The moderating effect of internationalization on the influence of changes in energy prices on Dutch non-finance companies’ stock price returns.

(1)

1

The moderating effect of internationalization on the

influence of changes in energy prices on Dutch non-finance

companies’ stock price returns.

This paper studies what the moderating effect of internationalization on the influence of changes in energy prices on Dutch non-finance companies’ stock price returns is. The sample of Dutch non-finance companies comprises 183 companies which constitute an unbalanced sample over the timeframe starting at January 1st_{, 1973 until April 1}st_{, 2016. It concludes that} energy industry and oil prices have a negative effect on stock price returns, while gas industry prices have a positive, yet small influence. This indicates that an increase in the energy industry price or the oil price leads to decreasing stock price returns. Since energy and oil are used in basically every company, for instance for lighting, production processes, and indirectly with transportation, it is sensible that an increase in these prices lowers (expected) profits and thus stock price returns. The same reasoning holds true for gas industry prices, however, companies tend to hedge better against gas industry prices than the other prices. Moreover, gas contracts are generally long-term contracts, so the effect of changes in gas prices may not show up immediately. The moderating effect of internationalization is mostly positive. If a company internationalizes its sales, its market becomes bigger, so there is more potential for making profits, which increases stock prices since stock prices reflect all (expected) profits.

Student number: s2309823

Name: Marloes Wiekens

Study programme: MSc International Financial Management

Supervisor: Dr. W. Westerman

(2)

2 1 Introduction

Klesch Aluminium Delfzijl is a Dutch melter of aluminium, which was previously called Aldel. The company was founded in 1966 near Slochteren, since there was explored a gas field, which provided cheap energy for Aldel. Aldel grew and became an important factor for the regional economy. In the early 90s, however, the market declined. In 1996, a new electricity contract was concluded in which electricity prices were higher than before, but it was guaranteed that Aldel was secured to stay into business at least until 2005. Then the troubles began. In 2009 Aldel was taken over by the private equity firm Klesch and Company Ltd. Although the region was heavily dependent on the employment of the firm, it had to close some departments in order to secure its existence. When, in 2010, the market improved, also the production of Aldel improved, and the future seemed bright. However, there was still one big problem for Aldel: the electricity prices in the Netherlands were way higher than the electricity prices in Germany. Since the countries are so close together, competition between the Dutch and German aluminium melters is prevalent. Therefore, Aldel made huge losses, but it could not go bankrupt due to the enormous demand for its products. To be able to stay in business, Aldel wanted to have an electricity line to Germany to be able to profit from German electricity prices, and managed to establish this. As of today, it profits as much as the German companies from the cheaper electricity and it still is in business because of this. Aldel is an example of a company which uses a massive load of electricity to produce its goods. Because of this, the electricity price is one of the most important factors when it comes to production costs. Therefore, a change in the price of electricity can have a huge impact on such a company. In the case of Aldel, it produced in the Netherlands and used electricity of the Dutch network with a relatively high price. However, it could not charge an exceptionally high price in its Dutch consumer market because there were some German firms who also sold similar products in the Dutch market. This led to the case that Aldel was making huge losses, which it could not cover up. When it succeeded to buy its electricity from the German market, it lost its disadvantage and could compete in the market again.

(3)

3

resulting performance. When a Dutch company, for example, produces and sells exclusively in the Netherlands, it has a relatively small market. In this market prices are set and known, and therefore it is hard to charge higher sales prices when energy prices increase. When a company, however, also sells in other countries, it has a larger potential market, which is less aware of the Dutch sales price. Therefore, it could be easier to charge a higher sales price when energy prices increase. In this case, (expected) profits are less influenced and harmed by energy price changes.

Most literature on the effect of energy prices are macroeconomic studies like the one by Chang et al. (2009), which state that low-growth economies are less willing to change to renewable energy as opposed to higher growth countries. Other studies are researching the rebound effect, like the one by Atkeson and Kehoe (1999) which finds that, in time-series data, energy use is inelastic with respect to price, while it is elastic in cross-section data. A different set of studies relates the increasing energy prices to technological investments and innovations, like the one by Linn (2008) which finds that entrants’ flexibility towards energy has not so much impact on energy consumption; rather input substitution and changes in industry composition explain the changes in energy consumption. Moreover, there are quite some studies concerning the environmental aspects, like the one by Hassouneh et al. (2012) which indicates that the increasing production of biofuel based mostly on feedstocks is linked to decreasing space for food production. However, for companies it is important to know what the effect of changes in energy prices is on their performance. Therefore, I study the effect of changes in energy prices on the stock price returns of non-finance companies in the Netherlands, moderated by the degree of internationalization.

(4)

4 2 Literature review

There are several factors influencing the energy prices and its components. To get an idea of what these factors are, Section 2.1 lists the most important ones.

2.1 Energy price

The influence that energy prices have, starts with the influences on the energy prices itself. There are several factors influencing the energy price. The most important distinction that has to be made is the distinction between gas and electricity.

2.1.1 Factors influencing gas prices

Oil price: almost sixty percent of all Dutch power plants, work on gas. In the Netherlands, the

price of gas is linked to the price of oil. Therefore, the price of oil is an important determinant for the energy prices. However, prices are set twice every year, so it could be that despite a fall in the oil price, the energy price is not decreasing at the same time (E-ON, 2012).

2.1.2 Factors influencing electricity prices

Coal price: most electricity produced in the Netherlands comes from coals. Therefore, a change

in the price of coals has a great impact on the price of electricity. When there are, for instance, heavy weather conditions, like the flooding in Australia in 2012, which restrict the supply of coals, and increases the coal price, the electricity price in the Netherlands rises (E-ON, 2012). 2.1.3 Factors influencing gas and electricity prices

Demand and supply: like basic economic theory suggests, energy prices also are driven by

supply and demand. When there is a decrease in supply, like in 2012 when Germany announced it stopped with nuclear energy, this has to be substituted with more expensive fossil fuels, leading to higher energy prices (E-ON, 2012).

Exchange rate: some energy companies buy their energy in another currency than euros. When,

for example, a company buys its energy in dollars, and the dollar depreciates against the euro, the consumers paying in euros will benefit (E-ON, 2012).

Weather: this is partly correlated with demand and supply. When there is a tough winter, there

is more demand for energy because people turn on their heating systems, so the price becomes higher. However, with a mild winter the opposite occurs. In a warm summer, people use their air-conditioning and more fridges, so the demand for energy increases, which increases the price. Also the reverse is true for summer (E-ON, 2012).

Economic developments: when there is an economic crisis, demand for energy becomes less,

(5)

5

growth, people are willing to spend more on energy, demand rises, and prices do so, too (E-ON, 2012).

Political unrest: when there is unrest in, for example, the Middle-East, local oil refineries

produce less than normal or even nothing at all, which cuts supply, and therefore, drives up the prices (E-ON, 2012).

2.2 Macro-economic effects

Now we know what causes changes in the price of energy, it is important to know what that does with stock price returns. However, most studies on the economic effect of changes in energy prices are macro-studies. As stated by Nandha and Faff (2008), a fluctuation in oil prices can affect the economy as a whole because of a transition of wealth. For example when the oil price rises, wealth is transferred from oil consumers to oil producers. Moreover, oil price fluctuations affect production costs, inflation rates and consumer confidence. Another study is done by Egger et al. (2013) who find that higher energy prices are observed in countries with higher income and higher energy inefficiency. Furthermore, it underlines the basic demand and supply theory that an increase in energy trade intensity increases the price of energy. However, interestingly, this holds to a varying degree in various countries. Moreover, if governments start to develop more alike policies regarding energy, local energy price will increase, but again, to a varying degree. Egger et al. (2013) also find some spill over effects, which means that a domestic energy policy can impact on the international market. They found that the transferal of shocks in the prices of energy depends on the similarity of policies regarding energy and international trade.

(6)

6

Although it seems a common believe in the literature that oil price increases have a negative influence on the macro economy, there are also opposing views. Chen et al. (1986), Huang et al. (1996), Wei (2003), Henriques and Sadorsky (2008), Aspergis and Miller (2009), and Al Janabi et al. (2010) do not find that oil price shocks have an influence on aggregate stock price returns. However, these studies all focus on the period before the 2008 financial crisis.

