The Impact of OPEC Announcements on Stock Returns

The Impact of OPEC Announcements

on Stock Returns

BACHELOR

THESIS WITHIN: Economics NUMBER OF CREDITS: 15 ECTS

PROGRAMME OF STUDY: International Economics

AUTHORS: Marcus Reilimo, Oliver Samer Haydar JÖNKÖPING May 2020

An event study of how information about oil production-cut decisions by

OPEC affect returns of different stocks listed on the London Stock Exchange

Bachelor Degree Project in Economics

Title: The Impact of OPEC Announcements on Stock Returns Authors: Marcus Reilimo, Oliver Samer Haydar

Tutor: Anna Nordén Date: 2020-05-18

Key terms: OPEC, Event study, CAPM, CAAR, EMH, Behavioural finance

Abstract

The purpose of this study is to investigate the effects of OPEC oil production cut announcements on stock returns of specified companies listed on the London Stock Exchange. Two categories are constructed from stocks of companies operating in oil, gas & mining sectors and companies operating in pharmaceutical, industrial engineering and industrial transportation sectors, respectively. The study is based on the theories of EMH and findings of behavioural finance and applies a CAPM model in the context of an event study methodology. Our findings show that in four out of five cases OPEC production cut announcements have significant effects on stocks in the chosen categories around the release of a supply cut announcement. The difference between post-announcement CAARs of the constructed categories is significant on one occasion. Organisations and investors can use these findings to better understand the impact of OPEC news announcements on the stock performance of companies in specified sectors.

Table of Contents

1.

Introduction ... 1

2.

Theory ... 5

2.1 Efficient Market Hypothesis ... 5

2.2 CAPM ... 5

2.2.1 CAAR... 6

2.3 Behavioural finance ... 6

3.

Literature review ... 7

3.1 News and the stock market ... 7

3.2 OPEC news announcements’ effects on stock markets ... 8

3.3 Oil prices and the stock market ... 9

3.3.1 CAARs ... 11

4.

Hypotheses ... 12

5.

Methodology ... 13

5.1 Event study methodology ... 13

5.2 Data gathering ... 13

5.2.1 CAPM: Model specification ... 14

5.3 OLS Regression model specification ... 15

6.

Empirical results ... 16

6.1 Descriptive statistics ... 16

6.2 Oil, gas & mining category ... 18

6.3 Oil input category ... 19

6.4 Visual analysis ... 19 6.5 Asymmetry analysis ... 22

7.

Discussion ... 22

8.

Conclusion ... 27

9.

Reference list ... 29

10.

Appendix ... 33

1. Introduction

_____________________________________________________________________________________ This part of our thesis introduces the reader to the background of the thesis, followed by the problem formulation and purpose. Furthermore, we state our two research questions, state key terms and explain the overall structure.

___________________________________________________________________

1.1 Background

Information communication technology has allowed for news and information to spread around the world with immediacy. In a matter of a few hours, news reports surrounding public events can have substantial effects on companies and investors. This especially pertains to news and information that concerns primary commodities, such as oil, which can substantially affect firms in all regards. This comes as oil is a vital input factor in the production processes of firms. As such, the price changes of oil are carefully monitored by companies, governments and especially investors, as fluctuations in the price of oil will naturally affect the world economy and stock markets. Specifically, oil supply volatility can affect the output decisions and profits of firms, which in turn can affect the ability of firms to create value for shareholders (Cunado & Perez de Gracia, 2003). Given this potential volatility of oil prices, it is in the best interest of shareholders to know the effects of oil price and supply changes on firms and stock returns, so as to maximize their returns.

Even though oil is a primary input in many production processes of firms, few countries supply the commodity. One consortium of countries that have a joint policy regarding the production and supply of oil is The Organization of the Petroleum Exporting Countries (OPEC). According to Kaufmann, Dees, Karadeloglou and Sanchez (2004) this power over oil production gives OPEC a considerable influence on crude oil supply and price. OPEC was founded in 1960 and has 13 member countries (Organization of Petroleum Exporting Countries [OPEC], n.d.). These countries produce approximately 40% of all crude oil in the world (Loutia, Mellos & Andriosopoulos, 2016).

Previous literature found that OPEC’s decisions to alter the level of oil supply affects oil prices in the UK (Guidi, Russell & Tarbert, 2006). The UK being an oil importing

country, the price of oil has significant effects on the economy (Scholtens & Yurtsever, 2012) and (Hamilton, 2011). The primary stock exchange in the UK is the London Stock Exchange (LSE), and as such we look at the effect of OPEC announcements on companies listed on the LSE. We opt for the LSE because London has been a major centre of finance for centuries and because few previous studies focused on the LSE, instead opting for major American or Asian stock markets. In the LSE there are the Financial Times Stock Exchanges (FTSE) 100 and FTSE 250 indices, which include the largest 100 and 101st-350th largest companies listed on the LSE, respectively. These indices include both British and international companies.

This thesis investigates how an announcement by OPEC to cut oil production affects stock returns in companies in the UK. More specifically, we investigate 30 specified companies in the FTSE 100 and FTSE 250. In order to investigate possible asymmetric (significantly different) effects on the stock returns of companies from different sectors which are all oil sensitive, we divide a group of 30 companies equally into two categories. The first category includes 15 companies that are either oil, gas or mining companies and the second category includes 15 companies in sectors that use oil as an input, including companies in the industrial transportation sector, industrial engineering sector and pharmaceutical sector. We choose these two categories as they allow us to capture any possible differences in the impact of OPEC announcements on these two categories.

1.2 Problem

One of OPEC’s stated missions is to provide a steady supply of oil to consumers at a stable price to ensure stable economic growth (OPEC, n.d.). Another important stated agenda of OPEC is to look after the interests of oil producing member countries by providing them a stable income (OPEC, n.d.). OPEC may follow these stated missions as it has the power to affect the price of oil by adjusting its supply of oil. OPEC decides about the level of oil production in its bi-annual meetings and in additional annual meetings held irregularly (OPEC, n.d.). As regards the second stated mission of OPEC, a potential tool for protecting the interests of member countries is altering the level of oil

Given the influence that OPEC has - controlling 40% of global crude oil - it is in the best interest of investors to be aware of the actions OPEC takes in regards to oil supply. This comes as investors want to know how this will affect the stock performance in order to maximize their profits. This is asserted by a number of studies find an increase in the price of oil to have a negative impact on the economy and stock market performance (Cunado & Perez de Gracia, 2003); (Scholtens & Yurtsever, 2012); (Xu, 2015); (Lee & Chiou, 2011) and (Cunado & Perez de Gracia, 2014).

1.3 Purpose

Giventhe few studies in the field of financial economics discussing the topic of OPEC news impact, this thesis investigates whether oil supply cut announcements by OPEC affect stock returns of stocks listed on the LSE. Past studies have found evidence that a supply cut has a more significant effect on the price of oil than a supply increase, these include Scholtens and Yurtsever (2012); Xu (2015); Broadstock, Wang and Zhang (2014) and Henriques and Sadorsky (2008) among others. The findings motivate us to focus on supply cut announcements by OPEC as opposed to oil demand shocks. Our purpose is to update investors with new information regarding OPEC news announcements’ effects on the stock market. We do this by considering the post 2008 financial crisis time frame up to the most recent OPEC oil cuts. In this time frame we found six oil supply cuts in the period of 2016 to 2019 upon which we base our study.

