Event Study Analysis

Event study analysis

“The relationship between elections and stock market returns”

Abstract

This master thesis research investigates the timing of election-induced events on the European financial stock markets. Positive stock market returns in periods with an election-induced event are found. Stock market returns are positively (less negatively) affected in periods ending with an election- induced event compared to the entire event-window. Furthermore, election-induced stock market returns on the European financial markets are positively correlated with the national stock market. Positive (negative) stock market returns following the election outcome on the national index will positively (negatively) affect the European stock returns. Moreover, we find that stock market returns are positively correlated with shifts in sentiment. Favourable (unfavourable) political news is positively (negatively) affecting stock market returns. Lastly, a negative relationship is found between economic performance and stock market returns during election periods. The abnormal stock market returns are positively (negatively) affected by elections before (after) the financial crisis. These results are in line with Pantzalis et al. (2000), Brown et al. (1998), Hibbs (1977), Lee et al. (2002) and Chan et al. (1996).

Keywords: Political elections, Stock market return, Financial Crisis, Europe.

Student name: Ronald Antonius Pieter van Eupen Administration number: 866381 E-mail address: Study program: MSc Finance CFA Track Supervisor: dr. D.A. (David) Hollanders Second reader: dr. F. (Fabio) Castiglionesi Date: August 29, 2017 2

Preface

This MSc thesis is the result of a graduation research under supervision of the Finance Department at Tilburg University. As part of the MSc Finance CFA track program, this research focussed around macroeconomic events and specifically the effect of political elections on stock market returns. This research has been practised in collaboration with the Tilburg School of Economics and Management (TiSEM). The basic objective of this research is to get more knowledge about investor behaviour and stock market returns during periods of elections.

In this MSc Finance, various effects and implications have been discussed with the help of an event based study regarding political elections and stock market returns. This research has contributed to help investors around the world in optimizing their asset allocation in times of uncertainty, as of elections.

Nothing of this would have been possible with the help of Tilburg School of Economics and Management. I would like to thank a few important people in particular. First, I would like to thank my supervisor dr. D.A. Hollanders with all the help and advice during my graduation. I have been more than grateful for the pleasant cooperation. Next, I would like to thank all the professors and faculty members of the Finance Department. Lastly and most important, I would like to thank my family and friends for all the support. They have kept me motivated and focussed during my graduation.

Ronald van Eupen Tilburg, August - 2017 3

Table of content

1. Introduction ...... 4 2. Literature review ...... 6 2.1 Stock market returns ...... 6 2.2 Stock market returns and political elections ...... 7 2.3 Spill-over effects of stock market returns ...... 9 3. Methodology ...... 11 3.1 Sample selection ...... 11 3.2 Event study methodology ...... 13 4. Data Analysis ...... 16 4.1 Constructing the index ...... 16 4.2 Identifying the missing data ...... 17 4.3 Political elections and stock market returns ...... 19 4.3.1. The Netherlands ...... 19 4.3.2. France ...... 27 4.3.3. Germany ...... 36 4.3.4. United Kingdom ...... 42 4.3.5. Spain ...... 50 4.3.6. Italy ...... 60 5. Robustness check...... 69 6. Conclusion ...... 71 6.1 Discussion ...... 72 7. Appendix ...... 73 7.1 Dutch parliamentary elections ...... 73 7.2 French presidential elections ...... 82 7.3 German federal elections ...... 91 7.4 United Kingdom general elections ...... 97 7.5 Spanish general elections ...... 106 7.5 Italian general elections ...... 115 8. References ...... 124 4

1. Introduction

March 16, 2017, German minister in the Chancellor’s Office Peter Altmaier tweeted the following: “Nederland oh Nederland jij bent een kampioen! Wij houden van Oranje om zijn daden en zijn doen! Gefeliciteerd met dit geweldig resultaat!”

This is one of many reactions after the election result of the Dutch general election on March 15, 2017. An important win by the conservative parties, where Euroscepticism is becoming increasingly popular across Europe. The Brexit of the United Kingdom and the presidential election of the United States of America was seen by many people as a sign of change. The rise of the “far-right” is not a false alarm, but the results show that a breakthrough of far-right is not unavoidable. This breakthrough is important as the Netherlands is the first important election in Europe in 2017, before the French presidential and the German federal election. Investors were pleased by the win of the Liberal Party VVD in the Netherlands as the AEX showed an 0.87% increase overnight whilst the election results were announced. However, the election results of the France and German elections are of greater importance for the stability in the EU, given their economic magnitude. In France, the first round of the presidential election is held on April 23th. Turbulent times are predicted by investors and analyst in case Mélenchon (far left) and Le pen (far right) won as they want to sever the connection of France with the European Union. A win by Emmanuel Macron gaining 24% of the votes is therefore a noteworthy revelation and characterized by a 4.13% increase overnight at the French stock index. Moreover, foreign markets showed positive stock market returns besides the effect of firms being cross listed. Elections have therefor an important impact on both the national stock market index as the foreign markets. Besides the economic impact, elections characterize democracy as first democracy depended on representatives for decision making, Woodruff (2005). Democracy originated more than 2400 years ago in Greece1. The definition of democracy is “rule by the people”, which means that the people govern their nation. Elections are mechanisms through which citizens participate directly in the political process and elect their fellow citizens into office for restricted periods of time, Alexander and Kaboyakgosi (2012). Elections are therefore major events that can change the direction of a country and should not be underestimated. It indirectly influences the financial landscape, affecting the optimal asset allocation of financial investors.

1http://www.civiced.org/pdfs/books/ElementsOfDemocracy/Elements_Subsection3.pdf 5

Because of this relationship, elections have an important weight on the financial sector. Moreover, unlike many other events that affect stock market returns, election events are known in advance. International financial investors have been struggling in showing returns over the years. The Financial Crisis of 2008 still has a major impact on international trade and investment streams around the globe. Setting a new direction through national and presidential elections can therefore be of great importance, especially after years of economic downturn. This study examines stock market behavior around political elections in different countries of the European Union and addresses the following questions. What is the spill-over effect of national elections on foreign stock markets? What is the influence of elections on stock markets returns after a period of growth? What is the influence of elections on stock market returns during a crisis? What is the influence of elections on stock market returns after a period of contraction? What is the duration of abnormal stock returns (before and after election)? This research explores these questions using a standard event study methodology based on King (2012). Abnormal return behavior around election dates across 6 European countries for the period around the financial crisis of 2008 are examined. To the best of the authors’ knowledge, it is the first study that rigorously quantifies the change in returns over time. Based on the question formulated above, this research will test whether elections have an influence on financial market returns and whether this influence changes over time around financial crisis.

Hence, the research question of this Master Finance thesis is:

“What is the effect of elections on stock market returns in a financial crisis?”

2. Literature review

The literature review section is separated in three sections. In the first section, research papers that discuss stock market returns are presented. Research papers about stock market returns and political elections are discussed in the second part. Lastly, papers that address spill-over effects on international stock markets are addressed.

2.1 Stock market returns

Even though no other research has performed a test to check whether the effect of elections on stock market returns changes over time in crisis periods, the empirical literature does provide numerous papers that address stock market returns. Particularly, the relation between volatility and stock market returns is an important topic within the academic literature. One of the most cited research papers in this area is performed by French et al. (1987), which investigated whether expected market risk premium (defined as expected return minus the risk-free rate) is positively related to volatility. They find a positive relation between the expected market risk premium and the predictable level of volatility. Also, a strong negative relation between the unpredictable component of stock market volatility and excess returns is found. However, if risk premiums are positively related to predictable volatility, a positive unexpected change in volatility increases future expected risk premiums. This indirectly means that elections should have a positive impact on stock market returns. An unexpected change in the election process which diminishes uncertainty has a positive effect on the predictable level of volatility. This in turn means an increase in future expected risk premiums, hence positive stock market returns. Merton (1980) elaborates this further and implies that the expected risk premium on the market is proportional to the variance of the market returns if investors demonstrates constant relative risk aversion. A percentage increase in post-event expected returns is equal to the percentage increase in the post-event return variance. In general, the increase in post-event return is bigger than the increase in the variance rate. Brown et al. (1988) have tested the uncertain information hypothesis (UIH) in explaining the response of investors to the arrival of unanticipated information. They find that when a large sample of favourable and unfavourable events are analysed separately, the immediate price change will be followed by positive returns during the post-event period.

Favourable or unfavourable news immediately affect stock market prices. As the uncertainty surrounding the event is reduced, this will consequently positively affect stock market returns, regardless of the event in dispute. However, unless we are in a perfect world, it is impossible to predict accurately the direction and magnitude of stock market returns for individual events on an ex ante basis because humans do not always act rational. The same was found by Werner et al. (1984). As uncertainty about the electoral outcome increases, it becomes harder to forecast the stock’s expected value. Hence, stock’s market prices are adjusted upward and downward over time in an attempt to equilibrate the stock’s expected value and posted price. An important study which this research is related to is Chan et al. (1996). It investigates the influence of political risk on stock price volatility in Hong Kong. They found elevated levels in volatility in cases where there is political news compared to situations where there is no political news. Hence, political news amplifies stock market volatility. This in turn is useful in this research as fluctuations in volatility can cause fluctuations as big in stock market returns as demonstrated by Merton (1980). Furthermore, favourable political news has a positive effect on stock market returns (Blue-index), while unfavourable news has a negative impact at the stock market (Blue-index). This seems logical, however as Chan et al. (1996) further elaborate this is not. Political news, favourable or unfavourable, only affects volatility of the Chinese orientated stocks but not the stock market returns of the Chinese orientated stocks due to marketwise and substitution effects. This contradicts this research paper, since it is expected that both national as well as a transnational stock market returns are affected by elections.

2.2 Stock market returns and political elections

The research question of this paper is about the impact of elections on financial stock market returns. Many papers in the academic literature have addressed the relationship between stock market returns and volatility as described above, however less papers have addressed the relationship between stock market performance and elections. Following Pantzalis et al. (2000), political elections are specifically important since elections provide inhabitants of a country the possibility to exert influence on the medium- and long-term course of that country. They hold the power to re-elect the incumbent or punish the incumbents by electing a new government setting a new course.

Therefore, elections have an indirect impact on the economy by which they are closely monitored by investors and other financial market participants. Consequently, financial markets react strongly on new information regarding decisions that impact monetary and fiscal policies. Based on the uncertain information hypothesis (UIH) formulated by Brown et al. (1988), stock market returns tend to positively react as uncertainty regarding the election outcome is reduced. However, if the election outcome is less certain or investors are not able to assess their asset allocations on the country’s new course, a surprise effect is expected. Positive price changes are then expected in the weeks following the election outcome. Indeed, as Pantzalis et al. (2000) have discovered, positive stock market reactions are measured in the two-week preceding the political election and this persists through the four-week period following the election. More specifically, the results suggest that a positive election effect is primarily concentrated in cases where the country’s economic performance is poor. This is particularly interesting as part of this research is focused around the question whether political elections are negatively correlated with economic performance. It is expected that the voting turn-out and the impact are higher after a period of economic downturn and vice versa. Lee et al. (2002) found that a shift in sentiment has a significant positive impact on stock market excess returns. Brown and Cliff (1999) have regressed sentiment with time and found weak evidence of short-run predictability but a strong correlation between sentiment and long-horizon (2–3 years) returns. As Bernard Baruch stated, what is important in market fluctuations are not the events themselves, but the humans’ reactions to those events. This is an important statement in behavioral finance, because empirical literature assumes that humans act rational. This statement is further grounded by an important law in experimental psychology formulated by Bayes. He states that humans correctly react to the arrival of new information. Nevertheless, research performed by Werner et al. (1984) have suggested that in violation of Bayes’ rule, most people tend to “overreact” to unexpected and dramatic news events. Humans, investors, overweight recent information and underweight prior data. The term overreaction carries with it an implicit comparison to some degree of reaction that is considered to be appropriate. Therefore, timing is an important variable in explaining stock market returns. Bullish shifts in sentiment are negatively related to the volatility of returns but are positively related with stock market excess returns, vice versa for a bearish shift in sentiment. Furthermore, the magnitude of the change in sentiment has a significant impact on stock market returns as well.

As the change in sentiment increases, the fluctuation in upward and downward revision of stock market returns becomes more severe. This is also the case in Bialkowski et al. (2008). A change in the political orientation of a country can cause a period of excessive volatility that continues for a longer period. Durnev (2011) has relate corporate investments to political uncertainty and found that investment-to-price sensitivity is 40% lower during election years than to non-election years. This effect is even stronger when it is harder in forecasting the election outcome. According to the “information view” of investments, elections are associated with uncertainty about future government policies. This may lower the information quality of stock prices. These results have implications for policymakers. In markets with high political uncertainty, the stock market is less able to guide capital to its best uses and lowers stock market performance. This suggests that it is important for policymakers to find mechanisms to reduce unnecessary political uncertainty. However, the anticipation left-wing or right-wing policymakers has significant influence on the volume of shares trading in the market as Werner et al. (1984) shows. The empirical results show a decrease in both trading volume as well as the mean and volatility of stock prices in the United States and Great Britain under left-wing administrations, vice versa for right-wing administrations. The result indicates a positive influence of right-wing party policies on stock market returns.

2.3 Spill-over effects of stock market returns

The previous parts of the literature review have focussed about the relationship between stock market returns and volatility and the effect of elections on stock market returns. However, as financial markets are more linked due to globalization, it is interesting to see the effect of stock market returns motivated by political elections on foreign markets. As already pointed out in the research paper of Chan et al. (1996), political news has a positive effect on British orientated stock returns, while unfavourable news has a negative impact on British orientated stock returns. This seems logical, however as Chan et al. (1996) further elaborate, this is not always the case. Political news, favourable or unfavourable, has a market wide effect. Substitution effects occur in which only the volatility of Chinese orientated stocks is affected but not the returns. This is particularly interesting as this research paper has the European Market as target market where national borders are less profound as in Hong Kong.

It is expected that the impact of a single nation within the European market has a much greater reach that impacts the European financial markets as a whole. This transnational impact is also find by Bialkowski et al. (2008), which has investigated the impact of elections across 27 industrialized OECD countries. Not only they find elevated stock market volatility around national elections, there appears to be a market inflation in unconditional variance. Second, uncertainty about election outcome has implications for risk averse investors. French et al. (1991) and Baxter et al. (1997) found a significant home bias. Investors are undiversified internationally. However, the financial crisis of 2008 have shown that financial markets are more interlinked than ever. Any wide spread fluctuations will increase systematic risk of all stocks listed. Hamao et al. (1990) have elaborate this further and find that unexpected changes in foreign stock market returns are associated with significant spillover effects on the conditional mean of the domestic market for both open-to-close and close-to-open returns. In their research focusing on the stock market of the United States, United Kingdom and Japan a significant spillover effect from the US and UK on the Japanese market is found. Hilliard (1979), Jaffe and Esterfiel (1985a, 1985b), Eun and Shim (1989), Barclay, Litzenbergen and Warner (1990) all find evidence of positive correlations in daily close-to-close returns across the financial stock markets. Therefore, it is expected that political elections have a significant spill-over effect on foreign financial markets.

Based on the literature discussed above, the following hypotheses are tested:

H1: Stock market returns are likely to be higher over election-induced periods than over periods without an election.

H2: Elections have a positive spill-over effect on foreign stock markets.

H3: The level of impact of elections on stock market returns is negatively correlated with economic performance.

3. Methodology

As mentioned in the introduction, the aim of this research paper is to study the impact of political elections on stock market returns. In particular, this research investigates the significance of time. The methodology part of this research paper consists of the selection criteria, the methodology and in the end the data is discussed.

3.1 Sample selection

As Pantzalis, Stangeland and Turtle (2000) have argued in their paper, political events exert significant influence on the financial markets. Particularly, investors are curious about political decisions that may impact monetary and fiscal policies. Therefore, investors are closely following political events and their outcomes. Asset allocations are revised as investors’ expectations are impacted by these political decisions. Political elections are amongst the most important political events. Especially, since voters have the opportunity to exert their voting power in choosing the course of a nation for the medium- and long term. Furthermore, the information coverage by the media, pollsters and analysts is tremendous. The process of filtering this information between politicians and the public has an indirect effect on the financial markets as it adjusts the thoughts of voters. Lastly, as the outcome of elections becomes more certain, financial market participants revise their asset allocations. For this reason, only political elections on a national level are used in this research as they exert the most power on the medium- and long-term course of a country and hence perform the most influence on the financial markets. Consequently, only general, presidential and federal elections are included. Next, a country sample is needed. Only countries of the European Union are used for this research. The European Union is a perfect market for my purposes, as the impact of elections is not limited to a national level, but also has a transnational impact as European member states are part of the European Union market. A good manner to measure spill-over effects. An important selection criteria for the sample selection is the scale of the economy. Therefore, only the most important economies of the European Union are chosen aligned by political elections. These consist of the countries listed in table 1.

The data that will be used can be seen in table 1 and will mainly be retrieved from DataStream, Yahoo Finance and Bloomberg. Market returns of national stock exchanges of the 6 biggest European economies by GDP are used2. In order to investigate whether the influence of elections on stock market returns changes by time, a categorization is made to distribute elections in a before, during or after financial crisis period. The election sample and the distribution of the political elections by time used for this research can also be found in table 1. Lastly, a created European market index is used as a benchmark model in order to measure the impact of national elections on the European level.

Table 1: event study sample

Country Type of Election Index Before During After The Netherlands General AEX 2006 2010 2017 France Presidential CAC 40 2007 2012 2017 Germany Federal DAX 2005 2009 2017 United Kingdom General FTSE 100 2005 2010 2015 Spain General IBEX 35 2008 2011 2015 Italy General FTSE MIB 2006 2008 2013

A few notes are in order. While the inhabitants of the United Kingdom have voted for leaving the European Union, this research will include the United Kingdom within the European Market for simplicity reasons. The general elections occurred during the time that the United Kingdom was a member state of the European Union. Also, the French presidential elections occur in two rounds. The uncertainty information hypothesis by Brown et al. (1988) states that uncertainty around elections is resolved prior to the actual election date which positively effects stock market returns. Most of the uncertainty surrounding the presidential elections is resolved after the first round as the number of electors is reduced. Much of the election result in round two is priced in. So, in order to calculate the clear impact of elections on stock market returns, only the first round of the French presidential elections is used.

2http://statisticstimes.com/economy/european-countries-by-gdp.php

3.2 Event study methodology

According to the efficient market hypothesis, stock prices reflect all available information. Using this hypothesis, an event study can be used to calculate how an event changes stock prospects by quantifying its impact on the stock price. An event study methodology is particularly suitable since the election dates differs in the sample as shown in table 1. By using an event study methodology, we use t=0 as instead of a calendar date. In this way, I can compare multiple election events starting at different moments in time. In following Pantzalis et al. (2000) and Bialkowski et al. (2008) an estimation period of 100 weeks will be used. Furthermore, the two-week pre-event window [-4, -3] is not included in the estimation window to prevent contamination by the event. The event study methodology is based on King (2012). The analysis differentiates between returns that would have been expected if the event did not take place (normal returns) and returns that were caused by the event (abnormal returns). The most used technique in an event-based research is the market model. I will use MacKinlay’s (1977) market model in my research which relates any excess return to the excess return of the market portfolio in order to calculate abnormal returns. First the daily stock market returns are modelled as follows:

R, = � + � ∗ � + �,

Where R, and � are the daily returns of the stock prices and the market portfolio and �, is the error-term. Further, under general conditions ordinary least squares (OLS) is a consistent estimation procedure for the market model parameters α and β. With the daily returns of stock prices the abnormal return can be calculated as the difference between the actual ex post return over the event window minus the normal return over the same event window:

AR, = R, − E(�,| �)

ARi,t, Ri,t, and E(Ri,t | Xi,t) are the abnormal, actual, and normal returns respectively for time period t. There are two ways of modeling the normal return. Since we use the market model, the normal return is defined as the expected return without conditioning on the event taking place, hence the market return. The market model assumes a stable linear relation between the market return and the security return, MacKinlay’s (1977).

This kind of analysis is performed for multiple events of the same type (here elections) and may show typical patterns of stock market returns. Abnormal returns associated with a point in time before or after the event are defined by the Average Abnormal Return (AAR). It allows to check the magnitude of the outperformance of the stock sample relatively towards the reference sample. AAR is defined as follows: 1 AAR = �� N ,

In the end, the total impact of the event over the time period, also known as the “event window”, has to be measured. The event window is defined by [T, T] and can be measured by adding up individual abnormal returns. This gives Cumulative Abnormal Return (CAR), as shown in the formula below:

CAR(,) = ��,

Next to the AR and the CAR, I will calculate the Cumulative Average Abnormal Return (CAAR) since we have an event study that holds multiple observations of individual events. Calculating AAR and CAAR is done since they give a good indication of the market reaction to the event (King, 2009 and 2012). It represents the mean values of identical events:

1 CAAR = �� (,)

To test whether the outperformance of the reference market by the chosen sample is significant is given by the distribution of the T-statistic. The T-stat will be calculated by the following method. First, I calculate the cumulative average return (CAR) and then I take the cumulative average abnormal return (CAAR). Next, I subtract the CAAR from the CAR and square them for before adding them up. The T-stat is then the CAAR divided by the standard deviation of the CAAR. Considering three election periods (before, during and after the crisis), I can investigate with the help of an event study methodology described above whether the optimism after a long period of growth has less impact on stock market returns than a long period of contraction.

According to the uncertain information hypothesis (UIH), the observed return over the period ending with an event-induced uncertainty should be higher than the average return over periods with no event-induced uncertainty. Therefore, I follow the uncertain information hypothesis (UIH) of Brown et al. (1988) to analyse the cumulative abnormal return (CAR) and see whether the observed returns are indeed higher than the average returns. In following King (2012) and Pantzalis et al. (2000), the event window is defined to be [-2, 0], the second week before the election, and ending at t=0, the election day. This event window is chosen because it includes the period with the most potential for uncertainty resolution leading up to an election. After that, the greatest degree of uncertainty resolution and hence the highest observed returns are expected immediately after the election (event) date due to extensive media coverage. So, it is expected that elections have a positive effect on stock market returns, see hypothesis formulated in the literature review section. The CAR’s are positive in the two-week period preceding the election. Hence:

H1a: CAR-2,0 > 0

H1b: CAR-2,4 > 0

If markets need time to assess election impacts following the election results, we would expect post-election positive abnormal returns. Therefore, a four-week period after the election is constructed to test whether this is true:

H1c: CAR1,4 > 0

In both these cases, it is expected that stock market returns are positively affected by the election results. Next, it is expected that stock market returns are negatively correlated by the economic business cycle. Abnormal stock market returns should be bigger after a period of contraction than after years of expansion as fluctuations stock market returns follow changes in sentiment. Final remarks have to be specified. Like Chan et al. (1996), we use open-to-open market returns. Open-to-open returns are a better measure of returns to cover the political news appearing in the newspapers, since some of the political news is announced during the trading hours of the previous day. Also, most of the election results are announced after closing of the financial markets and the official results are often announced during the following day.

4. Data Analysis

The data analysis chapter is separated in the following parts. In the first part, the creation of a European index is discussed. After that, the missing data is identified according to a four- step process described by Hair et al. (1995). Thereafter, the national stock indices are regressed with the European Market Index.

4.1 Constructing the index

The sample selection in section 4 has described the stock indexes used in this research. The purpose of this paper is to check for a transnational impact of national elections on the European stock markets. A European index has to specified to regress national stock markets in order to check if a transnational impact is measured. Multiple indices are investigated for this purpose, among which Stoxx50, Eurostoxx100, S&P Europe 350 and FTSE Eurotop 100. However, none of the described indexes are suitable for this research due to their short trading history. The Stoxx Europe 600 index is also investigated to measure the effect of elections of European member states on the European stock market. The Stoxx Europe 600 index is composed of 600 firms across 17 European countries, among which the 8 countries specified in table 1 could be used for this research. However, the companies that are used for this index are large, mid and small capitalization firms. In order to measure a clear impact on the national level, only small cap firms are used in this research as they are exposed to national level implications. Large cap companies are orientated on a more international level and are affected more by implications that act on the global level as Cheol et al. (2008) have discussed in their research paper. To overcome this difficulty of using an index to measure the impact of political elections on the European level, this research will construct an index based on the PWC Top 100 European and German companies3. PWC has identified 100 listed European companies and ranked them by market capitalization. This research has investigated and compared the European listed companies over a period starting in 2008 till 2014. Based on the article, a few assumptions have been made. To form a valid European Market Index, only stocks of European Union companies are used, meaning that Russian stocks are excluded. Next, only stocks with at least 7 years of trading history are used. For this reason, Continental AG is excluded. Lastly, companies that are purchased by an acquirer are not integrated.

3https://www.pwc.de/de/kapitalmarktorientierte-unternehmen/assets/healthcare-und-pharma-werte-sind-bei-anlegern-besonders-gefragt

In the end, an index is constructed that includes 80 European listed companies ranked by their market capitalization. In the following graphs, an overview is given which categorizes the companies by country as well as by industry.

4 1 Categorization by industry: 7 19

9 Financial Services Retail & Consumer

TMT Energy 10 17 Healthcare & Pharma Industrial Production

13 Automotive Transport & Logistics

European listed companies per country

UK Germany France Switzerland Spain Italy Sweden Netherlands Norway Ireland Denmark Belgium 0 2 4 6 8 10 12 14 16 18 20

4.2 Identifying the missing data

By following Hair et al (1995), a four-step process is used for identifying the missing data. Often the missing data are expected in the research design. In these cases, the data are termed Ignorable missing data. No specific remedy is implemented to tackle the blank spaces because the allowances for the missing data are inherent in the technique used. But in many cases, the missing data cannot be classified as ignorable. As this research uses yahoo finance and DataStream as main data entry, missing data processes are known and can be identified according to procedural factors. Even if there is little control over the missing data, the blank spaces cannot be ignored.

If the number of missing data is low enough in not affecting the outcome of the research, then any of the approaches for remedying missing data founded by Hair et al (1995) may be applied without biasing the results in any significant manner. On the other hand, if the missing data is concentrated in a small subset which can be substantially reduced by excluding the missing data, then the most efficient way is just to delete these subsets. The dataset is large enough to prevent a biasing effect by deleting the missing data. Generally, missing data under 10 percent can be ignored, except when the missing data occurs in a nonrandom manner, Hair et al., 2009, p. 46. The missing daily returns that are exceeding the 10% threshold are excluded for this reason. This means that at least 72 daily returns are integrated to form a single trading day in the index. Furthermore, outliers are excluded from the data sample. According to Hair et al. 2009, p. 66, an outlier is deleted if the value is aberrant and not representative of any observation without limiting its generalizability. In this research, a return is referred as an outlier if the daily return exceeds 50%. If this is the case, the return is investigated whether it is a return correction. When it is not a correction and no other country specific reason is found, then the return is excluded from the generalized dataset. However, this research wants to be even more precise by the “All-Available Data” method described by Hair et al (1995). This method suggests that only valid values are used to construct a generalized data set. Instead of mean reversion, which can bias the daily returns, the All-Available Data approach only includes daily returns of companies that show daily returns on that specific data. On days where no trading history occurred for that company, it is excluded from the generalized dataset. In this manner, the index is equally weighted and not biased by the missing data.

In order to analyse if the results are significant, this research will verify these results with the significance values4 described in the tables below. Value (-) 1.645 (-) 1.960 (-) 2.346 (-) 2.576 (-) 3.291 P - Value 0.10* 0.05** 0.025*** 0.01**** 0.0001*****

P – Values: Interpretation: p > 0.10 No evidence against the null hypothesis 0.05 < p < 0.10 Weak evidence against the null hypothesis 0.025 < p < 0.05 Moderate evidence against the null hypothesis 0.01 < p < 0.025 Good evidence against the null hypothesis 0.001 < p < 0.01 Strong evidence against the null hypothesis p < 0.001 Very strong evidence against the null hypothesis

4http://www.jcpcarchives.org/full/p-value-statistical-significance-and-clinical-significance-121.php

4.3 Political elections and stock market returns

In this section, the impact of political elections on the stock market returns are analysed. It will be categorized in the following sub chapters. Firstly, the elections are analysed on a country level. A differentiation is made between small and large cap companies. To measure the impact of elections on the national economy the national index is regressed with a small cap index. The same is done with middle and large cap companies to measure the impact of elections on international level. Furthermore, the country level analysis is categorized in before, between and after the Financial Crisis of 2008. To conclude this analysis, an overall comparison is given.

4.3.1. The Netherlands

The political system in the Netherlands is based on a constitutional monarchy since 18145. This means that the monarch’s power is regulated through a constitution. An important alteration was made to the constitution in 1848 which marks the beginning of a parliamentary democracy. However, after numerous political conflicts between the monarch and the parliament, the monarch’s power is inferior to the wills of the parliament since 1870. The power is with the country’s head of state but is very limited. In the end, the ministers are accountable to the parliament for everything what is happening with the fulfilment of their policies. These ministers are chosen by the political party based on a list structure which in the end is chosen by the voters. For about 125 years, the main political parties that have been dominating the Dutch parliament are Labour, Christian Democrats and Liberals. Since the end of World War II, the political landscape has changed with medium- and small-sized parties. The most important elections in the Netherlands are the parliamentary elections which happen every 4 years. However, they also can be held on an earlier date if the government decides to resign itself or is forced to resign by the parliament. The general elections of 2006, 2010 and 2017 are analysed.

5http://nimd.org/wp-content/uploads/2015/02/Dutch-Political-System.pdf

Before

The parliamentary election before the financial crisis were held on November 22, 2006 following the fall the cabinet Balkende II6. On June 29, D66 lost its trust in the government and decides to resign itself by the parliament. New elections were held and even with the fall of Balkende II, the Christian Democrats (CDA) still remained the largest party with 41 seats. Interesting was the increase of the Socialist party (SP) which went from 9 to 25 seats7. Originally the election was planned on May 2007, but due to the fall a new government led by Christian Democrats (CDA), Labour (PVDA) and Christen Union was already formed by February 22, 2007.

Given the results shown in table 2, a positive impact is measured on the national level with a 0.51% positive abnormal return. Despite the positive estimate of the election on the market returns, the measured result is not significant. A significant negative abnormal return is measured the day after national election. This could be due to the measuring standard used in this research. As this research is using open-to-open market returns, Like Chan et al. (1996), the election results are presented after the election trading day. The settled government led by the Christian Democrats was punished together with the other incumbent parties, whereas left-orientated parties won a substantial share of the votes8. As concluded by Akitoby and Stratmann (2007), financial markets give a premium to right wing regimes while penalizing left wing regimes that undertake current-spending-driven expansion. The fear of left-wing administrations resulted in negative abnormal returns the days following the election. In following Pantzalis et al. (2000), this research expects a positive effect of elections on stock market returns. As discussed in chapter 4.2, the cumulative abnormal return (CAR) is expected to be positive in the three event-windows. The observed returns during a period preceding an election should be higher than the average return over a period where no event- induced uncertainty exists. However, as shown in table 2a, the CAR is negative in the three time periods and hence we fail to reject hypothesis 1. As explained above, this could be due to the penalizing effect of the voters on the incumbent parties. The uncertainty of the political direction of the country has a negative impact on financial markets. Fuelled by this uncertainty, the CAR is negative over all three periods. Together with the result of table 2, hypothesis 1 is rejected in the before Financial Crisis period in the Netherlands.

6 https://www.revolvy.com/topic/Dutch%20general%20election,%202006&item_type=topic 7 https://www.volkskrant.nl/politiek/sp-wint-fors-cda-blijft-grootste~a803990/ 8 https://www.volkskrant.nl/binnenland/verkiezingsuitslag-toont-gespleten-nederland~a802294/

Table 2a: Cumulative Abnormal Return (CAR) over the event window. The cumulative abnormal stock return (CAR) is calculated using the next method: First I calculate the abnormal stock returns (AR’s) and then taking the sum of these individual AR’s give the CAR. Time period [-2,4] [-2,0] [1,4] CAR: -4.37% -1.39% -2.47%

European financial markets are positively affected by the Dutch parliamentary election, table 3. At the election day, the impact of the election has led to positive stock market returns on the UK and Spanish stock market, but negative on the French and Italian stock market. The parliamentary election has no effect on the German stock market. Despite the strongest return shown on the Italian stock market showed in the table, none of the data are significant to accept hypothesis 2. This result is also strengthened by the cumulative average abnormal return (CAAR), as shown in table 3a. Even if the event window is taking apart, the data shows a negative impact of political election on stock market returns. However, like Pantzalis et al. (2000), the greatest degree of uncertainty resolution is expected preceding the election day due to extensive media coverage. As seen from table 4a, the most favourable returns are measured immediately after the event, despite the negative direction of this result. Therefore, the uncertain information hypothesis (UIH) formulated by Brown et al. (1988) doesn’t apply here. In table 4, the average abnormal returns (AAR) are presented. This table shows a positive but no significant impact of the Dutch parliamentary election of 2006 on the foreign stock markets. This result is aligned by the results presented in table 3 and table 3a. Based on the tables, hypothesis 2 is rejected.