2.3 Meso-economic effects

Regarding studies concerning the effect on aggregate, country-level stock price returns, Jones and Kaul (1996) show that aggregated stock price returns in the USA, Canada, Japan and the UK react negatively to increasing oil prices. Papapetrou (2001) finds similar results for Greece. Moreover, Al-Mudhaf and Goodwin (1993) find that companies listed on the New York Stock Exchange holding a significant portion of assets in domestic oil production report positive stock price returns after an oil price increase. Sadorsky (2001) and Boyer and Filion (2007) find similar results for Canada, and El-Sharif et al. (2005) find the same for the UK.

2.4 Micro-economic effects

In general, it seems like oil price increases are a bad sign for macroeconomic growth, however, there are several studies arguing that oil price changes should also have an effect on stock markets. For example, Huang et al. (1996) argue that if the oil price has such an effect on the economy as a whole, it should also have an impact on (individual) stock prices. Moreover, Mussa (2000) suggests that oil prices influence economic activity, earnings, inflation rates and monetary policies, therefore changes in oil prices affect asset prices and the financial market. According to Jones et al. (2004), who sum up the literature until then, stock prices represent the present and expected profitability of a company. So if oil price changes have an effect on the (future) profitability of a firm, this should be directly translated in changes in the company’s stock price. Even when there are expected changes in the future oil price, this should be reflected in the current stock price. According to Nandha and Faff (2008), oil prices affect stock price returns, because oil is a basic driver of economic activity and there is also a general believe that oil prices drive stock price returns.

(7)

7

transport, paper and packaging and banking report negative sensitivity to oil price changes. Moreover, Al-Mudhaf and Goodwin (1993) find that companies holding a significant portion of assets in domestic oil production report positive stock price returns after an oil price increase. Hammoudeh and Li (2004) find similar results for US stock price returns for the oil and gas industries. Oberndorfer (2009) studies two specific industries, namely European energy corporations and European utility corporations. He finds that systematic risk is not the only driver of stock price returns in these industries. Besides other macro-economic variables, energy market developments play an important role in explaining stock price returns. Oil price changes have a positive effect on oil and gas stocks, while oil price volatility has a negative influence on both stocks. While oil and gas stock price returns increase when energy prices increase, Oberndorfer (2009) finds that stock price returns of utilities decrease with energy price increases. Although oil price changes do have an impact on energy and utility stocks, gas prices do not. This is not in line with a Canadian study where they report stock price return sensitivity to oil prices as well as gas prices, although the effect for gas prices also was smaller here (Boyer and Filion, 2007). Oberndorfer (2009) argues that their result is specifically surprising for the stocks of energy corporations since these rely more heavily on gas than on oil. One of the reasons behind this could be that gas contracts usually are long-term contracts, which link the price to the oil price to prevent the incentive for fuel switching. When the prices are linked, customers will have little incentive to switch form one resource to another, causing harm to the providers of the resource customers switched from. Another explanation could be that energy companies tend to hedge more against gas price risks than oil price risks (Haushalter, 2000). Besides oil and gas, there is yet another input for electricity generation: coal. Increases in coal prices have a negative effect of utility companies and energy companies since coal is an input cost. The effect of changes in coal prices is smaller than the effect of oil prices on the stock price returns of these two industries, although it is the most used energy resource in Europe. Therefore, it seems that stock market participants primarily focus on oil prices as the indicator for energy price developments.

(8)

‘pass-8

through’ effect. Additionally, financial markets are not only influenced directly, but also indirectly, for example by monetary policy, employment and consumer confidence.

Hypothesis 1a: Increases in energy prices lead to decreases in overall stock price returns. Hypothesis 1b: Increases in oil prices lead to decreases in overall stock price returns. Hypothesis 1c: Increases in gas prices lead to decreases in overall stock price returns.

2.5 Internationalization

We now know what may cause changes in energy prices, and what those changes do to the economy. The next step is to see what the moderating effect of internationalization is on the effect of changes in energy prices on stock price returns. A recent study of Sato et al. (2015) suggests that changing energy prices have a significant, but yet very small impact on imports. An increase in the energy price difference between the same sectors in two countries of 10% causes the import to change by only 0.2%. They also find, however, that this result is somewhat larger for more energy-intensive industries. Thus, they find a significant, yet very small, effect of changes in energy prices on imports and exports. However, it could also be the case that companies might still be able to import or export, but now with a lower profit margin.

Hypothesis 2: Increases in internationalization lead to increases in stock price returns.

2.6 Financial crises

(9)

9

January 1st_{, 1990 until December 31}st_{, 2007; the crisis period is from January 1}st_{2008 until June} 30th_{2009; and the post-crisis period is from July 1}st_{2009 until December 31, 2012. Tsai (2015)} finds that in the pre-crisis period, the increase of oil prices has a negative effect on stock prices. During and after the crisis period, US stock prices respond positively to an increase in the oil price. This means that in the recovery of the financial crisis, the stimulating effect of higher oil prices dominates the downside effect of higher marginal production costs. This is the most recent crisis

Hypothesis 3a: Crises lead to decreases in stock price returns.

Hypothesis 3b: The crisis following Black Monday leads to a decrease in stock price returns. Hypothesis 3c: The dotcom crisis leads to a decrease in stock price returns.

Hypothesis 3d: The subprime mortgage crisis leads to a decrease in stock price returns.

2.7 Asymmmetry effects

(10)

10 3 research objective and research questions

The research objective for this paper is to get clear what the influence is of changes in energy prices on Dutch non-finance companies, moderated by the degree of internationalization of those companies. To study the influence that energy prices have on companies, for example Nandha and Faff (2008) and Oberndorfer (2009) study the effect of oil prices on share prices. To be able to compare the results of this paper with earlier research papers, I follow their method. Moreover, the Netherlands is not only reliant on oil, but also on gas prices. Therefore, I also study the effect of using gas industry prices instead, and the effect of using energy industry prices, which combines the different types of energy. Furthermore, the effects of different types of industries and several crises are taken into consideration as well.

Therefore, the main research question I want to answer is:

1) What is the moderating effect of internationalization on the influence of changes in energy prices on stock price returns for non-finance firms in the Netherlands?

The sub-questions following this main question are:

1) What is the effect of changes in energy prices on stock price returns for non-finance firms in the Netherlands?

2) What is the moderating influence of internationalization on the effects of changes in energy prices on stock price returns for non-finance firms in the Netherlands?

3) What is the influence of different industries on the effects of changes in energy prices on stock price returns for non-finance firms in the Netherlands?

4) What is the influence of crises on the effects of changes in energy prices on stock price returns for non-finance firms in the Netherlands?

In these questions, I refer to changes in energy prices, however, the changes in oil and gas prices are studied separately as well.

4 Data

(11)

11

fluctuations. There is a plan to establish a European energy network, but this is not yet the case, so today companies are still reliant on regional or national energy networks (Dorsman et al., 2011). Since I study the effect that energy prices have on stock price returns, I use only listed companies in the sample. The companies were selected on the basis of their industry as referred to in NACE rev.2 main sections. The industries included are:

 A agriculture, forestry and fishing;  B mining and quarrying;

 C manufacturing;

 D electricity, gas, steam and air conditioning supply;

 E water supply; sewerage, waste management and remediation activities;  F construction;

 G wholesale and retail trade; repair of motor vehicles and motorcycles;  H transportation and storage;

 I accommodation and food service activities;  J information and communication;

 M professional, scientific and technical activities;  N administrative and support service activities;

 O public administration and defence; compulsory social security;  P education;

 Q human health and social work activities;  R arts, entertainment and recreation;  S other service activities;

 T activities of households as employers; undifferentiated goods- and services-producing activities of households for own use;

 U activities of extraterritorial organisations and bodies, the industries not included are:

 K financial and insurance activities;  L real estate activities.

(12)

12

over total sales as a measure of internationalization, and total assets. The energy price information is from Eurostat, which provides high quality European statistics. The energy price information entails the energy industry price, gas industry price and the oil price. Unfortunately, no data on coal prices is available. The GDP data is retrieved from the World Bank, which provides developmental data about countries around the world. Moreover, the Eurodollar exchange rate is retrieved from Yahoo! Finance, which provides financial information. Finally, the data for the crises is retrieved from The Economist. Since the oil crisis is too closely linked to the different energy prices, and in particular to the oil price, this crisis is not possible to study in this setting, because it creates a near singular matrix in the regressions, therefore, it is not possible to study the effect of this crisis.