We pose the following two research questions:

1. Do OPEC announcements to cut the level of oil supply have an impact on stock performances of companies operating in sectors that previously have been found to be sensitive to oil price changes?

2. Are the effects of OPEC supply-cut decisions different among sectors that previously have been found to be affected differently by oil price changes?

Given the results of these research questions, we attempt to determine whether OPEC oil cut announcements in the time frame of 2016 to 2019 have a significant effect on stock market returns of selected sectors listed on the LSE. Combining these two questions into one holistic answer, we want to determine whether and how the two categories differ in regard to their reaction to OPEC announcements.

In this thesis, we investigate whether the expected returns (ERs) of chosen stocks listed on the LSE are different from the actual returns within specified time windows around the release of new information regarding OPEC’s decisions to cut the level of oil supply. We do this by applying the Capital Asset Pricing Model (CAPM). Further, by applying the event study methodology, we investigate whether there is an accumulation of Average Abnormal Returns (AAR) obtained from CAPM outputs. We contextualise our findings in two contrasting areas regarding abnormal returns (AR), namely the Efficient Market Hypothesis (EMH) and the field of behavioural finance. The EMH expects investors to behave rationally, which would lead to an efficient capital market in which CAARs (Cumulative Average Abnormal Returns) are impossible to obtain (Malkiel, 2003). However, this idea stands in contrast to empirical findings in the field of behavioural finance, where research has found investors to behave irrationally - see for example Barberis and Thaler (2002) and Liu, Liu and Han (2019).

Our results can potentially help investors understand whether the effects of oil price changes are similar for companies in categories we investigate. This information is of great interest to investors who benefit from knowing the change in stock market prices before they decide to invest in the stocks of the relevant categories. As such, there is an interest in understanding whether these oil price cuts affect the returns of stocks of specified companies in specified categories listed on the LSE.

The remainder of this thesis is structured as follows: section 2 presents our theoretical framework, while section 3 places our thesis in the literature. This is followed by our methodology in section 4. In section 5, we present our empirical findings followed by a discussion. Finally, we conclude our thesis with final remarks and limitations.

2. Theory

_____________________________________________________________________________________ This part of our thesis aims to showcase the theories central to this thesis. Furthermore, it introduces the theoretical framework of the main model used in this thesis.

______________________________________________________________________

2.1

Efficient Market Hypothesis

EMH suggests that share prices efficiently and immediately absorb all information regarding economic conditions (Fama, 1970); as such, asset prices are always a fair representation of their market value. Essentially, EMH states that stock prices behave as random walks and thus cannot be predicted with the help of past prices (Ozdemir, 2008), i.e. looking at earnings, asset values and various ratios commonly used by investors (Malkiel, 2003). Investors should not be able to achieve ARs in the stock market; the only way to achieve higher returns is by buying riskier assets, not by stock picking (Malkiel, 2003).

Following the principles of EMH, any information about future effects on companies should be immediately reflected in the share price of that company and not when the announced impact occurs (Malkiel, 2003). Furthermore, news on one day should only affect stock prices on that particular day and the stock prices on the following day should only be affected by news released on that day and so on (Malkiel, 2003). Thus, from that viewpoint it should not be possible to earn excess returns over a period of multiple days or weeks.

2.2

CAPM

According to Armitage (1995) CAPM is a theory-based model that can be used to calculate the ER of a stock. The CAPM can help an investor choose an optimal portfolio in specified circumstances. Although there are several specifications of the CAPM (such as Fama-MacBeth Model, CCAPM and ICAPM), we make use of the traditional CAPM model developed by Sharpe (1964) and Lintner (1965), as theirs is arguably the most widely used CAPM model. According to Fama and French (2004) the Sharpe-Lintner CAPM model builds on the model of portfolio choice by Markowitz (1959).

2.2.1 CAAR

We use CAPM to derive the ARs of the stocks of the companies we include in our sample. According to Armitage (1995), an AR is defined as the actual return minus the ER (specification of mathematical definition in methodology section). A cumulative abnormal return (CAR) is the aggregate of all consecutive ARs. An AAR is the average of all ARs. These definitions result in “CAAR”: the aggregate of consecutive AARs within a specific event window.

2.3

Behavioural finance

Behavioural finance is a branch of behavioural economics that seeks to understand and explain the impact of human psychology on investors’ decisions and thus on stock market returns. EMH theory assumes all investors to be rational, however behavioural finance challenges the EMH and the hypothesis of rationality (Liu et al., 2019). According to Barberis and Thaler (2002) behavioural finance classifies irrational traders as “noise traders” and rational traders as “arbitrageurs”.

Momentum is one of the concepts used in behavioural finance to explain abnormal returns and can be an example of irrational behaviour of investors. An example of momentum is a bullish market where investors see the market perform well and as a consequence participate in the market, thus further accelerating the already bullish market, see for example Shiller (2003). In this scenario, stock prices can be driven up through “short-term” momentum. Momentum can also cause downward pressure on stocks. For example in a situation where poorly performing stocks are sold to cut losses, thus causing the stock price to fall further (Grinblatt & Moskowitz, 1999) and as such causing the price to no longer represent the true value of the stock. This is in contrast to the EMH which stipulates stock prices always to be a true representation of the stock’s true value. In the presence of momentum, abnormal or excess returns can be made in the short run, which violates the fundamentals of the EMH and implies that investors are irrational.

The herd effect is another concept within behavioural finance according to which investors do not trust their own information when making decisions and instead follow the “herd”, that is, other investors (Liu, Liu & Han, 2019). As investors somewhat blindly trust and follow the behaviour of other investors, the behaviour of an individual investor may be considered rational, since they are following their inherent instinct. However, the collective behaviour of the larger group of investors may be irrational (Liu, Liu & Han, 2019).

3. Literature review

_____________________________________________________________________________________ This section of our thesis reviews relevant literature in the context of the theories discussed above. Furthermore, we showcase the difference between our study and previous studies in the investigated field. Also, we formulate initial expectations of our results in the context of the previous studies.

______________________________________________________________________

3.1

News and the stock market

The effects of news are connected to the theory of EMH as EMH supports the idea that news regarding business conditions have an impact on stock prices and that future expectations are represented in current stock prices (Malkiel, 2003). Many research papers investigate the impact of different types of news on stock prices, such as firm specific profit warnings or macroeconomic factors. Commonly, the EMH can be tested through an event study by investigating the abnormal or excess return of a stock on and around the day of a news release; see for example Armitage (1995) and Boehmer, Masumeci and Poulsen (1991). In this research paper, we aim to investigate the effect of new information on stock prices in the framework of the EMH.

Individual investors often try to earn returns that are higher than market returns by picking stocks and acting on new information in the belief that they get an advantage in the stock market (Black, 1986). These investors act on “noises” in the stock market (news); Black (1986) calls them “noise traders”. De Long, Shleifer, Summers and Waldmann (1990) find that around the time of a news release or an event, noise traders temporarily increase

the risk and volatility of an asset with their speculative trading habits, thus also increasing the potential returns that noise traders can earn. Bernard and Thomas (1989) make the finding that positive earnings announcements cause a post-earnings announcement drift which results in excess returns, meaning that an investor or a trader is able to earn returns that outperform the market. Similarly, Chan (2003) finds that over and under reactions to public news are a source of ARs. In addition, Li et al. (2014) find that firm specific news has an informative effect on investors. On the other hand, they find that emotions and opinions spreading in social media cause irrational decisions among investors, resulting in market fluctuations and increased volatility. The effects of news investigated by previous research will be contextualised in the evaluation of this study.