Table 3a: Cumulative Average Abnormal Return (CAAR) over the event window. The cumulative average abnormal stock return (CAAR) is calculated using the next method: First I calculate the abnormal stock returns (AR’s) per country and then I take the average of these individual AR’s. Lastly all the individual AAR’s are added to form the CAAR. Time period [-2,4] [-2,0] [1,4] CAAR -1.74% -1.48% -0.21%

During

The Dutch parliamentary election during the Financial Crisis were held on June 9, 20109. Originally, this election was planned on May 11, 2011. However, an early election was necessary due to the collapse of Balkenende IV. The Social Democrats were forced to resign from the government as Labour refused to agree with an extension of the Dutch military actions in Afghanistan. This fall is known as the Uruzgan crisis. With this new election, the overall winner was the Freedoms Party led by Geert Wilders. Due to all the commotion, he gained 15 positions during this election. Another great win was the increase by 9 positions for VVD, which makes them therewith the biggest party. The incumbent government incurred a big loss. The Social Democrats lost 20 positions, Labour (PVDA) lost 3 positions and the Christian Union (CU) lost 1 position. This big loss of the Christian Democrats made Jan Peter Balkenende decide to retire from his position as leader of his party. On October 14, 2010, a new “minority” cabinet was formed after 127 days by VVD, CDA and supported by the Freedom’s Party (PVV).

The ASCX small cap index showed no reaction. The parliamentary election resulted in no significant abnormal stock market returns on the ASCX-index. A positive abnormal stock market return is measured at t=0, which indicates a positive effect on stock market returns. However, this research fails to reject hypothesis 1 due to lack of significance. In following Pantzalis et al. (2000), this research expects a positive effect of elections on stock market returns. Observed returns during a period preceding an election should be higher than the average return over a period where no event-induced uncertainty exists. As shown in table 3, this is the case. Most of the uncertainty is resolved in the two weeks preceding the election. The uncertain information hypothesis by Brown et al. (1988) is therefore applicable here. Conversely, a negative effect of a political election is measured over the entire event window. This is due to the negative abnormal stock market returns in the four weeks trailing the election. The loss of the incumbent parties and the rise of the freedoms party could be an explanation. Financial markets perform best in a stable political landscape as the last three governments were led by the Christian Democrats. As found by Alesina et al. (1996), political instability reduces growth. They found a negative and statistically significant effect of government change on growth. Also, the substantial increase of the Freedoms party PVV led by Geert Wilders reduces investors’ expectations on the economy as the PVV is anti-European.

9https://www.parlement.com/id/vib7f70ymd9f/tweede_kamerverkiezingen_2010 23

The economy and productivity of financial markets will decrease in countries led by populist governments as the world economy becomes fragmented and more prone to inflation10. This could be a reason of the substantial negative CAR in the weeks trailing the general election.

Table 5a: Cumulative Abnormal Return (CAR) over the event window. The cumulative abnormal stock return (CAR) is calculated using the next method: First I calculate the abnormal stock returns (AR’s) and then taking the sum of these individual AR’s give the CAR. Time period [-2,4] [-2,0] [1,4] CAR: -0.81% 1.12% -1.84%

The European Financial Markets are influenced by the Dutch parliamentary election of 2010, table 6. At the election day, the impact of the election has led to positive stock market returns on the French and Spanish stock market, but negative on the German, UK and Italian stock market. The Spanish stock market showed the largest abnormal return. However, the results lack significance. Therefore, the Dutch parliamentary election of 2010 has no significant impact on the foreign stock markets and hypothesis 2 is rejected here. This result is also strengthened by the cumulative average abnormal return (CAAR), as shown in table 6a. As Pantzalis et al. (2000) expects, the greatest degree of uncertainty resolution is expected preceding the election due to extensive media coverage. As seen from table 6a, this is not the case with a positive cumulative abnormal return of 2.73%. The uncertain information hypothesis (UIH) formulated by Brown et al. (1998) clarifies this finding. As the election results becomes more certain moving more towards the election day, the election-induced uncertainty is reduced. Risk-adjusted expected returns fall which increases the stock market returns which results in the CAR presented in table 6a. These findings are strengthened by table 7, which shows the Average Abnormal Returns of the international sample and showed no significant increase. Overall the impact on the Dutch parliamentary election was even negative as the CAAR value shows us.

Table 6a: Cumulative Average Abnormal Return (CAAR) over the event window. The cumulative average abnormal stock return (CAAR) is calculated using the next method: First I calculate the abnormal stock returns (AR’s) per country and then I take the average of these individual AR’s. Lastly all the individual AAR’s are added to form the CAAR. Time period [-2,4] [-2,0] [1,4] CAAR: -0.70% 2.73% -3.26%

10http://www.schroders.com/en/us/institutional/insights/economic-views3/is-populism-good-for-markets/

After

On March 15, 2017, the citizens of the Netherlands were able to perform their vote in choosing a new government. The tension in the financial markets is high, since the outcome of the Netherlands is perceived as a leading indicator for the upcoming elections in France and Germany in 201711. After Brexit and President-elect Donald Trump in the United States of America, this election was seen as an electoral test of the increasing populism on the European continent. This fear was justified as the Freedoms party led by Geert Wilders became the second largest party. However, the liberals clearly won the parliamentary election of 2017 even though they lost 8 positions. An interesting fact was the huge loss of positions by Labour which lost their seats towards other left-wing and middle parties.

On the national level, the parliamentary election resulted in no significant increase in stock market returns given the stock market returns measured at the ASCX-index. A positive abnormal return is measured at t=0, which indicates a positive effect on stock market returns. Moreover, in following Pantzalis et al. (2000), this research expects a positive effect of elections on stock market returns. Observed returns during a period preceding an election should be higher than the average return over a period where no event-induced uncertainty exists. However, like Pantzalis et al. (2000), this is only the case in the two weeks preceding the election, as shown in table 8a. The uncertain information hypothesis by Brown et al. (1988) is therefore partially applicable here. A negative effect of a political election is measured over the whole event window and the four weeks trailing the election outcome. Still, due to extensive media coverage and campaigning some uncertainty is resolved preceding the election given the positive CAR. Together with the results of table 8, we fail to reject hypothesis 1 as this value lacks significance.

Table 8a: Cumulative Abnormal Return (CAR) over the event window. The cumulative abnormal stock return (CAR) is calculated using the next method: First I calculate the abnormal stock returns (AR’s) and then taking the sum of these individual AR’s give the CAR. Time period [-2,4] [-2,0] [1,4] CAR: -0.93% 0.04% -0.73%

11http://www.npr.org/sections/parallels/2017/03/16/520363987/in-dutch-vote-first-of-3-key-european-elections-populism-takes- second-place

The European Financial Markets are influenced by the Dutch parliamentary election of 2010, table 9. The impact of the election has led to positive stock market returns on the French, Spanish and Italian stock market, but negative on the German and UK stock market. The French stock market showed the largest abnormal return. When the abnormal returns are combined in table 10, the election has a positive impact on stock market returns with average abnormal return of 0.19%. However, none of the shown data are significant to reject hypothesis 2: “Political elections have a positive spill-over effect on foreign stock markets”. But despite the lack of significance, the data shows the direction, which is in line with Pantzalis et al. (2000). This result is also strengthened by the cumulative average abnormal return (CAAR), as shown in table 9a. As Pantzalis et al. (2000) expects, the greatest degree of uncertainty resolution is expected starting immediately preceding the election day due to extensive media coverage. As seen from table 9a, this is the case. The uncertain information hypothesis by Brown et al. (1988) is valid. As the uncertainty surrounding the event is reduced, this will consequently positively affect stock market returns, regardless of the event in dispute as can be seen from table 9a. These findings are also strengthened by the results shown in table 10. The average abnormal returns (AAR) of the international sample are positive, but showed no significant increase. During the whole event-window, the parliamentary election showed a positive impact. But the data is in line with the findings described above.

Table 9a: Cumulative Average Abnormal Return (CAAR) over the event window. The cumulative average abnormal stock return (CAAR) is calculated using the next method: First I calculate the abnormal stock returns (AR’s) per country and then I take the average of these individual AR’s. Lastly all the individual AAR’s are added to form the CAAR. Time period [-2,4] [-2,0] [1,4] CAAR: 0.29% 0.90% -0.42%

Dutch parliamentary elections: conclusion

In order to investigate hypothesis 3, the before, between and after Financial Crisis elections described above are analysed. This research expects that political elections are negatively correlated with economic performance. This means that after years of economic growth the election result is less significant on stock market returns than after years of economic down turn. Moreover, during economic down turns, the sentiment of voter’s changes and will have a punishment effect on future elections. This is in line with Anderson (2000).

In his research, the results reveal that political context interacts with economic perceptions to affect voting behavior. When the institutional context clarifies who is in charge of policymaking and when the target of credit and blame is large, economic effects are stronger. Conversely, when clarity of responsibility is obscured and when the level of responsibility is low, governing parties are less affected by how citizens evaluate the nation’s economy. Stevenson (2001) agrees as people want policy to move left when the economy is expanding and right when contracting. The empirical analysis in this article generalizes the empirical finding to the larger set of western democracies and holds in countries with different cultures, institutions and historical experiences. Moreover, an indirect path of voting is found. The economic situation leads to preferences (left or right) which leads to particular voting behaviour. Taken together, these findings explain that voting behaviour based on economic performance is enhanced when accountability is simple. If we link these results to the results found by this research we can see positive cumulative abnormal return over the whole event window in the period before the Financial Crisis. Next, like Pantzalis et al. (2000), positive cumulative abnormal returns are measured in the period before and during the financial crisis period in the two weeks preceding the election. However, these values are not significant. The only value that is significant is the two weeks preceding the parliamentary election in the after financial crisis period. Despite the other values been positive, in the after financial crisis period the stock market returns are negatively correlated to the election outcome. The impact on stock market returns over the entire event window is less responsive before the financial crisis than it is afterwards. Even despite the lack of significance to reject hypothesis 3, this shows a pattern. Over the entire event-window, the magnitude of the election result is negatively correlated with economic performance. Moreover, the stock market returns react more negative moving towards the end of the Financial Crisis. This is in line with Stevenson (2001).

Table 11: Impact of the Dutch parliamentary elections on stock markets return during the event window. The cumulative average abnormal stock return (CAAR) is calculated using the next method: First I calculate the abnormal stock returns (AR’s) per country and then I take the average of these individual AR’s. Lastly all the individual AAR’s are added to form the CAAR. (*, Significant values are retrieved from table A: significant table on page 18.) Time period [-2,4] [-2,0] [1,4] Before -1.74% -1.48% -0.21% During -0.70% 2.73% -3.26% After 0.29% 0.90% -0.42%

4.3.2. France

The French Governmental system is based on the constitution of the Fifth Republic set out by General Charles de Gaulle since 195812. This constitution increased the power at the expense of the national assembly. In the new constitution, the accountability of the government is to the Parliament, which is made up of two chambers. Besides, the president of the Republic is accountable to the people. In 1962, a constitutional amendment was made that enabled direct popular elections of the president. By 1965, Charles de Gaulle became the first elected president by popular vote since 1848. Assigned by the president is the Prime Minister, which is the head of the government and forms the government. In the end, the president must approve the appointed ministers to form the national government. The Prime Minister oversees all actions by the ministers and ensures that political guidelines set out by the government are implemented. The most important elections in France are the presidential elections. The presidential elections are based upon two rounds. Eleven candidates, each backed by at least 500 mayors, MPs, MEPs or senators, have qualified for the first round13. Whoever reaches more than 50% of the votes wins election. However, this is rarely the case in the first round. In that case, the two candidates with the highest percentage of votes face each other two weeks later in a second and final round. The candidate with the highest percentage of votes wins.

Before

The French presidential election before the Financial Crisis were held on April 22, 200714. After the first round of the election, none of the candidates achieved more than 50% of the votes. Therefore, the top two contenders, Nicolas Sarkozy of the ruling Union for Popular Movement Party and Ségolène Royal leading the Socialist party, were sent to the second and final round. The second round was held of May 6, 2007. In this second round, Sarkozy defeated Royal by 53% versus 47% of the votes. Interestingly was the near-record 84% turn-out ratio for this presidential election. Fueled by voter’s fear of economic decline at home and diminishing influence of foreign affairs has led to a massive turnout. Sarkozy has understood these deeply rooted national frustrations and planned to reform the national economy through modernizations15.

13https://www.theguardian.com/world/2017/apr/04/french-elections-all-you-need-to-know 14http://www.euronews.com/2017/04/21/a-history-of-french-presidential-elections-through-nine-infographics 15http://www.washingtonpost.com/wpdyn/content/article/2007/04/13/AR2007041301401.html

Given the results shown in table 12, the first-round election shows a positive impact on the national level with a 1.33% increase in abnormal return. This value is very significant and provides strong evidence that hypothesis 1 is failed to be rejected. Besides, a very significant positive abnormal return was measured the day preceding the presidential election. The uncertain information hypothesis (UIH) formulated by Brown et al. (1998) clarifies this finding. As the election result becomes more certain moving more towards the election day, the election-induced uncertainty is reduced. Risk-adjusted expected returns fall which increases the stock market returns. The second round of the French presidential election shows a positive return as well. However, in contrast to the first round this value is not significant. As with the first round, the second-round findings are explained by the UIH. As described, a reduction of uncertainty will lead to positive stock market returns. The greatest degree of uncertainty reductions is associated in the first round of the presidential election as the number of candidates is the greatest. Based on the results of the first round, the election- induced uncertainty in the second round is reduced less as the election result is more certain than in the first round. This yields lower observed returns. As found by the research of Pantzalis et al. (2000), positive cumulative abnormal return (CAR) are measured in the two weeks ending with an event-induced uncertainty. According the UIH described above, the closer towards the election the more the election-induced uncertainty is reduced. This reduces risk-adjusted expected returns and increases stock prices. Henceforth, the greatest degree of uncertainty resolution and thus the highest observed returns are expected in the time period immediately preceding the election day as this is when media coverage and campaigning are at their peak. The CAR at [-2,0] justifies this in table 12a. Besides, as some uncertainty has been resolved in the week preceding the election, it is expected that the cumulative abnormal returns (CARs) remain positive in the time period following the election week. The positive CAR in the period [1,4] in table 12a shows this. Based on the finding in tables 12 and 12a, hypothesis 1 is failed to be rejected for the French Presidential election of 2007.

Table 12a: Cumulative Abnormal Return (CAR) over the event window. The cumulative abnormal stock return (CAR) is calculated using the next method: First I calculate the abnormal stock returns (AR’s) and then taking the sum of these individual AR’s give the CAR. Time period [-2,4] [-2,0] [1,4] CAR: 2.03% 2.73% 0.63%

If we extend our view to the international level, we see that the French Presidential Election of 2007 has led to positive abnormal stock market returns in our sample selection. However, only the positive abnormal stock market returns of the UK and Italy are significant. If we combine the results of table 13, the French presidential election have led to a positive impact on the international financial markets. This result is also strengthened by the Cumulative Average Abnormal Return (CAAR), as shown in table 13a. Like Pantzalis et al. (2000) expects, this research expects the greatest degree of uncertainty resolution immediately preceding the election day due to extensive media coverage. As seen from table 13a, this is not entirely the case. Despite the uncertain information hypothesis (UIH) formulated by Brown et al. (1988), most the uncertainty is resolved after the actual election day resulting in a positive cumulative abnormal return of 1.83%. However, despite the level of significance, a positive effect of the election outcome on stock market returns is measured, which is in line with Pantzalis et al. (2000). Table 14 shows the Average Abnormal Returns (AAR) for the international sample. The data is in line with the findings described above. Therefore, we fail to reject hypothesis 2.

Table 13a: Cumulative Average Abnormal Return (CAAR) over the event window. The cumulative average abnormal stock return (CAAR) is calculated using the next method: First I calculate the abnormal stock returns (AR’s) per country and then I take the average of these individual AR’s. Lastly all the individual AAR’s are added to form the CAAR. Time period [-2,4] [-2,0] [1,4] CAAR: 2.65% 1.33% 1.83%

During

The French presidential election during the financial crisis is held on April 22, 201216. After a close race between Sarkozy, Hollande and Le Pen, the first round was won by Francois Hollande with 28.6% of all votes. Despite the loss by Le Pen, it was a record-high score of 17.9% for a far-right candidate in the first round of the presidential election. Since no absolute majority was gained during the first round, a second round was necessary with a straight runoff between the incumbent president Sarkozy and runner up Hollande. Sarkozy, former head of the conservative UMP party, is the candidate of the right. During his period as president, France was strongly affected by the Financial Crisis.

16http://about-france.com/presidential-election.htm

Despite the unpopular reforms, he was stressing his success of the necessary reforms. Supported on the international level by German chancellor Angela Merkel and Barack Obama, Sarkozy made France competing on the global economy. Francois Hollande was the candidate from the Socialist party. Characterized by his party as soft, Hollande was determined to reduce the national budget deficit, creating more public jobs and raising the highest tax bracket to 75%. During the second round17, Francois Hollande won the power in France with 51.9% of the votes. A turning point of the lurch in European politics and German-led austerity measures. It was the first win for a left-wing president in almost 20 years.

Based on the results in table 15, the French presidential election of 2012 negatively impacted the national small cap index. Both during the first and second round of the election. Despite the negative impact of the election result on the national stock market returns, the measured result is not significant. However, significant abnormal returns were measured preceding the election event. Like Pantzalis et al. (2000), the greatest degree of uncertainty resolution is expected preceding the election day due to extensive media coverage. Moreover, a significant negative impact was measured the day after the final round. This could be due to the measuring standard used in this research. As this research is using open-to-open market returns, Like Chan et al. (1996), the election results are presented after the election trading day. But also, discussed by Akitoby and Stratmann (2007), financial markets give a premium to right wing regimes while penalizing left wing regimes that undertake current-spending- driven expansion. The fear of left-wing administrations resulted in negative abnormal returns the day(s) following the election. This is also in line with the Mukherjee and Leblang (2005), rational expectations of higher inflation under left-wing administrations lowers the volume of stocks traded in the stock market. The decline in trading volume leads to a decrease in the mean and volatility of stock prices not only during the incumbency of left-wing governments, but also when traders expect the left-wing party to win elections. The results of table 15a strengthen the findings described above as the election-induced stock market returns are negative in all time periods. Moreover, the stock market returns are less negative following the election when the outcome is certain. The UIH by Brown et al. (1988) is partly applicable as the election-induced uncertainty positively affects stock market returns. However, despite the negative impact of the presidential election of 2012 on the French stock market returns, none of the values are significant, hence we fail to reject hypothesis 1.

Table 15a: Cumulative Abnormal Return (CAR) over the event window. The cumulative abnormal stock return (CAR) is calculated using the next method: First I calculate the abnormal stock returns (AR’s) and then taking the sum of these individual AR’s give the CAR. Time period [-2,4] [-2,0] [1,4] CAR: -4.73% -3.46% -1.52%

The impact of the election on the foreign stock markets are shown in table 16. Interesting values are measured here. According to the uncertainty information hypothesis (UIH) formulated by Brown et al. (1988), uncertainty reduction is associated with positive observed stock market returns. As the election outcome becomes more certain, the financial market investors revise their probability distributions of policies to be implemented. However, a distinction has to be made between performance driven and policy driven elections, Harrington (1993). Policy driven elections are associated with a large amount of uncertainty resolution as the outcome becomes more certain. But elections driven by performance are not associated with this uncertainty reduction as the outcome becomes clear, since not much uncertainty regarding the effects of policy changes is resolved. Given the results based on table 16, the first round of the presidential election negatively affected the international stock market returns. Despite the lack of significance, all countries from the data sample showed a negative impact. During the second round of the election the stock market reacted negatively on the win of Hollande with strong significance in Spain and Italy. Interestingly is the significant positive affect on the German stock market, as Hollande is not well known in the international scope and wants to transform Europe's handling of the economic crisis by fighting back against German-led austerity measures17. Moreover, investor expectations of left-wing governments lower the trading volume in the financial markets. Besides, the fear of increased levels of inflation decreases the mean and volatility of stock market returns as Mukherjee (2005) discussed. However, as the official results become more certain, the German market reacts negatively during the trading days following the second election round. These findings described above are in line with the Cumulative Average Abnormal Return (CAAR) showed in table 16a. Like Pantzalis et al. (2000), this research expects positive stock market returns in the weeks preceding the election given the uncertainty information hypothesis (UIH). Despite the negative impact measured in the two-week preceding the election another underlying proposition of the UIH is valid.

17https://www.theguardian.com/world/2012/may/06/francois-hollande-wins-french-election

Price changes will be positive (nonnegative) as uncertainty is resolved around unfavorable (favorable) events, which is measured in the period [1,4] following the election results. The UIH is therefore partially applicable here. Based on the findings derived from tables 16, 16a and 17, only the Italian and Spanish stock markets provide strong evidence to reject hypothesis 2, but the positive stock market returns measured on the German index provides evidence that fails to reject hypothesis 2. However, combining the international abnormal stock market returns fails to reject hypothesis 2, based on table 17. Despite the lack of evidence to reject hypothesis 2, the results provide a strong signal of a negative impact of the French presidential election of 2012.

Table 16a: Cumulative Average Abnormal Return (CAAR) over the event window. The cumulative average abnormal stock return (CAAR) is calculated using the next method: First I calculate the abnormal stock returns (AR’s) per country and then I take the average of these individual AR’s. Lastly all the individual AAR’s are added to form the CAAR. Time period [-2,4] [-2,0] [1,4] CAAR: -0.40% -1.53% 0.49%

After

With the Dutch parliamentary election as general rehearsal, the French presidential election of 2017 was the second important election of 2017. The fear of the right-wing is no false alarm as Brexit surprised the international financial markets. To many, the populist surge is a response to the political failures of the incumbent political parties. Besides, it is also an emotional backlash to the European Union’s compromise machine between the center-left and the center-right that routinely ignores opposing voices18. With the lingering Euro crisis, the European sentiment of the voters changed and had a punishing effect on the future elections, Anderson (2000). A lower growth rate increases the support for extreme political platforms, Brückner et al. (2010). This punishment effect resulted in enormous percentage of votes for Le Pen as candidate for the nationalist party. Though, Macron won the first round by a small difference. However, result of the first round made clear the huge fragmentation of sentiment within France. The country known for its unity was divided more than ever. This fragmentation continued in the plans proposed by the candidates. Le Pen wants to abandon the Eurozone, suspend legal immigration and wants to lower the retirement age to 60.

18https://www.foreignaffairs.com/articles/europe/2016-06-03/rise-populism-europe

Macron wants to strengthen the European bands, cut corporate taxations to 25% and a public investment plan to renewable energy. Backed by the socialists and republicans, Macron won the election with his political party “En Marche” (On the Move) with 66.1% of the votes and became the youngest head of state with 39 years of age19. His clear win has been hailed by European leaders as it became a sincere sign of unity with the European Union20.

The win by Emmanuel Macron positively affected the French financial markets as can be seen from the event days. The markets responded very positive after the first round of the election and moderate positive after the second round of the presidential election. As pointed out by Pantzalis et al. (2000), the greatest degree of uncertainty resolution and hence the highest stock market returns are expected preceding the election. This is also find by this research with a positive cumulative abnormal return of 3.54% preceding the event, table 18a. The win of Macron is a relief for the financial markets. The consequences of a win by Le Pen are enormous. Global trade will be lower due to extra trade barriers. Production costs will rise as specialization and economies of scale are erased by these barriers. In the end, a more stagflationary economy is expected21. Even though the exit polls were pointing Emmanuel Macron as winner in the second round, some uncertainty regarding the outcome was in the market. The official result came nonetheless as a relief and resulted in a positive impact on the national level. Based on the findings described above, we fail to reject hypothesis 1.

Table 18a: Cumulative Abnormal Return (CAR) over the event window. The cumulative abnormal stock return (CAR) is calculated using the next method: First I calculate the abnormal stock returns (AR’s) and then taking the sum of these individual AR’s give the CAR. Time period [-2,4] [-2,0] [1,4] CAR: -0.54% 3.54% -1.07%

The findings of the national level can be extended to the international financial markets. As can be seen from table 19, all the countries of the data sample reacted positively on the election results after the first round. Moreover, the uncertainty information hypothesis (UIH) formulated by Brown et al. (1998) is valid given the cumulative abnormal return [-2,0] shown in table 19a. Focussing on the second round, only the Netherlands and UK reacted significantly positive after the final round.

19http://www.bbc.com/news/world-europe-38220690 20http://www.telegraph.co.uk/news/2017/05/08/eu-leaders-hail-emmanuel-macrons-emphatic-election-triumph-landslide/ 21http://www.schroders.com/en/us/institutional/insights/economic-views3/is-populism-good-for-markets

A pattern can be seen as the Northern European countries, represented by the Netherlands, Germany and the UK, reacted significantly more positively by the election results than the Southern European countries represented by Italy and Spain. This could be explained by the fight of the Southern European countries, Greece, Italy, Portugal and Spain (GIPS), on the austerity policies set in place by the European Union. The anti-Europe movement is more pronounced in southern Europe as a result of the large reformations necessary to improve the economic stability of southern Europe. These measurements primarily reform wages, social services and public ownership. This is in line with the findings by Stevenson (2001). People want the national policy to move left when the economy is expanding and right when the economy is contracting. Anderson (2000) found the same. Voters’ ability to express discontent based on economic performance is enhanced when accountability is simple. Overall, the presidential election resulted in positive stock market returns based on the findings of tables 19, 19a and 20. Therefore, this research fails to reject hypothesis 2.

Table 19a: Cumulative Average Abnormal Return (CAAR) over the event window. The cumulative average abnormal stock return (CAAR) is calculated using the next method: First I calculate the abnormal stock returns (AR’s) per country and then I take the average of these individual AR’s. Lastly all the individual AAR’s are added to form the CAAR. Time period [-2,4] [-2,0] [1,4] CAAR: 1.35% 2.62%* 1.08%

French presidential elections: conclusion

This research expects that during times of growth, the impact of elections on stock market returns is less severe than during times of contractions. Alesina et al. (1996) explains this with help of the economic business cycle. During economic down turns, the sentiment of the voters is negatively affected by the state of the economy and will have a punishment effect on the incumbent parties during future elections. Moreover, as Anderson (2000) found in his research, this effect is stronger when it is clear for the voters who to blame. The more sizeable the target, to stronger voters’ ability is to express discontent to the incumbent parties. The opposite is valid as well, governmental parties are less affected by a negative sentimental change of voters’ behaviour if the clarity of responsibility is low. Normally, people tend to vote “left” during expansions and “right” during times of contraction, Stevenson (2001).

A pattern is then found between economic performance and voting behaviour. The same is found by Harrington (1993). Voting behaviour is affected by both policy and performance of the incumbent government. If people are less able to differentiate between policies, a higher chance peoples voting behaviour is purely performance driven rather than policy driven. If we link these results to the results found by this research we can see positive cumulative average abnormal returns over the whole event window in the period before and after the Financial Crisis, see table 21. Moreover, like Pantzalis et al. (2000), significant positive cumulative average abnormal returns are measured in the period before and during the financial crisis period in the two weeks preceding the elections. The uncertainty information hypothesis (UIH) formulated by Brown et al. (1998) is valid. Interestingly is the positive CAAR measured the weeks following the event. In contrast to the Netherlands the French presidential elections have a positive effect on stock market returns after the election-event. One reason could be the second election round. Despite some uncertainty reduction happened in the first round, the outcome of the second round is not clear. So, some uncertainty reduction is happening around the second round, which positively affect stock market returns as the risk-adjusted expected returns fall. Based on the UIH, the magnitude of the French presidential elections increases with time in the two weeks preceding the event. Also, the magnitude of the presidential elections on the French Small Cap index significantly increased with time, tables 12,15 & 18. Based on these findings, this research fails to reject hypothesis 3. The magnitude of the French elections is negatively correlated with time.

Table 21: Impact of the French presidential elections on stock markets return during the event window. The cumulative average abnormal stock return (CAAR) is calculated using the next method: First I calculate the abnormal stock returns (AR’s) per country and then I take the average of these individual AR’s. Lastly all the individual AAR’s are added to form the CAAR. (*, Significant values are retrieved from table A: significant table on page 18.) Time period [-2,4] [-2,0] [1,4] Before 2.65% 1.33%* 1.83% During -0.40% -1.53% 0.49% After 1.35% 2.62%* 1.08%

4.3.3. Germany

The German federal elections are different to many other countries. As Germany is divided in 299 electoral districts, voters have two ballots in these federal elections22. With the first vote, the voter chooses his candidate. This candidate is chosen by his or her party. However, an independent candidate can also run for candidate after gaining at least 200 signatures. After choosing the candidate with the first vote, half of the 595 seats of Bundestag are filled. The second vote goes to the political party. With this last vote, the rest of the remaining seats are filled based on a percentage based division. Based on rakings, the voter cannot be sure which individuals further down on the party list will make it into the Bundestag. States with a larger population get to send more parliamentarians to the Bundestag than smaller ones. Only the top few names are sent to the parliament. As a measure of stability, the German election law states that political parties must obtain at least five percent of the proportional vote in order to be represent in the parliament. This prevents splintering parties, which makes decision making extremely difficult. Moreover, it also prevents extreme left and right wing political parties from entering of the parliament.

Before

The German federal election of 2005 was special for a couple of reasons. Angela Merkel became the first female chancellor in history. Normally, federal elections are happening every four years. However, Gerhard Schröder, the incumbent chancellor of the Social Democratic Party dissolved the parliament in 2005. The main reason for the early held election was the result of the region election in Northrhein-Westphalia23. Since, 1966 it had been government by the Social Democrats. This ended abruptly by the win of the Christian Democratic Union putting considerable pressure on Schröder to reform his agenda. After the loss in the state selection, Schröder asked his followers to withhold a motion of confidence in the Bundestag in order that it would fail and thus trigger an early federal election24. As it happened, an early election was held on September 18, 2005. After starting the campaign with a 21% lead over SPD, the election was ultimately won with only 1% difference favoring CDU with Angela Merkel becoming chancellor. On October 10, 2005, the new government was formed between CDU/CSU and SPD.

22http://www.dw.com/en/how-do-germans-elect-their-parliament/a-17018116 23https://www.sussex.ac.uk/webteam/gateway/file.php?name=epern-election-briefing-no-23.pdf&site=266 24https://www.revolvy.com/main/index.php?s=German%20federal%20election,%202005&item_type=topic

As can be seen from table 22, the German federal election had no significant effect on the national level. Even though the stock market returns were positively affected by the election outcome, the values lack significance. The same findings are presented in table 22a. The cumulative abnormal returns are positive in the two-week period preceding the election event. Even though the data is not significant, the direction of the change is in line with Pantzalis et al. (2000). The uncertainty information hypothesis (UIH) formulated by Brown et al. (1998) states that the greatest degree of uncertainty reduction is in the two-week period preceding the event as media coverage is at its peak. Moreover, the change of governmental structure is in line with Stevenson (2001). People want policy to move left when the economy is expanding and right when the economy is contracting. After the tech bubble, financial markets were recovering and the economy was expanding until the Financial Crisis of 2008 started. Based on the findings described above, the findings are in line with academic research, but this research fails to reject hypothesis 1 as the measured data lacks significance.

Table 22a: Cumulative Abnormal Return (CAR) over the event window. The cumulative abnormal stock return (CAR) is calculated using the next method: First I calculate the abnormal stock returns (AR’s) and then taking the sum of these individual AR’s give the CAR. Time period [-2,4] [-2,0] [1,4] CAR: -0.26% 0.78% -0.60%

Expanding the scope from the national to the international financial stock markets, the stock market returns measured on the international financial markets show the same pattern as table 23 and 24 exhibit. The election outcome resulted in no significant impact. Very significant values are measured at t=-3. A reason could be financial news regarding the fear of stagflation as inflationary level exceeds well above national targets. Together with a potential temporarily halt in interest rate hikes by the FED25, financial investors are alerted. Analysing the cumulative average abnormal return (CAAR) for the international markets suggest a positive influence of the German federal election on international stock market returns. Despite the lack of significance, this is in line with the findings by Pantzalis et al. (2000). Moreover, the data of table 23a shows that the degree of uncertainty reduction is the greatest following the election event. The uncertainty information hypothesis by Brown et al. (1998) is therefore partially applicable here. Like Pantzalis expectations, one reason could be that investors need time to assess the election impacts following election outcome.