Table 1: Explanation variables used

Meaning Source

SP_RETURN stock price return in percentages Datastream

EIP energy industry price Eurostat

GIP gas industry price Eurostat

OP oil price Eurostat

FSTS foreign sales to total sales (degree of internationalization) Datastream

GDP gross domestic product Worldbank

TA total assets Datastream

EURODOLLAR Eurodollar exchange rate Yahoo! Finance

INDUSTRY_B industry B Orbis

INDUSTRY_C industry C Orbis

etc. etc. etc.

BLACK_MONDAY crisis after Black Monday The Economist

DOTCOM Dotcom crisis The Economist

SUBPRIME subprime mortgage crisis The Economist

(13)

13

variables has a normal distribution. Again, this should not be a problem since there are so many observations. Most correlations are quite low, so this should not be a problem. The two high correlations (correlations ≥ |0.8|) are between the oil and the gas industry price, which are linked to each other, and the gas industry price and GDP. The correlations table can be found in Appendix B, Table B1.

Table 2: Descriptive statistics

SP_RETURN EIP GIP OP FSTS GDP TA EURODOLLAR

Observations 50478 61488 63684 92232 29607 92232 38154 90585 Mean 0.659 0.148 11.674 58.735 50.989 3.79E+10 2263968 1.186 Median 0.000 0.134 8.117 52.073 54.790 3.57E+10 324144 1.211 Maximum 17.144 0.204 21.090 110.421 98.490 5.32E+10 18155293 1.448 Minimum -11.410 0.105 5.541 25.535 0.000 2.36E+10 8111 0.849 Std. Dev. 7.315 0.031 5.750 28.404 32.615 1.07E+10 4585219 0.169 Skewness 0.437 0.444 0.518 0.530 -0.175 0.163 2.581 -0.434 Kurtosis 2.929 1.809 1.553 1.910 1.716 1.463 8.639 2.361 Jarque-Bera 1617.290 5655.082 8399.255 880.759 2185.542 9489.815 92909.000 4379.445 Probability 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000

Sample period: 01-01-1973 until 01-04-2016 Included amount of companies: 183

The industry variables are dummy variables, in which a 0 represents an industry in which the particular company is not operating, and a 1 representing the industry in which the particular company does operate. Similarly, the crises variables are dummy variables. Here, a 0 represents the period in which the crisis is not prevalent, and a 1 represents the period in which the particular crisis is prevalent. The crisis following Black Monday starts at October first, 1987, and ends at December first, 1988. The dotcom crisis starts at January first, 2001, and ends at December first, 2001. And finally, the subprime mortgage crisis starts at January first, 2008, and ends at April first, 2016.

5 Methodology

(14)

14

To prevent inflated results because of homoscedastic residuals, I used heteroscedasticity-consistent White standard errors (see Appendix C for non-winsorized data, and Appendix D for winsorized data). Thereafter, I use fixed cross-section effects and use a redundant fixed effects test to see if fixed effects are appropriate (see Appendix C for non-winsorized data, and Appendix D for winsorized data). Then I use random cross-section effects and use a Hausman test to see if random fixed effects are appropriate instead (see Appendix C for non-winsorized data, and Appendix D for winsorized data).

To begin, the most basic regressions are used:

𝑆𝑃_𝑅𝐸𝑇𝑈𝑅𝑁𝑖,𝑡 = 𝛼 + 𝛽1∗ 𝐸𝐼𝑃𝑖,𝑡 + 𝜀𝑖,𝑡, (1)

𝑆𝑃_𝑅𝐸𝑇𝑈𝑅𝑁𝑖,𝑡 = 𝛼 + 𝛽1∗ 𝐺𝐼𝑃𝑖,𝑡 + 𝜀𝑖,𝑡, (2)

𝑆𝑃_𝑅𝐸𝑇𝑈𝑅𝑁𝑖,𝑡 = 𝛼 + 𝛽1∗ 𝑂𝑃𝑖,𝑡+ 𝜀𝑖,𝑡, (3)

𝑆𝑃_𝑅𝐸𝑇𝑈𝑅𝑁𝑖,𝑡 = 𝛼 + 𝛽1∗ 𝐸𝐼𝑃𝑖,𝑡+ 𝛽2∗ 𝐺𝐼𝑃𝑖,𝑡+ 𝛽3∗ 𝑂𝑃𝑖,𝑡 + 𝜀𝑖,𝑡, (4)

where 𝑆𝑃_𝑅𝐸𝑇𝑈𝑅𝑁_𝑖,𝑡 is the stock price return of company i at time t, 𝐸𝐼𝑃_𝑖,𝑡 is the energy industry price for company i at time t, 𝐺𝐼𝑃_𝑖,𝑡 is the gas industry price for company i at time t, 𝑂𝑃_𝑖,𝑡 is the oil price for company i at time t, and 𝜀_𝑖,𝑡 is the error term for company i at time t. These regressions indicate how the stock price return reacts on the changes in energy industry prices, gas industry prices, oil prices, or all the prices combined, respectively. Hereafter, the internationalization factor is entered into the regressions: 𝑆𝑃_𝑅𝐸𝑇𝑈𝑅𝑁_𝑖,𝑡 = 𝛼 + 𝛽₁∗ 𝐸𝐼𝑃_𝑖,𝑡+ 𝛽₂∗ 𝐹𝑆𝑇𝑆_𝑖,𝑡+ 𝜀_𝑖,𝑡, (5)

𝑆𝑃_𝑅𝐸𝑇𝑈𝑅𝑁_𝑖,𝑡 = 𝛼 + 𝛽₁∗ 𝐺𝐼𝑃_𝑖,𝑡+ 𝛽₂∗ 𝐹𝑆𝑇𝑆_𝑖,𝑡+ 𝜀_𝑖,𝑡, (6)

𝑆𝑃_𝑅𝐸𝑇𝑈𝑅𝑁_𝑖,𝑡 = 𝛼 + 𝛽₁∗ 𝑂𝑃_𝑖,𝑡+ 𝛽₂ ∗ 𝐹𝑆𝑇𝑆_𝑖,𝑡+ 𝜀_𝑖,𝑡, (7)

(15)

15

Since the stock price returns are affected by other factors as well, the control variables are added to the regressions: 𝑆𝑃_𝑅𝐸𝑇𝑈𝑅𝑁_𝑖,𝑡 = 𝛼 + 𝛽₁∗ 𝐸𝐼𝑃_𝑖,𝑡 + 𝛽₂∗ 𝐹𝑆𝑇𝑆_𝑖,𝑡+ 𝛽₃∗ 𝐺𝐷𝑃_𝑖,𝑡+ 𝛽₄∗ 𝑇𝐴_𝑖,𝑡+ 𝛽₅ ∗ 𝐸𝑈𝑅𝑂𝐷𝑂𝐿𝐿𝐴𝑅_𝑖,𝑡+ 𝜀_𝑖,𝑡, (9) 𝑆𝑃_𝑅𝐸𝑇𝑈𝑅𝑁𝑖,𝑡 = 𝛼 + 𝛽₁∗ 𝐺𝐼𝑃_𝑖,𝑡+ 𝛽₂∗ 𝐹𝑆𝑇𝑆_𝑖,𝑡+ 𝛽₃∗ 𝐺𝐷𝑃_𝑖,𝑡+ 𝛽₄∗ 𝑇𝐴_𝑖,𝑡+ 𝛽₅ ∗ 𝐸𝑈𝑅𝑂𝐷𝑂𝐿𝐿𝐴𝑅𝑖,𝑡+ 𝜀𝑖,𝑡, (10) 𝑆𝑃_𝑅𝐸𝑇𝑈𝑅𝑁𝑖,𝑡 = 𝛼 + 𝛽1∗ 𝑂𝑃𝑖,𝑡+ 𝛽2∗ 𝐹𝑆𝑇𝑆𝑖,𝑡+ 𝛽3∗ 𝐺𝐷𝑃𝑖,𝑡+ 𝛽4∗ 𝑇𝐴𝑖,𝑡+ 𝛽5 ∗ 𝐸𝑈𝑅𝑂𝐷𝑂𝐿𝐿𝐴𝑅𝑖,𝑡+ 𝜀𝑖,𝑡, (11) 𝑆𝑃_𝑅𝐸𝑇𝑈𝑅𝑁𝑖,𝑡 = 𝛼 + 𝛽1∗ 𝐸𝐼𝑃𝑖,𝑡+ 𝛽2 ∗ 𝐺𝐼𝑃𝑖,𝑡 + 𝛽3∗ 𝑂𝑃𝑖,𝑡 + 𝛽4∗ 𝐹𝑆𝑇𝑆𝑖,𝑡+ 𝛽5∗ 𝐺𝐷𝑃𝑖,𝑡+ 𝛽6 ∗ 𝑇𝐴𝑖,𝑡+ 𝛽7 ∗ 𝐸𝑈𝑅𝑂𝐷𝑂𝐿𝐿𝐴𝑅𝑖,𝑡+ 𝜀𝑖,𝑡, (12) where 𝑆𝑃_𝑅𝐸𝑇𝑈𝑅𝑁𝑖,𝑡 is the stock price return of company i at time t, 𝐸𝐼𝑃𝑖,𝑡 is the energy industry price for company i at time t, 𝐺𝐼𝑃_𝑖,𝑡 is the gas industry price for company i at time t, 𝑂𝑃_𝑖,𝑡 is the oil price for company i at time t, 𝐹𝑆𝑇𝑆_𝑖,𝑡 is the degree of internationalization for company i at time t, 𝐺𝐷𝑃_𝑖,𝑡 is the gross domestic product for company i at time t, 𝑇𝐴_𝑖,𝑡 are the total assets of company i at time t, 𝐸𝑈𝑅𝑂𝐷𝑂𝐿𝐿𝐴𝑅_𝑖,𝑡 is the Eurodollar exchange rate for company i at time t, and 𝜀_𝑖,𝑡 is the error term for company i at time t. These regressions indicate how the stock price return reacts on the changes in energy industry prices, gas industry prices, oil prices, or all the prices combined, respectively, moderated by the degree of internationalization and corrected for the control variables’ influences.