Guidi et al. (2006) investigate the effect of OPEC decisions on oil prices and stock markets in the U.K and U.S. By applying the event study method, Guidi et al. (2006) find that stock prices are affected by new information regarding OPECs production quota changes as those changes are expected to affect current and future business conditions. As opposed to our thesis, their paper focuses on stock market indices whereas our paper focuses on specific industries. As such, we expect a clearer impact of news on stock returns. Nevertheless, our thesis builds on the findings of Guidi et al. (2006) who suggest investigating industry specific impacts. As such, we extend the current knowledge by investigating company category specific reactions to OPEC news announcements regarding oil supply cuts.

3.2

OPEC news announcements’ effects on stock markets

Studies investigating the effects of oil cuts or supply-side shocks on the stock market find on one hand that supply-shocks do not affect stock returns, while on the other hand, some studies find that supply-shocks affect the stock market. However, these studies look at the effect of oil cuts or shocks, while our study aims to investigate the effects of information about oil supply on stock markets. We discuss these studies because we assume investors to be rational profit-maximisers and to thus have this information. We also contextualise this in our discussion.

Numerous studies, such as Cunado and Perez de Gracia (2003), Scholtens and Yurtsever (2012) and Xu (2015) among others find that oil price increases have a negative effect on a number of industries. Lee and Chiou (2011) and Cunado and Perez de Gracia (2014), among others, find a negative relationship between oil prices and stock market returns. This connects well to our thesis, as when new information causes investors to believe that oil prices will increase, they will act accordingly in an attempt to maximize their profits in the stock market. Scholtens and Yurtsever (2012) find that oil, gas and mining industries are positively affected by an increase in oil price and Xu (2015) finds that the performance of oil and gas related stocks are positively affected by an increase in oil price. Historically, companies that had oil as an input responded negatively to an increase in the price of oil and oil, gas and mining companies responded positively; see Scholtens and Yurtsever (2012), Xu (2015), Broadstock et al. (2014) and Henriques and Sadorsky (2008). Based on these findings, we build our two categories: “oil, gas & mining” and “oil input”.

According to Kilian and Park (2009), cumulative stock returns are not significantly affected by oil disruptions. Similarly, Bastianin, Conti and Manera (2016) find that the stock market volatility is unaffected by oil price shocks. In addition, Filis, Degiannakis and Floros (2011) find no relation between stock markets and supply-side shocks. Filis et al. (2011) state that supply-side shocks do not affect the correlation between lagged oil prices and current stock prices. On the other hand, Kang, Ratti and Yoon (2014) state that the disruptions in the supply of oil are associated with positive effects on the covariance and volatility. In addition, Hamilton (2003) finds on the industry-level that there is a strong correlation between oil price shocks and output. These sets of findings seem to contrast each other, and thus present an ambiguity. This will be addressed in this research paper by means of testing the hypothesis of asymmetric returns in our two selected categories.

3.3

Oil prices and the stock market

Researchers who use the event study method to investigate OPEC news announcements find uniform effects on the stock market of oil supply cuts. All research papers listed below find evidence of negative effects on stock markets in case of oil cut announcements and positive effects on stock markets in case of OPEC oil supply increase news

announcements, respectively. In addition, all research papers discussed below find significant effects of OPEC news announcements on stock performance. However, this thesis will only discuss the results of the oil cut announcement, as that is where our focus lies.

Demirer and Kutan (2010) find that there is an asymmetric effect of OPEC and Strategic Petroleum Reserves news announcements on oil prices in that only OPEC’s announcements have a significant impact on oil spot and future prices, thus lending probable support to the influence of OPEC. In addition, Guidi et al. (2006) find that there are asymmetric effects of OPEC news announcements conferences for U.K and U.S. stock markets during conflict periods; noting that the effect for the U.K. is slightly stronger than that for the U.S. Our thesis, on the other hand, does not focus on capturing the differences between conflict periods and normal periods. Schmidbauer and Rösch (2012) find that there are significant effects after OPEC announcements on the West Texas Intermediate stock market; specifically, Schmidbauer and Rösch (2012) find that anticipation effects to cut oil production levels are stronger than anticipation effects to maintain or increase oil production levels, thus lending further support for investigating oil cut announcements.

Some researchers find significant ARs in stocks markets as a consequence of OPEC news announcements. Loutia et al. (2015) find evidence of ARs onoil prices for both positive and negative news announcements. Loutia et al. (2016) find negative returns for oil cut announcements and positive returns for oil increase announcements, but state that negative news announcements have a stronger impact on oil prices. Lin and Tamvakis (2009) find that quota cuts positively affect crude oil price returns, except when markets are weak. In our case, the UK market is bound in a strong economy, and as such should have an impact according to Lin and Tamavakis (2009). Taking a different example, Lauenstein and Simic (2017) find that during oil supply increase OPEC news announcements there are positive CARs in spot and future freight markets and that during oil supply decrease OPEC news announcements the CARs are negative.

prior to the official announcements. Cox, Dayanandan, Donker and Nofsinger (2017) also find information leakage, noting that “negative abnormal returns start to accumulate before the announcement, indicating information leakage.” These findings are in line with our results (please see discussion), as we have also found significant CAARs prior to the event date in certain circumstances.

In sum, the largest part of the body of literature agrees that OPEC production cuts increase both stock and oil prices and OPEC production increases have the opposite effect while some studies find production cuts to have a more significant impact on both oil and stock prices. However, these findings do not account for industry-specific effects, opting to investigate stock market or country effects. Therefore, our thesis contributes to the literature by addressing two different categories of companies - one consisting of oil, gas and mining companies and the other one consisting of companies that use oil as an input from the pharmaceutical, industrial transportation and industrial engineering sectors. Our findings are more specific than the findings discussed above and can help investors understand sector-specific impacts.

3.3.1 CAARs

A plethora of research papers use CARs to test their event study results. Examples of event studies investigating OPEC news announcements effects on oil prices using CARs include Lin and Tamavakis (2009) and Loutia et al. (2016) among others. Chan (2003) uses CARs to calculate stock price reactions to public news by companies and compares these reactions to companies that did not issue news. Cox et al. (2017) use CARs to investigate the effect of profit warnings on stock returns. We take a similar approach to the above mentioned studies, by aggregating ARs and using these to check for deviations from the CAPM model.

In our thesis, we specifically use CAARs as, unlike the above mentioned studies, we calculate the ARs of each stock for each day in our event window (a particular time period over which a total impact of an event is measured). We average those into AARs; the consecutive AARs within the event window are accumulated, thus resulting in CAARs. We thus follow research papers like Lauenstein and Simic (2017) and Cox et al. (2017)

who test the reactions of returns to news announcements using CAARs obtained from results from an event study. Oprean and Tănăsescu (2013) use CAARs to examine the effect of public news announcements on stock markets and Clark, Cornwell, and Pruitt (2008) use CAARs to calculate the effects of sponsorship announcements on shareholder wealth. However, we use CAARs specifically to combine the research of reactions in stock markets to public news and the method of CAARs in the event study context to make new findings regarding OPEC news announcements.