25http://www.telegraph.co.uk/finance/2922163/Falling-food-prices-peg-US-inflation.html

The financial markets could only partially resolve prior uncertainty regarding the election event. If there is a significant amount of uncertainty resolution following the election date, we would expect to observe post-election positive abnormal returns. As can be seen from table 23a, this is what this research finds. However, as Pantzalis et al. (2000), the results lack levels of significance. Based on the derived findings, we fail to reject hypothesis 2.

Table 23a: Cumulative Average Abnormal Return (CAAR) over the event window. The cumulative average abnormal stock return (CAAR) is calculated using the next method: First I calculate the abnormal stock returns (AR’s) per country and then I take the average of these individual AR’s. Lastly all the individual AAR’s are added to form the CAAR. Time period [-2,4] [-2,0] [1,4] CAAR: 1.71% 0.27% 1.48%

During

The federal election “during” the financial crisis was held on September 27, 2009. Angela Merkel was re-elected as Chancellor for a second term26. Despite the lingering financial crisis, the party led by Angela Merkel, Christian Democratic Union (CDU), was down by only 1.4 percentage points relative to the 2005 federal election. The biggest loss was incurred by incumbent left-wing party SPD, which was down by 11.1 percentage points to a historical low of 23.1. The biggest winner of the federal election was the Free Democratic Party (FDP) that gained 10 percentage points. Together with Merkels’ CDU a coalition of centre-right is formed.

According to Stevenson (2001), the election result is not a surprise. People tend to vote right during economic times of contraction and left when the economy is expanding. Especially when voting behaviour is affected by economic perceptions and thus priced in, Anderson (2000) and Lee et al. (2002). As pointed out by Mukherjee et al. (2005), Douglas and Hibbs (1977), Stevenson (2001), investors are best served by right-wing administrations. Therefore, positive stock market returns are expected on the national level. Especially, since the FDP is expected to strive for business-friendly measures like tax cuts and relaxing rules that protect employees from dismissal. Financial markets give premiums to right-wing administrations while penalizing left-wing administrations, Akitoby and Stratmann (2007). Douglas and Hibbs (1997) found a historical pattern between inflation/unemployment level trade-off between left and right-wing parties.

26http://www.spiegel.de/international/germany/germany-has-voted-merkel-wins-german-election-has-majority-for-center-right- government-a-651610.html

Low rates of inflation have prevailed in countries ruled by centre to right governments, whereas low levels of unemployment have prevailed in governments controlled by left-wing parties. Moreover, the expectation of lower inflation levels under right wing administrations increases the mean and volatility levels of stocks, which will generate higher stock market returns, as found by Mukherjee et al. (2005). The election outcome produced no significant stock market returns on the German small cap index, SDAX. However, a positive effect was measured during the entire event window. Moreover, the uncertainty information hypothesis formulated by Brown et al. (1998) is applicable here. This research expects positive stock market returns in the weeks preceding the election event like Pantzalis et al. (2000). The greatest uncertainty reduction and hence the greatest stock market returns are expected before the event as the media coverage and campaigning are at its peak. Given the results in table 25a, this indeed what this research finds. Based on the findings described above, we fail to reject hypothesis 1. The German federal election of 2009 positively affected the German small cap index returns.

Table 25a: Cumulative Abnormal Return (CAR) over the event window. The cumulative abnormal stock return (CAR) is calculated using the next method: First I calculate the abnormal stock returns (AR’s) and then taking the sum of these individual AR’s give the CAR. Time period [-2,4] [-2,0] [1,4] CAR: 2.54% 0.79% 1.43%

Expanding the scope towards the international stock markets, it is expected that the election result is positively received by the financial markets. Especially the continuation of the governmental policies raises growth figures as pointed out by Alesina et al. (1992). Economic growth rates are positively affected by low propensity of economic collapse. Moreover, investors favour the expectations of right-wing administrations while disfavour the possibility of left-wing administrations, Mukherjee et al. (2005), Douglas and Hibbs (1977), Stevenson (2001). The results exhibit in table 26a, show that this is the case. The federal election positively affect abnormal stock market returns over the entire event-window. However, the value is not significant to answer hypothesis 2. Despite the lack of significance, it provides a direction and magnitude of the impact of the election outcome. The strongest impact was measured at the Dutch financial market.

This is no surprise as the Dutch economy relies heavily on Germany. More the 20% of total exports is exported to Germany, making its most important export country27. The same rule is found for France. In contrast to the national level, the UIH does not prevail on the international financial markets given the negative stock market returns in the two weeks preceding the election event. The greatest degree of election-induced uncertainty reduction is expected in the weeks preceding the election due to extensive media coverage and campaigning, Pantzalis et al. (2000). However, the negative results imply that uncertainty resolution has partially occurred, especially given the positive stock market returns measured in the weeks following the election. As Pantzalis et al. (2000), investors sometimes need more time in order to resolve election-induced uncertainty, hence positive abnormal returns are expected after the election event. As the results of table 26a exhibit, this is indeed what this research finds. Despite the lack of significance, the values provide evidence of a positive direction and change of the election outcome on international stock market returns.

Table 26a: Cumulative Average Abnormal Return (CAAR) over the event window. The cumulative average abnormal stock return (CAAR) is calculated using the next method: First I calculate the abnormal stock returns (AR’s) per country and then I take the average of these individual AR’s. Lastly all the individual AAR’s are added to form the CAAR. Time period [-2,4] [-2,0] [1,4] CAAR: 0.12% -0.67% 0.87%

German federal elections: conclusion

The analysis of German federal elections on stock market returns is limited to before and during the financial crisis. New federal elections are planned on September 24, 2017. This research will use exit polls and the election results of the Netherlands and France to evaluate the performance of stock market return overtime. The exit polls exhibit a coalition of CDU/CSU and FDP that will remain a majority in the German Bundestag28. Both the Dutch and the French general election positively affected the stock market returns. The election outcome of the Netherlands was seen as a leading indicator for the French and German elections. The increase of far-right parties is a signal which should not be underestimated. The migration and the lingering euro crisis results a response by populist parties to the apparent political failures by the incumbent governments.

27http://www.worldsrichestcountries.com/top-dutch-exports.html 28http://www.express.co.uk/news/politics/824007/German-election-2017-who-will-win-latest-polls-Angela-Merkel-Martin-Schulz- Germany

Despite, as the Dutch and French elections have exhibits, the far-right parties have been tackled by the more “conservative” parties. Moreover, as a measure of stability, the German parties need at least five percent of the proportional vote in order to represent the voters in the parliament in contrast to the Netherland. Based on the polls and the election outcome of the Netherlands and France early 2017, it is expected that the German elections will result in positive stock market returns. Moreover, this research expects a conservative win. This is line with Stevenson (2001), people tend to vote right during economic expansions and tend to vote left during economic contractions. As the German average growth rate is 1.6% over a five-year period starting in 2011, the Germany economy has been growing significantly. Moreover, the unemployment rate has been driven downwards during the same period from 7% till 6.4% in 2015. Anderson (2000) found that the political context interacts with economic perceptions that will affect voting behaviour. Especially when the target of blame is clear, the ability of voters to express discontent based on the economic performance is enhanced. As the economy generates positive values, it is expected that the voters of Germany will choose a centre to right wing coalition29. However, given the results in table 28, hypothesis 3 is failed to be rejected. The magnitude of the election outcome was stronger before the financial crisis than during the financial crisis. This is against this research expectations that the election outcome will increase the magnitude of stock market returns over time. However, the results shown in table 28 demonstrate a positive effect of federal elections on stock market returns.

Table 28: Impact of the German federal elections on stock markets return during the event window. The cumulative average abnormal stock return (CAAR) is calculated using the next method: First I calculate the abnormal stock returns (AR’s) per country and then I take the average of these individual

AR’s. Lastly all the individual AAR’s are added to form the CAAR. (*, Significant values are retrieved from table A: significant table on page 18.) Time period [-2,4] [-2,0] [1,4] Before 1.71% 0.27% 1.48% During 0.12% -0.67% 0.87% After N/A N/A N/A

29http://www.focus-economics.com/countries/germany

4.3.4. United Kingdom

In the United Kingdom, new general elections can only happen after prorogation. A speech is written on behalf of the queen to formally end the session of the parliament. Even though the old government has been dissolved, the government departments will continue their routine tasks until a new government has been formed. During the general elections, each constituency lists candidates which will represent the local area in the House of Commons30. The candidate with the most votes becomes their MP. Usually the party with the most number of seats in the House of Commons is responsible of the formation of a new government31. The leader of this party becomes the Prime Minister of the country but the queen holds absolute power in appointing the Prime Minister guided by constitutional conventions. The Prime Minister is responsible for assigning ministers to the various governmental departments. After putting the ministers in place, the new government is formed and will have a term for the next five years as set by the Fixed-term Parliaments Act 2011.

Before

In the general election on May 5, 2005, Labour won 55.1% of the seats in the House of Commons with 35.2% of the public votes32. It became the thinnest majority in the government on public support. During the election, Conservatives focussed on the following key subjects: “crime, tax, immigrations, healthcare and clean hospitals”. Its biggest rival Labour focussed more on the economy, healthcare and education. With the phrase “Vision for a third term”, Labour continued having a majority of the seats in the House of Commons. Tony Blair, first candidate of Labour, remained Prime minister and the newly formed government was made official over the weekend with an official announcement on May 9, 2005.

The election outcome positively affects the British FTSE Small Cap index. This is against the findings Mukherjee et al. (2005). The anticipation of left-wing orientated administrations lowers trading volume. As a result, the decrease in trading volume decreases both the mean and volatility of stock prices. Douglas and Hibbs (1977) strengthen this finding. Left-wing orientated parties attach greater importance to full employment than to inflation. Moreover, the unemployment rate appeared to be driven downwards during Labour governments and driven upwards during Conservative governments. However, as can be seen from table 29 and table 29a, this research found positive stock market returns in response of a Labour victory. 30http://www.parliament.uk/about/how/elections-and-voting/general/ 31http://www.parliament.uk/about/how/role/parliament-government/ 32https://www.electoral-reform.org.uk/wp-content/uploads/2017/06/2005-UK-general-election.pdf

The uncertainty information hypothesis by Brown et al. (1998) is valid as the CAR during the period preceding the election event is significant positively affected. Additionally, the highest observed returns are expected during this period as media coverage and campaigning are at its peak. This is strengthened by the very strong return at t=-1, table 30. As campaigning is at its maximum the day before the election, the greatest degree of uncertainty reduction is measured here. Based on the findings described above, we fail to reject hypothesis 1.

Table 29a: Cumulative Abnormal Return (CAR) over the event window. The cumulative abnormal stock return (CAR) is calculated using the next method: First I calculate the abnormal stock returns (AR’s) and then taking the sum of these individual AR’s give the CAR. Time period [-2,4] [-2,0] [1,4] CAR: 0.87% 0.55%* 0.78%

Expanding the scope towards the international financial markets, no significant impact is measured on the stock market returns, as can be seen from tables 30, 30a and 31. During the entire event-window the stock market returns reacted negatively on the election results. This is in line with the findings of Mukherjee et al. (2005). As left-wing administrations focus more on low unemployment figures and less on inflation levels, Douglas and Hibbs (1977), trading volume decreases and results in a diminishing rate of the mean of volatility figures of stock market returns. In contrast to the national level, the uncertainty information hypothesis, Brown et al. (1998), does not hold. Taken together, the political election do not positively affect foreign markets. However, due to the lack of significance we fail to reject hypothesis 2.

Table 30a: Cumulative Average Abnormal Return (CAAR) over the event window. The cumulative average abnormal stock return (CAAR) is calculated using the next method: First I calculate the abnormal stock returns (AR’s) per country and then I take the average of these individual AR’s. Lastly all the individual AAR’s are added to form the CAAR. Time period [-2,4] [-2,0] [1,4] CAAR: -1.81% -0.19% -1.72%

During

After a 13-year term of Labour, Conservatives took over ruling Downing Street. In the month preceding the election event the conservative gained a solid lead in the exit polls. However, there were doubts of the readiness of David Cameron as leader of the Conservative Party.

After the election result of 2005, leader of the Conservative party Michael Howard resigned and his leadership was taken over by David Cameron during a Conservative leadership contest33. The doubts about the readiness of Cameron was translated in a decline of the Conservative lead in the exit polls. A novelty was introduced during the 2010 election in the form of televised debates between the three main parties: Labour, Conservative and Liberal Democrats. By an outstanding first television performance the Liberal Democrats took the lead in the polls. However, after a second and third debate this lead was then taken over by the Conservatives. The Conservatives won the British general election of 2010 significantly, gaining 97 seats with respect to 200534. Labour incurred the biggest loss by losing 91 seats in the Houses of Commons. However, as no majority was secured as 323 seats are needed to govern, it became unclear who became prime minister. In the following days, negotiations between Cameron and Clegg began. Cameron was chosen as the new Prime Minister after the resignation of outgoing Prime Minister Brown on the 10th of May.

With a clear win of the Conservative party, it is expected that the markets react positively on the election result. The anticipation of left-wing orientated administrations reduces the mean and volatility of stock market returns as a result of falling trading volumes, Mukherjee et al. (2005). The fear of higher inflation expectations by investor during terms of left-wing administrations is grounded by Douglas and Hibbs (1977). They find that during the post-war period, high rates of inflation have occurred in nations governed by Labour parties. A trade-off between unemployment levels and inflation levels was seen where Labour was responsible for governing low unemployment rates and high rates of inflation. The opposite is found for Conservatives. As a result, financial markets give a premium to right wing regimes while penalizing left wing regimes that undertake current-spending-driven expansion, Akitoby and Stratmann (2007). Despite findings of previous academic research, this research finds the opposite, given the results in table 32a. However, this could be explained by the lack of a governmental majority. Straight after the election result, negotiations started between the Conservative party and the Liberal Democrats about who became Prime Minister. As t=3 from table 32 shows, the official announcement of David Cameron becoming Prime Minster as leader of the Conservative party is positively received by the market. But despite this positive effect, hypothesis 1 is rejected given the significant negative stock market returns following the outcome and the negative CAR during all event-windows shown in table 32a.

33http://news.bbc.co.uk/2/hi/uk_news/politics/4502652.stm 34https://www.britannica.com/topic/British-general-election-of-2010

Table 32a: Cumulative Abnormal Return (CAR) over the event window. The cumulative abnormal stock return (CAR) is calculated using the next method: First I calculate the abnormal stock returns (AR’s) and then taking the sum of these individual AR’s give the CAR. Time period [-2,4] [-2,0] [1,4] CAR: -6.01% -5.71% -1.13%

Expanding the scope to the international financial markets, no significant impact is measured on the election day. Despite the lack of significance, the markets showed a small positive abnormal return of 0.44%, table 34. However, as discussed on the national level, the win by the Conservative party is accompanied by an absence of a majority. Table 33 shows significant positive stock market returns at t=3, the day at which David Cameron was officially announced as Prime Minister. This is in line with Douglas and Hibbs (1997) and Mukherjee et al. (2005) who find a lower trading volume if left-wing administrations are expected to win. Moreover, investors give premiums to right wing administrations while penalizing left-wing administrations, Akitoby and Stratmann (2007). However, a difference between northern and southern European stock market returns is visible given the magnitude at t=3 and the cumulative abnormal return. A possible explanation are the austerity policies set in place in the southern European countries to improve economic stability within the Eurozone. This heavily affects the southern European countries and more anti-Europe voices are noticeable. As leader of the Conservative party, previous academic research by Douglas and Hibbs (1997), showed the priority of price stability over full employment during Conservative governments. Therefore, no change is expected for the austerity measurement in southern Europe. Following the results in table 34, the uncertain information hypothesis (UIH) formulated by Brown et al. (1998) is visible. Despite the negative CAAR measured preceding the election event, the uncertainty reduction positively affects stock market returns, Pantzalis et al. (2000). Given the results in the entire event-window and the period following the event, most of the uncertainty has been resolved in the weeks preceding the election event by extensive media coverage like the television debates. Based on the findings derived from tables 33, 33a and 34, this research fails to reject hypothesis 2 as no significant values are measured that show a negative effect of the British general election on foreign financial markets.

Table 33a: Cumulative Average Abnormal Return (CAAR) over the event window. The cumulative average abnormal stock return (CAAR) is calculated using the next method: First I calculate the abnormal stock returns (AR’s) per country and then I take the average of these individual AR’s. Lastly all the individual AAR’s are added to form the CAAR. Time period [-2,4] [-2,0] [1,4] CAAR: -1.81% -0.19% -1.72%

After

During the election of 2015, the Conservative party led by David Cameron had won with an outright majority of 331/650 seats35. A gain of 24 seats in contrast to the 2010 general election. In contrast to most of the exit polls, David Cameron returned to Downing Street. The biggest hit was incurred by Labour, who were completely wiped out in Scotland by the Scottish National Party (SNP). SNP was also the biggest winner of the 2015 election as it gained 50 seats over the 6 seats gained at the 2010 election. An aggregate feeling of remoteness by the Scottish people found expression through the SNP. Due to the immense loss by the Labour party, Ed Miliband was set to resign his function as leader of Labour36.

The election results had no significant impact on the national level. Analysing table 35, the election showed a marginal negative result. Even though the election surprise with Cameron re-elected as Prime Minister did not influence stock market returns. Combining the results with table 35a concludes that the British general election of 2015 negatively affect stock market returns during the event-window. In contrast to Pantzalis et al. (2000), no positive stock market returns are measured as a result of the uncertain information hypothesis formulated by Brown et al. (1998). Moreover, the measured results are against the findings of Stevenson (2001). People are expected to vote left during periods of expansion and right during periods of contraction. However, this could be explained by Harrington (1993). In his research, he finds that voting behaviour is based on incumbents’ policy and performance. The less voters are able to distinguish which policy is best, the more easily they will vote purely on performance. As the economy has gradually improved every year since 201037, voters have less alternatives to express discontent with economic performance, Anderson (2000). Based on the findings described above, we fail to reject hypothesis 1.

35http://www.bbc.com/news/election-2015-scotland-32635871 36http://www.telegraph.co.uk/news/general-election-2015/11588781/who-won.html 37https://ig.ft.com/sites/numbers/economies/uk?mhq5j=e3

Table 35a: Cumulative Abnormal Return (CAR) over the event window. The cumulative abnormal stock return (CAR) is calculated using the next method: First I calculate the abnormal stock returns (AR’s) and then taking the sum of these individual AR’s give the CAR. Time period [-2,4] [-2,0] [1,4] CAR: -5.44% -2.32% -3.17%

The foreign financial markets are significantly influenced by the British general election of 2015. As can be seen from table 36, the election positively affected the Dutch stock market returns on the election day. However, significant negative stock market returns are measured the following day. This could be due to the measuring standard used in this research. As this research is using open-to-open market returns, Like Chan et al. (1996), the election results are presented after the election trading day. No one had expected the outcome of the election and even Cameron himself did not expect the outcome38. With a majority win of the Conservative party, we would expect positive stock market returns as described by Mukherjee et al. (2005). However, the huge win of the Scottish National Party (SNP) has led to a fragmented United Kingdom. The financial markets are influenced by this segregation as Cameron stands for an enormous challenge the coming five years. The uncertain information hypothesis (UIH) formulated by Brown et al. (1998) is applicable given the result shown in table 36a. The extensive media coverage and campaigning have reduced the election-event uncertainty, which positively affect stock market returns. Like Pantzalis et al. (2000), this research found a positive cumulative abnormal return in the two-week period preceding the election. However, a very significant cumulative abnormal stock market return is measured over the entire event-window and the weeks following the election. Together with the significant negative abnormal return measured at t=1, when the election outcome is official, the foreign markets are negatively affected by the British general election of 2015. Hence, hypothesis 2 is rejected as a result.

Table 36a: Cumulative Average Abnormal Return (CAAR) over the event window. The cumulative average abnormal stock return (CAAR) is calculated using the next method: First I calculate the abnormal stock returns (AR’s) per country and then I take the average of these individual AR’s. Lastly all the individual AAR’s are added to form the CAAR. Time period [-2,4] [-2,0] [1,4] CAAR: -2.19%* 0.85% -2.58%***

38https://www.wsj.com/articles/the-u-k-s-electoral-earthquake-1431095499

United Kingdom general elections: conclusion

The results derived from the analysis of the three time periods are used in order to study hypothesis 3: “Political elections are negatively correlated with economic performance”. This research expects that during times of growth, the impact of elections on stock market returns is less severe than during times of contractions. Alesina et al. (1996) explains this with help of the economic business cycle. During economic down turns, the sentiment of the voters is negatively affected by the state of the economy and will have a punishment effect on the incumbent parties during future elections. Moreover, as Anderson (2000) found in his research, this effect is stronger when it is clear for the voters who to blame. The more sizeable the target, to stronger voters’ ability is to express discontent to the incumbent parties. The opposite holds true as well, governmental parties are less affected by the negative sentiment change of voters if the clarity of responsibility is low. Normally, people tend to vote “left” during expansion times and “right” during times of contraction, Stevenson (2001). A pattern is then found between economic performance and voting behaviour. The same was found by Harrington (1993). Voting behaviour is affected by both policy and performance of the incumbent party. If people are less able to differentiate between policies, a higher chance peoples voting behaviour is purely performance driven rather than policy driven.

If we link these results to the results found by this research, a negative cumulative average abnormal return over the whole event window in the period before and after the Financial Crisis is found, see table 39. Like Pantzalis et al. (2000), a significant positive cumulative average abnormal return is measured in the two weeks preceding the election. The uncertainty information hypothesis (UIH) formulated by Brown et al. (1998) is valid. No significant impact is measured in the two-week period preceding the elections before and during the financial crisis. However, the result is in contrast with Mukherjee et al. (2005) who find positive stock market returns during right-wing administrations but also when traders expect right-wing administrations to win elections. Conversely for left-wing administrations. Table 38 shows that stock market returns are indifferent to the British elected government.

Overall, this research expects that the magnitude of the elections on stock market returns increases with time. As table 39 shows, this is indeed what this research finds. The strongest cumulative average abnormal return is measured after the financial crisis period. Based on these findings, this research fails to reject hypothesis 3. The magnitude of the British general elections is negatively correlated with time.

Table 38: Impact of the British general elections on stock markets return during the event window. The cumulative average abnormal stock return (CAAR) is calculated using the next method: First I calculate the abnormal stock returns (AR’s) per country and then I take the average of these individual

AR’s. Lastly all the individual AAR’s are added to form the CAAR. (*, Significant values are retrieved from table A: significant table on page 18.) Time period [-2,4] [-2,0] [1,4] Before -1.81% -0.19% -1.72% During -1.59% -0.01% -1.14% After -2.19%* 0.85% -2.58%***

4.3.5. Spain

The Spanish governmental system works like the governmental system of the United Kingdom. The parliament is based upon a lower house, congress, upper house and a senate. Every new law is passed through the congress before the senate will eventually decide whether new laws are put in place. A big difference with the UK is the ballot voting system. The Spanish people cannot vote for the candidates listed but are voting for the party in general. The names are listed and ordered by the parties themselves. This is known as a “closed list” proportional representation system39. After the elections are closed, the counting begins. The seats are divided proportionally according to the number of votes. As a party to be represent, a margin of 3% is needed to be part of the government. The first seat will go to the party with the most votes. The second seat goes to the next party with the highest number of votes. This happens within every constituency in order form a national government based on the division of the votes in the different constituencies. Three elections are analysed; March 2008, November 2011 and lastly December 2015.

Before

The Spanish general election of March 20008 occurred in the early days of the financial crisis. For many investors, the bankruptcy of Lehman Brothers (September 2008) is seen as the start of the global financial crisis. However, signs of overheating were already measured in the years preceding the crisis, especially the housing market. Spain entered the recession in the third quarter of 2008, when the Spanish economy began showing signs of economic slowdown after a decade of growth. The Spanish general election was won by the “Spanish Socialist Workers’ Party” (PSOE) with 169 seats (43.7%) in the Congress of Deputies40, just 7 seats short of absolute majority. Its main rival was the opposition conservative “People’s Party” (PP) led by Mariano Rojoy. Rojoy was able to obtain 153 seats (40.1%). Together they gained close to 83% of total votes. Leader of the PSOE Jose Luis Rodriguez Zapatero was sworn in for a second term as Prime Minister after a surprise victory in the general election of 2004.

39https://www.thespainreport.com/articles/533-151220135803-this-is-how-spain-s-electoral-system-works 40http://news.bbc.co.uk/1/hi/world/europe/7285885.stm 51

The win by the socialist party is positively received by the Spanish small cap index, table 39. After the election day, which occurred on Sunday, the small cap index showed a positive abnormal stock market return of 1.20%. This is against Stevenson (2001), Mukherjee et al. (2005) and Akitoby and Stratmann (2007), who find negative correlations between the win of left-wing administrations and stock market returns. Investors anticipate to higher inflation levels under the incumbency and potential win of left-wing parties, Douglas and Hibbs (1977). Especially when the inflation levels are at a 10-year high in 2008. The expectation of higher inflation levels lowers trading volume and negatively affect the mean and volatility levels of stocks. This will generate lower stock market returns, Mukherjee et al. (2005). As a result, financial markets will give a premium to right-wing administrations, while penalizing left-wing administrations, Akitoby and Stratmann (2007). Moreover, the voting behavior is against the findings by Stevenson (2001). Voters tend to vote left during economic expansion and vote right during economic contractions. After a decade of economic growth, the Spanish economy was sputtering and unemployment rates rose to the highest of the century. It is therefore expected that a more conservative party won the election. In contrast to academic research, the election outcome positively affected the stock market returns given the results in table 39 and 39a. The uncertain information hypothesis (UIH) formulated by Brown et al. (1998) is valid. Like Pantzalis et al. (2000), positive stock market returns are expected in the weeks preceding the election as media coverage and campaigning are at its peak. This reduces the election-induced risk as the outcome becomes more certain, hence higher observed returns are expected. As table 39a exhibits, this is indeed what this research finds. A few significant values have been measured after the election in reaction of the financial crisis. However, these values are measured after the election so the UIH is still valid. Concluding, based on the findings described above, hypothesis 1 is failed to be rejected. Positive abnormal stock market returns are measured following the election outcome.

Table 39a: Cumulative Abnormal Return (CAR) over the event window. The cumulative abnormal stock return (CAR) is calculated using the next method: First I calculate the abnormal stock returns (AR’s) and then taking the sum of these individual AR’s give the CAR. Time period [-2,4] [-2,0] [1,4] CAR: 7.30% 3.59% 4.91%

The international financial investors responded positively on the election outcome of the Spanish general election of 2008. Especially the stock market returns of the British and German index showed significant positive abnormal stock market returns. As explained on the national level, this is against the findings by Stevenson (2001), Mukherjee et al. (2005) and Akitoby and Stratmann (2007), who find negative correlations between the win of left-wing governments and stock market returns. However, the event-window exhibits a few very significant outcomes that influences the stock market reaction on the election result of the Spanish general election. These significant values can be attributed to the financial crisis. At t=2, the Federal Reserve Bank (FED) announced that it will pomp an extra capital injection in the financial markets of $200bn to boost the economy41. This positively affected the international stock markets. At t=5, the US Federal Reserve Bank agreed to guarantee $30bn assets of Bear Sterns after JP Morgan Chase has bought Bear Stearns on March 14, 200842. At t=9, the international financial markets are positively surprised by an increased acquisition price of JP Morgan Chase on Bear Sterns43. Investors were worried that the initial offer, that was only a fraction of the market value, was too low and would lead to extreme losses in market value of other banks. At t=14, the merger negotiations between Air France-KLM and Alitalia had been stopped after Maurizio Prato, chairman of Alitalia, resigned44. This has resulted in significant negative stock market returns in the Netherlands, France and Italia, table 40. Despite the influence of these important events, a positive impact of the election on the international stock market can be retrieved from table 40a. Like Pantzalis et al. (2000), positive stock market returns are expected as the outcome becomes more certain due to extensive media coverage and campaigning. The uncertain information hypothesis (UIH) formulated by Brown et al. (1998) is applicable here. Ultimately, despite the financial crisis influences, the international financial stock market returns are positively affected by the election outcome and like Pantzalis et al. (2000) positive abnormal stock market returns are measured preceding the election. Therefore, we fail to reject hypothesis 2.

Table 40a: Cumulative Average Abnormal Return (CAAR) over the event window. The cumulative average abnormal stock return (CAAR) is calculated using the next method: First I calculate the abnormal stock returns (AR’s) per country and then I take the average of these individual AR’s. Lastly all the individual AAR’s are added to form the CAAR. Time period [-2,4] [-2,0] [1,4] CAAR: 5.24%***** 1.32% 4.56%***

41https://www.volkskrant.nl/economie/verder-herstel-op-europese-beurzen~a959678/ 42https://www.usatoday.com/story/money/business/2013/09/08/chronology-2008-financial-crisis-lehman/2779515/ 43https://www.volkskrant.nl/economie/flinke-koerswinsten-op-europese-beurzen~a963787/ 44https://www.volkskrant.nl/economie/air-france-klm-wijst-alitalia-de-deur~a887072/

During

The Spanish general election was originally planned in March 2012. However, as the country was experiencing severe economic problems with soaring bond rates and high unemployment levels, new elections were called early on November 20, 2011. Spanish Prime Minister Zapatero wants to bring stability in politics and economy45. A newly formed government could increase the stability on the financial markets as investors know what the course direction will be. In terms of uncertain information hypothesis (UIH) formulated by Brown et al. (1998), the election-induced uncertainty is move forward in order to relieve the financial markets for this uncertainty. The Spanish incumbent political parties have incurred huge losses as voters have been struggling by high unemployment rates, austerity measures and piles of debt. The centre-right opposition party “Popular Party” (PP) led by Mariano Rojoy gained majority in the lower house with 44.6% of the seats. The incumbent “Spanish Socialist Workers’ Party” (PSOE) achieved 29% of the seats, a loss of 15%. Jose Luis Rodriguez Zapatero is replaced by Mariano Rojoy, leader of PP, after two terms as Prime Minister of Spain.

As expected by Douglas and Hibbs (1997), Stevenson (2001), Mukherjee et al. (2005) and Akitoby and Stratmann (2007), the Spanish small cap index is positively affected by win of the centre-right Popular Party (PP). As can be seen from the abnormal returns in table 42, the IBEX small cap index showed a positive abnormal return of 0.45%. Besides, as Anderson (2000) and Stevenson (2001) found, economic performance affects voters’ perceptions and behaviour. The Spanish economy performed poorly in the years preceding the election, with unemployment rates exceeding 20%. Normally, unemployment rates are best served by left- wing regimes and inflation rates are best served by right-wing regimes, Douglas and Hibbs (1977). However, both inflation and unemployment rates are significantly impacted by the financial crisis. During times of recession people tend to vote for right-wing administrations, whereas they vote left during economic expansions. Moreover, if it is clear who is leading the national government, the target of blaming is large and there are few alternative choices, the impact on the financial markets will be more significant, Anderson (2001). Consequently, it is no surprise that the election is won by the centre-right opposition “Popular Party”.

45https://www.theguardian.com/world/2011/jul/29/spain-early-election

Table 42a: Cumulative Abnormal Return (CAR) over the event window. The cumulative abnormal stock return (CAR) is calculated using the next method: First I calculate the abnormal stock returns (AR’s) and then taking the sum of these individual AR’s give the CAR. Time period [-2,4] [-2,0] [1,4] CAR: -2.39% -3.51% 1.57%

Given the results presented in table 42a, the election-induced uncertainty resolution is not valid. In contrast to the findings of Pantzalis et al. (2000), a more significant cumulative abnormal return is measured in the period following the election. However, as Pantzalis et al (2000) described, sometimes the election-induced uncertainty is partially resolved prior the election. In that case, the financial markets need time to process the election outcome. As resulted in table 42a, positive abnormal stock market returns are measured in the weeks following the election, strengthening the expectations formulated by Pantzalis et al. (2000). Hence, based on the initial impact of the election outcome and the positive stock market returns following the election outcome, we fail to reject hypothesis 1.