The industry in which the companies are operating may also have an effect on the stock price returns, so now the industries are added.

(16)

16

𝑆𝑃_𝑅𝐸𝑇𝑈𝑅𝑁_𝑖,𝑡

= 𝛼 + 𝛽₁∗ 𝐸𝐼𝑃_𝑖,𝑡 + 𝛽₂∗ 𝐹𝑆𝑇𝑆_𝑖,𝑡+ 𝛽₃∗ 𝐺𝐷𝑃_𝑖,𝑡+ 𝛽₄∗ 𝑇𝐴_𝑖,𝑡+ 𝛽₅

∗ 𝐸𝑈𝑅𝑂𝐷𝑂𝐿𝐿𝐴𝑅_𝑖,𝑡+ 𝛽₆∗ 𝐼𝑁𝐷𝑈𝑆𝑇𝑅𝑌_𝑄_𝑖,𝑡+ 𝛽₇∗ 𝐼𝑁𝐷𝑈𝑆𝑇𝑅𝑌_𝑅_𝑖,𝑡 + 𝛽₈ ∗ 𝐼𝑁𝐷𝑈𝑆𝑇𝑅𝑌_𝑆_𝑖,𝑡+∗ 𝜀_𝑖,𝑡, (14) where 𝑆𝑃_𝑅𝐸𝑇𝑈𝑅𝑁_𝑖,𝑡 is the stock price return of company i at time t, 𝐸𝐼𝑃_𝑖,𝑡 is the energy industry price for company i at time t, 𝐹𝑆𝑇𝑆_𝑖,𝑡 is the degree of internationalization for company i at time t, 𝐺𝐷𝑃_𝑖,𝑡 is the gross domestic product for company i at time t, 𝑇𝐴_𝑖,𝑡 are the total assets of company i at time t, 𝐸𝑈𝑅𝑂𝐷𝑂𝐿𝐿𝐴𝑅𝑖,𝑡 is the Eurodollar exchange rate for company i at time t, and all the industry variables are dummies for each company for each year, having a value of zero, when a company is not in that industry at that time, and being one, if it is, and 𝜀𝑖,𝑡 is the error term for company i at time t. These regressions are also run with the gas industry price (𝐺𝐼𝑃𝑖,𝑡) instead of the energy industry price, and with the oil price (𝑂𝑃𝑖,𝑡) instead of the energy industry price, thereafter, these three energy prices are also combined. These regressions indicate how the stock price return reacts on the changes in energy industry prices, gas industry prices, oil prices, or all the prices combined, respectively, moderated by the degree of internationalization and corrected for the control variables’ influences and the industries in which the companies operate.

There are some crises in the timeframe which could also have an effect on the stock price returns. The oil crisis is not taken into account here, because this creates a near singular matrix in the regressions.

𝑆𝑃_𝑅𝐸𝑇𝑈𝑅𝑁_𝑖,𝑡

= 𝛼 + 𝛽₁∗ 𝐸𝐼𝑃_𝑖,𝑡+ 𝛽₂∗ 𝐹𝑆𝑇𝑆_𝑖,𝑡+ 𝛽₃∗ 𝐺𝐷𝑃_𝑖,𝑡 + 𝛽₄∗ 𝑇𝐴_𝑖,𝑡+ 𝛽₅

(17)

17

oil price (𝑂𝑃_𝑖,𝑡) instead of the energy industry price, thereafter, these three energy prices are also combined. These regressions indicate how the stock price return reacts on the changes in energy industry prices, gas industry prices, oil prices, or all the prices combined, respectively, moderated by the degree of internationalization and corrected for the control variables’ influences and the crises in the time period.

(18)

18

degree of internationalization and corrected for the control variables’ influences, the industries in which the companies operate and the crises in the time period.

6 Results

As mentioned in the data section, the variables do not have a normal distribution and there are enormous outliers. Therefore, the data is winsorized. In order to give outputs which are as representative as possible, I provide the outputs for the winsorized data here. However, for completeness, all regressions and tests done with the winsorized data are also done with the data which was not winsorized, and are included in Appendix C.

All residuals show huge non-normality when not using White standard errors, therefore, White standard errors are used. The regressions without White standard errors and the results for the residuals normality tests can be found in Appendix C for non-winsorized data, and Appendix D for winsorized data. Moreover, some Models use fixed or random cross-section effects. The regressions for fixed as well as random cross-section effects, and the test results for the redundant fixed effects tests and Hausman tests are listed in Appendix C for non-winsorized data, and Appendix D for winsorized data.

6.1 Basic regressions

(19)

19

regression, but is highly significant there. This indicates that when the Eurodollar exchange rate increases by €1, the share price return falls by 1.798% (Model 12).

6.2 Industry effects

(20)

20

Table 4: Regression outcomes for the dependent variable SP_RETURN, with White standard errors

Model 1a _{Model 2}a _{Model 3}b _{Model 4}b _{Model 5}a _{Model 6}a _{Model 7}a _{Model 8}a _{Model 9}a _{Model 10}a _{Model 11}a _{Model 12}a

C 1.614*** 1.244*** 1.163*** 2.207*** 1.961*** 1.703*** 1.745*** 2.676*** 4.407*** 3.844*** 1.983*** 5.979*** EIP -5.417*** -7.014*** -4.162** -7.010*** -1.454 -5.909** GIP -0.039*** 0.069*** -0.038*** 0.104*** 0.036 0.204*** OP -0.008*** -0.020*** -0.009*** -0.027*** -0.006*** -0.028*** FSTS -0.005* -0.004 -0.005 -0.004 -0.002 -0.002 -0.002 -0.002 GDP BLN -0.017 -0.052*** -0.232** -5.440*** TA BLN -48.600 -57.200* -78.000*** -4.470 EURODOLLAR -1.751 -0.618 0.583 -1.798*** R^2 0.014 0.013 0.001 0.003 0.012 0.012 0.011 0.016 0.014 0.013 0.012 0.017 Adjusted R^2 0.009 0.009 0.001 0.003 0.005 0.005 0.005 0.008 0.007 0.006 0.005 0.009

(21)