4. Hypotheses

_____________________________________________________________________________________ In this section of our thesis we formulate our two hypotheses we aim to test in this thesis.

______________________________________________________________________ We test the two following hypotheses:

i)

H0: OPEC announcements to cut oil supply do not have a significant effect on returns of

stocks listed on the LSE

H1: OPEC announcements to cut oil supply have a significant effect on returns of stocks

listed on the LSE

ii)

H0: There is no significant asymmetry in the impact of OPEC announcements between

the oil, gas & mining category and oil-input category

5. Methodology

_____________________________________________________________________________________ In this section of the thesis we describe the process of gathering and processing our data, followed by a specification of our two models.

5.1

Event study methodology

We follow Henderson (1990) who provides steps for conducting a general event study. The steps are: 1. defining the data the market receives; 2. taking the difference in the return of the news vs no-news events and measuring this difference; 3. aggregating the ARs found in the prior step and 4. finally conducting a statistical test to investigate whether the ARs are significant and for how long. Following the steps provided by Henderson (1990), our first step is measuring the effect of news announcement by OPEC to cut oil prices; step two is the difference between the expectation of CAPM and the actual returns (specifications below); step three is aggregating the data by means of CAARs and step four is checking whether the data is significant at the 10%, 5% and 1% levels of significance and over what time horizons (specifications below).

5.2

Data gathering

In order to investigate the different companies, we choose a total of 30 companies and divide them into two categories that we name the “oil, gas & mining” category and the “oil input” category. Each of those two categories include 15 companies. The basis for choosing the sectors from which the companies are picked are previous findings about oil price and OPEC announcement effects on companies operating in those industries, such as Scholtens and Yurtsever (2012), Xu (2015), Broadstock et al. (2014) and Henriques and Sadorsky (2008). The oil, gas & mining category consists of oil, gas and mining companies listed on the LSE. The “oil input” category consists of pharmaceutical, industrial transportation and industrial engineering companies listed on the LSE. A list of all the companies used in the investigation of this thesis can be found in appendix 1.

Daily stock price data for each event are gathered online from Yahoo finance. FTSE 100 daily price data is also retrieved from Yahoo finance. U.K. 10-year gilt yield data is retrieved from the website of the United Kingdom Debt Management Office. The data is used to calculate the expected returns, using the CAPM model and the daily ARs. Following Guidi et al. (2006), we use an estimation period of 100 days from 130 to t-30 which we use to calculate the standard error of each individual stock and standard deviation of the AARs, where the letter “t” stands for the day of the announcement. We do this to consider the most recent market conditions. Following Cox et al. (2017), we create our event window from t-30 to t+30 and we calculate the CAARs for windows of various lengths within these event windows to capture differences in the magnitude and significance of CAARs. The period of focus in this study is the post financial crisis era, i.e. the time period after 2008. From 2008 to 2015, there were no oil production cut decisions by OPEC. From 2016 until 2019, there were six oil production cut

announcements, of which exclude one in July 2019 because it coincided with a market correction. As such, this study includes five oil production cut announcements on the dates of 28th September, 2016; 10th December; 2016, 25th May; 2017; 7th December 2018 and on the 12th September, 2019. These oil production cut announcements were retrieved from OPEC's official website. We compare the ARs we find for each of the categories with each other to check if there are significant differences in the ARs for each selected category.

5.2.1 CAPM: Model specification

The ER using CAPM is calculated as follows:

E(Rit) = Rft + [E(Rmt) - Rft]

where E(Rit) is the ER on asset i for time t and Rft, represents a

measure of the risk-free rate of interest, which is the U.K 10-year Gilt bond in this thesis.

E(Rmt) is the expected market return, which is the daily FTSE all index return is this

ARit= Rit - {Rft + [E(Rmt) - Rft]}

where the (ARit) is calculated by deducting E(Rit) from the actual return (Rit) Armitage

(1995).

After calculating the AR for each stock on each day of the event window, we calculate the AARs of all stocks on each day of the event window:

AARit =



where N is the number of ARs included. Finally, we compute the CAARs for different event windows:

CAARit =



A sided t-test is used to check the significance level of each CAAR. The t-test is two-sided as we do not know whether the CAAR will be positive or negative and we want to test whether it is significantly different from 0. The t-statistic is given by:

tCAAR = √(N) × (CAAR ÷

σ

where “N” is the number of days in the event window and “

σ

5.3

OLS Regression model specification

A pooled OLS regression is used to test the significance of independent variables that have been previously found to be significant in explaining the effect of new information on a share’s ARs. Jackson and Madura (2003) and Cox et al. (2017) find that the size of a company explains some of the impact that a profit warning has on the shares’ ARs. Cox et al. (2017) also find the profitability of a company to be a significant explanatory

variable in explaining the (AARs) of a company on the day of the announcement. Similarly, Clark et al. (2008) find market value of equity to be a significant variable in explaining the impact of a sport sponsorship announcement on the sponsoring company’s share's CAR 10 days after the announcement (t+10). In addition to testing the significance of independent variables previously found to be significant, we test whether the effect on a share’s CAR over the time-period 0 to t+30 is different between the two selected categories.

We run the following regression:

CAAR[0, +30]i = MrktCapi + ROAi + Doilinput

where CAAR[0, +30]i is the CAAR of the company “i” 30 days after the announcement

and serves as the dependent variable. MrktCapi is the company’s Market Capitalisation

chosen based upon previous findings that the size of a firm and market value of equity to be significant explanatory variables. ROAi is a 5-year average return on assets chosen on

the basis of (Cox et al., 2017) who found it to be a significant explanatory variable. Finally, Doilinput is a dummy variable that takes on value 1 if the company is in the oil input category. This regression is run for each separate event in the statistical software EViews Version 11. The results for each event are present in appendix 3.

6. Empirical results

_____________________________________________________________________________________ This section of our thesis describes and analyses the results from our two models in the context of our hypotheses. Furthermore, we present key statistics of our variables as well as provide a visual presentation of the results.

______________________________________________________________________

6.1

Descriptive statistics

The period of focus offers in total 6 oil production cut announcements of which we have excluded one production cut announcement on the 1st of July 2019 due to it coinciding

have been within our event window. The market correction would yield biased results which are not in line with general economic conditions and as such would skew our findings.

The results and t-test values for each category are presented in tables 1, 2, 3, 4 and 5 in appendix 2. The descriptive statistics table below provides an insight into the variables used to calculate the ARs, AARs, CARs and CAARs.