Expanding the scope towards the international financial markets, the same result is found. Despite the Dutch AEX index, all the other indexes produced positive stock market returns following the election outcome. No significant values are measured, but the sign and magnitude provide evidence of a positive impact on the financial markets as investors are relieved by a stable political government. A higher political instability will negatively influence growth rates, Alesina et al. (1992). Moreover, left-wing parties attach greater importance to full employment than to inflation levels, whereas right-wing parties attach greater importance to price stability relative to unemployment levels, Douglas and Hibbs (1977). Investors’ expectations anticipate to the incumbency and possible win of left-wing parties. Trading volume decreases which negatively affect the mean and volatility figures of stocks and ultimately leads to lower stock market returns, Mukherjee et al. (2005). Hence, the positive impact of the Spanish general election is in line with previous academic research. A few significant events occurred during the event-window. At t=-4, news came out that the Greek national economy has been dropped 5.2% with respect to previous year. Moreover, fear for a recession occurred in the Netherlands as the economy decreased three quarters in a row46. At t=-5, the Italian government passed through another round of austerity measures and Berlusconi accepted his defeat and finally stepped down as Prime Minister of Italy47.

46https://www.volkskrant.nl/economie/recessie-dreigt-in-nederland-economie-krimpt-in-derde-kwartaal~a3033966/ 47https://www.volkskrant.nl/economie/griekse-economie-krimpt-met-5-2-procent~a3034205/

This cleared the road for Mario Monti to install a government based of professionals and intellectuals. These values are more heavily induced by economic news regarding the financial crisis. Nevertheless, a negative impact of the Spanish election outcome is measured during all time periods, table 40a. Despite the centre-right coalition, the cumulative abnormal stock returns are negative impacted by the news, not only preceding the election event but also following the event. However, given the lack of significance, we fail to reject hypothesis 2.

Table 43a: Cumulative Average Abnormal Return (CAAR) over the event window. The cumulative average abnormal stock return (CAAR) is calculated using the next method: First I calculate the abnormal stock returns (AR’s) per country and then I take the average of these individual AR’s. Lastly all the individual AAR’s are added to form the CAAR. Time period [-2,4] [-2,0] [1,4] CAAR: -3.19% -2.02% -0.12%

After

The Spanish general election after the financial crisis was held on December 20, 2015. After a long period of austerity measures set by the European Union to reform the Spanish economy, it was starting to grow in 201448. However, this could not help the centre-right Popular Party (PP) continuing a political majority. Despite the lack of majority, the PP remained the largest party in the government with 29% of the votes. The main rival of the PP, the socialist party (PSOE), became second with 22% of the votes. A remarkable result was achieved by the newly formed anti-austerity party Podemos that gained 21% of the votes in barely 2 years of existence. The fear of the far-right parties is not misplaced given the result by Podemos49. The lingering Euro crisis and migration problems are the foundation of the populist surge. Leader of Podemos, Pablo Iglesias, called his victory a historic day for Spain as a new political era is started. However, none of the political parties could secure a majority coalition, leading to new elections in 2016. Since this research is measuring the impact of elections on stock market returns, the election result of the 2015 general election is analysed.

48https://tradingeconomics.com/spain/gdp-growth 49www.foreignaffairs.com/articles/europe/2016-06-03/rise-populism-europe

The news of the election outcome is negatively interpreted by the Spanish small cap index. A very significant negative abnormal stock market return of -2.11% is measured at the election day, table 45. This result is strengthened by table 45a. The cumulative abnormal stock market return shows a negative result in every time-period. Despite the negative value during the two-week period preceding the election, the uncertain information hypothesis (UIH) formulated by Brown et al. (1998) is valid. The election-induced uncertainty has positively affected the stock market returns compared to the entire event-window. Like Pantzalis et al. (2000), positive stock market returns are expected as the risk surrounding the election outcome is reduced. Given the result in table 45a, this is not the case. In contrast to Douglas and Hibbs (1997), Stevenson (2001), Mukherjee et al. (2005) and Akitoby and Stratmann (2007), no significant positive abnormal stock market return is measured. Hence, hypothesis 1 is rejected. The Spanish general election of 2015 did not positively affect the Small Cap stock market returns. This could partially be explained by the rise of Podemos in a very short period of time. Populist parties are bad for the economy as the economy becomes more fragmented and inflation prone50, while investors favour stable inflation levels, Mukherjee et al. (2005) and Douglas and Hibbs (1977). Given the results presented in table 45a, strong negative results are measured. Hence, hypothesis 1 is rejected. The Spanish general election negatively affected the stock market returns on the Spanish small cap index.

Table 45a: Cumulative Abnormal Return (CAR) over the event window. The cumulative abnormal stock return (CAR) is calculated using the next method: First I calculate the abnormal stock returns (AR’s) and then taking the sum of these individual AR’s give the CAR. Time period [-2,4] [-2,0] [1,4] CAR: -5.12% -1.97% -5.26%

The European international financial markets showed a very strong negative return. This could be explained by the rise of the far-right populist party Podemos. The populist parties are bad for the economy and globalization standards50. International trade agreements resulted in increased productivity standards due to specialization and economies of scale. Increasing trade barriers to protect the domestic market will reduce productively and growth. The markets become more fragmented and more prone to inflation10. The fragmented election result increases political instability, as coalitions are harder to form. Especially since it is not clear how the populist parties want to govern, except stopping mass migration.

50http://www.bbc.com/news/world-europe-36381019

Alesina et al. (1992) have found that countries experiencing periods of high propensity of a government collapse generate significant lower growth rates than in countries with a low propensity of a government collapse. Conversely, low economic growth rates do not increase the propensity of a governmental collapse. However, voting behavior is affected by economic perceptions that interact in a political context, Stevenson (2001). Moreover, when the institutional context clarifies which regime is in control, when the target of credit and blame is large, and when citizens have fewer viable alternative choices, economic effects are stronger. Voters’ ability to express discontent with economic performance is enhanced if accountability of is simple, Anderson (2000). The produced excess returns are correlated with shifts in sentiment, Lee et al. (2002). People tend to vote left (right) during economic expansions (contraction), which means that a “sentiment” is a systematic risk that is priced in. More favorable (unfavorable) changes in sentiment will yield higher (lower) observed returns. As found on the national level, some election-induced uncertainty is resolved in the two-week period preceding the election. The uncertain information hypothesis (UIH) formulated by Brown et al. (1998) is observable. However, positive stock market returns are expected as the election-induced risk is reduced due to extensive media coverage and campaigning, Pantzalis et al. (2000). Observing the results in table 46a, very strong significant negative stock market returns are measured which challenges the findings by Pantzalis et al. (2000). Given the results presented in tables 46 and 46a, this research can conclude that the Spanish general election does not positively affect the international financial markets. Hypothesis 2 is therefore rejected.

Table 46a: Cumulative Average Abnormal Return (CAAR) over the event window. The cumulative average abnormal stock return (CAAR) is calculated using the next method: First I calculate the abnormal stock returns (AR’s) per country and then I take the average of these individual AR’s. Lastly all the individual AAR’s are added to form the CAAR. Time period [-2,4] [-2,0] [1,4] CAAR: -7.08%** -3.47%***** -6.17%*

Spanish general elections: conclusion

This research expects that during times of economic growth, voters’ perceptions are less affected by economic performance and will have less impact on voting behaviour. Greater changes in stock market returns are expected during times of economic contraction, both in direction as in magnitude. Alesina et al. (1996) found that voting behaviour is negatively affected by the state of the economy and will have a punishment effect on the incumbent parties in future elections. Moreover, this effect is stronger when it is clear for the voters who to blame, Anderson (2000). The more sizeable the target, to stronger voters’ ability to express discontent to the incumbent parties. The results presented in table 48 strengthen this verdict, the economic performance positively affect voting behaviour before the financial crisis. The opposite of Andersons’ (2000) is valid as well, governmental parties are less affected by the negative changes in sentiment of voters if the clarity of responsibility is low. According to Stevenson (2001), people tend to vote left during expansions and right during contractions. Governmental parties pursue different plans. Investors are best served by right-wing regimes given their focus on inflation levels. Left-wing regimes focus more on full employment, Douglas and Hibbs (1977). Investors expectations of higher inflation under left-wing administrations lowers the volume of stocks traded in the financial markets. This negatively affects the mean and volatility of stock prices and will generate lower returns, Mukherjee et al. (2005). Hence, investors will give a premium to right wing regimes with their focus on economic stability while penalizing left wing regimes, Akitoby and Stratmann (2007). Evaluating the results presented in table 48 provide strong evidence of a “timing” effect of national elections on stock market returns. In the period before the financial crisis, the election outcome positivity affected the stock market returns in the three time periods, whereas continuing towards the end of the financial crisis the magnitude of the impact becomes greater and more negative. Moreover, like Pantzalis et al. (2000), this research expects a positive cumulative abnormal return (CAR) in the weeks preceding the election- event given the extensive media coverage and campaigning. Moreover, the time-period before the financial crisis strengthen this expectation. Despite the significant negative return in the same period after the financial crisis, the uncertain information hypothesis (UIH) formulated by Brown et al. (1998) leads to some election-induced uncertainty reduction which positively affect stock market returns.

Moreover, table 48 exhibits a pattern that correlates stock market returns with the economic business cycle. During times of economic expansion, stock market returns are positive, while after times of economic contraction, stock market returns are negative due to a penalizing effect of voters. Given the significant results, this research fails to reject hypothesis 3. The election-induced magnitude of stock market returns is negatively correlated with economic performance. This is in line with Alesina et al. (1996), Anderson (2000) and Stevenson (2001).

Table 48: Impact of the Spanish general elections on stock markets return during the event window. The cumulative average abnormal stock return (CAAR) is calculated using the next method: First I calculate the abnormal stock returns (AR’s) per country and then I take the average of these individual

AR’s. Lastly all the individual AAR’s are added to form the CAAR. (*, Significant values are retrieved from table A: significant table on page 18.) Time period [-2,4] [-2,0] [1,4] Before 5.24%***** 1.32% 4.56%*** During -3.19% -2.02% -0.12% After -7.08%** -3.47%***** -6.17%*

4.3.6. Italy

The current Italian political system is based on a republic51. After World War II, the Italian people voted against the monarchy during a constitutional referendum on June 2, 1946. The parliament consists out of a lower house, upper house, Chamber of Deputies and the Senate. Both the lower and upper house representatives are chosen over a five-year term. The Italian voting system is based on a party list system where the candidates are ranked according to their level of priority. A coalition is then formed by individual parties which must agree on appointing a leader. The party with the largest vote is responsible to form a government approved by the president of the republic. The President is elected by both houses of parliament and by the representatives of 58 regions. The President is elected for a term of seven years. The power of the President is used to appoint the council of minister and the Prime Minister who is in charge of the composition of the cabinet and advices the president.

Before

The Italian general election of 2006 is known as interesting. It was the closest election race in history, but rather remarkable was the response by outgoing Prime Minister Silvio Berlusconi who refused to accept the result weeks after the result52. The election was won by the centre- left opposition candidate Romano Prodi. Berlusconi refused to accept Prodi as winner of the general election in the aftermath of the election. Berlusconi kept the government unstable as he denounced a double check of the votes. Eventually on May 5, Berlusconi resigned as Prime Minister. After the election of Giorgio Napolitano as President of Italian Republic, the formation of a new government started on the 16th of May. The formation occurred the following day where a new centre-left government was sworn in.

Table 49a: Cumulative Abnormal Return (CAR) over the event window. The cumulative abnormal stock return (CAR) is calculated using the next method: First I calculate the abnormal stock returns (AR’s) and then taking the sum of these individual AR’s give the CAR. Time period [-2,4] [-2,0] [1,4] CAR: -2.54% -1.95% -1.86%

51http://www.understandingitaly.com/profile-content/government.html 52https://www.sussex.ac.uk/webteam/gateway/file.php?name=epern-election-briefing-no-25.pdf&site=266

The Italian election of 2006 had a significant negative impact on the Italian small cap index. This could be explained by the election victory of the centre-left (Unione) led by Romano Prodi. The anticipation of a left-wing coalition winning the political election results in a decline in trading volume, Mukherjee et al. (2005). This is due to the fact that inflation levels are likely to increase under left-wing administrations. Research done by Hibbs (1977) agrees and found that left-wing parties attach greater value to full employment than to stable inflation levels in contrast to right-wing parties. Right-wing parties attach greater value towards price stability, hence inflation rates, than to full employment. Investors anticipate to these possible alternatives and the fear of left-wing administrations results in lower stock market returns. The findings shown in table 30a strengthen these results. The anticipation of a left-wing administration lowers the cumulative abnormal return during the two-week period preceding the event as well as the entire event-window. Despite the expected positive stock market returns in the two weeks preceding the election, the uncertain information hypothesis (UIH) formulated by Brown et al. (1998) is not valid, table 49a. Some uncertainty is resolved in the weeks following the election as less negative returns are measured here. However, given the strong negative abnormal returns, hypothesis 1 is rejected.

Tables 50 till 51 exhibit the impact of the Italian general election on the foreign stock markets. The stock markets generated a negative return following the Italian election outcome. The national indices of the Netherlands and the United Kingdom show a significant negative result. The international stock market returns are in line with the FTSE MIB returns in Milan. Investor expectations of a left-wing administrations lowers the total trading volume which reduces the mean and volatility of the stocks, Mukherjee et al. (2005). These anticipations are justified by the Hibbs (1997). History has shown that left-wing coalitions focus more on full employment than inflation levels. The deposition of Berlusconi as Prime Minister of Italy is also in line with the academic research done by Anderson (200). The Italian election reveals that political context interacts with economic perception affecting voting behavior. Moreover, the ability of people to express discontent during elections is enhanced as economic performance is easily accountability to the responsible administrations. A punishment effect is expected in these events. Given the election outcome of the Italian election of 2006, this is indeed what this research finds. The change in the political landscape is also in line with the findings of Stevenson (2001).

With an expanding economy from 2002 till 200853, the aggregate policy tends to move leftish, whereas the aggregate policy tends to move rightist in a contracting phase. Looking in more detail at the findings of table 50a, this research finds no uncertainty reduction in the weeks preceding the election event. So, in contrast to Pantzalis et al. (1998), the risk-adjusted returns are not positively affected by the extensive media coverage and campaigning. Instead, more uncertainty is measured as the cumulative average abnormal return (CAAR) is strongly significant. However, uncertainty reduction is measured in the weeks following the event. Even though the CAAR is negative, the magnitude is less negative than the two-week period preceding the election and the entire event-window. Based on these findings and the results presented in tables 50, 50a and 51, hypothesis 2 is rejected. The Italian general election of 2006 negatively affects foreign stock market returns.

Table 50a: Cumulative Average Abnormal Return (CAAR) over the event window. The cumulative average abnormal stock return (CAAR) is calculated using the next method: First I calculate the abnormal stock returns (AR’s) per country and then I take the average of these individual AR’s. Lastly all the individual AAR’s are added to form the CAAR. Time period [-2,4] [-2,0] [1,4] CAAR: -0.57% -0.80%** -0.31%

During

Only 2 years after the previous election, the Italians had to elect a new Parliament. Reasons for the early held election are attributed to two main factors54. First, the majority in the parliament by the incumbent parties was very fragmented. The ideologically between the communists, Christian democrats and Liberal parties were not on the same line. Second, the majority in the parliament was small. This hindered a smooth governmental process of passing laws and debating about policy alterations. Things started to escalate when Walter Veltroni, leader of the Democratic party, questioned the authority and leadership role of Prime Minister Romano Prodi. Under the pressure of a proposed referendum, the relationship between the governmental parties was tense. After the allowance of the referendum by the Constitutional Court early 2008, Celmente Mastelle (Leader of UDEUR Populars) announced his withdrawal from the government. This paved the way for the centre-right coalition led by Berlusconi. With solid a majority in both the Lower House and the Senate, the formation was simplified.

53http://www.focus-economics.com/countries/Italy 54https://www.sussex.ac.uk/webteam/gateway/file.php?name=epern-election-briefing-no-41.pdf&site=266

The re-election of Berlusconi after two years of absence negatively impacted the stock market returns of the national index, FTSE MIB Small Cap Index. Despite the significant negative impact, the outcome of the election is in line with Stevenson (2001). The economic business cycle affects voting behavior. More specific, a positive correlation is found between excess stock market returns and shifts in sentiment, Lee et al. (2002). People tend to vote for left-wing parties under economic expansions, and vice versa during economic contractions. The election was held during times of high economic pressure. The worldwide economy was contracting as the housing crisis of 2007 continued to affect banks and hedge funds that were left with worthless subprime mortgages on their balance sheets. Moreover, during economic down turns, the sentiment of voter’s changes and will have a punishment effect on the future elections, Alesina et al. (1996). Especially if voter’s ability to express discontent is enhanced as the accountability of blaming the incumbent parties is simplified, Anderson (2000). Despite the strong negative impact on stock markets returns after the outcome, the election resulted in positive stock market returns during the event-window as table 52a exhibits. This finding can be grounded by Mukherjee et al. (2005). Investors expectation of lower inflation levels during right-wing administrations increases the trading volume leading to higher mean and volatility levels. Additionally, in following Pantzalis et al. (1998), this research expects positive stock market returns in the weeks preceding the election-event. As can be seen from table 52a, the same result is measured. The uncertain information hypothesis (UIH) formulated by Brown et al. (1988) is valid. Even though the general election negatively impacted the stock market returns on the national level at t=0, the election positively affected the abnormal stock market returns during all event-windows. Based on these findings, this research fails to reject hypothesis 1.

Table 52a: Cumulative Abnormal Return (CAR) over the event window. The cumulative abnormal stock return (CAR) is calculated using the next method: First I calculate the abnormal stock returns (AR’s) and then taking the sum of these individual AR’s give the CAR. Time period [-2,4] [-2,0] [1,4] CAR: 6.68% 3.71% 0.57%

Extending the scope towards the international stock markets, the same pattern is found in the two-week period preceding the election. Hence, as is the case on the FTSE MIB Small Cap index, the uncertainty reduction following extensive media coverage and campaigning positively affected stock market returns on foreign indices, Pantzalis et al. (2000). Despite the uncertain information hypothesis (UIH), the stock market returns are negatively affected in the weeks following the election result, table 53a. Moreover, the international stock market returns are negative at t=0, especially the British stock market shows a strong negative return after the election result. Against the findings at the national level, the international markets exhibit no significant change in stock market returns over the entire event-window. The results of the international stock markets are against Mukherjee et al. (2005). Investor expectations of low inflation levels instead of low unemployment under right- wing administrations did not positively affect the stock markets in the weeks following the election outcome. Especially the strong majority in both lower and upper house should help the government in policy making decisions as well as quickly intervene in urgent situations like the Naples garbage emergency54. This should increase the political stability. As Alesina et al. (1992) pointed out, economic growth is negatively influenced by an increased propensity of a governmental collapse. Especially during the early financial crisis period in which the elections was held, a stable political governmental system positively effects economic stability. Given the negative abnormal stock market returns following the election outcome and the negative cumulative abnormal return measured following the election and during the entire period, hypothesis 2 is rejected. The general election of 2008 did not positively affect stock returns.

Table 53a: Cumulative Average Abnormal Return (CAAR) over the event window. The cumulative average abnormal stock return (CAAR) is calculated using the next method: First I calculate the abnormal stock returns (AR’s) per country and then I take the average of these individual AR’s. Lastly all the individual AAR’s are added to form the CAAR. Time period [-2,4] [-2,0] [1,4] CAAR: -0.11% 2.88%** -3.41%*****

After

With the re-election of Silvio Berlusconi as Prime Minister of Italy in 200855, a fourth Berlusconi cabinet started. It was expected to last the full five years given the large majority in both houses of parliament. However, after a few setbacks in local and regional elections together with the worsening economic crisis, the governmental majority was depleted. After the approval of austerity measures put in place by the European Union, Berlusconi pledged to step down. The position was taken over by Mario Monty on the same day. Likewise, a new cabinet had to be installed. Instead of politicians, a new technocratic cabinet was composed out of professionals and intellectuals. The Monti-government accomplished its goal of restoring financial markets confidence as the debt-spreads went progressively down. Even though Berlusconi resigned as Prime Minister in the 2011, Berlusconi was running for another term. However, the 2013 general election was won by the centre-left led by Pier Luigi Bersani. The centre-right coalition led by Berlusconi had lost its majority.

The election result is positively received by Italian stock market, FTSE MIB, at the election day. However, as can be seen from table 55a, the election result has negatively impacted the stock market returns during the entire event-window. This is in line with Mukherjee et al. (2005) and Douglas and Hibbs (1977). The anticipation of left-wing- administrations lowers stock market returns. The historical focus on unemployment levels rather than inflation levels, reduces trading volume. The reduction in trading volume reduces the mean and volatility of stocks and leads to lower stock market returns. Investors are best served by right-wing administrations. Consequently, financial markets are giving a premium to right-wing governments, while penalizing left-wing governments, Akitoby and Stratman (2007). Moreover, the election outcome is in line with Stevenson (2001). People vote left when the economy is expanding and right when the economy is contracting. The general election of 2013 was held in the aftermath of the financial crisis and the economy was slightly recovering. Additionally, there is a correlation between excess returns and shifts in sentiment, Lee et al. (2001). A bullish change in sentiment leads to an increase of stock market volatility with higher future excess returns, and vice versa. The same is found by Anderson (2000), the economic perceptions affect voting behaviour. Besides, as the accountability of blaming the incumbent policy makers is simplified, the economic effects are stronger.

55http://cadmus.eui.eu/bitstream/handle/1814/29550/Garzia%202013%20West%20European%20Politics.pdf;sequence=2 66

Given the turbulent economic and political time under the fourth Berlusconi cabinet, voting behaviour had a penalizing effect on the governmental system as the ability to express discontent was simplified. However, the uncertain information hypothesis (UIH) formulated by Brown et al. (1998) positively affects stock market returns preceding the election, table 55a. Based on the findings described above, we fail to reject hypothesis 1. The Italian general election of 2013 did positively influences the stock market returns on a national level.

Table 55a: Cumulative Abnormal Return (CAR) over the event window. The cumulative abnormal stock return (CAR) is calculated using the next method: First I calculate the abnormal stock returns (AR’s) and then taking the sum of these individual AR’s give the CAR. Time period [-2,4] [-2,0] [1,4] CAR: -4.47% -0.40% -3.01%

The international financial stock market returns reacted positively on the outcome of the election. Especially the northern-European countries showed significant positive returns. Alesina et al. (1996) found a negative correlation between political stability and economic growth. Countries with a high propensity of a government collapse generate significant lower economic growth rates compared to countries with stable governments. Given the turbulent economic and governmental times in Italy under Berlusconi, the financial market investors are relieved by the win of a left-wing coalition. The impact of the election on the entire event- window is in line with Dougals and Hibbs (1977). The win of a left-wing coalition improves the Italian economic and governmental stability, however left-wing administrations have greater attention to full employment then to stable inflation levels. The expectation of higher inflation levels under left-wing coalitions reduces trading volume in the stock markets. Lower trading volume will lead to lower mean and volatility figures for stocks and produces lower stock market returns, Mukherjee et al. (2005). Very significant negative returns are measured two days preceding the election event. This can be explained by financial troubles of Imtech56. An amortization expense of 300 million euro is made in the German and Polish part of Imtech. Furthermore, this research expects positive stock market returns preceding the election given the uncertain information hypothesis (UIH) formulated by Brown et al. (1998). Given the extensive media coverage and campaigning, some election-induced uncertainty is resolved. The risk-adjusted return will drop and higher stock market returns are expected. Even though a negative return is measured, the abnormal return is positively affected by the election.

56https://www.volkskrant.nl/economie/problemen-technische-dienstverlener-imtech-groter-dan-gedacht~a3400784/ 67

However, in contrast to Pantzalis et al. (2000), this research finds a significant negative impact of the Italian general election on international stock market returns given the results of table 56a. Extending the UIH in the weeks following the election, positive stock market returns are expected as the election-induced uncertainty is resolved. Given the results of table 56a, negative stock market returns are measured as well. Hence, the Italian general election of 2013 negatively affected the international financial markets. Based on these findings, hypothesis 2 is rejected.

Table 56a: Cumulative Average Abnormal Return (CAAR) over the event window. The cumulative average abnormal stock return (CAAR) is calculated using the next method: First I calculate the abnormal stock returns (AR’s) per country and then I take the average of these individual AR’s. Lastly all the individual AAR’s are added to form the CAAR. Time period [-2,4] [-2,0] [1,4] CAAR: -4.17%**** -1.68%* -1.84%*

Italian general elections: conclusion

In order to answer hypothesis 3, a division is made between the different time periods. General elections are investigated before, during and after the financial crisis of 2008. Table 58 exhibits the accumulated results of the cumulative average abnormal return during the different time periods. This research expects that the magnitude of election outcome is increased by time. This means that after the financial crisis, the outcome of the election is more important to the voters and will have a significant effect on the financial markets. This could be explained by Harrington (1993). In his research, he finds that voting behaviour is based on incumbents’ policy and performance. The less voters are able to distinguish which policy is best, the more easily they will vote purely on performance. Moreover, as the accountability of blaming the incumbent policy makers is simplified, the economic effects are stronger as Anderson (2000) found. A correlation between stock market returns and voting behaviour is established. Especially when shifts in sentiment is priced in, Lee et al. (2001). Bullish change in voters’ sentiment will lead to higher excess return as stock market volatility increases. The opposite is valid as well, bearish changes in sentiment decreases stock market volatility and will generate lower excess returns.

Table 58 shows values that supports hypothesis 3. A small negative impact was found in the time-period before the financial crisis was affecting the financial markets. Moreover, this research finds stronger effects of the election outcome during and after the financial crisis. A correlation between “timing” and change in magnitude is found. However, like Pantzalis et al. (2000), this research expects positive stock market returns preceding the election-event. Based on the uncertain information hypothesis (UIH) formulated by Brown et al. (1998), the reduction in election-induced uncertainty positively affect stock market returns. However, this is only the case during the financial crisis, table 58. Furthermore, as expected by Pantzalis et al. (2000), uncertainty-reduction has partially resolved and markets needs time to assess the election results. Positive CAR is then expected in the weeks following the election. However, based on the findings showed in table 58, this is not the case. The UIH is therefore only applicable in the 2008 general election. Concluding, based on the findings described by table 58, this research fails to reject hypothesis 3. Even though the negative influence of the election outcome on stock market returns, the elections produced stronger changes in stock market returns continuing the time spectrum.

Table 58: Impact of the Italian general elections on stock markets return during the event window. The cumulative average abnormal stock return (CAAR) is calculated using the next method: First I calculate the abnormal stock returns (AR’s) per country and then I take the average of these individual

AR’s. Lastly all the individual AAR’s are added to form the CAAR. (*, Significant values are retrieved from table A: significant table on page 18.) Time-period [-2,4] [-2,0] [1,4] Before -0.57% -0.80%** -0.31% During -0.11% 2.88%** -3.41%***** After -4.17%**** -1.68%* -1.84%*

5. Robustness check

One of the short-comings of an event-study analysis is that the results are highly sensitive to the method used in performing an event-study. In order to validate the values of the data analysis, I check the robustness of the results with the European Market Index discussed. Moreover, I check the robustness of the abnormal stock market returns according to the mean R-Square analysis after including the European Market Index.

First, as discussed in chapter 5.1, a European Market Index is created for the purpose to validate the data measured in the data analysis. The European Market index contains the 80 largest listed firms on the European Financial markets. By regressing election events directly on the European Market index, I will overcome country-fixed effects as the listed firms are more sensitive to events on the international level. A purer effect should be measured. This is indeed what is found, a positive cumulative abnormal stock market returns (CAR) becomes more positive and a negative CAR return becomes less negative. Overall, the election-induced events have a more positive impact on the stock market returns when the election event is regressed with the European Market Index, as can be seen from table 59.

Table 59: Impact of national elections on the European Market Index. Cumulative Abnormal Return (CAR): First I calculate the abnormal stock returns (AR’s) and then taking the sum of these

individual AR’s give the CAR. (N/A, the election-event in Germany is planned in September 2017). United Country: Netherlands France German Spain Italy Kingdom

[-2,4] 2.21% 1.38% N/A -2.78% -5.12% -4.60% After [-2,0] 1.35% 2.34% N/A 0.25% -1.97% -1.88% [1,4] 0.94% 0.78% N/A -2.80% -5.26% -1.81%

[-2,4] -3.71% -5.64% -0.06% -6.72% -2.39% 1.57% During [-2,0] -0.41% -4.14% -0.75% -3.18% -3.51% 2.79% [1,4] -3.47% -1.98% 0.57% -3.65% 1.57% -1.43% Time Period

[-2,4] -1.90% 2.33% 1.03% -0.56% 7.30% -0.09% Before [-2,0] -1.32% 0.98% 0.47% -0.07% 3.59% -0.75% [1,4] -0.41% 2.07% 0.75% -0.54% 4.91% 0.00%

However, as presented in table 59, the election events did not positively affect stock market returns in every time-period. Like Pantzalis et al. (2000), this research expects positive stock market returns in the two-week period preceding the election event. Following the uncertain information hypothesis (UIH) formulated by Brown et al. (1998), when the uncertainty regarding the election-induced event is reduced, the risk-adjusted expected returns fall which positively affects stock market returns. The extensive media coverage and campaigning by the candidates reduces the risk factor. By regression the European Market Index, the average election-induced event positively affect stock market returns before and after the financial crisis. Moreover, it could be expected that the election-induced uncertainty has partially resolved prior the election and that financial markets need time to assess the election outcome. Positive stock market returns are then expected in the weeks following the election. After including the European Market Index, the average stock market returns are only positive before the financial crisis. Overall, “timing” has negatively affect stock market returns as the magnitude of the financial markets reaction increases and the direction altered negatively.

Second, I check the robustness of the abnormal stock market returns according to the mean R-Square. The mean R-Square in five countries is positively affected by the inclusion of the European Market Index given the increase showed in table 60. Only the mean R-square of the United Kingdom is negatively changed by 1 percentage point. Overall, I find that the mean R- Square increases with 12 percentage points from 61% to 73%, after the European Market Index is included to the regression analysis. This means that more of the variation in stock market returns is explained by the regression analysis.

Table 60: Mean R-Square of the regression analysis. Mean R-Square without European Market Index 61% Mean R-Square with European Market Index 73%

Without European With European Mean R-Square Change Market Index Market Index Netherlands 66% 84% + 22% France 71% 92% + 21% Germany 64% 72% + 8% United Kingdom 34% 33% - 1% Italy 64% 78% + 14% Spain 65% 77% + 12%

6. Conclusion

This research paper has investigated the impact of election-induced events on the European financial stock markets. Moreover, this studied the relationship of elections on stock market returns and economic performance, which no other research performed. Unlike many other events that affect stock market returns, election events are known in advance. Only the result of the election is not certain. As a result, stock market returns of an election-induced event are likely to be higher than without the event, Brown et al. (1998). This research found positive stock market returns in periods with an election-induced event. Moreover, in following Pantzalis et al. (2000), positive abnormal stock market returns are found in the two-week period preceding the election. The uncertainty reduction positively affects stock market returns. The uncertain information hypothesis (UIH) formulated by Brown et al. (1998), is valid. Positive (less negative) stock market returns are measured in the period ending with an election-induced event compared to the entire event-window. Furthermore, the election-induced stock market returns on the European financial markets are positively correlated with the stock market returns on national stock market index. Positive (negative) stock market returns following the election outcome on the national index will positively (negatively) affect the European stock market returns. Moreover, stock market returns are positively correlated with shifts in sentiment. Like Lee et al. (2002), sentiment is a systematic risk that is priced. This is in line Chan et al. (1996), favourable (unfavourable) political news is positively (negatively) affecting stock market returns. Lastly, the financial markets show positive (less negative) stock market returns after a period of economic expansion (contraction). Hence, a negative relationship is found between economic performance and stock market returns during election periods. The abnormal stock market returns are positively (negatively) affected by elections before (after) the financial crisis. These values suggest that voting behaviour is affected by the economy, Stevenson (2001). Again, sentiment is a systematic risk that is priced. Bullish changes in sentiment yields higher excess returns whereas bearish changes in sentiment yields lower excess returns, Lee et al. (2002). Moreover, the magnitude of the change is positively correlated with economic business cycle. This research finds strong significant negative stock market returns after the financial crisis and positive stock market returns before the financial crisis.

6.1 Discussion

This research aims to help investors’ optimal asset allocations during times of elections. This analysis is the first that combines the effect of elections on stock market returns during financial crisis periods. As such, it provides important insights in the link of changes in economic state and investors reaction to election-induced events. As find by Pantzalis et al. (2000), increased levels of uncertainty exist the weeks preceding the election. Increased excess returns are expected as the uncertainty surrounding the event is resolved as the outcome becomes more certain moving towards the election day. However, this research paper has only focussed on national elections affecting the European financial markets, hence a certain bias is created. This research is limited to the stock market returns of six European countries. Therefore, it is recommended to extend the scope of this research. First, by including more European countries to capture the total European market reaction and extending the sample selection to other countries in the world in order to capture the total investors response to election outcomes. Secondly, by including more elections during several economic recession periods so we can generalize the results.