21

Model 13 Model 14 Model 15 Model 16 Model 17 Model 18 Model 19 Model 20

C 2.755*** 3.843*** 2.890*** 4.038*** 1.010 1.960*** 4.961*** 6.115*** EIP -2.634 -2.222 -7.188*** -6.798*** GIP 0.056** 0.059** 0.227*** 0.230*** OP -0.004 -0.003 -0.027*** -0.027*** FSTS 0.002 0.003* 0.002 0.003* 0.002 0.003* 0.002 0.003* GDP BLN -0.012 -0.013 -0.065*** -0.066*** -0.034*** -0.034*** -0.062*** -0.065*** TA BLN 10.200 11.100 8.880 9.220 7.610 9.140 10.100 10.700 EURODOLLAR -1.123*** -1.684*** -0.799* -0.808* 0.453 0.444 -2.016*** -2.018*** INDUSTRY_B 1.673*** 1.529** 1.363** 1.640*** INDUSTRY_C 1.226*** 1.162** 1.031** 1.193** INDUSTRY_E -0.420 -0.432 -0.282 -0.493 INDUSTRY_F 1.084** 1.048** 0.805* 1.053** INDUSTRY_G 0.937* 0.922* 0.795* 0.905* INDUSTRY_H 1.410*** 1.320** 1.212** 1.379*** INDUSTRY_I 1.252 1.110 0.883 1.156 INDUSTRY_J 1.089** 0.983** 0.923** 1.057** INDUSTRY_M 1.041** 1.001** 0.811* 1.019** INDUSTRY_N 1.955*** 1.848*** 1.666*** 1.919*** INDUSTRY_Q -1.886** -1.931** -1.756** -1.855** INDUSTRY_R -0.744 -0.662 -0.618 -0.703 INDUSTRY_S -2.053* -2.156** -1.838* -2.328** R^2 0.003 0.002 0.002 0.002 0.002 0.002 0.005 0.005 Adjusted R^2 0.002 0.002 0.002 0.002 0.002 0.002 0.005 0.004

(22)

22 6.3 Crises effects

Table 6 introduces the several crises. All indicators for the several energy prices have the same sign (negative for energy industry price and oil price, and positive for the gas industry price) and similar significance as before. Still, internationalization is not significant, and now changing between positive and negative values. GDP and the Eurodollar exchange rate show similar significance levels as before, and still show negative values. Total assets show varying results, the significant result being negative. The crisis following Black Monday does not show any significant results, nor are the results consistent. However, the dotcom crisis and subprime mortgage crisis show almost everywhere highly significant results, these results being negative.

6.4 Combined effects

(23)

23

Model 21b _{Model 22}a _{Model 23}a _{Model 24}b

C 3.871*** 5.203*** 2.962*** 6.362*** EIP -2.194 -7.264*** GIP 0.080** 0.255*** OP -0.008*** -0.026*** FSTS 0.003 -0.001 -0.002 0.003 GDP BLN -0.003 -0.042** -0.012 -0.065*** TA BLN 12.200 41.500 -72.3** 12.200 EURODOLLAR -1.964 -2.332*** -0.519 -2.355*** BLACK_MONDAY 0.158 -0.018 -0.110 0.095 DOTCOM -1.131** -2.481*** -2.048*** -1.104** SUBPRIME -0.278* -0.671*** 0.068 -0.507** R^2 0.003 0.016 0.014 0.005 Adjusted R^2 0.002 0.009 0.007 0.005

(24)

24

Model 25 Model 26 Model 27 Model 28 Model 29 Model 30 Model 31 Model 32

C 2.623*** 3.720*** 4.206*** 5.283*** 2.016*** 2.936*** 5.192*** 6.323*** EIP -2.645 -2.226 -7.803*** -7.441*** GIP 0.101*** 0.104*** 0.259*** 0.262*** OP -0.005* -0.005* -0.026*** -0.026*** FSTS 0.003 0.003* 0.003 0.003* 0.002 0.003* 0.003 0.003* GDP BLN 0.002 0.001 -5.190*** -0.536*** 0.021** -0.021** -0.060*** -0.006*** TA BLN 11.100 11.900 10.200 10.500 80.700 9.600 10.800 11.300 EURODOLLAR -2.024*** -2.003*** -2.498*** -2.495*** -0.669 -0.676 -2.430*** -2.427*** INDUSTRY_B 1.654*** 1.501** 1.348** 1.621*** INDUSTRY_C 1.192** 1.102** 1.009** 1.171** INDUSTRY_E -0.398 -0.482 -0.364 -0.497 INDUSTRY_F 1.054** 0.989** 0.775 1.038** INDUSTRY_G 0.905* 0.872* 0.761 0.893* INDUSTRY_H 1.379*** 1.252** 1.180** 1.356*** INDUSTRY_I 1.247 1.158 0.903 1.178 INDUSTRY_J 1.058** 0.917* 0.893* 1.034** INDUSTRY_M 1.004** 0.945* 0.784* 1.000** INDUSTRY_N 1.934*** 1.824*** 1.653*** 1.914*** INDUSTRY_Q 1.825** -1.824** -1.780** -1.775** INDUSTRY_R -0.731 -0.624 -0.567 -0.718 INDUSTRY_S -2.020* -2.222** -1.937* -2.306** BLACK_MONDAY 0.176 0.193 -0.019 0.005 -0.071 -0.047 0.108 0.122 DOTCOM -1.145*** -1.137*** -2.476*** -2.462*** -2.055*** -2.047*** -1.109** -1.096** SUBPRIME -0.253 -0.232 -0.642*** -0.641*** 0.096 0.100 -0.504** -0.500** R^2 0.003 0.003 0.006 0.005 0.004 0.004 0.006 0.005 Adjusted R^2 0.003 0.002 0.005 0.005 0.004 0.003 0.005 0.005

(25)

25 7 Discussion and recommendations

According to Hypothesis 1a, increases in energy prices lead to decreases in stock price returns. Although the literature does not focus on the general energy price, the summarized coefficient signs in Table 8 show indeed that the coefficients are always negative. This means that an increase in the energy industry price causes stock price returns to fall. This is a sensible outcome, since energy is used in basically every company. For example, manufacturing companies use energy in their production processes, and service companies use energy to run their computers and turn on the lights. Most research has been done on oil prices. For example Faff and Brailsford (1999), Boyer and Filion (2007) and Oberndorfer (2009) find that for industries except the energy-related industries itself, stock price returns react negatively to oil price increases. From Table 8, Hypothesis 1b is confirmed; increases in oil prices lead to decreases in overall stock price returns. This result confirms the believe that the oil prices influence economic activity, and as a result they affect asset prices and the financial market (Mussa, 2000). Moreover, oil is used in a lot of industries, not only directly, but also indirectly, for instance in transportation. Contrary to prior literature, Hypothesis 1c cannot be confirmed. If there is any effect of gas industry prices on stock price returns, then there is a very small positive return, instead of a negative return (Boyer and Filion, 2007). However, there are very few significant results, so the effect of gas industry prices seems to be minimal. This is probably the case because gas contract usually are long-term contracts, linked to the oil price, to prevent the motive for fuel switching (Oberndorfer, 2009). Therefore, the effect does not show up immediately. Another reason is that companies tend to hedge better against gas price risk, than for example against oil price risk (Haushalter, 2000). This study focuses on energy industry prices, gas industry prices, and oil prices. Further research could see if lagged variables help when using gas industry prices. Since gas contract are generally long-term, it could well be that the result shows up later on in the stock price returns. Moreover, as mentioned in the literature, coal is another input for electricity prices. Although stock price participants seem to primarily focus on oil prices, it could be interesting to see what the effect is of coal prices as well.

(26)

26

health and social work activities, and other service activities, report significantly lower stock price returns than on average. In this study, these industries are used to see how they affect stock price returns. For further research, it is also interesting to see how the different energy prices, internationalization and the crises affect these industries. For example, Tsai (2015) finds some asymmetry effect for changes in oil prices on different industries’ stock price returns. This could be elaborated by studying the effect of several crises and adding the internationalization aspect.

Table 8: Coefficient signs regression results Tables 4-7

All coefficients Significant coefficients EIP GIP OP FSTS EIP GIP OP FSTS Basic regressions Table 4 + 0 4 0 0 0 2 0 0 - 6 2 6 8 6 2 6 1 Industries added Table 5 + 0 4 0 8 0 0 0 4 - 4 0 4 0 2 0 2 0 Crises added Table 6 + 0 2 0 2 0 0 0 0 - 2 0 2 2 1 0 2 0 Everything combined Table 7 + 0 4 0 8 0 0 0 4 - 4 0 4 0 2 0 4 0

These are the coefficient signs for all the regression results from Tables 4-7 summarized per table.

Hypothesis 2 is confirmed. Internationalization seems to have a positive influence on stock price returns when looking at the coefficient signs in Table 8. The effect is not as strong and significant as, for example, the results for oil prices. This was already expected, because Sato et. al (2015) find similar, but also weak results. This outcome means that, in general, when companies become more international, so they have a larger percentage of their total sales in foreign countries, their stock price returns increase. This is probably due to the fact that stock prices, and therefore stock price returns reflect the present and expected profitability of a firm (Jones et. al, 2004). If a company internationalizes its sales, its market becomes bigger, so there is more potential for making profits, which increases share prices.