The minimum and maximum values of the Rit variable show that in all 5 events the returns are heterogeneous within our categories, as the minimum returns are negative and the maximum returns positive in each case. The minimum and maximum values of the betas show that there is less heterogeneity in the oil input category in terms of stock betas. The betas are of the same sign in the oil input category in all but one of the cases while the betas in the oil, gas & mining category are different in sign in three out of five cases. The standard deviations of returns in each category are relatively low in each case. This indicates that while the minimum and maximum values are of opposite sign and far away from the mean value, the majority of the remaining values are close to the mean. The high standard deviations of the beta variables indicate heterogeneity of the betas in both categories. Table 1 N Min ogm Max ogm Mean ogm SD ogm Min oi Max oi Mean oi SD oi Event 1 Rit 465 -8.18% 9.84% 0.40% 0.024 -3.94% 8.73% -0.01% 0.015 Rft 31 0.77% 1.34% 1.09% 0.17 0.77% 1.34% 1.09% 0.18 b 15 -0.31 3.23 1.5 0.81 0.46 1.7 0.94 0.34 E(Rmt) 31 -1.47% 1.31 0.04% 0.0074 -1.47% 1.32 0.04% 0.0074 Event 2 Rit 465 -5.2% 8.57% 0.14% 0.015 -5.99% 4.53% 0.09% 0.0001 Rft 31 3.35% 3.64% 3.5% 0.0012 3.35% 3.64% 3.5% 0.0013 b 15 -0.19 0.064 -0.07 0.133 -0.24 0.04 -0.08 0.076 E(Rmt) 31 -4.67% 3.71% 0.15% 0.017 -4.67% 3.71% 0.15% 0.017

Event 3 Rit 465 -6.44% 8.03% -0.09% 0.025 -4.03% 8.47% -0.06% 0.014 Rft 31 1.04% 1.24% 1% 0.00074 1.04% 1.24% 101% 0.00075 b 15 -0.15 2.60 1.28 1.51 0.20 2.27 1.23 0.684 E(Rmt) 31 -0.98% 0.86% -0.07% 0.0043 -0.98% 0.86% -0.07% 0.0044 Event 4 Rit 465 -8.97% 14.75% 0.15% 0.025 -5.35% 6.12% 0.13% 0.021 Rft 31 1.27% 1.28% 1.275% 0.00005 1.27% 1.28% 1.275% 0.00006 b 15 0 3.7 1.56 0.903 -15 -30 3.8 1.57 E(Rmt) 31 -1.43% 2.2% 0.08% 0.0096 0.57 1.98 1.02 0.480 Event 5 Rit 465 -5.49% 10.34 -0.03% 0.031 -5.15% 11.11% -0.01% 0.0185 Rft 31 0.50% 0.53% 0.517% 0.0002 0.50% 0.53% 0.517% 0.0003 b 15 0.21 1.95 1.06 0.49 0.45 1.56 1.11 0.301 E(Rmt) 31 -2.96% 1.37% 0.004% 0.0077 -2.96% 1.37% 0.004% 0.0078

Legend: Ogm stands for oil, gas & mining category and Oi stands for oil-input category.

6.2

Oil, gas & mining category

The CAAR results we obtain show that the effects of an oil production cut on oil, gas & mining companies after OPEC announcements are only significant in 1 of 5 events. For the September 2016 event all but one of the CAAR date intervals are significant. However, in December 2016, May 2017, December 2018 and September 2019, there are only a few significant date intervals (appendix 2). In the events of September 2016, December 2016 and December 2018, the effects are significant and positive (appendix 2). We find that the effect on CAARs is always larger in magnitude and more significant in the event window beginning five days prior to the event [-5, 0] than the one starting on the day of the event (appendix 2). Furthermore, events in September 2016 and December 2018 exhibit significant CAAR values in the [-5, 0] event window. A notable finding is that all significant post-announcement CAARs for the oil, gas & mining category are

the effect of an oil shock on energy related stocks is positive whenever it is significant. Looking at the [-20, +20] intervals there are two occasions of significantly negative CAARs in the oil, gas & mining category (appendix 2).

6.3

Oil input category

The oil input sector exhibits both positive and negative CAARs at different levels of significance. In the events of September 2016, December 2016, May 2017, and September 2019, we see three CAAR values in the range [-5, +5], [-5, +10], [-5,+20] and [-5,+30] as being significant for a total of 14 such cases (appendix 2). We only find six significant CAAR values ranging between [-20, +20] (appendix 2). However, in most events the effects on the oil input category are insignificant and rather random. For example, the CAARs in September 2016 are positive and September 2019 are weakly negative, while the CAARs in December 2016 and May 2017 are almost identically positive (appendix 2). In December 2018, we observe both negative and positive results. Interestingly, all the significant CAAR values for December 2016 and May 2017 are positive, while all the significant CAAR values in September 2016 are weakly negative (appendix 2). It is questionable whether the negative effects of September 2016 and 2019 in oil input companies might be specific to the month.

6.4

Visual analysis

The following figures depict the CAAR values of each category in each individual event with time on the horizontal axis and CAAR on the vertical axis. As the main window in the visual analysis we use the event window starting 20 days prior to the event (t-20) and reaching to 20 days after the event (t+20) in order to capture any information leakage and post announcement effects similarly to Demirer and Kutan (2010).

All graphs below indicate different trend movements of the CAARs of oil input companies from the CAARs of oil, gas & mining companies, except in December 2016, where in the post-announcement period the oil, gas & mining category’s CAAR was consistently above the oil input companies by about 3 to 5 percentage points. Furthermore, as can be seen from graphs 1 through 5, the oil, gas and mining category’s CAARs start to drift up or down multiple days prior to the event announcement, especially in the events of May 2017, December 2018 and September 2019. The oil input sector’s

CAARs also indicate some pre-announcement movement in CAAR data in the events of May 2017, December 2018 and September 2019. However, only December 2018 and September 2019 seem to exhibit significant drifts (appendix 2). The CAAR values of the [-5, 0] event window are not significant for the oil input category (appendix 2).

-4.00% -2.00% 0.00% 2.00% 4.00% 6.00% 8.00% 10.00% 12.00% 14.00% -20 -18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 14 16 18 20 C AAR time

Dec-16

oil, gas & mining oil input

-5.00% 0.00% 5.00% 10.00% 15.00% -20 -18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 14 16 18 20 C AA R time

Sep-16

oil, gas & mining oil input

Figure 1

-15.00% -10.00% -5.00% 0.00% 5.00% -20 -18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 14 16 18 20 C AAR time

May-17

oil, gas & mining oil input

-12.00% -10.00% -8.00% -6.00% -4.00% -2.00% 0.00% 2.00% -20 -18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 14 16 18 20 C AAR time

Dec-18

oil, gas & mining oil input

-0.04 -0.02 0 0.02 0.04 0.06 0.08 -20 -18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 14 16 18 20 C AAR time

Sep-19

oil, gas & mining oil input

Figure 3

Figure 4

6.5

Asymmetry analysis

The focus of the results of our second model is on the first event, which is the oil production cut announcement released on the 28th of September 2016. The reason for focusing on this event is that it is the most significant one both in terms of our CAPM model and the regression model used to compare the effects between the two categories. A table of descriptive statistics of event 1 is presented below and a similar table for all the events is in Appendix 3 for closer examination.

Table 2: Descriptive Statistics

Variable Coefficient Coefficient t-statistic Significance

Event 1

Constant 0.115380 2.450844** 0.0213**

Market Capitalisation 9.97E-06 0.043477 0.9657 Return on Assets 0.490039 0.728466 0.4728 Oil input dummy -0.171918 -3.266065*** 0.0031*** F-statistic 3.624982**

R-squared 0.294914

Adjusted R-squared 0.213558

The F-statistic shows that the model is significant at the 95% level. Neither Market Capitalisation nor Return on Assets are significant as they have been in previous studies. However, the oil input dummy variable is highly significant and negative. We find that the difference between the CAARs of the two categories on day t+30 is significant and that the oil input category’s CAAR in the interval [0, +30] (appendix 2) is on average 17.2 percentage points lower than that of the oil, gas & mining categories.