7. Appendix

7.1 Dutch parliamentary elections

Before

Table 2: The Impact of the Dutch parliamentary election of 2006 the national level. The small-cap ASCX index is used in order to measure the election impact on a national level.

One-day Stock market returns are calculated with the following formula: (Rit/Rit-1) -1. (*, Significant values are retrieved from table A: significant table on page 18.) Day AR AR t-test T = 20 0.77% 1.30 T = 19 -0.43% -0.72 T = 18 -0.37% -0.63 T = 17 0.69% 1.16 T = 16 0.75% 1.26 T = 15 0.54% 0.91 T =14 -0.46% -0.78 T = 13 -0.02% -0.03 T = 12 -0.02% -0.03 T = 11 -0.68% -1.14 T = 10 0.27% 0.45 T = 9 -0.07% -0.12 T = 8 -0.54% -0.90 T = 7 -0.76% -1.28 T = 6 -0.82% -1.38 T = 5 0.44% 0.74 T =4 -0.37% -0.62 T = 3 -0.35% -0.60 T = 2 -0.51% -0.85 T = 1 -1.04% -1.75* T = 0 0.51% 0.86 T = -1 -0.08% -0.13 T = -2 -0.47% -0.78 T = -3 -0.48% -0.81 T = -4 -0.66% -1.11 T = -5 -0.16% -0.28 T = -6 0.23% 0.39 T = -7 -0.12% -0.19 T = -8 -0.67% -1.13 T = -9 0.66% 1.11 T = -10 -0.15% -0.26

Table 3: The impact of the Dutch parliamentary election of 2006 on the European level. The AEX index is regressed in an event-study methodology with the CAC^40, DAX, FTSE 100, IBEX 35 and the FTSE MIB index respectively. (*, Significant values are retrieved from table A: significant table on page 18.) Country France Germany UK Spain Italy AR T-Stat AR T-Stat AR T-Stat AR T-Stat AR T-Stat

T = 20 0.48% 1.45 0.62% 1.56* 1.06% 1.90** 0.63% 1.62* 0.22% 0.54

T = 19 0.14% 0.43 -0.27% -0.69 -0.38% -0.68 0.21% 0.53 0.05% 0.11 T = 18 -0.22% -0.67 -0.32% -0.81 -0.36% -0.65 -0.01% -0.01 -0.35% -0.86 T = 17 0.39% 1.19 0.30% 0.75 0.60% 1.07 0.67% 1.72* 0.78% 1.91** T = 16 0.29% 0.90 0.75% 1.89** 1.19% 2.12** 0.67% 1.72* 0.63% 1.55* T = 15 0.83% 2.55 0.62% 1.57* 0.83% 1.48* 0.72% 1.84** 1.01% 2.47** T =14 0.03% 0.10 0.06% 0.16 0.25% 0.44 -0.12% -0.30 0.20% 0.48 T = 13 -0.64% -1.95** -0.37% -0.94 0.23% 0.41 -0.12% -0.32 -0.02% -0.04 T = 12 -0.22% -0.68 -0.37% -0.94 -0.16% -0.29 -0.18% -0.47 -0.10% -0.25 T = 11 -0.12% -0.38 -0.56% -1.41 -0.74% -1.32 -0.18% -0.45 -0.73% -1.80* T = 10 -0.32% -0.98 -0.41% -1.02 0.29% 0.51 -0.36% -0.92 0.24% 0.60 T = 9 -0.44% -1.34 -0.27% -0.67 0.11% 0.19 -0.83% -2.15** 0.00% 0.01 T = 8 0.36% 1.11 -0.07% -0.16 -0.75% -1.35 0.21% 0.55 -0.34% -0.82 T = 7 -0.06% -0.18 -0.29% -0.73 -0.62% -1.11 -0.36% -0.92 -0.40% -0.97 T = 6 -0.73% -2.22*** -0.76% -1.91** -0.29% -0.52 -0.80% -2.05** -0.47% -1.16 T = 5 0.30% 0.92 0.71% 1.79* 1.07% 1.92** 0.31% 0.79 0.30% 0.73 T =4 -0.06% -0.18 0.21% 0.52 -0.79% -1.40 0.30% 0.78 -0.21% -0.52 T = 3 -0.06% -0.18 0.30% 0.76 -0.46% -0.83 0.42% 1.09 -0.12% -0.29 T = 2 0.30% 0.91 -0.49% -1.25 -0.59% -1.05 -0.38% -0.99 -0.44% -1.07 T = 1 -0.24% -0.72 -0.65% -1.63* -0.18% -0.32 -0.78% -2.01** -0.61% -1.48* T = 0 -0.15% -0.44 0.00% 0.01 0.38% 0.67 0.05% 0.12 -0.05% -0.13 T = -1 -0.58% -1.77* -0.74% -1.86** -0.02% -0.04 -0.59% -1.51* -0.45% -1.11 T = -2 0.24% 0.72 -0.31% -0.79 -0.43% -0.76 -0.31% -0.81 -0.06% -0.14 T = -3 0.16% 0.49 -0.35% -0.88 -0.43% -0.76 -0.50% -1.29 -0.34% -0.82 T = -4 0.11% 0.32 -0.34% -0.87 -0.33% -0.59 -0.02% -0.06 -0.17% -0.42 T = -5 0.27% 0.83 0.21% 0.52 0.42% 0.74 -0.06% -0.16 0.30% 0.73 T = -6 -0.63% -1.93** -0.34% -0.85 0.42% 0.75 -0.17% -0.45 -0.06% -0.14 T = -7 -0.30% -0.92 -0.08% -0.19 0.13% 0.23 0.13% 0.34 -0.55% -1.35 T = -8 -0.47% -1.43 -0.55% -1.39 -0.65% -1.16 -0.42% -1.07 -0.19% -0.46 T = -9 -0.19% -0.58 0.09% 0.21 0.42% 0.75 -0.15% -0.38 -0.21% -0.50 T = -10 0.07% 0.22 0.02% 0.06 -0.03% -0.06 0.05% 0.14 0.33% 0.80

Table 4: The average abnormal stock return (AAR) together with the cumulative average abnormal return (CAAR) and T-stats per day during the six-week period. The average abnormal returns are calculated by first calculating the abnormal stock returns per day for each stock market index in the sample during the event-window. Then the average abnormal stock market returns from all stock indexes on a specific date during this event-window. The cumulative average abnormal stock return (CAAR) is the sum of all average abnormal returns (AAR’s). (*, Significant values are retrieved from table A: significant table on page 18.) AAR T-STAT T = 20 0.60% 1.41 T = 19 -0.05% -0.06 T = 18 -0.25% -0.60 T = 17 0.55% 1.33 T = 16 0.71% 1.64* T = 15 0.80% 1.98** T = 14 0.08% 0.18 T = 13 -0.18% -0.57 T = 12 -0.21% -0.53 T = 11 -0.47% -1.07 T = 10 -0.11% -0.36 T = 9 -0.29% -0.79 T = 8 -0.12% -0.14 T = 7 -0.34% -0.78 T = 6 -0.61% -1.57* T = 5 0.54% 1.23 T = 4 -0.11% -0.16 T = 3 0.02% 0.11 T = 2 -0.32% -0.69 T = 1 -0.49% -1.23 T = 0 0.05% 0.05 T = -1 -0.48% -1.26 T = -2 -0.18% -0.36 T = -3 -0.29% -0.65 T = -4 -0.15% -0.32 T = -5 0.23% 0.53 T = -6 -0.16% -0.52 T = -7 -0.13% -0.38 T = -8 -0.45% -1.10 T = -9 -0.01% -0.10 T = -10 0.09% 0.23 CAAR: -1.74% -1.27

During

Table 5: The Impact of the Dutch parliamentary election of 2010 on the national level. The small-cap ASCX index is used in order to measure the election impact on a national level. One-day Stock market returns are calculated with the following formula: (Rit/Rit-1) -1. (*, Significant values are retrieved from table A: significant table on page 18.) Day AR T-test T = 20 0.05% 0.05 T = 19 -0.26% -0.26 T = 18 -0.23% -0.22 T = 17 -0.46% -0.45 T = 16 -2.21% -2.16*** T = 15 -0.63% -0.62 T =14 -0.51% -0.50 T = 13 -0.73% -0.71 T = 12 -1.57% -1.54* T = 11 0.31% 0.31 T = 10 -0.98% -0.96 T = 9 -0.29% -0.29 T = 8 0.88% 0.87 T = 7 0.45% 0.44 T = 6 -0.34% -0.33 T = 5 1.93% 1.89** T =4 0.32% 0.31 T = 3 0.72% 0.71 T = 2 1.24% 1.21 T = 1 0.37% 0.36 T = 0 0.09% 0.09 T = -1 0.90% 0.88 T = -2 -2.55% -2.50**** T = -3 -0.22% -0.21 T = -4 2.24% 2.19*** T = -5 0.00% 0.00 T = -6 -1.47% -1.44 T = -7 1.72% 1.69* T = -8 1.26% 1.24 T = -9 2.09% 2.05** T = -10 -2.94% -2.88****

Table 6: The impact of the Dutch parliamentary election of 2010 on the European level. The AEX index is regressed in an event-study methodology with the CAC^40, DAX, FTSE 100, IBEX 35 and the FTSE MIB index respectively. (*, Significant values are retrieved from table A: significant table on page 18.) Country France Germany UK Spain Italy AR T-Stat AR T-Stat AR T-Stat AR T-Stat AR T-Stat

T = 20 0.01% 0.01 0.29% 0.29 -0.42% -0.26 -0.49% -0.47 -0.04% -0.03 T = 19 -0.02% -0.03 -0.23% -0.23 -0.16% -0.10 -1.17% -1.11 0.54% 0.45 T = 18 -0.14% -0.20 -0.01% -0.01 -0.69% -0.43 -0.51% -0.49 -0.38% -0.32 T = 17 -0.47% -0.66 -0.87% -0.85 0.17% 0.11 -2.42% -2.30*** -1.31% -1.08 T = 16 -0.55% -0.79 -1.09% -1.08 -2.16% -1.34 -0.67% -0.64 -0.81% -0.67 T = 15 -0.17% -0.24 -0.94% -0.93 -0.58% -0.36 0.57% 0.55 0.21% 0.18 T =14 0.13% 0.18 0.10% 0.10 -0.53% -0.33 0.42% 0.40 0.76% 0.63 T = 13 -0.18% -0.26 -0.51% -0.50 -0.24% -0.15 -0.98% -0.93 -1.01% -0.83 T = 12 0.31% 0.43 -0.96% -0.94 -1.61% -1.00 0.66% 0.63 -0.27% -0.22 T = 11 0.38% 0.54 0.22% 0.22 0.99% 0.61 -0.02% -0.02 0.05% 0.04 T = 10 0.13% 0.18 -0.37% -0.37 -0.32% -0.20 0.22% 0.21 0.46% 0.38 T = 9 0.26% 0.37 -0.12% -0.12 -1.28% -0.79 -0.01% -0.01 0.69% 0.57 T = 8 0.09% 0.13 0.53% 0.52 1.69% 1.05 -1.09% -1.04 0.05% 0.04 T = 7 0.05% 0.07 -0.13% -0.13 0.23% 0.15 -0.43% -0.40 0.15% 0.12 T = 6 -0.12% -0.17 -0.31% -0.30 -0.72% -0.45 0.53% 0.51 0.15% 0.13 T = 5 -0.22% -0.31 0.86% 0.84 2.09% 1.30 -0.42% -0.40 -0.88% -0.73 T =4 -0.11% -0.16 0.35% 0.34 -0.27% -0.17 1.59% 1.51 -0.18% -0.15 T = 3 -0.60% -0.85 0.19% 0.19 0.30% 0.18 -2.61% -2.49**** -1.13% -0.93 T = 2 -0.55% -0.77 1.06% 1.05 2.17% 1.35 -1.49% -1.42 -0.41% -0.34 T = 1 0.25% 0.36 -0.42% -0.41 -0.15% -0.09 -0.32% -0.30 0.22% 0.18 T = 0 0.18% 0.26 -0.14% -0.14 -0.02% -0.01 0.99% 0.94 -0.22% -0.18 T = -1 -0.01% -0.02 0.43% 0.43 1.45% 0.90 0.49% 0.47 -0.32% -0.26 T = -2 0.45% 0.63 -1.38% -1.35 -3.07% -1.91** 0.08% 0.08 0.71% 0.59 T = -3 0.02% 0.03 -0.18% -0.18 -0.63% -0.39 1.40% 1.33 0.30% 0.25 T = -4 0.45% 0.64 1.16% 1.14 3.15% 1.96 0.91% 0.87 0.92% 0.76 T = -5 0.02% 0.03 0.00% 0.00 -0.01% -0.01 0.38% 0.36 1.33% 1.10 T = -6 0.44% 0.63 -1.13% -1.11 -1.29% -0.80 0.07% 0.07 -0.02% -0.02 T = -7 -0.19% -0.27 0.90% 0.89 1.50% 0.93 0.22% 0.21 -0.58% -0.48 T = -8 0.30% 0.42 0.48% 0.47 0.44% 0.28 1.59% 1.52* 0.41% 0.34 T = -9 0.34% 0.49 1.15% 1.13 3.29% 2.05** 0.56% 0.53 0.76% 0.62 T = -10 -0.32% -0.45 -1.31% -1.29 -3.82% -2.38*** 0.70% 0.66 0.28% 0.23

Table 7: The average abnormal stock return (AAR) together with the cumulative average abnormal return (CAAR) and T-stats per day during the six-week period. The average abnormal returns are calculated by first calculating the abnormal stock returns per day for each stock market index in the sample during the event-window. Then the average abnormal stock market returns from all stock indexes on a specific date during this event- window. The cumulative average abnormal stock return (CAAR) is the sum of all average abnormal returns (AAR’s). (*, Significant values are retrieved from table A: significant table on page 18.) AAR T-STAT T = 20 -0.13% -0.09 T = 19 -0.21% -0.20 T = 18 -0.34% -0.29 T = 17 -0.98% -0.96 T = 16 -1.06% -0.90 T = 15 -0.18% -0.16 T = 14 0.18% 0.20 T = 13 -0.58% -0.53 T = 12 -0.37% -0.22 T = 11 0.32% 0.28 T = 10 0.02% 0.04 T = 9 -0.09% 0.00 T = 8 0.25% 0.14 T = 7 -0.03% -0.04 T = 6 -0.09% -0.06 T = 5 0.29% 0.14 T = 4 0.28% 0.28 T = 3 -0.77% -0.78 T = 2 0.16% -0.03 T = 1 -0.08% -0.05 T = 0 0.16% 0.17 T = -1 0.41% 0.30 T = -2 -0.64% -0.39 T = -3 0.18% 0.21 T = -4 1.32% 1.07 T = -5 0.34% 0.30 T = -6 -0.38% -0.25 T = -7 0.37% 0.26 T = -8 0.64% 0.60 T = -9 1.22% 0.96 T = -10 -0.89% -0.64 CAAR: -0.70% -0.61

After

Table 8: The Impact of the Dutch parliamentary election of 2017 the national level. The small-cap ASCX index is used in order to measure the election impact on a national level.

One-day Stock market returns are calculated with the following formula: (Rit/Rit-1) -1. (*, Significant values are retrieved from table A: significant table on page 18.) Day AR T-test T = 20 -0.37% -0.52 T = 19 -0.74% -1.04 T = 18 0.92% 1.29 T = 17 0.77% 1.08 T = 16 -1.24% -1.74* T = 15 0.51% 0.72 T =14 -0.87% -1.22 T = 13 -0.22% -0.31 T = 12 0.04% 0.06 T = 11 0.12% 0.17 T = 10 0.29% 0.41 T = 9 0.83% 1.16 T = 8 -0.93% -1.31 T = 7 0.50% 0.70 T = 6 -0.23% -0.32 T = 5 -0.21% -0.30 T =4 0.37% 0.51 T = 3 -0.44% -0.62 T = 2 -0.51% -0.71 T = 1 0.45% 0.63 T = 0 0.24% 0.34 T = -1 -0.63% -0.88 T = -2 0.00% 0.00 T = -3 0.95% 1.33 T = -4 -0.21% -0.30 T = -5 -047% -0.66 T = -6 0.14% 0.19 T = -7 -0.23% -0.32 T = -8 -0.94% -1.32 T = -9 0.66% 0.93 T = -10 0.52% 0.73

Table 9: The impact of the Dutch parliamentary election of 2017 on the European level. The AEX index is regressed in an event-study methodology with the CAC^40, DAX, FTSE 100, IBEX 35 and the FTSE MIB index respectively. (*, Significant values are retrieved from table A: significant table on page 18.) Country France Germany UK Spain Italy AR T-Stat AR T-Stat AR T-Stat AR T-Stat AR T-Stat T = 20 -0.19% -0.49 0.13% 0.19 0.23% 0.25 -0.18% -0.26 0.13% 0.12 T = 19 0.42% 1.06 0.38% 0.58 -0.37% -0.41 0.81% 1.18 0.69% 0.63 T = 18 0.34% 0.86 0.20% 0.30 0.48% 0.53 0.42% 0.61 0.56% 0.51 T = 17 0.02% 0.04 0.58% 0.87 0.96% 1.05 -0.24% -0.35 0.10% 0.09 T = 16 0.10% 0.25 0.35% 0.53 -0.76% -0.83 -0.41% -0.59 -0.07% -0.07 T = 15 -0.01% -0.02 0.12% 0.17 0.07% 0.08 -0.33% -0.47 0.19% 0.17 T =14 0.03% 0.09 0.10% 0.15 -0.56% -0.61 0.65% 0.94 0.67% 0.61 T = 13 -0.44% -1.13 -0.45% -0.67 0.83% 0.91 -0.22% -0.31 -0.28% -0.25 T = 12 0.15% 0.38 0.00% 0.00 0.19% 0.21 0.24% 0.35 -0.05% -0.05 T = 11 -0.24% -0.62 -0.22% -0.33 -0.21% -0.23 0.35% 0.51 0.46% 0.42 T = 10 -0.13% -0.34 -0.51% -0.76 0.27% 0.29 0.02% 0.03 -0.23% -0.21 T = 9 -0.18% -0.45 -0.02% -0.03 1.09% 1.19 -0.31% -0.45 0.01% 0.01 T = 8 -0.12% -0.31 -0.53% -0.80 -1.13% -1.23 -0.24% -0.35 -0.34% -0.31 T = 7 -0.02% -0.04 -0.40% -0.60 0.47% 0.52 0.02% 0.03 -0.25% -0.22 T = 6 -0.78% -1.98 -0.60% -0.90 0.06% 0.07 -1.12% -1.62 -1.53% -1.40 T = 5 -0.03% -0.08 0.48% 0.72 -0.93% -1.02 -0.09% -0.13 0.20% 0.18 T =4 0.04% 0.10 0.20% 0.30 0.43% 0.47 0.00% 0.00 0.31% 0.28 T = 3 0.25% 0.64 0.17% 0.26 0.21% 0.23 -0.29% -0.42 0.88% 0.80 T = 2 -0.13% -0.33 0.28% 0.41 -0.82% -0.89 -1.08% -1.57 -1.06% -0.97 T = 1 -0.01% -0.03 -0.24% -0.35 0.83% 0.91 -0.59% -0.86 -1.27% -1.16 T = 0 0.35% 0.88 -0.28% -0.42 -0.12% -0.13 0.41% 0.59 0.60% 0.55 T = -1 -0.01% -0.03 -0.13% -0.19 -0.01% -0.01 0.48% 0.70 0.11% 0.10 T = -2 0.22% 0.55 0.80% 1.19 0.11% 0.12 0.50% 0.72 0.04% 0.04 T = -3 0.27% 0.68 0.34% 0.51 1.23% 1.34 -0.72% -1.04 0.01% 0.01 T = -4 0.26% 0.66 0.43% 0.64 0.45% 0.50 -0.04% -0.06 0.52% 0.47 T = -5 0.36% 0.91 0.02% 0.03 -0.25% -0.28 -0.03% -0.04 0.03% 0.02 T = -6 0.20% 0.52 -0.04% -0.06 0.15% 0.17 -0.39% -0.56 0.42% 0.38 T = -7 -0.31% -0.78 0.62% 0.93 0.33% 0.36 -0.44% -0.63 -0.71% -0.65 T = -8 -0.35% -0.91 -0.16% -0.24 -0.60% -0.66 0.01% 0.01 -0.46% -0.42 T = -9 -0.10% -0.25 0.12% 0.18 0.53% 0.58 -0.05% -0.08 -0.36% -0.32 T = -10 -0.11% -0.27 0.02% 0.04 0.52% 0.57 -0.41% -0.59 0.12% 0.11

Table 10: The average abnormal stock return (AAR) together with the cumulative average abnormal return (CAAR) and T-stats per day during the six-week period. The average abnormal returns are calculated by first calculating the abnormal stock returns per day for each stock market index in the sample during the event-window. Then the average abnormal stock market returns from all stock indexes on a specific date during this event-window. The cumulative average abnormal stock return (CAAR) is the sum of all average abnormal returns (AAR’s). (*, Significant values are retrieved from table A: significant table on page 18.) AAR T-STAT T = 20 0.02% -0.04 T = 19 0.39% 0.61 T = 18 0.40% 0.56 T = 17 0.28% 0.34 T = 16 -0.16% -0.14 T = 15 0.01% -0.01 T = 14 0.18% 0.24 T = 13 -0.11% -0.29 T = 12 0.11% 0.18 T = 11 0.03% -0.05 T = 10 -0.12% -0.20 T = 9 0.12% 0.05 T = 8 -0.47% -0.60 T = 7 -0.03% -0.06 T = 6 -0.79% -1.17 T = 5 -0.08% -0.07 T = 4 0.20% 0.23 T = 3 0.25% 0.30 T = 2 -0.56% -0.67 T = 1 -0.26% -0.30 T = 0 0.19% 0.29 T = -1 0.09% 0.12 T = -2 0.33% 0.52 T = -3 0.22% 0.30 T = -4 0.32% 0.44 T = -5 0.02% 0.13 T = -6 0.07% 0.09 T = -7 -0.10% -0.15 T = -8 -0.31% -0.44 T = -9 0.03% 0.02 T = -10 0.03% -0.03 CAAR: 0.29% 0.11

7.2 French presidential elections

Before

Table 12: The Impact of the French Presidential election of 2007 the national level. The small-cap CAC small index is used in order to measure the election impact on a national level. One-day

Stock market returns are calculated with the following formula: (Rit/Rit-1) -1. The blue line is the first election round, yellow is the second election round. (*, Significant values are retrieved from table A: significant table on page 18.) Day AR T-STAT T = 20 -0.09% -0.15 T = 19 -0.09% -0.15 T = 18 0.96% 1.57 T = 17 -0.55% -0.90 T = 16 -0.02% -0.03 T = 15 0.44% 0.73 T = 14 -1.35% -2.20*** T = 13 1.84% 3.01***** T = 12 -1.40% -2.30*** T = 11 -0.60% -0.98 T = 10 0.04% 0.06 T = 9 0.40% 0.65 T = 8 -0.06% -0.10 T = 7 0.32% 0.53 T = 6 0.79% 1.30 T = 5 -0.43% -0.70 T = 4 -0.79% -1.29 T = 3 0.95% 1.56 T = 2 -0.47% -0.78 T = 1 -0.58% -0.95 T = 0 1.33% 2.18*** T = -1 1.53% 2.51**** T = -2 -1.28% -2.09 T = -3 0.20% 0.33 T = -4 0.42% 0.69 T = -5 0.59% 0.97 T = -6 -0.01% -0.02 T = -7 -0.11% -0.17 T = -8 -0.04% -0.06 T = -9 0.22% 0.37 T = -10 -0.14% -0.23

Table 13: The impact of the French Presidential election of 2007 on the European level. The CAC^40 index is regressed in an event-study methodology with the AEX, DAX, FTSE 100, IBEX 35 and the FTSE MIB index respectively. The blue line is the first election round, yellow is the second election round. (*, Significant values are retrieved from table A: significant table on page 18.) Netherlands Germany UK Spain Italy AR T-Stat AR T-Stat AR T-Stat AR T-Stat AR T-Stat

T = 20 -0.45% -1.41 -0.50% -1.35 0.20% 0.33 -0.17% -0.46 0.53% 1.41 T = 19 0.05% 0.15 -0.11% -0.29 0.04% 0.07 0.27% 0.71 0.31% 0.82 T = 18 0.08% 0.25 -0.15% -0.39 0.75% 1.26 0.64% 1.71 1.80% 4.77***** T = 17 -0.85% -2.63**** -0.42% -1.13 -0.59% -0.98 -0.88% -2.34*** -0.89% -2.37*** T = 16 0.39% 1.22 -0.17% -0.47 0.12% 0.20 -0.70% -1.87 0.37% 0.98 T = 15 0.15% 0.48 0.00% 0.01 0.52% 0.86 -0.05% -0.14 -0.16% -0.42 T =14 -0.34% -1.04 -0.33% -0.88 -1.07% -1.79* -0.59% -1.57* -0.36% -0.96 T = 13 0.67% 2.07 0.50% 1.34 1.60% 2.66**** 0.21% 0.56 0.63% 1.67 T = 12 0.09% 0.28 -0.45% -1.21 -1.30% -2.17*** -0.70% -1.85** -0.12% -0.32 T = 11 -0.39% -1.21 -0.52% -1.38 -0.43% -0.72 -0.53% -1.40 -0.69% -1.82** T = 10 0.81% 2.52**** 0.67% 1.79* 0.57% 0.95 0.42% 1.13 0.36% 0.95 T = 9 0.36% 1.12 0.27% 0.72 0.25% 0.42 -0.56% -1.49* 0.50% 1.32 T = 8 -0.16% -0.49 -0.19% -0.50 -0.51% -0.85 -0.02% -0.05 0.36% 0.95 T = 7 -0.17% -0.54 0.03% 0.09 0.08% 0.13 0.35% 0.92 0.40% 1.05 T = 6 0.61% 1.88** 0.47% 1.26 1.18% 1.96** 0.98% 2.61**** 0.73% 1.93** T = 5 0.30% 0.94 -0.33% -0.88 -0.11% -0.19 0.72% 1.91** -0.12% -0.32 T =4 -0.07% -0.22 -0.87% -2.33*** -0.73% -1.22 0.25% 0.68 -0.39% -1.05 T = 3 1.10% 3.42***** 0.37% 0.98 1.05% 1.75* 0.21% 0.56 0.24% 0.65 T = 2 -0.22% -0.69 0.40% 1.08 0.19% 0.32 1.72% 4.58 -0.06% -0.15 T = 1 0.10% 0.30 -0.39% -1.05 -0.40% -0.66 0.24% 0.63 -0.41% -1.08 T = 0 0.17% 0.52 0.26% 0.70 1.02% 1.69* 0.11% 0.28 1.02% 2.72**** T = -1 0.53% 1.64* 0.75% 1.99** 1.45% 2.41*** 0.69% 1.84** 0.70% 1.87** T = -2 -0.35% -1.09 0.31% 0.82 -0.92% -1.54* -0.37% -0.99 -0.14% -0.36 T = -3 0.05% 0.15 -0.12% -0.31 0.20% 0.32 0.58% 1.54 -0.30% -0.80 T = -4 0.02% 0.06 -0.71% -1.90** 0.09% 0.14 0.44% 1.17 0.09% 0.25 T = -5 -0.53% -1.64* -0.02% -0.05 0.57% 0.94 0.33% 0.89 0.48% 1.28 T = -6 0.45% 1.40 0.24% 0.63 0.26% 0.44 0.62% 1.65 0.16% 0.43 T = -7 0.32% 0.98 -0.06% -0.16 -0.22% -0.37 -0.09% -0.24 -0.02% -0.05 T = -8 -0.69% -2.13** -0.70% -1.88** -0.04% -0.07 -0.09% -0.24 -0.51% -1.34 T = -9 -0.01% -0.04 -0.05% -0.13 0.06% 0.11 0.13% 0.34 0.10% 0.27 T = -10 0.11% 0.33 -0.04% -0.09 0.02% 0.03 0.18% 0.47 0.14% 0.36

Table 14: The average abnormal stock return (AAR) together with the cumulative average abnormal return (CAAR) and T-stats per day during the six-week period. The average abnormal returns are calculated by first calculating the abnormal stock returns per day for each stock market index in the sample during the event-window. Then the average abnormal stock market returns from all stock indexes on a specific date during this event-window. The cumulative average abnormal stock return (CAAR) is the sum of all average abnormal returns (AAR’s). The blue line is the first election round, yellow is the second election round. (*, Significant values are retrieved from table A: significant table on page 18.) AAR T-STAT T = 20 -0.08% -0.30 T = 19 0.11% 0.29 T = 18 0.63% 1.52* T = 17 -0.73% -1.89** T = 16 0.00% 0.02 T = 15 0.09% 0.16 T = 14 -0.54% -1.25 T = 13 0.72% 1.66 T = 12 -0.50% -1.05 T = 11 -0.51% -1.31 T = 10 0.57% 1.47* T = 9 0.16% 0.42 T = 8 -0.10% -0.19 T = 7 0.14% 0.33 T = 6 0.79% 1.93** T = 5 0.09% 0.29 T = 4 -0.36% -0.83 T = 3 0.60% 1.47 T = 2 0.41% 1.03 T = 1 -0.17% -0.37 T = 0 0.52% 1.18 T = -1 0.82% 1.95** T = -2 -0.30% -0.63 T = -3 0.08% 0.18 T = -4 -0.01% -0.05 T = -5 0.17% 0.28 T = -6 0.35% 0.91 T = -7 -0.02% 0.03 T = -8 -0.41% -1.13 T = -9 0.05% 0.11 T = -10 0.08% 0.22 CAAR: 2.65% 0.66

During

Table 15: The Impact of the French Presidential election of 2012 the national level. The small-cap CAC small index is used in order to measure the election impact on a national level. One-day Stock market returns are calculated with the following formula: (Rit/Rit-1) -1. The blue line is the first election round, yellow is the second election round. (*, Significant values are retrieved from table A: significant table on page 18.) Day AR T-STAT T = 20 -0.25% -0.37 T = 19 1.49% 2.19***

T = 18 0.62% 0.92

T = 17 -1.98% -2.91****

T = 16 1.39% 2.04** T = 15 -1.68% -2.47**** T = 14 0.03% 0.05 T = 13 -0.38% -0.56 T = 12 -1.12% -1.66* T = 11 0.27% 0.40 T = 10 -1.39% -2.04**

T = 9 -0.07% -0.10

T = 8 -0.56% -0.82

T = 7 -0.65% -0.96 T = 6 -0.31% -0.45 3.23**** T = 5 2.20% * T = 4 -1.16% -1.71* T = 3 0.94% 1.39 T = 2 1.43% 2.10** T = 1 -0.09% -0.14 T = 0 -0.25% -0.36 T = -1 -1.67% -2.46**** T = -2 -0.81% -1.20

T = -3 1.95% 2.88****

T = -4 1.36% 2.00**

T = -5 -1.88% -2.76**** T = -6 0.07% 0.10 T = -7 0.61% 0.90 T = -8 -0.89% -1.30 T = -9 -0.86% -1.26 T = -10 -1.10% -1.62*

Table 16: The impact of the French Presidential election of 2012 on the European level. The CAC^40 index is regressed in an event-study methodology with the AEX, DAX, FTSE 100, IBEX 35 and the FTSE MIB index respectively. (*, Significant values are retrieved from table A: significant table on page 18.) Netherlands Germany UK Spain Italy AR T-Stat AR T-Stat AR T-Stat AR T-Stat AR T-Stat