(27)

27

would decrease with any crisis. In a crisis, investors become afraid of a descending share price. Therefore, demand falls, supply increases, so the share prices drop. Regarding the results, Hypothesis 3a is thus only partly confirmed by the dotcom and the subprime mortgage crises, but it is not by the crisis following Black Monday. To see if the crisis after Black Monday really has no effect on stock price returns, further research could alter the timeframe in which the crisis is supposed to have an effect. Furthermore, when taking a more international view, there are more crises at which could be looked at, such as the Asian crisis in 1997.

A recommendation for further research on this topic is to broaden the scope. This study focuses on Dutch non-financial companies, and is thus specific for the Dutch market. This decision limits the sample and is focused on only one small country and economy. Therefore, it is hard to generalize results, if possible at all. Moreover, an important source of energy in the Netherlands is coal, for which data is lacking. Having access to this data could alter the results. If further research could extend the scope of the research, the conclusions would be more generalizable. However, special care needs to be drawn to the different energy prices, then. The markets for energy prices are seen as world markets, however, due to, for example, geographical constraints, the prices companies actually pay are not the same across the world. Further research could then also compare the effects of the different energy prices on different areas. For instance, some countries rely more heavily on gas, while other countries are more reliant on oil.

8 Summary and conclusions

(28)

28

The second sub-question asks for the moderating effect of internationalization. The overall results for internationalization are not as clear-cut as the results for, for example, the oil price. However, the significant results provide evidence that internationalization has a positive effect on stock price returns. This means that companies with a larger percentage of their total sales abroad, on average have higher stock price returns. Thus, the moderating influence of internationalization on the effects of changes in energy prices on stock price returns for non-finance firms in the Netherlands is positive.

The third sub-question asks for the effect of the different industries on stock price returns. The industry in which a company operates in is crucial for its stock price return. Only three industries report insignificant coefficients (water supply, sewerage, waste management and remediation activities, accommodation and food service activities, and arts entertainment and recreations). All other industries report significant values, of which most industries have a positive impact on stock price returns (mining and quarrying, manufacturing, construction, wholesale and retail trade, repair of motor vehicles and motorcycles, transportation and storage, information and communication, professional, scientific and technical activities, and administrative and support service activities). Two industries have a negative impact on stock price returns: human health and social work activities, and other service activities. Thus, different industries have a determinative influence on the effects of changes in energy prices on stock price returns for non-finance firms in the Netherlands.

The last sub-question is examining the influence of crises on the effects of changes in energy prices on stock price returns for non-finance firms in the Netherlands. This study tested for three crises: the crisis following Black Monday, the dotcom crisis, and the subprime mortgage crisis. Of these crises, two provided significant negative values, indicating that in a crisis period, stock price returns fall. Only the crisis following Black Monday delivered insignificant results. Overall, crises have a negative influence on the effects of changes in energy prices on stock price returns for non-finance firms in the Netherlands

(29)

29

positive, yet very small influence on stock price returns for Dutch non-finance firms. Furthermore, the moderating effect of internationalization is positive. Moreover, firm industries and financial crises have a significant impact on stock price returns for Dutch non-finance firms. Most industries report a positive influence, but two industries show a negative influence. The crises have an exclusively negative impact on stock price returns for Dutch non-finance firms.

At the moment, most companies are hedging against the risk of changing gas industry prices, however, from the results it becomes clear that the energy industry price as well as the oil price have highly significant negative influence on stock price returns. This indicates that if the energy industry price or the oil price increases, there is a really large chance that stock price returns of companies fall. Therefore, it could be wise for companies to look into the possibilities to better hedge against energy industry price risk and oil price risk. In this way, stock price returns are better protected against changes in these prices. Furthermore,

(30)

30 9 References

Al Janabi, M. A. M., Hatmi-J, A. and Irandoust, M. (2010). An empirical investigation of the informational efficiency of the GCC equity markets: evidence from bootstrap simulation.

International review of financial analysis, 19, 47-54.

Al-Mudhaf, A. and Goodwin, T. H. (1993). Oil shocks and oil stocks: evidence form the 1970s. Applied economics, 25, 181-190.

Aspergis, N. and Miller, S. M. (2009). Do structural oil-market shocks affect stock prices? Energy economics, 31, 575-595.

Atkeson, A., and Kehoe, P. J. (1999). Models of energy use: putty versus putty-clay. The American economic review, vol. 89, no. 4.

Bachmeier, L. (2008). Monetary policy and the transmission of oil shocks. Journal of

macroeconomics, 30, 1738-1755.

Boyer, M. M. and Filion, D. (2007). Common and fundamental factors in stock returns of Canadian oil and gas companies. Energy economics, 29 (3), 428-453.

Chang, T., Hung, C., and Lee, M. (2009). Threshold effect of the economic growth rate on the renewable energy development from a change in energy price: Evidence from OECD countries. Energy policy, 36.

Chen, N. F., Roll, R. and Ross, S. A. (1986). Economic forces and the stock market.

Journal of business, 59, 383-403.

Ciner, C. (2001). Energy shocks and financial markets: nonlinear linkages. Studies in

nonlinear dynamics and econometrics, 5 (3), 203-212.

(31)

31

Dorsman, A., Westerman, W., Karan, M. B., and Arslan, Ö. (2011). Financial aspects in energy: A European perspective. Springer, Heidelberg, Dordrecht, London, New York. ISBN: 978-3-642-19708-6.

Egger, P. H., Kesina, M., and Nigai, S. (2013). Contagious energy prices. The world

economy.

El-Sharif, I., Brown, D., Burton, B, Nixon, B. and 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.

Faff, R. and Brailsford, T. (1999). Oil price risk and the Australian stock market. Journal

of energy finance & development, 4, 69-87.

Gisser, M. and Goodwin, T. H. (1986). Crude oil and the macroeconomy: tests of some popular notions. Journal of money, credit and banking, 18, 95-103.

Guidi, M., Russell, A. and Tarbert, H. (2006). The effect of OPEC policy decisions on oil and stock prices. OPEC review, 30, 1-18.

Hamilton, J. D. (1983). Oil and the macroeconomy since World War II. Journal of

political economy, 91, 228-248.

Hamilton, J. D. (2003). What is an oil shock? Journal of economics, 113, 363-398.

Hamilton, J. D. (2008). Oil and the macroeconomy. In: Durluf, Steven, N., Blume, Lawrence, E. (Eds.). In the new Palgrave dictionary of economics, second edition Palgrave Macmillan, Houndsmills, UK and New York.

Hammoudeh, S. and Choi, K. (2007). Characteristics of permanent and transitory returns in oil-sensitive emerging stock markets: The case of GCC countries. Journal of

(32)

32

Hammoudeh, S. and Li, H. (2004). Risk-return relationships in oil-sensitive stock markets. Finance letters, 2 (3), 10-15.

Hassouneh, I., Serra, T., Goodwin, B. K., and Gill, J. M. (2012). Non-parametric and parametric modeling of biodiesel, sunflower oil, and crude oil price relationships. Energy

economics, 34.

Haushalter, G. D. (2000). Financing policy, basis risk, and corporate hedging: evidence form oil and gas producers. Journal of finance, 22, 107-152.

Henriques, I. and Sadorsky, P. (2008). Oil prices and the stock prices of alternative energy companies. Energy economics, 30, 998-1010.

http://magazine.eon.nl/artikelen/2012/10/factoren-die-energieprijs-beinvloeden/

(2012). Retrieved: 20-01-2016.

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

IEA (2004). Analysis of the impact of high oil prices on the global economy.

International Energy Agency report, May 2004. Available at:

http://library.iea.org/dbtw-wpd/textbase/papers/2004/high_oil_prices.pdf.

Jones, C. and Kaul, G. (1996). Oil and the stock markets. Journal of finance, 51, 463-491.

Jones, D. W., Leiby, P. N. and Paik, I. K. (2004). Oil price shocks and the macroeconomy: what has been learned since 1996. Energy journal, 25 (2), 1-32.

Lardic. S. and Mignon, V. (2008). Oil prices and economic activity: An asymmetric cointegration approach. Energy economics, 30, 847-855.

Linn, J. (2008). Energy prices and the adoption of energy-saving technology. The

(33)

33

Miller, J. I. and Ratti, R, A. (2009). Crude oil and stock markets: Stability, instability, and bubbles. Energy economics, 31, 559-568.