7. Discussion

_____________________________________________________________________________________ This section of our thesis answers the research questions posed at the beginning of the thesis. Furthermore, we discuss and interpret the results of our two categories found in both research questions in the context of the literature. Also, we consider several factors that may affect our results.

In events 1 and 2 the oil, gas & mining category’s returns are positive, as could be expected of rational traders. This comes as previous supply cuts have been found to benefit oil, gas & mining companies (we refer the reader to section 3.3 of the literature review). However, in event 2 the oil input category exhibits significant positive CAARs which is against the rational expectation, since oil supply cuts and increased prices have been found to negatively affect most industries, especially the industry that uses oil as an input; see for example (Cunado & Perez de Gracia, 2003); (Scholtens & Yurtsever, 2012) and (Xu, 2015). Our event study results show that there are significant ARs which indicates that the new information is not always immediately absorbed into the stock prices. A possible explanation for the CAARs can be over and under reactions to the news as found by Chan (2003), both of which can cause a post announcement drift instead of the effect being immediately represented in the stock prices. Similarly, Bernard and Thomas (1989) find a post announcement drift that results in excess returns.

El-Sharif, Brown, Burton, Nixon and Russel (2005) find that the mining sector and the oil & gas sector exhibit similar reactions to oil price changes. This findings is interesting, as it lends support to the notion that these industries can be grouped and analysed together. A large body of previous research suggests that an increase in the price of oil has a negative impact on multiple industries and a positive impact on energy companies (Scholtens & Yurtsever, 2012); (Xu, 2015); (Broadstock et al., 2014) and (Henriques & Sadorsky, 2008). Our results are in line with these findings in that any significant post-announcement CAARs for the oil, gas & mining sector are positive. See for example Guidi et al. (2006) who find that OPEC’s decisions to cut oil supply increase oil prices both in the U.S and UK. Nandha and Faff (2008) find in their analysis of several sectors that higher oil prices only affect oil, gas & mining companies positively. We cannot confirm these findings through our research either, as we have found both positive and negative significant effects of oil supply cuts on the oil input category. El-Sharif et al. (2005) state that changes in oil equity values, as compared to other input costs and operational costs, have the most significant effects on variability in the oil & gas sector. These findings might explain some of the observed differences in our data, as the oil, gas & mining category is more significantly affected, both in terms of of number of statistically significant CAAR intervals and magnitude, by OPEC news announcements

than the oil-input category, as can be observed looking at the CAARs and the asymmetry analysis.

In the oil-input category, we expect oil cut announcements to have a significant effect on stock performance. Tahmoorespour, Rezvani, Safari and Randjbaran (2018) find in their research that pharmaceutical companies in the U.K. were negatively affected by oil production cuts. According to Tahmoorespour et al. (2018), this is because the U.K. has a public health sector, meaning that U.K. pharmaceutical companies cannot immediately increase their share prices in times of supply shocks. This in turn leads to the sector losing in attractiveness to investors, resulting in lower share prices. We cannot confirm through our analysis, however, that oil supply cuts affect the share prices of these companies negatively, as we have found no post-announcement significant CAARs for our oil input category in any event. Furthermore, according to Nandha and Faff (2008) oil impacts a wide range of by-products which are used in almost any sector, including the health sector. This implies that stockholders’ anticipation of any detrimental shocks to the health- and also industrial production- sectors might lower the share price. These findings cannot be confirmed either through our research, as we have not found any evidence of significant post announcement CAARs on stock returns of companies in the oil input category. Regarding the industrial transportation sector, El-Sharif et al. (2005) state that there is a weak relationship between oil price movements and returns on stocks. The authors further state that fuel costs are important to transportation firms, as these might be detrimental to firms in this sector in case of oil fluctuations. Narayan and Sharma (2011) support this idea by stating that firms in the transportation sector are more likely to depend on oil. These findings relate well to the findings of the medical/pharmaceutical sector as both sectors depend on oil for their activities. Nevertheless, we cannot find evidence of such effects, either, as we do not have significant CAARs.

Our CAAR results show that the oil input category exhibits both positive and negative CAARs in the [-5, 0] window or [-20, 0] window (appendix 2), that are both weakly and strongly significant. Thus, our CAAR results of the oil input category do not clearly oppose or support the findings of previous studies. Heterogeneity of companies chosen

strongly negative effect in e.g. a pharmaceutical company might affect the overall results significantly in that it has a strong counter effect on other companies’ CAARs.

Overall, the oil-input category unexpectedly exhibits little significant negative CAARs in any of the five events. This suggests that oil output cuts by OPEC in the recent years haven’t had as significant of an impact on stock price performance as there have been in previous years.

In addition to calculating post announcement CAAR values of [0, +5;+10;+20;+30], we calculated pre-announcement CAAR values, starting from negative values and going to 0, e.g. [-5, 0] (appendix 2). We do this because previous research has suggested the presence of either information leakage or speculative trading before important news announcements. For example, Cox et al. (2017) and Schmidbauer and Rösch (2012), have found information leakage prior to news releases. De Long et al. (1990) suggest that announcement CAARs are due to speculative habits of investors.We find significant pre-announcement CAARs in our research. In the case of both the oil, gas & mining category and the oil input category, we found significant pre-announcement CAARs in the [-20, +20] window and in the oil, gas & mining in the [-5, 0] CAAR interval (appendix 2). This may be due to noise traders habits. De Long et al. (1990) find that the risk and volatility of a stock rises temporarily around an event date. According to De Long et al. (1990), the cause of a rise in risk and volatility are the speculative trading habits of “noise traders” who wish to earn excess returns in the market by acting on news or by anticipating the impact of news on the stock market.

As behavioral finance suggests, investors and traders do not always act rationally (Barberis & Thaler, 2002). As such, it is important to consider the effects of behavioural finance and spreading of information through e.g. social media. We do this qualitatively in the below section. One important aspect, hereby, to consider is that with the digitalisation of the world, news, information, opinions and emotion can be spread in the social-media sphere almost instantly. This can affect investors’ behaviour and cause them to act irrationally. Which may cause market fluctuations and increased volatility as noted by Li et al. (2014). This may offer another explanation for the different observed results for the different events. Spreading of information, noise trading as found by De Long et

al. (1990), and over- and underreaction to news combined with the herd effect concept from behavioural finance may offer an explanation to the variation of the CAARs between the five events, as well as the pre-announcement CAARs. According to Liu et al. (2019) investors tend to follow the behaviour of other investors instead of relying on their own information and judgement. This can cause a “herd effect” which may lead to the larger group of investors acting irrationally. Our suspicion is thus that the “herd effect” is strengthened by the effects of social media on investors Li et al. (2014) and thus causes irrational or unexpected behaviour, leading to significant CAARs in the stock market. This is an aspect that is not considered by earlier research papers, probably because social media was not as prominent. It is, however, a necessary aspect to consider in our time period which coincides with the ascendance of social media, because it may explain some of the difference in our findings to the findings of previous research. However, it may be the case in some instances, similarly to the findings of Li et al. (2014) and Liu et al. (2019), that the information investors gain from communicating with each other may cause the investors to collectively act differently than expected e.g. sell oil stocks in the event of an oil supply cut only because all other investors are doing so as well.