T = 20 -0.13% -0.31 0.31% 0.52 -0.86% -0.87 -0.23% -0.24 -1.08% -1.29 T = 19 0.48% 1.14 0.18% 0.29 1.50% 1.52* 1.93% 2.07** -0.73% -0.87 T = 18 0.48% 1.15 0.29% 0.49 1.09% 1.11 -2.05% -2.21*** 1.22% 1.46 T = 17 -0.37% -0.87 -0.44% -0.74 -1.91% -1.94** 1.08% 1.16 0.83% 0.99 T = 16 0.73% 1.74* 0.40% 0.66 1.47% 1.49 1.05% 1.13 0.28% 0.34 T = 15 -0.42% -1.00 0.06% 0.10 -1.83% -1.86** 1.25% 1.34 2.28% 2.73**** T =14 0.25% 0.61 -0.52% -0.87 0.38% 0.39 0.27% 0.29 0.58% 0.69 T = 13 -0.73% -1.74* -0.98% -1.64* -0.72% -0.73 0.43% 0.46 -0.36% -0.43 T = 12 -0.50% -1.19 -0.07% -0.12 -1.04% -1.05 -1.64% -1.76* -0.69% -0.82 T = 11 -0.46% -1.11 -0.76% -1.28 0.11% 0.11 1.52% 1.64* 0.86% 1.03 T = 10 -0.56% -1.34 -0.80% -1.34 -0.90% -0.91 -1.14% -1.23 -0.06% -0.07 T = 9 -0.52% -1.24 1.08% 1.80** 0.16% 0.16 -3.41% -3.66***** -1.53% -1.82** T = 8 -0.42% -1.00 -0.16% -0.26 -1.05% -1.07 -0.77% -0.83 0.69% 0.82 T = 7 0.22% 0.53 0.70% 1.17 -0.32% -0.32 2.54% 2.72**** 2.59% 3.10***** T = 6 -0.17% -0.41 -0.45% -0.75 -0.45% -0.46 0.96% 1.04 0.57% 0.68 T = 5 1.26% 3.02***** -0.11% -0.18 2.13% 2.16*** -1.54% -1.66* -1.57% -1.88** T =4 -0.94% -2.24*** -0.91% -1.52* -1.63% -1.66* 2.00% 2.15** 0.76% 0.91 T = 3 0.89% 2.13** 0.88% 1.46 1.90% 1.93** 0.41% 0.44 -0.92% -1.10 T = 2 1.30% 3.11 1.19% 1.98** 1.66% 1.69* -0.36% -0.39 -0.27% -0.32 T = 1 -0.38% -0.91 -0.08% -0.14 -0.60% -0.61 0.31% 0.34 0.65% 0.78 T = 0 -0.47% -1.12 -0.38% -0.64 -0.85% -0.87 -1.32% -1.41 -0.16% -0.19 T = -1 -1.30% -3.11***** -1.24% -2.07** -2.32% -2.36*** 0.27% 0.29 0.70% 0.84 T = -2 -0.35% -0.84 -0.72% -1.21 -0.60% -0.60 2.39% 2.56**** 0.93% 1.11 T = -3 0.42% 1.01 0.14% 0.23 1.48% 1.51* -0.41% -0.45 -1.86% -2.23*** T = -4 0.49% 1.17 -0.31% -0.52 0.55% 0.56 1.71% 1.83 0.04% 0.05 T = -5 -1.18% -2.82**** -0.24% -0.40 -2.08% -2.11** 0.63% 0.68 1.40% 1.67* T = -6 0.08% 0.18 0.17% 0.28 -0.08% -0.08 2.36% 2.54**** 0.11% 0.13 T = -7 0.23% 0.55 -0.10% -0.16 0.73% 0.74 -1.06% -1.14 -0.91% -1.09 T = -8 -0.22% -0.53 -1.02% -1.70* -0.91% -0.92 -0.09% -0.09 1.81% 2.17*** T = -9 -0.65% -1.55 0.00% 0.00 -1.71% -1.74* -0.26% -0.28 0.49% 0.59 T = -10 -0.53% -1.28 -0.43% -0.71 -0.98% -0.99 -0.32% -0.34 0.27% 0.32

Table 17: The average abnormal stock return (AAR) together with the cumulative average abnormal return (CAAR) and T-stats per day during the six-week period. The average abnormal returns are calculated by first calculating the abnormal stock returns per day for each stock market index in the sample during the event-window. Then the average abnormal stock market returns from all stock indexes on a specific date during this event-window. The cumulative average abnormal stock return (CAAR) is the sum of all average abnormal returns (AAR’s). The blue line is the first election round, yellow is the second election round. (*, Significant values are retrieved from table A: significant table on page 18.) AAR T-STAT T = 20 -0.40% -0.44 T = 19 0.67% 0.83 T = 18 0.21% 0.40 T = 17 -0.16% -0.28 T = 16 0.78% 1.07 T = 15 0.27% 0.26 T = 14 0.19% 0.22 T = 13 -0.47% -0.82 T = 12 -0.79% -0.99 T = 11 0.25% 0.08 T = 10 -0.69% -0.98 T = 9 -0.84% -0.95 T = 8 -0.34% -0.47 T = 7 1.15% 1.44 T = 6 0.09% 0.02 T = 5 0.03% 0.29 T = 4 -0.14% -0.47 T = 3 0.63% 0.97 T = 2 0.70% 1.21 T = 1 -0.02% -0.11 T = 0 -0.64% -0.84 T = -1 -0.78% -1.28 T = -2 0.33% 0.20 T = -3 -0.05% 0.02 T = -4 0.50% 0.62 T = -5 -0.29% -0.60 T = -6 0.53% 0.61 T = -7 -0.22% -0.22 T = -8 -0.08% -0.21 T = -9 -0.43% -0.60 T = -10 -0.40% -0.60 CAAR: -0.40% -0.06

After

Table 18: The Impact of the French Presidential election of 2017 the national level. The small-cap CAC small index is used in order to measure the election impact on a national level. One-day Stock market returns are calculated with the following formula: (Rit/Rit-1) -1. The blue line is the first election round, yellow is the second election round. (*, Significant values are retrieved from table A: significant table on page 18.) Day AR T-STAT T = 20 -0.56% -1.00 T = 19 0.08% 0.15 T = 18 0.04% 0.07 T = 17 -0.92% -1.65* T = 16 -0.55% -0.98 T = 15 -0.46% -0.83 T = 14 0.41% 0.73 T = 13 -0.31% -0.56 T = 12 0.02% 0.04 T = 11 -0.56% -1.00 T = 10 -0.91% -1.63* T = 9 1.05% 1.89** T = 8 0.17% 0.30 T = 7 0.16% 0.29 T = 6 -0.17% -0.31 T = 5 -0.13% -0.23 T = 4 -0.43% -0.78 T = 3 -0.10% -0.18 T = 2 -0.54% -0.97 T = 1 -0.38% -0.68 5.41**** T = 0 3.01% * T = -1 1.09% 1.96** T = -2 -0.16% -0.30 T = -3 -0.92% -1.64 T = -4 0.21% 0.37 T = -5 -0.38% -0.68 T = -6 0.93% 1.67* T = -7 -0.69% -1.24 T = -8 0.36% 0.65 T = -9 0.92% 1.66* T = -10 -0.84% -1.50*

Table 19: The impact of the French Presidential election of 2017 on the European level. The CAC^40 index is regressed in an event-study methodology with the AEX, DAX, FTSE 100, IBEX 35 and the FTSE MIB index respectively. (*, Significant values are retrieved from table A: significant table on page 18.) Netherlands Germany UK Spain Italy AR T-Stat AR T-Stat AR T-Stat AR T-Stat AR T-Stat

T = 20 -0.02% -0.05 0.18% 0.29 -0.49% -0.54 0.36% 0.58 -0.12% -0.11 T = 19 0.21% 0.55 0.27% 0.44 0.47% 0.51 -0.49% -0.80 0.98% 0.93 T = 18 0.00% 0.00 -0.23% -0.38 0.15% 0.16 -0.04% -0.07 -0.94% -0.90 T = 17 -0.24% -0.62 -0.53% -0.87 -1.14% -1.25 0.53% 0.86 0.55% 0.52 T = 16 -0.01% -0.03 0.25% 0.41 -0.80% -0.87 -0.10% -0.17 -0.09% -0.08 T = 15 -0.22% -0.58 0.03% 0.05 -0.36% -0.40 -0.39% -0.64 -0.51% -0.49 T =14 0.32% 0.85 -0.27% -0.45 0.17% 0.19 -0.23% -0.38 0.08% 0.07 T = 13 -0.07% -0.17 -0.14% -0.23 -0.45% -0.49 1.23% 2.00** 0.15% 0.14 T = 12 -0.06% -0.17 -0.09% -0.15 -0.05% -0.05 0.03% 0.05 -0.54% -0.51 T = 11 -0.10% -0.26 -0.07% -0.12 -0.35% -0.38 0.69% 1.13 -0.12% -0.11 T = 10 -0.92% -2.42 -0.51% -0.83 -0.73% -0.80 -0.19% -0.32 -0.04% -0.04 T = 9 0.57% 1.49* 0.61% 1.00 1.52% 1.67* 0.03% 0.05 -0.31% -0.30 T = 8 0.75% 1.99** 0.28% 0.47 0.58% 0.63 -0.20% -0.32 -0.97% -0.93 T = 7 -0.34% -0.90 -0.21% -0.34 0.25% 0.28 -0.51% -0.83 -0.31% -0.29 T = 6 -0.12% -0.31 0.09% 0.15 0.00% 0.00 -0.20% -0.33 0.23% 0.22 T = 5 -0.01% -0.02 -0.15% -0.25 0.50% 0.55 -0.37% -0.61 -0.17% -0.16 T =4 0.14% 0.38 -0.01% -0.01 0.10% 0.11 0.38% 0.61 0.62% 0.59 T = 3 0.33% 0.87 0.28% 0.46 -0.01% -0.01 0.90% 1.47* 0.09% 0.08 T = 2 0.27% 0.71 -0.10% -0.16 -0.11% -0.12 -0.23% -0.38 -0.17% -0.16 T = 1 -0.49% -1.29 -1.15% -1.89** -0.75% -0.82 -0.51% -0.83 -2.41% -2.31*** T = 0 2.31% 6.09***** 1.91% 3.14***** 4.05% 4.43***** 1.05% 1.71* 2.38% 2.27*** T = -1 0.95% 2.51**** 0.89% 1.47* 1.38% 1.51* 1.07% 1.74* 0.73% 0.70 T = -2 0.32% 0.84 0.47% 0.77 0.28% 0.30 -0.60% -0.98 -1.04% -0.99 T = -3 -1.05% -2.77**** -0.82% -1.35 -0.80% -0.88 -0.92% -1.50* -0.17% -0.17 T = -4 -0.11% -0.28 -0.23% -0.38 -0.14% -0.15 -0.39% -0.64 0.31% 0.30 T = -5 0.06% 0.16 -0.18% -0.29 -0.39% -0.43 0.40% 0.65 0.71% 0.68 T = -6 0.23% 0.60 0.34% 0.56 0.44% 0.48 0.01% 0.01 0.33% 0.31 T = -7 -0.46% -1.22 -0.05% -0.07 -0.81% -0.88 0.37% 0.61 0.27% 0.26 T = -8 -0.27% -0.71 -0.12% -0.19 0.17% 0.19 0.08% 0.13 0.23% 0.22 T = -9 0.06% 0.16 0.59% 0.97 0.95% 1.04 -0.26% -0.43 0.09% 0.09 T = -10 -0.17% -0.44 0.24% 0.39 -0.88% -0.96 -0.54% -0.88 -0.18% -0.18

Table 20: The average abnormal stock return (AAR) together with the cumulative average abnormal return (CAAR) and T-stats per day during the six-week period. The average abnormal returns are calculated by first calculating the abnormal stock returns per day for each stock market index in the sample during the event-window. Then the average abnormal stock market returns from all stock indexes on a specific date during this event-window. The cumulative average abnormal stock return (CAAR) is the sum of all average abnormal returns (AAR’s). The blue line is the first election round, yellow is the second election round. (*, Significant values are retrieved from table A: significant table on page 18.) AAR T-STAT T = 20 -0.02% 0.03 T = 19 0.29% 0.33 T = 18 -0.21% -0.24 T = 17 -0.17% -0.27 T = 16 -0.15% -0.15 T = 15 -0.29% -0.41 T = 14 0.01% 0.06 T = 13 0.14% 0.25 T = 12 -0.14% -0.17 T = 11 0.01% 0.05 T = 10 -0.48% -0.88 T = 9 0.48% 0.78 T = 8 0.09% 0.37 T = 7 -0.22% -0.42 T = 6 0.00% -0.05 T = 5 -0.04% -0.10 T = 4 0.25% 0.34 T = 3 0.32% 0.57 T = 2 -0.07% -0.02 T = 1 -1.06% -1.43 T = 0 2.34% 3.53 T = -1 1.00% 1.58 T = -2 -0.11% -0.01 T = -3 -0.75% -1.33 T = -4 -0.11% -0.23 T = -5 0.12% 0.15 T = -6 0.27% 0.39 T = -7 -0.14% -0.26 T = -8 0.02% -0.07 T = -9 0.29% 0.37 T = -10 -0.31% -0.41 CAAR: 1.35% 1.16

7.3 German federal elections

Before

Table 22: The Impact of the German federal election of 2005 the national level. The small-cap SDAX index is used in order to measure the election impact on a national level.

One-day Stock market returns are calculated with the following formula: (Rit/Rit-1) -1. (*, Significant values are retrieved from table A: significant table on page 18.) Day AR T-test T = 20 0.59% 1.22 T = 19 -0.24% -0.49 T = 18 -0.34% -0.70 T = 17 -0.34% -0.71 T = 16 -0.29% -0.60 T = 15 0.32% 0.66 T =14 -0.78% -1.60 T = 13 -1.32% -2.70**** T = 12 0.42% 0.85 T = 11 0.17% 0.35 T = 10 -0.25% -0.51 T = 9 0.00% 0.00 T = 8 1.17% 2.41*** T = 7 -0.10% -0.20 T = 6 0.61% 1.25 T = 5 1.32% 2.72**** T =4 0.39% 0.81 T = 3 -1.54% -3.16***** T = 2 -0.57% -1.17 T = 1 -0.27% -0.56 T = 0 0.44% 0.89 T = -1 0.16% 0.32 T = -2 -0.27% -0.56 T = -3 -1.51% -3.11***** T = -4 -0.89% -1.83** T = -5 0.60% 1.23 T = -6 0.01% 0.03 T = -7 -0.03% -0.07 T = -8 1.36% 2.80**** T = -9 0.97% 1.99** T = -10 -0.05% -0.10

Table 23: The impact of the German federal election of 2005 on the European level. The DAX index is regressed in an event-study methodology with the AEX, CAC^40, FTSE 100, IBEX 35 and the FTSE MIB index respectively. (*, Significant values are retrieved from table A: significant table on page 18.) Country Netherlands France UK Spain Italy AR T-Stat AR T-Stat AR T-Stat AR T-Stat AR T-Stat

T = 20 -0.03% -0.05 -0.01% -0.03 0.57% 1.25 0.64% 1.30 0.47% 1.02

T = 19 -0.35% -0.64 0.21% 0.42 0.15% 0.33 -0.24% -0.50 -0.32% -0.68 T = 18 0.16% 0.29 -0.10% -0.21 -0.25% -0.55 -0.13% -0.27 -0.33% -0.71 T = 17 0.09% 0.16 -0.22% -0.45 -0.37% -0.81 -0.17% -0.34 0.01% 0.02 T = 16 0.15% 0.27 -0.03% -0.06 -0.26% -0.56 -0.41% -0.83 0.02% 0.04 T = 15 0.17% 0.31 0.09% 0.18 0.53% 1.16 0.37% 0.75 0.30% 0.65 T =14 -0.07% -0.13 -0.29% -0.58 -0.23% -0.50 -0.80% -1.65* -0.40% -0.86 T = 13 -0.65% -1.19 -0.34% -0.69 -0.93% -2.04** -0.79% -1.63* -0.43% -0.93 T = 12 0.90% 1.66* 0.59% 1.18 0.57% 1.25 0.73% 1.50* 1.24% 2.68**** T = 11 0.07% 0.14 0.23% 0.46 0.26% 0.57 0.16% 0.34 0.40% 0.86 T = 10 -0.24% -0.45 0.22% 0.44 0.09% 0.19 0.06% 0.13 0.31% 0.66 T = 9 -0.09% -0.16 -0.23% -0.46 0.22% 0.47 -0.02% -0.04 -0.07% -0.14 T = 8 0.86% 1.58* 0.78% 1.56* 0.94% 2.07** 0.70% 1.44 0.94% 2.02** T = 7 0.03% 0.06 -0.05% -0.11 0.11% 0.23 0.06% 0.13 0.07% 0.16 T = 6 0.37% 0.69 0.11% 0.22 0.50% 1.10 0.56% 1.15 0.36% 0.79 T = 5 0.96% 1.76* 0.62% 1.24 1.21% 2.66**** 0.77% 1.58* 1.20% 2.59**** T =4 0.51% 0.95 0.12% 0.23 0.05% 0.10 -0.26% -0.53 0.32% 0.69 T = 3 -0.91% -1.67* -0.88% -1.75* -1.18% -2.58**** -1.24% -2.55**** -1.30% -2.81**** T = 2 0.21% 0.39 0.26% 0.52 0.08% 0.18 0.05% 0.09 0.10% 0.21 T = 1 -0.45% -0.82 -0.37% -0.74 -0.08% -0.17 -0.18% -0.37 0.05% 0.11 T = 0 -0.11% -0.20 0.17% 0.33 0.07% 0.16 0.03% 0.07 0.05% 0.11 T = -1 0.06% 0.12 -0.10% -0.20 -0.12% -0.25 0.05% 0.10 -0.01% -0.02 T = -2 -0.15% -0.27 -0.10% -0.19 -0.16% -0.35 -0.20% -0.41 -0.45% -0.97 T = -3 -1.18% -2.17*** -1.15% -2.30*** -1.37% -3.01***** -1.42% -2.90***** -1.33% -2.87**** T = -4 -0.61% -1.12 -0.60% -1.20 -0.92% -2.02** -0.56% -1.15 -0.52% -1.12 T = -5 0.21% 0.38 0.07% 0.14 0.42% 0.91 0.15% 0.30 0.19% 0.41 T = -6 0.21% 0.39 0.24% 0.48 0.37% 0.81 -0.04% -0.08 -0.08% -0.16 T = -7 0.28% 0.51 0.41% 0.82 0.17% 0.37 0.05% 0.11 0.20% 0.43 T = -8 0.94% 1.73* 0.67% 1.33 1.32% 2.90***** 0.66% 1.35 1.06% 2.29*** T = -9 0.55% 1.01 0.48% 0.95 0.92% 2.01** 0.41% 0.83 0.75% 1.61* T = -10 0.53% 0.97 0.39% 0.77 0.25% 0.55 -0.01% -0.01 0.21% 0.45

Table 24: The average abnormal stock return (AAR) together with the cumulative average abnormal return (CAAR) and T-stats per day during the six-week period. The average abnormal returns are calculated by first calculating the abnormal stock returns per day for each stock market index in the sample during the event-window. Then the average abnormal stock market returns from all stock indexes on a specific date during this event-window. The cumulative average abnormal stock return (CAAR) is the sum of all average abnormal returns (AAR’s). (*, Significant values are retrieved from table A: significant table on page 18.) AAR T-STAT T = 20 0.33% 0.70 T = 19 -0.11% -0.21 T = 18 -0.13% -0.29 T = 17 -0.13% -0.28 T = 16 -0.11% -0.23 T = 15 0.29% 0.61 T = 14 -0.36% -0.74 T = 13 -0.63% -1.29 T = 12 0.81% 1.65* T = 11 0.22% 0.47 T = 10 0.09% 0.19 T = 9 -0.04% -0.07 T = 8 0.84% 1.73 T = 7 0.04% 0.09 T = 6 0.38% 0.79 T = 5 0.95% 1.97** T = 4 0.15% 0.29 T = 3 -1.10% -2.27*** T = 2 0.14% 0.28 T = 1 -0.21% -0.40 T = 0 0.04% 0.09 T = -1 -0.02% -0.05 T = -2 -0.21% -0.44 T = -3 -1.29% -2.65**** T = -4 -0.64% -1.32 T = -5 0.21% 0.43 T = -6 0.14% 0.29 T = -7 0.22% 0.45 T = -8 0.93% 1.92** T = -9 0.62% 1.28 T = -10 0.27% 0.55 CAAR: 1.71% 1.01

During

Table 25: The Impact of the German federal election of 2009 the national level. The small cap SDAX index is used in order to measure the election impact on a national level.

One-day Stock market returns are calculated with the following formula: (Rit/Rit-1) -1. (*, Significant values are retrieved from table A: significant table on page 18.) Day AR T-test T = 20 -0.22% -0.21 T = 19 -0.13% -0.13 T = 18 0.51% 0.48 T = 17 -0.46% -0.44 T = 16 1.31% 1.25 T = 15 -1.30% -1.24 T =14 -0.11% -0.11 T = 13 1.33% 1.27 T = 12 -0.28% -0.27 T = 11 0.72% 0.68 T = 10 0.03% 0.03 T = 9 0.19% 0.18 T = 8 0.07% 0.06 T = 7 1.12% 1.07 T = 6 0.42% 0.40 T = 5 -0.51% -0.49 T =4 -1.35% -1.29 T = 3 -1.35% -1.29 T = 2 -0.09% -0.09 T = 1 1.87% 1.78* T = 0 -0.31% -0.30 T = -1 -0.80% -0.77 T = -2 -0.10% -0.10 T = -3 0.12% 0.11 T = -4 0.74% 0.70 T = -5 -0.45% -0.43 T = -6 0.32% 0.30 T = -7 0.78% 0.74 T = -8 -0.40% -0.38 T = -9 1.27% 1.21 T = -10 -0.35% -0.34

Table 26: The impact of the German federal election of 2009 on the European level. The DAX index is regressed in an event-study methodology with the AEX, CAC^40, FTSE 100, IBEX 35 and the FTSE MIB index respectively. (*, Significant values are retrieved from table A: significant table on page 18.) Country Netherlands France UK Spain Italy AR T-Stat AR T-Stat AR T-Stat AR T-Stat AR T-Stat

T = 20 0.48% 0.38 0.44% 0.41 -1.09% -0.80 0.42% 0.39 1.22% 1.03 T = 19 -1.17% -0.93 -1.13% -1.07 0.66% 0.48 -1.80% -1.68* 1.19% 1.00 T = 18 1.32% 1.04 1.12% 1.06 -0.65% -0.47 0.17% 0.16 -0.11% -0.10 T = 17 0.12% 0.09 0.17% 0.17 0.23% 0.17 0.67% 0.62 0.33% 0.28 T = 16 0.28% 0.22 0.01% 0.01 -0.11% -0.08 0.15% 0.14 0.60% 0.50 T = 15 -0.52% -0.42 -0.20% -0.18 -0.82% -0.60 -0.40% -0.38 -0.65% -0.55 T =14 -0.08% -0.07 -0.24% -0.22 0.50% 0.36 -0.06% -0.06 -0.05% -0.04 T = 13 0.90% 0.72 0.68% 0.65 -0.30% -0.22 0.27% 0.25 0.04% 0.03 T = 12 -1.35% -1.06 -0.61% -0.58 0.71% 0.52 -0.36% -0.33 0.53% 0.44 T = 11 0.12% 0.09 0.29% 0.28 -0.05% -0.04 1.20% 1.12 -0.57% -0.48 T = 10 -1.04% -0.82 -0.19% -0.18 0.31% 0.23 0.80% 0.75 0.07% 0.06 T = 9 0.72% 0.57 1.10% 1.04 -0.51% -0.37 0.74% 0.69 -0.59% -0.50 T = 8 -1.15% -0.91 -0.79% -0.75 1.08% 0.79 0.24% 0.22 0.75% 0.63 T = 7 0.49% 0.39 0.41% 0.39 0.12% 0.09 0.64% 0.60 -0.22% -0.19 T = 6 -0.20% -0.15 -0.30% -0.29 0.03% 0.02 -1.39% -1.29 0.04% 0.03 T = 5 -0.22% -0.17 0.20% 0.19 0.18% 0.13 -0.51% -0.48 0.73% 0.61 T =4 0.58% 0.46 0.18% 0.17 -0.95% -0.69 0.22% 0.21 -0.86% -0.72 T = 3 -0.65% -0.51 -0.19% -0.18 -0.31% -0.23 -0.06% -0.05 -0.39% -0.32 T = 2 -0.06% -0.05 0.01% 0.01 -0.03% -0.02 0.31% 0.29 -0.45% -0.38 T = 1 -0.02% -0.02 -0.20% -0.19 0.88% 0.65 0.15% 0.14 1.01% 0.84 T = 0 1.06% 0.84 0.53% 0.50 -0.42% -0.31 0.16% 0.15 -0.94% -0.79 T = -1 -0.42% -0.33 0.07% 0.06 -0.04% -0.03 -0.36% -0.33 0.02% 0.01 T = -2 0.13% 0.11 0.00% 0.00 -0.31% -0.23 -0.35% -0.32 -1.04% -0.87 T = -3 0.09% 0.07 0.48% 0.45 0.17% 0.13 0.11% 0.10 0.30% 0.25 T = -4 0.29% 0.23 -0.18% -0.17 0.54% 0.39 -0.31% -0.28 1.63% 1.37 T = -5 -0.65% -0.51 -0.49% -0.46 -0.40% -0.29 -0.56% -0.53 0.23% 0.19 T = -6 0.86% 0.68 0.54% 0.51 -0.97% -0.71 0.60% 0.56 -0.63% -0.52 T = -7 -0.16% -0.13 -0.45% -0.43 -0.18% -0.13 0.07% 0.07 0.00% 0.00 T = -8 -0.74% -0.59 -0.66% -0.63 -0.40% -0.29 -0.99% -0.92 -0.21% -0.18 T = -9 -0.40% -0.32 -0.62% -0.59 0.52% 0.38 -0.72% -0.67 0.76% 0.64 T = -10 0.96% 0.76 0.74% 0.70 -0.51% -0.38 0.30% 0.28 -0.79% -0.66

Table 27: The average abnormal stock return (AAR) together with the cumulative average abnormal return (CAAR) and T-stats per day during the six-week period. The average abnormal returns are calculated by first calculating the abnormal stock returns per day for each stock market index in the sample during the event-window. Then the average abnormal stock market returns from all stock indexes on a specific date during this event-window. The cumulative average abnormal stock return (CAAR) is the sum of all average abnormal returns (AAR’s). (*, Significant values are retrieved from table A: significant table on page 18.) AAR T-STAT T = 20 0.33% 0.31 T = 19 -0.50% -0.48 T = 18 0.38% 0.35 T = 17 0.26% 0.24 T = 16 0.28% 0.22 T = 15 -0.55% -0.45 T = 14 -0.02% -0.03 T = 13 0.42% 0.36 T = 12 -0.27% -0.24 T = 11 0.25% 0.23 T = 10 0.00% 0.01 T = 9 0.34% 0.32 T = 8 0.00% -0.03 T = 7 0.41% 0.34 T = 6 -0.33% -0.31 T = 5 0.01% 0.01 T = 4 -0.25% -0.18 T = 3 -0.35% -0.28 T = 2 -0.05% -0.03 T = 1 0.45% 0.35 T = 0 0.08% 0.08 T = -1 -0.21% -0.17 T = -2 -0.32% -0.27 T = -3 0.24% 0.21 T = -4 0.35% 0.28 T = -5 -0.37% -0.31 T = -6 0.12% 0.13 T = -7 -0.06% -0.06 T = -8 -0.58% -0.51 T = -9 -0.09% -0.11 T = -10 0.16% 0.16 CAAR: 0.12% 0.10

7.4 United Kingdom general elections

Before

Table 29: The Impact of the UK general election of 2005 the national level. The FTSE Small Cap index is used in order to measure the election impact on a national level.

One-day Stock market returns are calculated with the following formula: (Rit/Rit-1) -1. (*, Significant values are retrieved from table A: significant table on page 18.) Day AR T-test T = 20 0.71% 1.97** T = 19 -0.61% -1.67* T = 18 -0.09% -0.24 T = 17 -0.40% -1.11 T = 16 0.25% 0.68 T = 15 -0.39% -1.08 T =14 -0.27% -0.74 T = 13 0.23% 0.63 T = 12 0.03% 0.09 T = 11 0.10% 0.27 T = 10 0.88% 2.41*** T = 9 0.16% 0.43 T = 8 -0.10% -0.29 T = 7 -0.22% -0.60 T = 6 0.20% 0.56 T = 5 -0.23% -0.64 T =4 -0.22% -0.60 T = 3 -0.20% -0.54 T = 2 0.27% 0.76 T = 1 0.21% 0.58 T = 0 0.47% 1.29 T = -1 1.18% 3.25***** T = -2 -0.09% -0.24 T = -3 0.30% 0.82 T = -4 0.25% 0.68 T = -5 -0.91% -2.50**** T = -6 -0.29% -0.80 T = -7 0.26% 0.71 T = -8 0.38% 1.06 T = -9 -0.24% -0.65 T = -10 -0.76% -2.09**

Table 30: The impact of the UK general election of 2005 on the European level. The FTSE 100 index is regressed in an event-study methodology with the AEX, CAC^40, DAX, IBEX 35 and the FTSE MIB index respectively. (*, Significant values are retrieved from table A: significant table on page 18.) Country Netherlands France German Spain Italy AR T-Stat AR T-Stat AR T-Stat AR T-Stat AR T-Stat

T = 20 -0.32% -0.35 -0.09% -0.10 -0.42% -0.51 0.34% 0.47 0.25% 0.46

T = 19 0.18% 0.19 -0.32% -0.39 -0.17% -0.21 -0.42% -0.59 -0.27% -0.51 T = 18 -0.36% -0.39 -0.51% -0.63 -0.89% -1.07 -0.72% -1.00 -0.44% -0.82 T = 17 -0.03% -0.03 0.35% 0.43 -0.27% -0.32 0.15% 0.20 0.07% 0.14 T = 16 0.15% 0.16 -0.28% -0.34 -0.84% -1.00 -0.03% -0.04 -0.05% -0.10 T = 15 -0.56% -0.60 -0.33% -0.40 -0.21% -0.25 -0.24% -0.33 -0.55% -1.03 T =14 0.05% 0.05 0.05% 0.06 0.29% 0.34 0.13% 0.18 0.13% 0.24 T = 13 0.18% 0.19 0.38% 0.47 -0.41% -0.49 0.35% 0.49 -0.45% -0.83 T = 12 -0.27% -0.30 -0.32% -0.39 -0.47% -0.56 0.10% 0.14 0.74% 1.37 T = 11 -0.17% -0.18 0.02% 0.03 -0.29% -0.34 0.11% 0.16 0.45% 0.83 T = 10 -0.02% -0.03 -0.13% -0.17 -0.64% -0.77 0.04% 0.05 -0.19% -0.36 T = 9 0.02% 0.02 0.38% 0.46 0.26% 0.31 0.22% 0.31 0.90% 1.68* T = 8 -0.29% -0.32 -0.69% -0.84 -0.28% -0.33 -0.20% -0.28 -0.22% -0.41 T = 7 -0.39% -0.42 -0.05% -0.06 -0.46% -0.55 -0.70% -0.98 -0.55% -1.03 T = 6 0.29% 0.31 0.59% 0.72 0.57% 0.69 0.61% 0.85 0.44% 0.81 T = 5 -0.43% -0.47 -1.01% -1.23 -0.86% -1.02 -0.72% -1.00 -0.91% -1.69* T =4 0.35% 0.38 0.41% 0.50 0.94% 1.12 0.26% 0.36 0.44% 0.82 T = 3 -0.18% -0.19 -0.08% -0.09 -0.06% -0.07 -0.20% -0.27 -0.17% -0.31 T = 2 0.28% 0.30 0.14% 0.17 0.08% 0.10 0.15% 0.21 0.13% 0.25 T = 1 -0.53% -0.58 -0.10% -0.12 -0.30% -0.36 0.41% 0.58 0.57% 1.05 T = 0 -0.29% -0.31 -0.12% -0.14 -0.16% -0.19 -0.03% -0.05 0.12% 0.22 T = -1 1.17% 1.27 0.92% 1.13 0.70% 0.84 0.70% 0.97 0.96% 1.78* T = -2 -0.23% -0.25 -0.51% -0.62 -0.59% -0.71 -0.50% -0.69 -0.42% -0.78 T = -3 -0.44% -0.47 -0.70% -0.85 -0.99% -1.18 -0.30% -0.42 -0.54% -1.01 T = -4 1.19% 1.29 1.09% 1.33 1.06% 1.26 0.38% 0.52 0.76% 1.42 T = -5 -0.03% -0.03 -0.82% -1.00 -0.81% -0.96 -0.86% -1.20 -0.80% -1.48 T = -6 -0.05% -0.05 0.28% 0.34 0.27% 0.32 -0.12% -0.17 0.07% 0.13 T = -7 -0.16% -0.17 -0.17% -0.21 -0.55% -0.65 -0.16% -0.22 0.01% 0.01 T = -8 0.92% 1.00 0.66% 0.80 0.45% 0.54 0.96% 1.34 0.93% 1.72* T = -9 -1.18% -1.28 -0.94% -1.15 -1.40% -1.67* -1.00% -1.40 -1.49% -2.77**** T = -10 0.72% 0.78 0.15% 0.19 0.70% 0.83 0.04% 0.05 0.20% 0.38

Table 31: The average abnormal stock return (AAR) together with the cumulative average abnormal return (CAAR) and T-stats per day during the six-week period. The average abnormal returns are calculated by first calculating the abnormal stock returns per day for each stock market index in the sample during the event-window. Then the average abnormal stock market returns from all stock indexes on a specific date during this event-window. The cumulative average abnormal stock return (CAAR) is the sum of all average abnormal returns (AAR’s). (*, Significant values are retrieved from table A: significant table on page 18.) AAR T-STAT T = 20 -0.05% 0.00 T = 19 -0.20% -0.30 T = 18 -0.59% -0.78 T = 17 0.05% 0.08 T = 16 -0.21% -0.27 T = 15 -0.38% -0.52 T = 14 0.13% 0.17 T = 13 0.01% -0.03 T = 12 -0.04% 0.05 T = 11 0.03% 0.10 T = 10 -0.19% -0.25 T = 9 0.36% 0.56 T = 8 -0.34% -0.44 T = 7 -0.43% -0.61 T = 6 0.50% 0.68 T = 5 -0.78% -1.08 T = 4 0.48% 0.64 T = 3 -0.13% -0.19 T = 2 0.16% 0.21 T = 1 0.01% 0.12 T = 0 -0.10% -0.10 T = -1 0.89% 1.20 T = -2 -0.45% -0.61 T = -3 -0.59% -0.79 T = -4 0.89% 1.17 T = -5 -0.66% -0.93 T = -6 0.09% 0.12 T = -7 -0.21% -0.25 T = -8 0.78% 1.08 T = -9 -1.20% -1.66 T = -10 0.36% 0.45 CAAR: -1.81% -0.78

100

During

Table 32: The Impact of the UK general election of 2010 the national level. The FTSE Small Cap index is used in order to measure the election impact on a national level.