Mollick, A. V. and Assefa, T. A. (2013). U.S. stock returns and oil prices: The tale from daily data and the 2008-2009 financial crisis. Energy economics, 36, 1-18.

Mork, K. A. (1989). Oil and the macroeconomy when prices go up and down: An extension of Hamilton’s results. Journal of political economics, 97 (3), 740-744.

Mussa, M. (2000). The impact of higher oil prices on the global economy. Internaltional

Monetary Fund 2000, December 8, 2000. Available at:

http://www.imf.org/external/pubs/ft/oil/2000/.

Nandha, M. and Faff, R. (2008). Does oil move equity prices? A global view. Energy

economics, 30, 986-997.

Oberndorfer, U. (2009). Energy prices, volatility, and the stock market: Evidence from the Eurozone. Energy policy, 37, 5787-5795.

Papapetrou, E. (2001). Oil price shocks, stock market, economic activity and employment in Greece. Energy economics, 23, 511-532.

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

Sadorsky, P. (2001). Risk factors in stock returns of Canadian oil and gas companies.

Energy economics, 23, 17-28.

(34)

34

The Economist (2014). http://www.economist.com/news/essays/21600451-finance-not-merely-prone-crises-it-shaped-them-five-historical-crises-show-how-aspects-today-s-fina

Retrieved: 25-05-2016.

Tsai, C. L. (2015), How do U.S. stock returns respond differently to oil price shocks pre-crisis, within the financial crisis, and post-crisis? Energy economics, 50, 47-62.

Wei, C. (2003). Energy, the stock market, and the putty-clay investment model. The

American economic review, 93, 311-323.

Zhang, D. (2008). Oil shock and economic growth in Japan: A non-linear approach.

(35)

35

Appendix A: Data without winsorization

Table A1: Descriptive statistics for non-winsorized data

SP EIP GIP OP FSTS GDP TA EURODOLLAR

Observations 50661 61488 63684 92232 29607 92232 38154 90585 Mean 1545.559 0.148 11.669 58.648 51.217 37900000000 3780001 1.185 Median 10.626 0.134 8.117 52.073 54.790 35700000000 324144 1.211 Maximum 8500008.000 0.223 21.770 117.090 421.100 53900000000 397000000 1.577 Minimum 0.001 0.100 4.935 17.532 0.000 22800000000 50 0.669 Standard deviation 80835.400 0.032 5.800 29.376 33.588 10800000000 17023523 0.181 Skewness 83.329 0.530 0.517 0.525 0.380 0.163 13.760 -0.507 Kurtosis 7455.627 2.073 1.574 2.012 7.531 1.474 245.902 2.876 Jarque-Bera 117000000000 5077 8235 7993 26037 9350 95001310 3941 Probability 0 0 0 0 0 0 0 0

Sample period: 01-01-1973 until 01-04-2016 Included amount of companies: 183

Appendix B: Correlations winsorized data

Table B1: Corralation matrix winsorized data

EIP SP_RETURN GIP OP FSTS EURODOLLAR GDP TA

(36)

36

Appendix C: Regression outcomes and tests for non-winsorized data

Table C1: Regression outcomes for the dependent variable SP_RETURN, without White standard errors

Model C1 Model C2 Model C3 Model C4 Model C5 Model C6 Model C7 Model C8 Model C9 Model C10 Model C11 Model C12

C 3.657* 0.829 0.426 4.975** 1.721*** 1.428*** 1.641*** 2.769*** 4.955*** 4.299*** 1.878*** 6.792*** EIP -14.292 -37.408** -5.109** -10.852*** -3.268 -9.743*** GIP 0.038 0.165 -0.045*** 0.192*** 0.043 0.298*** OP 0.012 0.001 -0.012*** -0.042*** -0.010*** -0.042*** FSTS 0.002 0.002 0.001 0.003 0.002 0.002 0.002 0.002 GDP BLN -0.009 -0.062*** -0.031*** -0.059** TA BLN -4.560 -4.840 -3.650 -4.060 EURODOLLAR -2.521*** -0.955 0.807 -2.330*** R^2 0.000 0.000 0.000 142.000 0.000 0.000 0.001 0.004 0.002 0.001 0.001 0.004 Adjusted R^2 0.000 0.000 0.000 0.000 0.000 0.000 0.001 0.003 0.002 0.001 0.001 0.004

* Significant at the 10%-level. ** Significant at the 5%-level. *** Significant at the 1%-level.

Table C2: Residual normality tests

Skewness 138.703 141.121 152.985 134.971 4.275 4.063 3.729 4.375 4.387 4.161 3.747 4.372

Kurtosis 20546.850 21308.040 25186.220 19430.180 101.819 95.055 80.815 104.751 104.743 97.436 80.850 104.712

Jarque-Bera 6.27E₊₁₁ _7.53E₊₁₁ _1.26E₊₁₂ _5.67E₊₁₁ _9412008 _8502006 _6583634 _9654907 _9633400 _8651484 _6578096 _9627092

(37)

37

Table C3: Regression outcomes for the dependent variable SP_RETURN, with White standard errors

C 3.657** 0.829* 0.426 4.975** 1.721*** 1.428*** 1.641*** 2.769*** 4.955*** 4.299*** 1.878*** 6.792*** EIP -14.292* -37.408 -5.109*** -10.852*** -3.268 -9.743*** GIP 0.038 0.165* -0.045*** 0.192*** 0.043 0.298*** OP 0.012 0.001 -0.012*** -0.042*** -0.010*** 0.042*** FSTS 0.002 0.002 0.001 0.003 0.002 0.002 0.002 0.002 GDP BLN -0.009 -0.062*** 0.031*** -0.059** TA BLN -4.560 -4.840 -3.650 -4.060 EURODOLLAR -2.521*** -0.955 0.807 -2.330*** R^2 0.000 0.000 0.000 0.000 0.000 0.000 0.001 0.004 0.002 0.001 0.001 0.004 Adjusted R^2 0.000 0.000 0.000 0.000 0.000 0.000 0.001 0.003 0.002 0.001 0.001 0.004

Table C4: Regression outcomes for the dependent variable SP_RETURN, with White standard errors and fixed cross-section effects

C 5.256** 2.096*** 1.449*** 6.059** 2.200*** 1.903*** 2.018*** 3.407*** 6.379*** 4.759*** 2.600*** 7.344*** EIP -24.535* -33.965 -6.877*** -11.278*** -3.055 -9.468*** GIP -0.056*** 0.065 -0.075*** 0.152*** 0.009 0.261*** OP -0.004 -0.003 -0.016*** -0.042*** -0.014*** -0.043*** FSTS -0.002 0.001 -0.001 0.001 0.003 0.003 0.002 0.003 GDP BLN -0.045*** -0.075*** -0.052*** 0.073*** TA BLN 12.100 13.200 8.880 13.100 EURODOLLAR -2.481*** -0.610 1.148** -1.970 R^2 0.021 0.021 0.024 0.024 0.009 0.009 0.009 0.014 0.012 0.014 0.009 0.015 Adjusted R^2 0.017 0.017 0.020 0.019 0.002 0.002 0.000 0.007 0.004 0.003 0.003 0.007

Table C5: Redundant fixed effects tests for cross-section fixed effects

Chi-square 824.857 855.399 1145.645 875.727 211.936 209.299 204.953 233.047 229.449 217.552 209.391 229.726

d.f. 182.000 182.000 178.000 178.000 167.000 167.000 164.000 164.000 163.000 163.000 163.000 163.000

(38)

38

Table C6: Regression outcomes for the dependent variable SP_RETURN, with White standard errors and random cross-section effects

C 5.283* 2.111*** 1.779*** 6.267** 1.788*** 1.471*** 1.675*** 2.883*** 5.207*** 4.379*** 1.937*** 6.899*** EIP -21.396* -35.132 -5.547** -11.039*** -3.297 -9.768*** GIP -0.022 0.098 -0.005*** 0.184*** 0.039 0.292*** OP 0.002 -0.001 -0.013*** -0.042*** -0.010*** -0.042*** FSTS 0.003 0.002 0.001 0.002 0.000 0.002 0.002 0.002 GDP BLN -0.015 -0.064*** -0.033*** -0.061** TA BLN -4.250 -4.580 -3.650 -3.720 EURODOLLAR -2.514*** -0.911 0.843 -2.277*** R^2 0.000 0.000 0.000 0.000 0.000 0.001 0.001 0.004 0.002 0.001 0.001 0.004 Adjusted R^2 0.000 0.000 0.000 0.000 0.000 0.000 0.001 0.004 0.002 0.001 0.001 0.004