The OLS regression analysis results indicate that the oil input dummy variable is significant and negative in events 1 and 4 but not significant in any other event. In event 1 the oil, gas & mining category exhibits significant excess returns while the oil-input category exhibits significantly negative CAARs in the [-5, +20] and [-5, +30] windows and significantly positive CAARs in the [-20, +20] window (appendix 2). However, in event 4 neither of the categories exhibit any significant CAARs in the [0, +30] window (appendix 2), which means that the difference between the two categories was not significant, regardless of the significant dummy variable. Thus, the results of this study regarding the asymmetrical impacts on the two categories remains limited as the difference in CAAR [0, +30] is significant in only one of the 5 events (appendix 2). Thus, there are no general findings we can provide.

We reject our first hypothesis of no significant effect on the returns of stocks listed on the LSE in four of five events. Furthermore, an equally weighted portfolio consisting of the

significant CAARs that an investor would have earned (net of taxes and transaction costs) would have been positive. Since we find significant CAARs in four of the five events (appendix 2), we cannot state that our results support the immediate absorbance of new information into stock prices as stipulated by the EMH. As such, the answer to our first research question is that around the release of the announcement significant CAARs can be found in most of the cases, thus we find a significant impact.

As regards our second research question as well as second hypothesis, the results show that the difference was significant only once which means that the null hypothesis of no asymmetry between the categories cannot be rejected in four of the five events.

8. Conclusion

_____________________________________________________________________________________ This section of our thesis summarises the findings of our thesis. Furthermore, we summarise the methodological approach used in our thesis. Finally, we give suggestions for further research into this area.

______________________________________________________________________ In this thesis, we investigate by means of a standard CAPM model whether five distinct OPEC oil production cut decision announcements, over the range of 2016 to 2019, have an impact on the stock performance of certain companies listed on the LSE. We test this by means of CAARs and whether they exhibit significant values based on 1%, 5% and 10% levels of significance. We test the results of companies in two categories; specifically, oil, gas & mining companies and companies that use oil as an input ( specifically: medical/pharmaceutical, industrial transport and industrial engineering companies). These two categories are chosen based on finding by previous studies. We also test the research question regarding whether the output we obtain from our CAPM models shows an asymmetry across the two broad sectors.

The main findings of the first research questions is that for the oil, gas & mining category the effects of OPEC news announcements mostly results in positive returns of varying magnitudes on the stock prices of these companies. For the oil input companies, the main finding is that most t-statistics are random; they are either positive, negative and

significant or insignificant. We also compare the results of the two categories and find, based on an OLS regression, no significant differences between the two categories, except for the event of December 2016, where the difference between the CAARs of the two categories are significantly different.

As for the EMH, we find significant accumulation of AARs which means that the new information is not always immediately absorbed into stock prices. This means that we cannot state that our results support that aspect of the EMH.

In light of the Covid-19 pandemic present during the time we wrote this thesis, we would recommend future authors investigating the effects of OPEC oil cut announcements on stock markets. Furthermore, as there was a simultaneous shock in the demand for and supply of oil due to competition between Russia and Saudi Arabia, we recommend future research to contextualise 2020 in the light of these events. Also, future research discussing OPEC oil cut announcement effects on stock different sectors of the economy should aim to minimize the heterogeneity of companies in their research.

9. Reference list

Armitage, S. (1995). Event Study Methods And Evidence On Their Performance.

Journal of Economic Surveys, 9(1), 25-52. doi:10.1111/j.1467-6419.1995.tb00109.x

Barberis, N., & Thaler, R. (2002, September). A Survey of Behavioral Finance (Working paper No. 9222). doi:10.3386/w9222

Bastianin, A., Conti, F., & Manera, M. (2016). The impacts of oil price shocks on stock market volatility: Evidence from the G7 countries. Energy Policy, 98, 160-169.

doi:10.1016/j.enpol.2016.08.020

Bernard, V. L., & Thomas, J. K. (1989). Post-Earnings-Announcement Drift: Delayed Price Response or Risk Premium? Journal of Accounting Research, 27, 1-36.

doi:10.2307/2491062

Black, F. (1986). Noise. The Journal of Finance, 41(3), 528-543. doi:10.1111/j.1540-6261.1986.tb04513.x

Boehmer, E., Masumeci, J., & Poulsen, A. B. (1991). Event-study methodology under conditions of event-induced variance. Journal of Financial Economics, 30(2), 253-272. doi:10.1016/0304-405x(91)90032-f

Broadstock, D. C., Wang, R., & Zhang, D. (2014). Direct and indirect oil shocks and their impacts upon energy related stocks. Economic Systems, 38(3), 451-467.

doi:10.1016/j.ecosys.2014.02.002

Chan, W. S. (2003). Stock price reaction to news and no-news: Drift and reversal after headlines. Journal of Financial Economics, 70(2), 223-260.

doi:10.1016/s0304-405x(03)00146-6

Clark, J. M., Cornwell, T. B., & Pruitt, S. W. (2008). The impact of title event sponsorship announcements on shareholder wealth. Market Lett, 20(2), 169–182. doi: 10.1007/s11002-008-9064-z

Cox, R. A., Dayanandan, A., Donker, H., & Nofsinger, J. (2017). The Bad, the boom and the bust: Profit warnings over the business cycle. Journal of Economics and

Business, 89, 13-19. doi:10.1016/j.jeconbus.2016.09.001

Cuñado, J., & Gracia, F. P. (2003). Do oil price shocks matter? Evidence for some European countries. Energy Economics, 25(2), 137-154. doi:10.1016/s0140-9883(02)00099-3

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

De Long, J. B., Shleifer, A., Summers, L. H., & Waldmann, R. J. (1990). Noise Trader Risk in Financial Markets. Journal of Political Economy, 98(4), 703-738.

doi:10.1086/261703

Demirer, R., & Kutan, A. M. (2010). The behavior of crude oil spot and futures prices around OPEC and SPR announcements: An event study perspective. Energy

Economics, 32(6), 1467-1476. doi:10.1016/j.eneco.2010.06.006

El-Sharif, I., Brown, D., Burton, B., Nixon, B., & Russell, 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. doi:10.1016/j.eneco.2005.09.002

Fama, E. F. (1970). EFFICIENT CAPITAL MARKETS: A REVIEW OF THEORY AND EMPIRICAL WORK. The Journal of Finance, 25(2), 383-417.

doi:10.2307/2325486

Fama, E. F., & French, K. R. (2004). The Capital Asset Pricing Model: Theory and Evidence. Journal of Economic Perspectives, 18(3), 25–46. doi:

10.1257/0895330042162430

Filis, G., Degiannakis, S., & Floros, C. (2011). Dynamic correlation between stock market and oil prices: The case of oil-importing and oil-exporting countries.