One-day Stock market returns are calculated with the following formula: (Rit/Rit-1) -1. (*, Significant values are retrieved from table A: significant table on page 18.) Day AR T-test T = 20 0.71% 0.83 T = 19 -0.20% -0.23 T = 18 -0.55% -0.64 T = 17 -0.10% -0.12 T = 16 2.67% 3.13***** T = 15 1.32% 1.54* T =14 -1.19% -1.39 T = 13 -0.41% -0.48 T = 12 0.16% 0.18 T = 11 -0.81% -0.94 T = 10 -2.18% -2.55**** T = 9 1.03% 1.21 T = 8 0.26% 0.30 T = 7 -2.61% -3.05** T = 6 0.60% 0.70 T = 5 0.55% 0.65 T =4 -0.58% -0.68 T = 3 3.59% 4.19***** T = 2 -1.23% -1.44 T = 1 -1.33% -1.56* T = 0 -0.84% -0.98 T = -1 -1.86% -2.18*** T = -2 -1.09% -1.28 T = -3 0.35% 0.41 T = -4 -0.22% -0.26 T = -5 -2.00% -2.34*** T = -6 0.10% 0.12 T = -7 0.45% 0.53 T = -8 -0.31% -0.36 T = -9 -1.11% -1.30 T = -10 0.82% 0.96

101

Table 33: The impact of the UK general election of 2010 on the European level. The FTSE 100 index is regressed in an event-study methodology with the AEX, CAC^40, DAX, IBEX 35 and the FTSE MIB index respectively. (*, Significant values are retrieved from table A: significant table on page 18.) Country Netherlands France German Spain Italy AR T-Stat AR T-Stat AR T-Stat AR T-Stat AR T-Stat

T = 20 1.26% 0.67 1.27% 0.71 1.10% 0.84 2.35% 1.44 1.52% 0.88 T = 19 -2.12% -1.12 -1.89% -1.06 -2.12% -1.62* -1.89% -1.16 -2.05% -1.19 T = 18 -0.30% -0.16 -0.29% -0.17 -0.22% -0.17 0.01% 0.00 0.88% 0.51 T = 17 0.76% 0.40 1.06% 0.60 0.11% 0.09 0.99% 0.61 1.03% 0.60 T = 16 1.31% 0.69 1.12% 0.63 1.09% 0.83 1.01% 0.62 0.11% 0.06 T = 15 1.16% 0.61 1.32% 0.74 1.12% 0.85 2.18% 1.34 1.18% 0.69 T =14 -3.85% -2.04** -3.68% -2.07** -3.52% -2.69**** -3.77% -2.32*** -3.72% -2.17** T = 13 2.54% 1.34 2.40% 1.35 2.65% 2.02** 3.63% 2.23*** 3.64% 2.12** T = 12 -0.76% -0.40 -0.63% -0.36 -0.49% -0.37 -1.64% -1.01 -0.31% -0.18 T = 11 0.43% 0.23 -0.15% -0.09 0.66% 0.51 -0.44% -0.27 0.36% 0.21 T = 10 -2.21% -1.17 -2.22% -1.25 -1.97% -1.51* -3.05% -1.87** -2.51% -1.46 T = 9 1.82% 0.96 1.58% 0.89 1.49% 1.14 1.21% 0.74 1.68% 0.98 T = 8 -1.31% -0.70 -1.20% -0.68 -1.39% -1.07 -1.66% -1.02 -1.57% -0.91 T = 7 -0.82% -0.43 -0.31% -0.17 -0.81% -0.62 0.97% 0.60 0.45% 0.26 T = 6 1.68% 0.89 1.99% 1.12 1.11% 0.85 2.93% 1.80* 2.86% 1.67* T = 5 -0.42% -0.22 -0.76% -0.43 -1.70% -1.30 -0.84% -0.52 -0.64% -0.37 T =4 -1.20% -0.63 -1.13% -0.63 -2.27% -1.74* -0.18% -0.11 -1.17% -0.68 T = 3 3.73% 1.97** 4.40% 2.47**** 3.80% 2.91**** -1.47% -0.91 -0.60% -0.35 T = 2 -4.00% -2.12** -5.13% -2.88**** -2.92% -2.23*** -1.82% -1.12 -2.94% -1.71* T = 1 0.53% 0.28 0.60% 0.34 0.11% 0.08 0.83% 0.51 2.21% 1.29 T = 0 0.33% 0.17 0.39% 0.22 -0.03% -0.02 0.75% 0.46 0.76% 0.44 T = -1 -0.27% -0.14 -0.43% -0.24 -0.70% -0.53 0.39% 0.24 0.77% 0.45 T = -2 -1.03% -0.54 -1.08% -0.61 -1.39% -1.06 -0.46% -0.28 -0.79% -0.46 T = -3 0.08% 0.04 0.08% 0.04 0.02% 0.02 -1.11% -0.68 0.14% 0.08 T = -4 -0.06% -0.03 0.01% 0.01 0.07% 0.06 0.67% 0.41 0.46% 0.27 T = -5 -0.78% -0.41 -0.07% -0.04 -0.37% -0.29 0.39% 0.24 0.42% 0.25 T = -6 0.61% 0.32 0.65% 0.36 0.36% 0.27 1.43% 0.88 1.20% 0.70 T = -7 0.32% 0.17 -0.03% -0.01 -0.44% -0.34 -0.11% -0.07 0.14% 0.08 T = -8 -0.79% -0.42 -0.29% -0.17 -0.57% -0.44 0.18% 0.11 0.15% 0.09 T = -9 -0.54% -0.29 -0.04% -0.02 -0.43% -0.33 0.51% 0.31 0.04% 0.03 T = -10 -0.09% -0.05 0.28% 0.16 -0.05% -0.04 0.62% 0.38 -0.31% -0.18

102

Table 34: The average abnormal stock return (AAR) together with the cumulative average abnormal return (CAAR) and T-stats per day during the six-week period. The average abnormal returns are calculated by first calculating the abnormal stock returns per day for each stock market index in the sample during the event-window. Then the average abnormal stock market returns from all stock indexes on a specific date during this event-window. The cumulative average abnormal stock return (CAAR) is the sum of all average abnormal returns (AAR’s). (*, Significant values are retrieved from table A: significant table on page 18.) AAR T-STAT T = 20 1.50% 0.91 T = 19 -2.01% -1.23 T = 18 0.01% 0.00 T = 17 0.79% 0.46 T = 16 0.93% 0.57 T = 15 1.39% 0.85 T = 14 -3.71% -2.26 T = 13 2.97% 1.81 T = 12 -0.77% -0.46 T = 11 0.17% 0.12 T = 10 -2.39% -1.45 T = 9 1.56% 0.94 T = 8 -1.43% -0.87 T = 7 -0.10% -0.07 T = 6 2.11% 1.27 T = 5 -0.87% -0.57 T = 4 -1.19% -0.76 T = 3 1.97% 1.22 T = 2 -3.36% -2.01 T = 1 0.85% 0.50 T = 0 0.44% 0.25 T = -1 -0.05% -0.05 T = -2 -0.95% -0.59 T = -3 -0.16% -0.10 T = -4 0.23% 0.14 T = -5 -0.08% -0.05 T = -6 0.85% 0.51 T = -7 -0.03% -0.03 T = -8 -0.26% -0.16 T = -9 -0.09% -0.06 T = -10 0.09% 0.05 CAAR: -1.59% -0.34

103

After

Table 35: The Impact of the UK general election of 2015 the national level. The FTSE Small Cap index is used in order to measure the election impact on a national level.

One-day Stock market returns are calculated with the following formula: (Rit/Rit-1) -1. (*, Significant values are retrieved from table A: significant table on page 18.) Day AR T-test T = 20 -1.09% -3.16***** T = 19 0.23% 0.67 T = 18 -0.47% -1.36 T = 17 -0.65% -1.89** T = 16 -0.82% -2.37*** T = 15 -0.01% -0.02 T =14 1.08% 3.15***** T = 13 -1.13% -3.29***** T = 12 0.05% 0.15 T = 11 0.07% 0.22 T = 10 0.02% 0.06 T = 9 0.17% 0.48 T = 8 -0.15% -0.43 T = 7 -0.16% -0.48 T = 6 0.03% 0.08 T = 5 0.01% 0.03 T =4 -1.08% -3.12***** T = 3 -0.51% -1.49 T = 2 1.62% 4.72***** T = 1 -0.33% -0.97 T = 0 -0.05% -0.15 T = -1 -0.94% -2.72**** T = -2 0.52% 1.52* T = -3 0.01% 0.03 T = -4 -1.33% -3.85***** T = -5 -0.77% -2.22*** T = -6 0.21% 0.60 T = -7 0.12% 0.35 T = -8 0.31% 0.90 T = -9 -0.42% -1.23 T = -10 0.01% 0.04

104

Table 36: The impact of the British general election of 2015 on the European level. The FTSE 100 index is regressed in an event-study methodology with the AEX, CAC^40, DAX, IBEX 35 and the FTSE MIB index respectively. (*, Significant values are retrieved from table A: significant table on page 18.) Country Netherlands France German Spain Italy AR T-Stat AR T-Stat AR T-Stat AR T-Stat AR T-Stat

T = 20 -0.83% -1.22 -0.60% -0.78 -0.61% -0.75 -0.54% -0.55 -0.12% -0.11 T = 19 0.45% 0.66 0.41% 0.53 0.17% 0.21 0.85% 0.87 0.61% 0.54 T = 18 0.44% 0.65 0.07% 0.10 0.20% 0.25 -0.37% -0.37 -0.68% -0.61 T = 17 -0.11% -0.16 -0.82% -1.06 -0.55% -0.67 -0.81% -0.83 -0.31% -0.28 T = 16 -0.14% -0.20 0.63% 0.82 0.43% 0.53 -0.22% -0.22 -0.19% -0.17 T = 15 0.12% 0.17 0.33% 0.43 0.21% 0.25 0.11% 0.11 -0.07% -0.06 T =14 0.34% 0.51 0.44% 0.58 0.72% 0.89 0.50% 0.51 0.05% 0.04 T = 13 -0.83% -1.22 -1.00% -1.30 -0.12% -0.15 -1.16% -1.17 -1.65% -1.47 T = 12 0.46% 0.67 0.54% 0.71 0.41% 0.51 1.92% 1.95** 1.43% 1.28 T = 11 -0.15% -0.22 -0.33% -0.44 -0.38% -0.47 -0.44% -0.44 0.07% 0.06 T = 10 -0.01% -0.01 -0.09% -0.11 0.34% 0.42 0.02% 0.02 -0.08% -0.07 T = 9 -0.58% -0.85 -0.60% -0.79 -0.58% -0.71 -0.28% -0.29 -0.79% -0.71 T = 8 -0.21% -0.31 -0.55% -0.72 -1.44% -1.77* -0.37% -0.37 -0.42% -0.37 T = 7 -0.08% -0.12 0.50% 0.65 0.51% 0.62 0.35% 0.35 1.08% 0.97 T = 6 -0.82% -1.20 -1.25% -1.63* -1.54% -1.89** -0.54% -0.55 -1.14% -1.02 T = 5 0.93% 1.38 0.93% 1.22 1.55% 1.90** 0.52% 0.53 0.02% 0.02 T =4 -1.34% -1.98** -1.53% -1.99** -1.30% -1.59* -1.64% -1.67* -1.41% -1.26 T = 3 -0.22% -0.32 0.57% 0.75 0.37% 0.45 0.29% 0.29 0.12% 0.11 T = 2 1.04% 1.53* 1.54% 2.00** 1.09% 1.33 1.21% 1.22 1.27% 1.14 T = 1 -1.78% -2.62**** -1.50% -1.95** -2.09% -2.57**** -1.57% -1.59* -2.59% -2.31*** T = 0 1.03% 1.52* 0.35% 0.46 0.69% 0.85 -0.14% -0.14 0.34% 0.31 T = -1 0.19% 0.28 0.59% 0.77 0.27% 0.33 0.96% 0.97 0.92% 0.82 T = -2 -0.01% -0.01 0.11% 0.14 -0.14% -0.17 0.05% 0.05 0.06% 0.06 T = -3 0.38% 0.57 -0.13% -0.17 -0.27% -0.33 -0.02% -0.02 -0.28% -0.25 T = -4 0.33% 0.48 0.61% 0.79 1.19% 1.46 0.29% 0.29 1.04% 0.93 T = -5 -0.39% -0.57 -0.01% -0.01 -0.13% -0.16 -0.89% -0.91 -0.65% -0.58 T = -6 -0.03% -0.04 -0.38% -0.49 -0.51% -0.62 -0.33% -0.34 -0.78% -0.70 T = -7 0.13% 0.19 0.24% 0.31 -0.41% -0.51 -0.20% -0.20 -0.16% -0.14 T = -8 0.42% 0.61 0.50% 0.65 0.88% 1.08 -0.02% -0.02 0.20% 0.18 T = -9 -0.23% -0.33 -0.63% -0.83 -0.01% -0.01 -0.36% -0.37 -0.61% -0.55 T = -10 -0.45% -0.66 0.02% 0.03 0.13% 0.16 0.00% 0.00 0.45% 0.41

105

Table 37: The average abnormal stock return (AAR) together with the cumulative average abnormal return (CAAR) and T-stats per day during the six-week period. The average abnormal returns are calculated by first calculating the abnormal stock returns per day for each stock market index in the sample during the event-window. Then the average abnormal stock market returns from all stock indexes on a specific date during this event-window. The cumulative average abnormal stock return (CAAR) is the sum of all average abnormal returns (AAR’s). (*, Significant values are retrieved from table A: significant table on page 18.) AAR T-STAT T = 20 -0.54% -0.68 T = 19 0.50% 0.56 T = 18 -0.07% 0.00 T = 17 -0.52% -0.60 T = 16 0.10% 0.15 T = 15 0.14% 0.18 T = 14 0.41% 0.51 T = 13 -0.95% -1.06 T = 12 0.95% 1.02 T = 11 -0.25% -0.30 T = 10 0.04% 0.05 T = 9 -0.57% -0.67 T = 8 -0.60% -0.71 T = 7 0.47% 0.50 T = 6 -1.06% -1.26 T = 5 0.79% 1.01 T = 4 -1.44% -1.70* T = 3 0.23% 0.26 T = 2 1.23% 1.44 T = 1 -1.91% -2.21*** T = 0 0.46% 0.60 T = -1 0.59% 0.64 T = -2 0.02% 0.01 T = -3 -0.06% -0.04 T = -4 0.69% 0.79 T = -5 -0.41% -0.45 T = -6 -0.40% -0.44 T = -7 -0.08% -0.07 T = -8 0.40% 0.50 T = -9 -0.37% -0.42 T = -10 0.03% -0.01 CAAR: -2.19% -1.57

106

7.5 Spanish general elections

Before

Table 39: The Impact of the Spanish general election of 2008 the national level. The Spanish IBEX Small Cap index is used in order to measure the election impact on a national level. One-day Stock market returns are calculated with the following formula: (Rit/Rit-1) -1. (*, Significant values are retrieved from table A: significant table on page 18.) Day AR T-test T = 20 -0.95% -1.14 T = 19 -0.24% -0.29 T = 18 0.76% 0.90 T = 17 0.00% 0.00 T = 16 0.10% 0.12 T = 15 2.27% 2.70**** T =14 -1.42% -1.69* T = 13 -1.44% -1.72* T = 12 0.29% 0.35 T = 11 -0.43% -0.51 T = 10 -0.34% -0.41 T = 9 2.77% 3.30***** T = 8 -0.80% -0.95 T = 7 1.31% 1.56* T = 6 0.95% 1.13 T = 5 -2.04% -2.43*** T =4 1.08% 1.28 T = 3 -2.11% -2.51**** T = 2 2.75% 3.28***** T = 1 1.21% 1.44 T = 0 1.20% 1.43 T = -1 -0.56% -0.66 T = -2 0.61% 0.73 T = -3 0.25% 0.30 T = -4 -0.82% -0.98 T = -5 1.17% 1.39 T = -6 0.09% 0.10 T = -7 0.00% 0.00 T = -8 1.06% 1.26 T = -9 -0.14% -0.16 T = -10 0.72% 0.86

107

Table 40: The impact of the Spanish general election of 2008 on the European level. The IBEX 35 index is regressed in an event-study methodology with the AEX, CAC^40, DAX, FTSE 100 and the FTSE MIB index respectively. (*, Significant values are retrieved from table A: significant table on page 18.) Netherlands France Germany United Kingdom Italy AR T-Stat AR T-Stat AR T-Stat AR T-Stat AR T-Stat

T = 20 -1.22% -1.91** -0.79% -1.38 -0.71% -1.27 -0.97% -1.17 -1.55% -2.67****

T = 19 0.14% 0.21 0.28% 0.49 -0.25% -0.45 -0.97% -1.16 0.23% 0.39 T = 18 -0.50% -0.78 0.52% 0.91 0.34% 0.60 0.41% 0.49 -0.02% -0.03 T = 17 -0.21% -0.33 -0.17% -0.30 -0.05% -0.08 -0.19% -0.23 -0.34% -0.58 T = 16 0.23% 0.35 0.02% 0.04 -0.08% -0.13 -0.20% -0.24 -0.72% -1.24 T = 15 0.98% 1.53* -0.27% -0.46 1.48% 2.64**** 2.87% 3.45***** 1.04% 1.79* T =14 -1.74% -2.72**** -2.01% -3.50***** -1.03% -1.83** -1.23% -1.48 -1.25% -2.16*** T = 13 -0.71% -1.11 0.17% 0.29 -0.76% -1.36 -1.19% -1.44 -0.70% -1.21 T = 12 0.67% 1.05 0.33% 0.58 0.21% 0.37 0.93% 1.12 0.20% 0.34 T = 11 -0.53% -0.83 -0.33% -0.57 0.12% 0.21 0.15% 0.19 0.19% 0.32 T = 10 0.66% 1.03 0.81% 1.42 -0.78% -1.39 -1.96% -2.36*** 0.38% 0.65 T = 9 0.86% 1.34 0.10% 0.17 2.67% 4.76***** 4.67% 5.62***** 1.66% 2.86**** T = 8 -0.30% -0.47 0.37% 0.64 -1.14% -2.03** -1.74% -2.09** 0.89% 1.54* T = 7 -0.14% -0.22 -0.92% -1.61* -0.50% -0.88 0.00% 0.00 0.20% 0.34 T = 6 1.02% 1.59* 0.50% 0.88 3.25% 5.79***** 3.92% 4.72***** 0.79% 1.36 T = 5 0.01% 0.02 0.89% 1.54* -1.40% -2.50**** -2.51% -3.02***** 0.05% 0.09 T =4 1.08% 1.68* 0.83% 1.44 1.61% 2.88**** 2.11% 2.54**** 1.81% 3.13***** T = 3 -0.87% -1.35 -0.31% -0.53 -2.07% -3.68***** -3.47% -4.17***** -0.48% -0.82 T = 2 1.64% 2.56**** 1.38% 2.40*** 2.72% 4.85***** 3.54% 4.26***** 1.52% 2.63**** T = 1 1.29% 2.01** 0.85% 1.48 0.85% 1.52* 1.29% 1.55* 1.13% 1.95** T = 0 -0.12% -0.18 0.66% 1.15 0.77% 1.37 0.67% 0.81 1.26% 2.17*** T = -1 0.31% 0.48 0.56% 0.97 -0.87% -1.55* -1.50% -1.81** -0.31% -0.54 T = -2 1.18% 1.84** 1.05% 1.83** 0.82% 1.46 1.12% 1.35 0.79% 1.36 T = -3 -1.05% -1.64* -0.82% -1.42 -0.08% -0.15 -0.89% -1.07 -0.65% -1.12 T = -4 -1.24% -1.94** -1.18% -2.05** -0.66% -1.18 0.00% 0.01 -0.75% -1.30 T = -5 0.91% 1.42 0.85% 1.48 0.27% 0.48 -0.76% -0.91 0.35% 0.60 T = -6 0.47% 0.74 0.67% 1.17 0.14% 0.25 -0.39% -0.47 -0.26% -0.45 T = -7 0.61% 0.95 0.61% 1.06 -0.22% -0.39 -0.35% -0.42 0.06% 0.11 T = -8 0.94% 1.47 1.12% 1.96* 0.60% 1.07 0.98% 1.18 0.67% 1.16 T = -9 -0.15% -0.24 -0.97% -1.69* -0.06% -0.11 -0.46% -0.55 -0.08% -0.13 T = -10 -0.59% -0.93 -0.55% -0.96 1.10% 1.97* 1.10% 1.33 0.92% 1.59*

108

Table 41: The average abnormal stock return (AAR) together with the cumulative average abnormal return (CAAR) and T-stats per day during the six-week period. The average abnormal returns are calculated by first calculating the abnormal stock returns per day for each stock market index in the sample during the event-window. Then the average abnormal stock market returns from all stock indexes on a specific date during this event-window. The cumulative average abnormal stock return (CAAR) is the sum of all average abnormal returns (AAR’s). (*, Significant values are retrieved from table A: significant table on page 18.) AAR T-STAT T = 20 -1.05% -1.68* T = 19 -0.11% -0.10 T = 18 0.15% 0.24 T = 17 -0.19% -0.31 T = 16 -0.15% -0.24 T = 15 1.22% 1.79* T = 14 -1.45% -2.34*** T = 13 -0.64% -0.97 T = 12 0.47% 0.69 T = 11 -0.08% -0.14 T = 10 -0.18% -0.13 T = 9 1.99% 2.95***** T = 8 -0.38% -0.48 T = 7 -0.27% -0.47 T = 6 1.90% 2.87**** T = 5 -0.59% -0.77 T = 4 1.49% 2.33*** T = 3 -1.44% -2.11** T = 2 2.16% 3.34***** T = 1 1.08% 1.70* T = 0 0.65% 1.06 T = -1 -0.36% -0.49 T = -2 0.99% 1.57* T = -3 -0.70% -1.08 T = -4 -0.77% -1.29 T = -5 0.32% 0.61 T = -6 0.13% 0.25 T = -7 0.14% 0.26 T = -8 0.86% 1.36 T = -9 -0.34% -0.54 T = -10 0.40% 0.60 CAAR: 5.24% 3.71*****

109

During

Table 42: The Impact of the Spanish general election of 2011 the national level. The Spanish IBEX Small Cap index is used in order to measure the election impact on a national level. One-day Stock market returns are calculated with the following formula: (Rit/Rit-1) -1. (*, Significant values are retrieved from table A: significant table on page 18.) Day AR T-test T = 20 -1.73% -2.10** T = 19 1.11% 1.35 T = 18 -0.55% -0.67 T = 17 -1.55% -1.88** T = 16 -1.70% -2.06** T = 15 2.12% 2.57**** T =14 -3.02% -3.66***** T = 13 -0.49% -0.60 T = 12 1.24% 1.50 T = 11 0.12% 0.14 T = 10 0.74% 0.90 T = 9 1.12% 1.36 T = 8 3.92% 4.75***** T = 7 -0.03% -0.04 T = 6 1.77% 2.14** T = 5 1.80% 2.19*** T =4 -0.28% -0.34 T = 3 0.15% 0.18 T = 2 -1.83% -2.22*** T = 1 -1.78% -2.16*** T = 0 0.45% 0.54 T = -1 -0.16% -0.20 T = -2 0.67% 0.81 T = -3 -0.62% -0.75 T = -4 -2.64% -3.20***** T = -5 2.99% 3.63***** T = -6 0.82% 1.00 T = -7 -3.10% -3.77***** T = -8 1.27% 1.54* T = -9 0.56% 0.67 T = -10 -3.74% -4.54*****

110

Table 43: The impact of the Spanish general election of 2011 on the European level. The IBEX 35 index is regressed in an event-study methodology with the AEX, CAC^40, DAX, FTSE 100 and the FTSE MIB index respectively. (*, Significant values are retrieved from table A: significant table on page 18.) Country Netherlands France Germany United Kingdom Italy AR T-Stat AR T-Stat AR T-Stat AR T-Stat AR T-Stat

T = 20 -1.22% -1.33% -0.44% -0.55 -0.92% -1.04 -2.05% -2.03** -0.82% -1.01

T = 19 0.14% 0.71% 0.78% 0.97 0.69% 0.78 1.03% 1.02 0.27% 0.34 T = 18 -0.50% -0.12% 0.67% 0.83 -0.72% -0.82 -0.18% -0.18 0.83% 1.03 T = 17 -0.21% -1.51%* -0.63% -0.78 -0.93% -1.05 -2.01% -1.99** -0.72% -0.89 T = 16 0.23% -1.12% -0.91% -1.13 -0.70% -0.79 -1.55% -1.54* 0.52% 0.64 T = 15 0.98% 1.40% 0.49% 0.61 0.99% 1.12 2.20% 2.19*** -0.58% -0.72 T =14 -1.74% -1.88%** -0.43% -0.53 -1.03% -1.17 -3.18% -3.16***** 1.39% 1.72* T = 13 -0.71% -0.62% -0.52% -0.65 -0.45% -0.51 -0.93% -0.93 0.84% 1.04 T = 12 0.67% 0.27% 0.21% 0.26 0.41% 0.47 1.40% 1.39 -0.19% -0.24 T = 11 -0.53% 0.48% 1.09% 1.36 1.46% 1.65* 0.18% 0.18 0.25% 0.31 T = 10 0.66% 0.91% 0.63% 0.79 1.31% 1.48 1.11% 1.10 -0.45% -0.55 T = 9 0.86% 0.69% 0.40% 0.49 0.35% 0.39 0.85% 0.84 -0.10% -0.12 T = 8 -0.30% 1.83%** 1.13% 1.41 1.32% 1.49 3.90% 3.87***** 0.00% 0.00 T = 7 -0.14% -0.72% -0.39% -0.48 -0.69% -0.78 -0.67% -0.67 0.13% 0.16 T = 6 1.02% 0.75% -0.10% -0.12 0.71% 0.80 1.36% 1.35 -0.31% -0.38 T = 5 0.01% 0.62% -0.02% -0.03 0.23% 0.26 1.64% 1.63* 0.62% 0.76 T =4 1.08% -0.42% -0.46% -0.57 -0.28% -0.32 -1.09% -1.08 -0.40% -0.49 T = 3 -0.87% -0.28% -0.57% -0.71 -0.68% -0.77 0.11% 0.10 0.24% 0.30 T = 2 1.64% -1.58%* -0.82% -1.02 -1.20% -1.35 -2.89% -2.86**** 0.22% 0.28 T = 1 1.29% -1.29% -1.05% -1.31 -1.20% -1.35 -1.54% -1.52* 0.14% 0.18 T = 0 -0.12% 1.12% 1.01% 1.26 1.14% 1.28 1.18% 1.18 0.80% 0.99 T = -1 0.31% 0.43% 0.95% 1.18 0.19% 0.21 -0.04% -0.04 1.02% 1.26 T = -2 1.18% 0.43% 0.37% 0.46 0.65% 0.74 0.75% 0.74 0.21% 0.26 T = -3 -1.05% -0.65% -0.20% -0.25 -0.78% -0.88 -1.26% -1.25 -0.98% -1.21 T = -4 -1.24% -1.97%** -1.43% -1.78 -1.69% -1.91** -3.33% -3.30***** 0.70% 0.87 T = -5 0.91% 2.28%*** 1.39% 1.74* 1.44% 1.63* 3.09% 3.07***** -0.49% -0.60 T = -6 0.47% -0.32% -0.54% -0.67 -0.76% -0.86 0.99% 0.98 -1.65% -2.04** T = -7 0.61% -1.15% -0.74% -0.92 -1.23% -1.38 -3.78% -3.75***** 1.41% 1.75* T = -8 0.94% 0.44% 0.07% 0.09 0.58% 0.66 1.23% 1.22 -0.05% -0.06 T = -9 -0.15% 0.28% 0.00% 0.00 -0.10% -0.12 0.49% 0.48 -2.03% -2.51**** T = -10 -0.59% -2.83%**** -1.11% -1.38 -1.44% -1.62* -4.04% -4.01***** -0.12% -0.15

111

Table 44: The average abnormal stock return (AAR) together with the cumulative average abnormal return (CAAR) and T-stats per day during the six-week period. The average abnormal returns are calculated by first calculating the abnormal stock returns per day for each stock market index in the sample during the event-window. Then the average abnormal stock market returns from all stock indexes on a specific date during this event-window. The cumulative average abnormal stock return (CAAR) is the sum of all average abnormal returns (AAR’s). (*, Significant values are retrieved from table A: significant table on page 18.) AAR T-STAT T = 20 -1.11% -1.27 T = 19 0.70% 0.81 T = 18 0.09% 0.14 T = 17 -1.16% -1.33 T = 16 -0.75% -0.85 T = 15 0.90% 1.00 T = 14 -1.03% -1.11 T = 13 -0.33% -0.37 T = 12 0.42% 0.45 T = 11 0.69% 0.82 T = 10 0.70% 0.80 T = 9 0.44% 0.50 T = 8 1.64% 1.83** T = 7 -0.47% -0.54 T = 6 0.48% 0.52 T = 5 0.62% 0.69 T = 4 -0.53% -0.60 T = 3 -0.24% -0.29 T = 2 -1.25% -1.40 T = 1 -0.99% -1.13 T = 0 1.05% 1.23 T = -1 0.51% 0.63 T = -2 0.48% 0.55 T = -3 -0.78% -0.89 T = -4 -1.54% -1.74* T = -5 1.54% 1.76* T = -6 -0.46% -0.60 T = -7 -1.10% -1.16 T = -8 0.45% 0.49 T = -9 -0.27% -0.36 T = -10 -1.91% -2.17*** CAAR: -3.19% -1.04

112

After

Table 45: The Impact of the Spanish general election of 2015 the national level. The Spanish IBEX Small Cap index is used in order to measure the election impact on a national level. One-day Stock market returns are calculated with the following formula: (Rit/Rit-1) -1. (*, Significant values are retrieved from table A: significant table on page 18.) Day AR T-test T = 20 -0.72% -0.74 T = 19 3.02% 3.13***** T = 18 -0.06% -0.06 T = 17 -0.98% -1.02 T = 16 2.33% 2.42*** T = 15 -1.35% -1.40 T =14 -0.63% -0.66 T = 13 -0.94% -0.98 T = 12 0.58% 0.60 T = 11 0.38% 0.39 T = 10 -1.33% -1.38 T = 9 1.17% 1.22 T = 8 -1.60% -1.66* T = 7 -1.33% -1.38 T = 6 0.32% 0.33 T = 5 -1.64% -1.70* T =4 -0.07% -0.07 T = 3 0.20% 0.21 T = 2 0.66% 0.69 T = 1 -1.16% -1.20 T = 0 -2.11% -2.19*** T = -1 -0.70% -0.73 T = -2 1.25% 1.29 T = -3 2.00% 2.08** T = -4 -1.23% -1.28 T = -5 0.13% 0.13 T = -6 0.74% 0.76 T = -7 -1.29% -1.34 T = -8 -0.36% -0.37 T = -9 -0.53% -0.55 T = -10 0.15% 0.15