Table C7: Hausman tests for random cross-section effects

Chi-square 0.113 0.000 0.000 0.000 3.217 9.967 10.094 0.000 0.000 0.000 0.000 0.000

d.f. 1.000 1.000 1.000 3.000 2.000 2.000 2.000 4.000 5.000 5.000 5.000 7.000

(39)

39

Model C13 Model C14 Model C15 Model C16 Model C17 Model C18 Model C19 Model C20

C 3.686*** 4.908*** 3.154** 4.317*** 0.956 1.866*** 5.634*** 6.812*** EIP -4.007 -3.389 -10.466*** -9.845*** GIP 0.043 0.046 0.297*** 0.300*** OP -0.010*** -0.009*** -0.042*** -0.042*** FSTS 0.001 0.002 0.001 0.002 0.000 0.002 0.001 0.002 GDP BLN -0.006 -0.007 -0.062*** -0.063*** -0.031*** -0.031*** -0.056** -0.059** TA BLN -4.180 -4.770 -4.630 -5.060 -3.670 -3.840 -3.600 -4.290 EURODOLLAR -2.558*** -2.516*** -0.969 -0.969 0.808 0.797 -2.363*** -2.351*** INDUSTRY_B 1.631 1.416 1.213 1.546 INDUSTRY_C 1.468 1.298 1.169 1.350 INDUSTRY_E -3.104 -3.093 -2.774 -2.972 INDUSTRY_F 1.160 1.046 0.683 1.052 INDUSTRY_G 0.985 0.888 0.660 0.852 INDUSTRY_H 1.433 1.257 1.141 1.337 INDUSTRY_I 1.426 1.150 0.743 1.248 INDUSTRY_J 1.211 0.992 0.906 1.111 INDUSTRY_M 1.223 1.043 0.730 1.106 INDUSTRY_N 2.179** 1.924** 1.654* 2.070** INDUSTRY_Q -2.551* -2.545* -2.248 -2.390 INDUSTRY_R -0.647 -0.449 -0.442 -0.544 INDUSTRY_S -2.010 -2.031 -1.608 -2.305 R^2 0.002 0.002 0.001 0.001 0.002 0.001 0.005 0.004 Adjusted R^2 0.002 0.002 0.001 0.001 0.001 0.001 0.004 0.004

Skewness 4.390 4.386 4.164 4.160 3.750 3.746 4.376 4.371

Kurtosis 104.918 104.752 97.577 97.444 81.018 80.853 104.890 104.722

Jarque-Bera 9666539 9635072 8677340 8652962 6606343 6578623 9660689 9629024

(40)

40

C 3.686*** 4.908*** 3.154*** 4.317*** 0.956 1.866*** 5.634*** 6.812*** EIP -4.007 -3.389 -10.466*** -9.845*** GIP 0.043 0.046 0.297*** 0.300*** OP -0.100*** -0.009*** -0.042*** -0.042*** FSTS 0.001 0.002 0.001 0.002 0.000 0.002 0.001 0.002 GDP BLN -0.006 -0.007 -0.062*** -0.063*** -0.031*** -0.031*** -0.056** -0.059** TA BLN -4.180 -4.770 -4.630 5.060 -3.670 -3.840 -3.600 -4.290 EURODOLLAR -2.558*** -2.516*** -0.969 -0.969 0.808 0.797 -2.363*** -2.351*** INDUSTRY_B 1.631** 1.416* 1.213* 1.546** INDUSTRY_C 1.468** 1.298** 1.169** 1.350** INDUSTRY_E -3.104 -3.093 -2.774 -2.972 INDUSTRY_F 1.160* 1.046 0.683 1.052 INDUSTRY_G 0.985 0.888 0.660 0.852 INDUSTRY_H 1.433** 1.257* 1.141* 1.337* INDUSTRY_I 1.426 1.150 0.743 1.248 INDUSTRY_J 1.211* 0.992 0.906 1.111* INDUSTRY_M 1.223* 1.043 0.730 1.106* INDUSTRY_N 2.179*** 1.924*** 1.654** 2.071*** INDUSTRY_Q -2.551** -2.545** -2.248* -2.390** INDUSTRY_R -0.647 -0.449 -0.442 -0.544 INDUSTRY_S -2.100* -2.031* -1.608 -2.305** R^2 0.002 0.002 0.001 0.001 0.002 0.001 0.004 Adjusted R^2 0.002 0.002 0.001 0.001 0.001 0.001 0.004

(41)

41

Model C21 Model C22 Model C23 Model C24

C 4.576*** 5.893*** 3.310*** 7.205*** EIP -2.337 -8.875** GIP 0.097** 0.328*** OP -0.013*** -0.041*** FSTS 0.002 0.002 0.001 0.002 GDP BLN 0.006 -0.045* -0.017 -0.063** TA BLN 4.000 -4.280 -3.520 -3.780 EURODOLLAR -2.724*** -3.086*** -0.652 -2.808*** BLACK_MONDAY -0.556 -0.743 -0.929* -0.627 DOTCOM -1.664*** -3.579*** -3.051*** -1.628*** SUBPRIME -0.459* -0.881*** 0.172 -0.502 R^2 0.002 0.004 0.003 0.005 Adjusted R^2 0.002 0.004 0.003 0.004

Skewness 4.396 4.201 3.760 4.383

Kurtosis 104.816 97.911 80.995 104.808

Jarque-Bera 9647417 8739472 6632754 9645437

(42)

42

C 4.576*** 5.893*** 3.310*** 7.205*** EIP -2.337 -8.875** GIP 0.097** 0.328*** OP -0.013*** -0.041*** FSTS 0.002 0.002 0.001 0.002 GDP BLN 0.006 -0.045* -0.017 -0.063** TA BLN 4.000 4.280 -3.520 -3.780 EURODOLLAR -2.724*** -3.086*** -0.652 -2.808*** BLACK_MONDAY -0.556 -0.743 -0.929* -0.627 DOTCOM -1.664** -3.579*** -3.051*** -1.628** SUBPRIME -0.459* -0.881*** 0.172 -0.502 R^2 0.002 0.004 0.003 0.005 Adjusted R^2 0.002 0.004 0.003 0.004

Table C14: Regression outcomes for the dependent variable SP_RETURN, with White standard errors and fixed cross-section effects

C 5.454*** 6.341*** 3.850*** 7.727*** EIP -2.463 -8.702** GIP 0.077* 0.301*** OP -0.016*** -0.041*** FSTS 0.003 0.004 0.001 0.003 GDP BLN -0.023 -0.062** -0.036*** -0.080*** TA BLN 16.300 17.900 10.700 15.800 EURODOLLAR -2.407*** -2.685*** -0.233 -2.408*** BLACK_MONDAY -0.650 -0.785 -1.012** -0.710 DOTCOM -1.624** -3.565*** -3.046*** -1.645** SUBPRIME -0.787*** -1.089*** -0.003 -0.692** R^2 0.013 0.014 0.011 0.015 Adjusted R^2 0.005 0.006 0.005 0.008

(43)

43

Table C15: Redundant fixed effects tests for cross-section fixed effects

Chi-square 234.986 221.116 209.854 232.468

d.f. 163.000 163.000 163.000 163.000

Probability 0.000 0.002 0.008 0.000

Table C16: Regression outcomes for the dependent variable SP_RETURN, with White standard errors and random cross-section effects

C 4.759*** 5.983*** 3.370*** 7.320*** EIP -2.430 -8.893** GIP 0.094** 0.322*** OP -0.013*** -0.041*** FSTS 0.000 0.002 0.001 0.002 GDP BLN 0.001 -0.048* -0.019 -0.066** TA BLN 3.600 -3.920 -3.510 -3.360 EURODOLLAR -2.681*** -3.037*** -0.607 -2.750*** BLACK_MONDAY -0.575 -0.752 -0.943* -0.644 DOTCOM -1.656*** -3.576*** -3.050*** -1.631** SUBPRIME -0.500* -0.897*** 0.166 -0.519 R^2 0.002 0.004 0.003 0.005 Adjusted R^2 0.002 0.004 0.003 0.005

Table C17: Hausman tests for random cross-section effects

Chi-square 19.868 0.000 0.000 0.000

d.f. 8.000 8.000 8.000 10.000