International Review of Financial Analysis, 20(3), 152-164.

doi:10.1016/j.irfa.2011.02.014

Guidi, M. G., Russell, A., & Tarbert, H. (2006). The effect of OPEC policy decisions on oil and stock prices. OPEC Review, 30(1), 1-18. doi:10.1111/j.1468-0076.2006.00157.x

Hamilton, J. D. (2011). Historical Oil Shocks (Working paper No. 16790). doi:10.3386/w16790

Hamilton, J. D. (2003). What is an oil shock? Journal of Econometrics, 113(2), 363-398. doi:10.1016/s0304-4076(02)00207-5

Henderson, G. V. (1990). Problems and Solutions in Conducting Event Studies. The

Journal of Risk and Insurance, 57(2), 282-306. doi:10.2307/253304

Henriques, I., & Sadorsky, P. (2008). Oil prices and the stock prices of alternative energy companies. Energy Economics, 30(3), 998-1010.

doi:10.1016/j.eneco.2007.11.001

Jackson, D., & Madura, J. (2003). Profit Warnings and Timing. Financial Review, 38(4), 497–513. doi: 10.1111/1540-6288.00057

Jensen, M. C., Black, F., & Scholes, M. S. (2006, June 13). The Capital Asset Pricing

Model: Some Empirical Tests [Scholarly project]. In SSRN. Retrieved 2020, from

https://papers.ssrn.com/sol3/papers.cfm?abstract_id=908569

Kaufmann, R. K., Dees, S., Karadeloglou, P., & Sanchez, M. (2004). Does OPEC Matter? An Econometric Analysis of Oil Prices. The Energy Journal, 25(4), 67-90. doi:10.5547/issn0195-6574-ej-vol25-no4-4

Kilian, L., & Park, C. (2009). The Impact Of Oil Price Shocks On The U.s. Stock Market. International Economic Review, 50(4), 1267-1287. doi:10.1111/j.1468-2354.2009.00568.x

Lauenstein, P., & Simic, A. K. (2017). Information processing in freight and freight forward markets: An event study on OPEC announcements. International Journal of

Financial Markets and Derivatives, 6(2), 149-181. doi:10.1504/ijfmd.2017.088002

Lee, Y.-H., & Chiou, J.-S. (2011). Oil sensitivity and its asymmetric impact on the stock market. Energy, 36(1), 168–174. doi: 10.1016/j.energy.2010.10.057

Li, Q., Wang, T., Li, P., Liu, L., Gong, Q., & Chen, Y. (2014). The effect of news and public mood on stock movements. Information Sciences, 278, 826-840.

doi:10.1016/j.ins.2014.03.096

Lin, S. X., & Tamvakis, M. (2010). OPEC announcements and their effects on crude oil prices. Energy Policy, 38(2), 1010-1016. doi:10.1016/j.enpol.2009.10.05

Lintner, J. (1965). The Valuation of Risk Assets and the Selection of Risky Investments in Stock Portfolios and Capital Budgets. The Review of Economics and Statistics, 47(1), 13. doi: 10.2307/1924119

Liu, X., Liu, B., & Han, X. (2019). Analysis of Herd Effect of Investor’s Behavior from the Perspective of Behavioral Finance. Proceedings of the 2019 International

Conference on Management, Education Technology and Economics (ICMETE 2019), 82. doi: 10.2991/icmete-19.2019.133

Loutia, A., Mellios, C., & Andriosopoulos, K. (2016). Do OPEC announcements influence oil prices? Energy Policy, 90(C), 262-272. doi:10.1016/j.enpol.2015.11.025

Malkiel, B. G. (2003). The Efficient Market Hypothesis and Its Critics. Journal of

Economic Perspectives, 17(1), 59-82. doi:10.1257/089533003321164958

Moskowitz, T. J., & Grinblatt, M. (1999). Do Industries Explain Momentum? The

Journal of Finance, 54(4), 1249-1290. doi:10.1111/0022-1082.00146

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

Economics, 30(3), 986-997. doi:10.1016/j.eneco.2007.09.003

Narayan, P. K., & Sharma, S. S. (2011). Investment and Oil Price Volatility. Journal of

Banking & Finance, 35, 3253-3262. doi:10.2139/ssrn.1107154

Oprean, C., & Tănăsescu, C. (2013). The Speed of Incorporating Information into Prices.

Procedia Economics and Finance, 6, 675–682. doi: 10.1016/s2212-5671(13)00188-3

Organization of Petroleum Exporting Countries (n.d.). Press releases. Retrieved 10 February, 2020, from https://www.opec.org/opec_web/en/press_room/28.htm

Ozdemir, Z. A. (2008). Efficient market hypothesis: evidence from a small open-economy. Applied Economics, 40(5), 633–641. doi: 10.1080/00036840600722315

Randjbaran, E., Tahmoorespour, R., Rezvani, M., & Safari, M. (2018). The Impact of Oil Price Fluctuations on Industry Stock Returns: Evidence from International Markets.

Journal of Management Info, 5(1), 1-12. doi:10.31580/jmi.v5i1.26

Schmidbauer, H., & Rösch, A. (2012). OPEC news announcements: Effects on oil price expectation and volatility. Energy Economics, 34(5), 1656-1663.

doi:10.1016/j.eneco.2012.01.006

Scholtens, B., & Yurtsever, C. (2012). Oil price shocks and European industries. Energy

Economics, 34(4), 1187-1195. doi:10.1016/j.eneco.2011.10.012

Sharpe, W. F. (1964). Capital Asset Prices: A Theory of Market Equilibrium under Conditions of Risk. The Journal of Finance, 19(3), 425. doi: 10.2307/2977928

Shiller, R. J. (2003). From Efficient Markets Theory to Behavioral Finance. Journal of

Economic Perspectives, 17(1), 83-104. doi:10.1257/089533003321164967

Xu, B. (2015). Oil prices and UK industry-level stock returns. Applied Economics,

10. Appendix

Appendix 1: Company list

Letters in brackets state companies’ stock market ticker

Oil, gas and mining companies Companies that use oil as an input in production

Royal Dutch Shell PLC (RDSb) Total (TTA)

British Petroleum PLC (BP) Lukoil (LKOH)

Cairn Energy PLC (CNE) Premier Oil PLC (PMO) Tullow Oil PLC (TLW) Glencore PLC (GLEN) EVRAZ plc (EVRE)

Anglo American PLC (AAL) Antofagasta PLC (ANTO) Rio Tinto Group (RIO) Schlumberger (SCL) Rosneft (ROSNq) Centrica PLC (CNA)

Industrial transportation companies:

Clarkson (CKN)

Royal Mail PLC (RMG)

Signature Aviation PLC (SIGSI) James Fisher & Sons plc (FSJ)

Industrial engineering companies:

Spirax-Sarco Engineering PLC (SPX) Bodycote PLC (BOY)

IMI (IMI)

Rotork PLC (ROR) Weir Group PLC (Weir)

Hill & Smith Holdings PLC (HILS)

Healthcare/pharmaceutical companies:

AstraZeneca PLC (AZN) GlaxoSmithKline PLC (GLK) Smith & Nephew PLC (SN)

Mediclinic International PLC (MDCM) UDG Healthcare plc (UDG)

Appendix 2: CAAR outputs

***, **, * present 1%, 5% and 10% level of significance, respectively

Event 1: Sep-16

Event window CAARogm tstat CAARoi t-stat

CAAR [-5, +5] 5.87%** 2.63 -0.89% -0.36 CAAR [-5, +10] 10.27%** 2.23 -0.71% -0.34 CAAR [-5. +20] 11.35%*** 2.54 -3.62%*** -2.50 CAAR [-5, +30] 12.65%*** 2.83 -3.07%*** -2.50 CAAR [0, +5] 4.34% 0.97 -0.17% -0.05 CAAR [0, +10] 8.37%** 1.96 0.02% 0.01 CAAR [0, +20] 9.82%** 2.20 -1.43% -0.80 CAAR [0, +30] 11.12%** 2.49 -0.26% -0.17 CAAR [-5, 0] 4.85%*** 3.2 -1.52% -0.50 CAAR [-20, +20] 12.37%*** 10.56 -2.88%*** -2.46