113

Table 46: The impact of the Spanish general election of 2015 on the European level. The IBEX 35 index is regressed in an event-study methodology with the AEX, CAC^40, DAX, FTSE 100 and the FTSE MIB index respectively. (*, Significant values are retrieved from table A: significant table on page 18.) Netherlands France Germany United Kingdom Italy AR T-Stat AR T-Stat AR T-Stat AR T-Stat AR T-Stat

T = 20 -0.01% -0.02 0.07% 0.12 1.19% 1.70* 0.62% 0.80 1.03% 1.64*

T = 19 0.70% 1.19 0.67% 1.22 0.74% 1.06 3.25% 4.23***** -1.84% -2.92**** T = 18 -0.66% -1.13 -0.14% -0.25 -0.71% -1.01 0.42% 0.55 0.94% 1.49 T = 17 -0.75% -1.28 -0.89% -1.61* -0.27% -0.39 -3.35% -4.35***** 0.79% 1.26 T = 16 -0.30% -0.51 0.04% 0.08 -0.42% -0.60 1.34% 1.74* 1.87% 2.97***** T = 15 -1.16% -1.98** -1.01% -1.83** -0.83% -1.19 -2.02% -2.62**** -0.24% -0.39 T =14 -0.56% -0.96 -0.28% -0.51 -0.48% -0.69 -0.65% -0.84 -0.02% -0.03 T = 13 -0.34% -0.58 -0.07% -0.12 0.58% 0.84 -1.81% -2.36*** -0.47% -0.75 T = 12 -0.32% -0.54 -0.61% -1.10 -0.89% -1.28 1.11% 1.44 -0.21% -0.33 T = 11 0.07% 0.12 0.48% 0.87 0.02% 0.02 0.49% 0.63 0.31% 0.49 T = 10 -0.84% -1.44 -1.38% -2.49**** -1.10% -1.57* -2.39% -3.10***** -0.81% -1.28 T = 9 0.61% 1.05 0.22% 0.40 2.38% 3.42***** 2.11% 2.74**** -0.78% -1.24 T = 8 -1.04% -1.78* -0.56% -1.01 -2.70% -3.87***** -3.40% -4.42***** 1.69% 2.68**** T = 7 -0.67% -1.15 -0.28% -0.51 -0.43% -0.62 -1.34% -1.74* -1.02% -1.62* T = 6 0.13% 0.23 0.15% 0.28 0.87% 1.25 0.96% 1.25 1.18% 1.88** T = 5 -0.72% -1.22 -0.42% -0.75 0.48% 0.69 -1.82% -2.37*** -0.38% -0.60 T =4 -0.19% -0.32 -0.35% -0.63 -0.47% -0.68 0.04% 0.05 -0.14% -0.23 T = 3 -0.16% -0.28 0.49% 0.88 0.15% 0.21 0.04% 0.05 -0.12% -0.20 T = 2 -0.10% -0.17 0.24% 0.43 0.34% 0.49 0.27% 0.36 0.42% 0.66 T = 1 -0.56% -0.95 -0.37% -0.68 -0.68% -0.98 -0.59% -0.77 -0.45% -0.72 T = 0 -2.55% -4.35***** -2.59% -4.69***** -2.57% -3.68***** -2.46% -3.19***** -2.60% -4.12***** T = -1 0.14% 0.25 0.54% 0.99 -0.84% -1.21 -1.41% -1.83** -0.02% -0.03 T = -2 0.11% 0.19 -0.42% -0.76 -0.17% -0.24 1.24% 1.61* 0.10% 0.15 T = -3 0.85% 1.45 1.00% 1.81** 0.88% 1.26 1.87% 2.43*** -0.15% -0.23 T = -4 -0.47% -0.80 -0.80% -1.45 -0.39% -0.56 -1.15% -1.49 0.19% 0.31 T = -5 -0.18% -0.31 -0.40% -0.73 0.08% 0.12 -0.30% -0.39 0.18% 0.28 T = -6 0.34% 0.57 -0.08% -0.14 -0.29% -0.42 0.19% 0.25 0.16% 0.25 T = -7 -0.31% -0.52 0.47% 0.86 0.42% 0.60 -0.80% -1.04 0.15% 0.24 T = -8 -0.90% -1.54* -0.71% -1.29 -0.51% -0.73 -1.07% -1.39 0.04% 0.06 T = -9 -0.62% -1.06 -1.17% -2.12** -1.42% -2.03** -0.98% -1.27 -0.57% -0.90 T = -10 0.60% 1.03 0.38% 0.68 0.42% 0.60 1.17% 1.52* 0.02% 0.03

114

Table 47: The average abnormal stock return (AAR) together with the cumulative average abnormal return (CAAR) and T-stats per day during the six-week period. The average abnormal returns are calculated by first calculating the abnormal stock returns per day for each stock market index in the sample during the event-window. Then the average abnormal stock market returns from all stock indexes on a specific date during this event-window. The cumulative average abnormal stock return (CAAR) is the sum of all average abnormal returns (AAR’s). (*, Significant values are retrieved from table A: significant table on page 18.) AAR T-STAT T = 20 0.58% 0.85 T = 19 0.70% 0.95 T = 18 -0.03% -0.07 T = 17 -0.89% -1.27 T = 16 0.51% 0.73 T = 15 -1.05% -1.60* T = 14 -0.40% -0.60 T = 13 -0.42% -0.59 T = 12 -0.18% -0.36 T = 11 0.27% 0.43 T = 10 -1.30% -1.98** T = 9 0.91% 1.27 T = 8 -1.20% -1.68* T = 7 -0.75% -1.13 T = 6 0.66% 0.97 T = 5 -0.57% -0.85 T = 4 -0.22% -0.36 T = 3 0.08% 0.13 T = 2 0.24% 0.35 T = 1 -0.53% -0.82 T = 0 -2.55% -4.01***** T = -1 -0.32% -0.37 T = -2 0.17% 0.19 T = -3 0.89% 1.34 T = -4 -0.52% -0.80 T = -5 -0.12% -0.20 T = -6 0.06% 0.10 T = -7 -0.01% 0.03 T = -8 -0.63% -0.98 T = -9 -0.95% -1.48 T = -10 0.52% 0.77 CAAR: -7.08% -1.84**

115

7.5 Italian general elections

Before

Table 49: The Impact of the Italian general election of 2006 the national level. The Italian FTSE Small Cap index is used in order to measure the election impact on a national level. One-day Stock market returns are calculated with the following formula: (Rit/Rit-1) -1. (*, Significant values are retrieved from table A: significant table on page 18.) Day AR T-test T = 20 -0.06% -0.13 T = 19 0.14% 0.29 T = 18 0.15% 0.31 T = 17 0.48% 0.98 T = 16 0.58% 1.17 T = 15 -1.09% -2.23*** T =14 0.95% 1.94** T = 13 -0.69% -1.41 T = 12 -0.91% -1.85** T = 11 0.58% 1.18 T = 10 0.06% 0.13 T = 9 -0.29% -0.58 T = 8 -0.71% -1.45 T = 7 0.57% 1.17 T = 6 0.20% 0.41 T = 5 0.33% 0.67 T =4 -0.40% -0.82 T = 3 0.27% 0.55 T = 2 -1.52% -3.09***** T = 1 0.76% 1.55* T = 0 -1.27% -2.58**** T = -1 -0.37% -0.75 T = -2 1.15% 2.34*** T = -3 -0.56% -1.15 T = -4 0.09% 0.18 T = -5 -0.13% -0.27 T = -6 0.43% 0.87 T = -7 0.78% 1.58* T = -8 -0.68% -1.38 T = -9 -1.15% -2.33*** T = -10 -0.24% -0.48

116

Table 50: The impact of the Italian general election of 2006 on the European level. The FTSE MIB index is regressed in an event-study methodology with the AEX, CAC^40, DAX, FTSE 100 and the IBEX 3 index respectively. (*, Significant values are retrieved from table A: significant table on page 18.) Country Netherlands France German UK Spain AR T-Stat AR T-Stat AR T-Stat AR T-Stat AR T-Stat

T = 20 0.17% 0.35 0.08% 0.17 0.01% 0.03 0.22% 0.44 0.08% 0.17 T = 19 0.66% 1.38 0.36% 0.79 0.38% 0.75 -0.05% -0.11 0.33% 0.75 T = 18 0.31% 0.65 0.43% 0.94 0.33% 0.66 0.53% 1.06 0.32% 0.72 T = 17 -0.30% -0.63 -0.18% -0.39 -0.53% -1.05 0.20% 0.41 -0.18% -0.41 T = 16 0.34% 0.70 -0.01% -0.02 -0.34% -0.68 0.48% 0.97 0.48% 1.09 T = 15 0.37% 0.77 0.01% 0.01 0.15% 0.30 -0.41% -0.83 -0.74% -1.66* T =14 0.14% 0.29 -0.09% -0.19 0.44% 0.87 0.58% 1.17 0.45% 1.03 T = 13 0.36% 0.74 -0.19% -0.40 0.14% 0.28 -0.29% -0.58 -0.37% -0.84 T = 12 -0.36% -0.75 -0.14% -0.29 -0.22% -0.45 -0.53% -1.06 -0.46% -1.03 T = 11 0.39% 0.81 0.33% 0.71 0.31% 0.61 0.38% 0.77 0.49% 1.10 T = 10 0.29% 0.60 0.13% 0.27 0.13% 0.26 0.29% 0.59 -0.11% -0.24 T = 9 -0.13% -0.27 -0.32% -0.70 -0.55% -1.10 0.09% 0.18 -0.20% -0.46 T = 8 -0.83% -1.72* -0.59% -1.28 -0.36% -0.71 -0.93% -1.89** -0.93% -2.10** T = 7 0.25% 0.52 -0.25% -0.54 -0.29% -0.59 0.82% 1.67* 0.15% 0.35 T = 6 -0.15% -0.31 0.28% 0.62 -0.59% -1.17 0.04% 0.08 0.09% 0.21 T = 5 -0.24% -0.51 -0.22% -0.48 0.30% 0.59 0.49% 0.98 0.09% 0.21 T =4 -0.20% -0.42 -0.40% -0.87 -0.49% -0.99 -0.64% -1.30 -0.33% -0.74 T = 3 0.37% 0.77 0.57% 1.24 0.40% 0.79 0.54% 1.09 0.80% 1.80* T = 2 -0.28% -0.59 -0.18% -0.39 -0.08% -0.16 -1.09% -2.20*** -0.70% -1.57* T = 1 0.28% 0.58 0.31% 0.66 -0.15% -0.30 0.27% 0.55 0.56% 1.26 T = 0 -0.82% -1.71* -0.36% -0.77 -0.01% -0.01 -1.02% -2.06** -0.51% -1.15 T = -1 -0.34% -0.71 -0.22% -0.48 -0.31% -0.62 -0.41% -0.83 -0.19% -0.43 T = -2 1.20% 2.50**** 0.99% 2.15*** 0.96% 1.91** 0.95% 1.92** 0.95% 2.14** T = -3 -0.32% -0.66 0.06% 0.12 -0.35% -0.69 -0.23% -0.46 -0.35% -0.80 T = -4 0.33% 0.69 0.29% 0.62 -0.27% -0.53 -0.34% -0.69 0.30% 0.68 T = -5 -0.09% -0.19 -0.35% -0.77 -0.24% -0.48 0.40% 0.80 -0.40% -0.89 T = -6 0.69% 1.44 0.23% 0.49 0.13% 0.25 0.12% 0.24 0.63% 1.43 T = -7 0.38% 0.80 0.09% 0.20 0.26% 0.52 0.84% 1.70* 0.44% 0.99 T = -8 -0.17% -0.36 -0.17% -0.37 -0.27% -0.54 -0.32% -0.64 -0.11% -0.25 T = -9 -0.93% -1.93** -0.64% -1.39 -0.42% -0.83 -0.80% -1.62* -0.80% -1.80* T = -10 -0.40% -0.83 -0.48% -1.04 -0.51% -1.02 -0.63% -1.28 -0.46% -1.05

117

Table 51: The average abnormal stock return (AAR) together with the cumulative average abnormal return (CAAR) and T-stats per day during the six-week period. The average abnormal returns are calculated by first calculating the abnormal stock returns per day for each stock market index in the sample during the event-window. Then the average abnormal stock market returns from all stock indexes on a specific date during this event-window. The cumulative average abnormal stock return (CAAR) is the sum of all average abnormal returns (AAR’s). (*, Significant values are retrieved from table A: significant table on page 18.) AAR T-STAT T = 20 0.11% 0.23 T = 19 0.34% 0.71 T = 18 0.38% 0.81 T = 17 -0.20% -0.42 T = 16 0.19% 0.41 T = 15 -0.12% -0.28 T = 14 0.30% 0.63 T = 13 -0.07% -0.16 T = 12 -0.34% -0.72 T = 11 0.38% 0.80 T = 10 0.15% 0.30 T = 9 -0.22% -0.47 T = 8 -0.73% -1.54* T = 7 0.14% 0.28 T = 6 -0.06% -0.11 T = 5 0.08% 0.16 T = 4 -0.41% -0.86 T = 3 0.54% 1.14 T = 2 -0.47% -0.98 T = 1 0.25% 0.55 T = 0 -0.54% -1.14 T = -1 -0.29% -0.61 T = -2 1.01% 2.12** T = -3 -0.24% -0.50 T = -4 0.06% 0.15 T = -5 -0.14% -0.31 T = -6 0.36% 0.77 T = -7 0.40% 0.84 T = -8 -0.21% -0.43 T = -9 -0.72% -1.52 T = -10 -0.50% -1.04 CAAR: -0.57% -0.53

118

During

Table 52: The Impact of the Italian general election of 2006 the national level. The Italian FTSE Small Cap index is used in order to measure the election impact on a national level. One-day Stock market returns are calculated with the following formula: (Rit/Rit-1) -1. (*, Significant values are retrieved from table A: significant table on page 18.) Day AR T-test T = 20 0.02% 0.02 T = 19 0.40% 0.46 T = 18 -0.27% -0.31 T = 17 -0.72% -0.82 T = 16 -0.43% -0.49 T = 15 -0.34% -0.38 T =14 0.16% 0.19 T = 13 1.33% 1.51* T = 12 -0.24% -0.27 T = 11 -0.76% -0.87 T = 10 1.19% 1.36 T = 9 0.81% 0.92 T = 8 -0.41% -0.47 T = 7 0.14% 0.16 T = 6 -0.57% -0.65 T = 5 1.11% 1.27 T =4 0.04% 0.05 T = 3 1.12% 1.28 T = 2 -0.35% -0.40 T = 1 0.72% 0.82 T = 0 -2.39% -2.73**** T = -1 0.46% 0.52 T = -2 0.58% 0.66 T = -3 0.41% 0.47 T = -4 -0.37% -0.43 T = -5 1.03% 1.17 T = -6 -0.07% -0.08 T = -7 1.51% 1.72* T = -8 3.91% 4.46***** T = -9 -0.34% -0.39 T = -10 -1.01% -1.15

119

Table 53: The impact of the Italian general election of 2013 on the European level. The FTSE MIB index is regressed in an event-study methodology with the AEX, CAC^40, DAX, FTSE 100 and the IBEX 3 index respectively. (*, Significant values are retrieved from table A: significant table on page 18.) Country Netherlands France German UK Spain AR T-Stat AR T-Stat AR T-Stat AR T-Stat AR T-Stat

T = 20 0.59% 1.12 -0.68% -1.49 -0.18% -0.30 0.16% 0.18 0.31% 0.51

T = 19 0.25% 0.48 -0.17% -0.37 -0.09% -0.15 0.29% 0.32 -0.38% -0.62 T = 18 -0.65% -1.24 0.22% 0.49 -0.14% -0.23 0.26% 0.30 -0.31% -0.51 T = 17 -1.13% -2.13** -0.36% -0.77 -1.02% -1.72* -1.08% -1.21 -1.65% -2.71**** T = 16 -0.57% -1.09 0.02% 0.04 -0.61% -1.02 -0.98% -1.10 0.23% 0.37 T = 15 -0.22% -0.41 -0.25% -0.54 -0.11% -0.18 -0.10% -0.11 0.07% 0.12 T =14 -0.22% -0.42 0.35% 0.77 -0.50% -0.84 -0.92% -1.03 -0.77% -1.27 T = 13 -0.08% -0.15 -0.27% -0.59 0.35% 0.59 1.53% 1.71* 0.05% 0.09 T = 12 -0.34% -0.65 0.16% 0.35 0.13% 0.22 -0.03% -0.03 0.59% 0.98 T = 11 -1.25% -2.36*** -0.74% -1.61* -0.74% -1.25 -0.80% -0.90 -1.09% -1.79* T = 10 1.22% 2.31 0.44% 0.95 0.49% 0.82 0.85% 0.95 0.96% 1.57* T = 9 0.50% 0.95 0.07% 0.15 -0.08% -0.13 1.22% 1.36 0.24% 0.39 T = 8 -0.46% -0.88 -1.10% -2.39*** -0.94% -1.58* -0.91% -1.02 -0.60% -0.99 T = 7 0.28% 0.53 -0.07% -0.14 0.20% 0.34 0.13% 0.15 0.34% 0.56 T = 6 0.59% 1.12 0.69% 1.50* 0.13% 0.21 -0.74% -0.83 0.75% 1.24 T = 5 -1.22% -2.32*** -0.54% -1.18 -0.64% -1.08 0.36% 0.41 -0.89% -1.47 T =4 0.15% 0.29 0.10% 0.21 0.36% 0.60 0.83% 0.93 0.45% 0.74 T = 3 -0.50% -0.95 -0.28% -0.62 -0.18% -0.30 -0.32% -0.36 0.01% 0.02 T = 2 -1.06% -2.01** -0.61% -1.33 -0.74% -1.25 -0.89% -0.99 -1.91% -3.14***** T = 1 -0.45% -0.85 -0.35% -0.76 0.44% 0.73 1.32% 1.48 0.27% 0.45 T = 0 0.29% 0.56 0.41% 0.89 -0.50% -0.83 -1.96% -2.20*** -0.33% -0.54 T = -1 -0.01% -0.01 -0.08% -0.17 -0.01% -0.02 0.47% 0.53 0.19% 0.31 T = -2 0.64% 1.21 0.50% 1.08 0.59% 1.00 0.27% 0.30 1.02% 1.68* T = -3 0.38% 0.72 0.85% 1.85** 0.92% 1.55* 0.66% 0.74 1.48% 2.43*** T = -4 -0.17% -0.33 -0.02% -0.04 -0.57% -0.96 -1.27% -1.42 -0.38% -0.62 T = -5 -0.44% -0.83 0.69% 1.51* 0.56% 0.94 0.66% 0.73 0.31% 0.51 T = -6 0.11% 0.22 0.16% 0.35 0.29% 0.48 0.14% 0.16 0.26% 0.43 T = -7 1.14% 2.17*** 0.93% 2.04** 0.88% 1.48 0.79% 0.88 0.99% 1.62* T = -8 0.29% 0.55 -0.97% -2.12** 1.12% 1.88** 2.57% 2.88**** 0.12% 0.20 T = -9 -0.53% -1.01 -0.81% -1.76* 0.27% 0.44 0.07% 0.08 1.14% 1.88** T = -10 -0.11% -0.20 0.83% 1.81** -0.20% -0.33 -0.63% -0.70 0.38% 0.63

120

Table 54: The average abnormal stock return (AAR) together with the cumulative average abnormal return (CAAR) and T-stats per day during the six-week period. The average abnormal returns are calculated by first calculating the abnormal stock returns per day for each stock market index in the sample during the event-window. Then the average abnormal stock market returns from all stock indexes on a specific date during this event-window. The cumulative average abnormal stock return (CAAR) is the sum of all average abnormal returns (AAR’s). (*, Significant values are retrieved from table A: significant table on page 18.) AAR T-STAT T = 20 0.04% 0.00 T = 19 -0.02% -0.07 T = 18 -0.12% -0.24 T = 17 -1.05% -1.71* T = 16 -0.38% -0.56 T = 15 -0.12% -0.23 T = 14 -0.41% -0.56 T = 13 0.32% 0.33 T = 12 0.10% 0.17 T = 11 -0.92% -1.58* T = 10 0.79% 1.32 T = 9 0.39% 0.55 T = 8 -0.80% -1.37 T = 7 0.18% 0.29 T = 6 0.28% 0.65 T = 5 -0.59% -1.13 T = 4 0.38% 0.55 T = 3 -0.25% -0.44 T = 2 -1.04% -1.74* T = 1 0.25% 0.21 T = 0 -0.42% -0.42 T = -1 0.11% 0.13 T = -2 0.60% 1.05 T = -3 0.86% 1.46 T = -4 -0.48% -0.68 T = -5 0.36% 0.57 T = -6 0.19% 0.33 T = -7 0.95% 1.64* T = -8 0.63% 0.68 T = -9 0.03% -0.07 T = -10 0.06% 0.24 CAAR: -0.11% -0.05

121

After

Table 55: The Impact of the Italian general election of 2006 the national level. The Italian FTSE Small Cap index is used in order to measure the election impact on a national level. One-day Stock market returns are calculated with the following formula: (Rit/Rit-1) -1. (*, Significant values are retrieved from table A: significant table on page 18.) Day AR T-test T = 20 1.29% 1.80* T = 19 -0.58% -0.82 T = 18 0.92% 1.28 T = 17 -0.68% -0.95 T = 16 1.82% 2.54**** T = 15 -2.91% -4.05***** T =14 0.88% 1.23 T = 13 -0.98% -1.37 T = 12 -0.79% -1.09 T = 11 -0.37% -0.52 T = 10 -0.24% -0.33 T = 9 0.04% 0.05 T = 8 -0.37% -0.52 T = 7 1.80% 2.50**** T = 6 1.02% 1.42 T = 5 -1.30% -1.81* T =4 -0.23% -0.33 T = 3 1.25% 1.74* T = 2 -2.67% -3.72***** T = 1 -1.95% -2.72**** T = 0 1.06% 1.48 T = -1 -1.89% -2.63**** T = -2 -1.50% -2.09** T = -3 1.71% 2.39*** T = -4 -0.34% -0.47 T = -5 -0.55% -0.77 T = -6 -0.76% -1.06 T = -7 -0.09% -0.13 T = -8 1.45% 2.01** T = -9 -0.73% -1.02 T = -10 1.23% 1.72*

122

Table 56: The impact of the Italian general election of 2013 on the European level. The FTSE MIB index is regressed in an event-study methodology with the AEX, CAC^40, DAX, FTSE 100 and the IBEX 3 index respectively. (*, Significant values are retrieved from table A: significant table on page 18.) Netherlands France German UK Spain AR T-Stat AR T-Stat AR T-Stat AR T-Stat AR T-Stat

T = 20 1.52% 2.31*** 0.92% 1.34 1.39% 1.59* 1.98% 2.14** 0.93% 1.07

T = 19 -0.09% -0.14 0.43% 0.63 -0.04% -0.04 -0.40% -0.43 0.30% 0.35 T = 18 1.09% 1.65* 0.73% 1.08 1.19% 1.36 1.20% 1.30 1.02% 1.17 T = 17 -1.01% -1.54* -0.48% -0.71 -0.92% -1.05 -0.70% -0.75 0.27% 0.31 T = 16 1.33% 2.01** 1.06% 1.55 1.42% 1.62* 2.06% 2.23*** 0.81% 0.93 T = 15 -2.27% -3.43***** -1.59% -2.32**** -2.26% -2.57**** -3.27% -3.54**** -0.96% -1.10 T =14 1.25% 1.89** 1.47% 2.16*** 1.20% 1.37 1.49% 1.61* 0.68% 0.78 T = 13 -0.83% -1.25 -0.99% -1.46 -1.04% -1.18 -0.47% -0.51 -0.87% -1.00 T = 12 -0.56% -0.84 -0.65% -0.95 -0.44% -0.50 -0.50% -0.54 -0.18% -0.21 T = 11 0.21% 0.31 0.29% 0.43 0.11% 0.13 0.06% 0.07 0.91% 1.05 T = 10 -0.27% -0.41 -0.26% -0.39 0.05% 0.06 -0.17% -0.19 -1.48% -1.70* T = 9 0.16% 0.24 -0.18% -0.27 0.06% 0.07 0.33% 0.36 -0.07% -0.08 T = 8 -0.37% -0.56 0.06% 0.08 -0.36% -0.41 -0.09% -0.10 0.33% 0.38 T = 7 1.22% 1.84** 0.99% 1.45 0.90% 1.03 1.57% 1.70* 0.95% 1.09 T = 6 0.18% 0.27 -0.34% -0.49 -0.33% -0.38 0.93% 1.00 -1.11% -1.27 T = 5 -1.71% -2.58**** -1.39% -2.03** -1.34% -1.53* -2.10% -2.27*** -1.13% -1.30 T =4 -0.46% -0.70 -0.31% -0.46 -0.53% -0.60 -0.48% -0.52 -0.62% -0.71 T = 3 1.23% 1.86** 0.19% 0.29 0.93% 1.06 1.28% 1.38 -0.14% -0.16 T = 2 -3.35% -5.07***** -3.35% -4.91***** -2.40% -2.74**** -2.31% -2.50**** -3.44% -3.95***** T = 1 -0.11% -0.16 0.71% 1.05 -1.03% -1.17 -1.53% -1.65* 1.47% 1.69* T = 0 1.06% 1.61* 0.16% 0.24 0.66% 0.76 1.43% 1.55* -0.07% -0.08 T = -1 -1.37% -2.08** -1.13% -1.66* -1.46% -1.67* -1.46% -1.58* -1.35% -1.55* T = -2 -0.65% -0.98 -0.68% -0.99 -1.04% -1.19 -1.66% -1.80* -0.39% -0.44 T = -3 0.99% 1.50* 0.23% 0.34 0.45% 0.52 1.32% 1.43 0.20% 0.23 T = -4 -0.35% -0.53 -0.54% -0.79 -0.64% -0.73 -0.37% -0.40 -0.18% -0.21 T = -5 -0.25% -0.38 -0.42% -0.62 -0.26% -0.30 -0.64% -0.69 0.65% 0.74 T = -6 -0.84% -1.26 -0.34% -0.50 -0.34% -0.39 -0.69% -0.75 -0.47% -0.54 T = -7 -0.05% -0.07 -0.15% -0.22 -0.37% -0.42 -0.28% -0.31 -0.29% -0.33 T = -8 1.08% 1.64* 0.44% 0.65 0.83% 0.94 1.05% 1.14 -0.51% -0.59 T = -9 -0.59% -0.89 -0.51% -0.75 -0.76% -0.87 -0.88% -0.95 0.34% 0.39 T = -10 0.92% 1.40 0.26% 0.38 0.95% 1.09 0.96% 1.04 -0.41% -0.47

123

Table 57: The average abnormal stock return (AAR) together with the cumulative average abnormal return (CAAR) and T-stats per day during the six-week period. The average abnormal returns are calculated by first calculating the abnormal stock returns per day for each stock market index in the sample during the event-window. Then the average abnormal stock market returns from all stock indexes on a specific date during this event-window. The cumulative average abnormal stock return (CAAR) is the sum of all average abnormal returns (AAR’s). (*, Significant values are retrieved from table A: significant table on page 18.) AAR T-STAT T = 20 1.35% 1.69* T = 19 0.04% 0.07 T = 18 1.05% 1.31 T = 17 -0.57% -0.75 T = 16 1.34% 1.67 T = 15 -2.07% -2.59**** T = 14 1.22% 1.56* T = 13 -0.84% -1.08 T = 12 -0.47% -0.61 T = 11 0.32% 0.40 T = 10 -0.43% -0.52 T = 9 0.06% 0.06 T = 8 -0.09% -0.12 T = 7 1.13% 1.42 T = 6 -0.13% -0.17 T = 5 -1.53% -1.94** T = 4 -0.48% -0.60 T = 3 0.70% 0.88 T = 2 -2.97% -3.83***** T = 1 -0.09% -0.05 T = 0 0.65% 0.81 T = -1 -1.36% -1.71* T = -2 -0.88% -1.08 T = -3 0.64% 0.80 T = -4 -0.42% -0.53 T = -5 -0.19% -0.25 T = -6 -0.53% -0.69 T = -7 -0.23% -0.27 T = -8 0.58% 0.76 T = -9 -0.48% -0.61 T = -10 0.54% 0.69 CAAR: -4.17% -2.87 124

8. References

Alesina, A., Özler, S., Roubini, N., Swagel P., “Political Instability and Economic Growth”, Journal of Economic Growth 1(2), 1996, No. 4173, pp.189-211.

Alexander K., Kaboyakgosi, G,. “A Fine Balance: Assessing the Quality of Governance in Botswana”, 2012

Anderson, J.C., “Economic Voting and Political Context: A Comparative Perspective”, Electoral Studies, 2000, Vol. 19, pp. 151-170.

Andrew, W.L., “What is an index?”, The Journal of Portfolio Management, Vol. 42, No. 2, winter 2006.

Akitoby, B., Stratmann, T., “Fiscal Policy and Financial Markets”, IMF Working Paper, Fiscal Affairs Department, 2006.

Bialkowski, J., Gottschalk, K., Wisniewski, T.P., “Stock market volatility around national elections”, Journal of Banking & Finance, Vol. 32, 2008, pp. 1941-1953.

Brown, K.C., Harlow, W.V., Tinic, S.M., “Risk aversion, uncertain information, and market e ciency”, Journal of Financial Economics, No. 22, 1988, pp. 355-385.

Cheol, S., Huang, W., Lai, S., “International Diversification with Large- and Small-Cap Stocks” Journal of Financial and Quantitative Analysis, Vol 43, No. 2, June 2008, pp. 489-524.

Douglas, A., Hibbs Jr., “Political Parties and Macroeconomic Policy”, The American Political Science Review, Vol. 71, No. 4, 1977, pp. 1467-1487.

Durnev, A., “The real effects of political uncertainty: Elections and investments sensitivity to stock prices”, McGill University, 2011.

French, K.R., Schwert, G.W., Stambauch, R.F., “Expected stock returns and volatility”, Journal of Financial Economics, Vol. 19, 1987, pp. 3-29.

Gemmill, G., “Political risk and market efficiency: Tests based in British stock and options markets in the 1987 election”, Journal of Banking and Finance, Vol. 16, 1992, pp. 211- 231.

Hair, J.F., Jr., Anderson, R.E., Tatham, R.L. and Black, W.C, “Multi-variate Data Analysis. Englewood Cliffs”, 7th edition, NJ: Prentice Hall, 1995.

125

Harrington, J.E., “Economic policy, economic performance and elections”, The American Economic Review, Vol. 83, 1993, 27-42.

King, M. R., “Time to Buy or Just Buying Time? The Market Reaction to Bank Rescue Packages", 2009, BIS Papers No 288 (September).

King, M. R., “The cross-border contagion and competition effects of bank bailouts announced in October 2008”, 2012, Available at SSRN 2019113. Retrieved from: http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2019113

Leblang, D., Mukherjee, B., “Government Partisanship, Election, and the Stock Market: Examining American and British Stock Returns, 1930-2000”, American Journal of Political Science, Vol 49, No. 4, 2005, pp 780-802.

Lee, W.Y., Jian, C.X., Indor, D.C., “Stock market volatility, excess returns, and the role of investor sentiment”, Journal of Banking & Finance, Vol. 26, 2002, pp. 2277-2299.

MacKinlay, A. Craig, “Event studies in economics and finance”, Journal of Economic Literature, Vol. 35, 1997, pp. 13-39.

Pantzalis, C., Stangeland, D.A., Turtle, H.J., “Political elections and the resolution of uncertainty: The international evidence”, Journal of Banking & Finance, Vol. 24, 2000, pp. 1575-1604.

Stevenson, R.T., “The Economy and Policy Mood: A Fundamental Dynamic of Democratic Politics?”, American Journal of Political Science, Vol 45, No. 3, July 2001, pp. 620-633.

Wei, K.C.J., Chan, Y.-c., “Political risk and Stock price volatility: The case of Hong Kong”, Pacific- Basin Finance Journal, Vol. 4, 1996, pp. 259-275.

Werner, F., De Bondt, M., Thaler, R., “Does the Stock Market Overreact?”, The Journal of Finance, Vol. 40, No. 3, 1985, pp. 793-805.

Yasushi, H., Masulis, R.W., Ng, V., “Correlations in Price Changes and Volatility across International Stock Markets”, The Review of Financial Studies, Vol. 3, No. 2, 1990, pp. 281-307.

Woodruff, P., “First Democracy: The Challenge of an Ancient Idea”, Oxford University Press, Inc., 2005.