40: France

0 10000 20000

ISEQ: Ireland

0 2000 4000 6000

BSI20: Belgium

(23)

23

Figure A.2: The exports and imports of crude oil by countries to/from U.S., 2002-2016

Notes: the X-axis represents the name of countries and the Y-axis represents the amount of crude oil that imported and exported from/to U.S. The amount of the crude oil is thousand barrels and it is the average amount during the sample period of 2002-2016. Source: Energy Information Administration (EIA)

0 10000 20000 30000 40000 50000 60000

Exports and imports of crude oil

(24)

24

Figure A.3: Estimation results for the multiple regression (2) with Brent oil, 2002-2016

Index Brent oil Euronext100 Crisis dummy R2 N

AEX .00991* (.00362) 1.03835* (.00503) .00016 (.00014) .9188 3768 ATX .01200 (.00895) .79128* (.01245) -.00064*** (.00035) .5179 3768 BSI20 .01906*** (.01099) .08565* (.01529) -.00002 (.00043) .0091 3768 CAC40 .00781* (.00209) 1.07564* (.00291) -.00008 (.00008) .9732 3768 DAX30 .00753 (.00560) .99883* (.00779) .00022 (.00022) .8140 3768 IBEX35 .00397 (.00632) .95760* (.00879) -.00042*** (.00024) .7592 3768 ISEQ .02239* (.00866) .76087* (.01204) .00038 (.00034) .5154 3768 PSI20 -.00485 (.00734) .64487* (.01021) -.00030 (.00028) .5145 3768

Notes: Numbers in parentheses are robust standard errors.

Referenties

GERELATEERDE DOCUMENTEN

In dit onderzoek zal het volgende worden onderzocht: Wat waren voor de Nederlandse oorlogsvrijwilligers van het 1 ste bataljon Jagers redenen en motivaties om zich aan te

An alarming finding from our study is that a large proportion of COVID- 19 trials test the same treatments or drugs, creating a thicket of redundant, uncoordinated, and

Structural change in the economy and a change in public opinion during the COVID-19 crisis jointly imply that government choices regarding investments, regulation and taxes can

Meningkatnya volume perdagangan telah memicu terjadinya konflik internal perebutan kendali dagang dalam kesultanan yang memunculkan dualisme kekuasaan: Pangeran Ratu di keraton

The  last  two  chapters  have  highlighted  the  relationship  between  social  interactions   and  aspiration  formation  of  British  Bangladeshi  young  people.

CONCLUSIONS: Among PsA patients receiving their first biologic, disease severity and outcomes differed within 5EU, with patients in the UK with relatively higher burden and poorer

The figure shows that this simple constitutive assumption is not valid away form the center of the channel (as discussed in Section IV), since the shear stress and strain rate

Unfortunately,  these  results  are  not  new:  limited  use  is  a  common  problem  in  PHR  evaluations  [27].  Several  recent  systematic  reviews  focusing