The effect of the U.S. presidential elections on the stock market of the top trading partners of the U.S.

The effect of the U.S. presidential

elections on the stock market of the top

trading partners of the U.S.

Bachelor Thesis Finance

University of Amsterdam

Faculty of Economics and Business

Author: Colin Beek Student ID: 10563202 Supervisor: L. Zou Date: 31th _{January 2018}

~ 2 ~

Statement of Originality

This document is written by Colin Beek who declares to take full responsibility for the contents of this document. I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it. The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

~ 3 ~

Abstract

This study investigates the effect of the presidential election in the United States on the countries with the most export to the United States. The sample contains 30 countries with data from 1984 till 2016. The results of this study show no significant effect on all the trading partners combined. The results did show significance for thirteen out of the thirty countries. The amount of export as a percentage of the GDP per capita of a country is found to be significant for these countries, but with very little explanatory power.

~ 4 ~

Index

Data and Methodology ... 10

Data ... 10 Methodology ... 11 Results ... 14 Conclusion ... 18 References ... 19 Appendix ... 21

Table 4: Summary of top exporting countries to the U.S. and used market indices ... 21

Table 5: Summary of Returns(in %) per index ... 22

Table 6: Cumulative abnormal returns per country ... 23

~ 5 ~

Introduction

Political events have a big influence on the financial market. Because of political decision making new information becomes available. The financial markets try to incorporate this information in the stock prices. The presidential election is a unique event, since after the election the voters have no more influence on the decision making in politics. Therefore, it is interesting for investors to keep an eye on these political events and the results.

The 58th _{quadrennial American presidential election, which was held on November 8}th

2016, was won by the republican candidate Donald John Trump. By beating the democratic presidential candidate Hilary Clinton, Donald Trump became the 45th president of the United States of America. The outcome of the 2016 election is viewed as one of the most shocking outcomes in modern political history. Since Donald Trump, a business man and a reality television star who has no government experience, defeated the former secretary of state. Trump ran a controversial campaign that focused on immigration control and included among others building a wall along the United States border with Mexico and a proposal to ban Muslims from entering the United States. He also made comments about withdrawing from several trade and international agreements like the Trans Pacific Partnership(TPP), NAFTA, the Paris Climate Accord, The U.S.-Cuba deal and NATO. Since being elected Donald Trump has made progress or successfully fulfilled withdrawment or adjustment of most of these agreements and comments.

On the night of the election, as the results came in, the markets went wild. Futures for the benchmark S&P 500 and the Dow Jones Industrial average indices fell more than four per cent(Kiersz, 2016). The stock markets recovered quickly and since the election of Donald J. Trump the S&P 500 has performed exceptionally well. Between the election of Donald J. Trump on November 8th _{2016 and September 2017 the S&P 500 has added 2.04 trillion dollars}

in market value (Imbert, 2017).

The presidential election of the United States is a large world event. The election outcome could change trading regimes, since the president can alter import tariffs as well as cancel trade agreements without much intervention from the US Congress. Also, the United States of America is one of the biggest countries in the world concerning import and export. Since the current literature about the effect of elections on market indexes their return and volatility focuses mainly on the domestic market this thesis will be focused on the effect of the

~ 6 ~

U.S. election outcome on the economy of the countries with a strong trade-relationship with the United States.

For investors it is important to know how the markets respond to certain events, so they can alter their trading strategy. This thesis will try and answer the following question:

Do the political elections in the United States of America have an abnormal effect on the market index return of the current top trade partners of the U.S.A.?

In this study the daily stock returns of the thirty current top trade partners of the United States will be used. The top trade partners are selected by their total amount of export to the United States. Following MacKinlay(1997) an event study will be used to investigate if cumulative abnormal returns of market indexes of the top trade partners are significantly different from zero at election day. After the event study a cross-section analysis will be conducted to research which factors influence the cumulative abnormal returns.

This thesis is organized as following. Section 2 will discuss relevant literature. Section 3 will explain data used for this study. Section 4 covers the methodology. In section 5 the results will be stated and discussed. And finally, section 6 will cover the conclusion and additionally the limitations of this research and suggestions for further research.

~ 7 ~

Literature review

This section will cover existing literature about the relationship between political events and market indexes. Following this literature several hypotheses will be formed.

Huang(1985) researched the pattern of common stock returns over the four-year election cycle as well as over different administrations. Their research shows a strong pattern in the four years of the election cycle. The existence of an election cycle is much more apparent when combined returns are shown between the first and second year and the third and fourth year. The combined return is notably higher in the third and fourth year compared to the combined first two years of the election cycle. The pattern is especially large in the most recent 20-year period of their research, were the annual difference is exceeding 24 per cent(Huang, 1985). F. Siokis and P. Kapopoulos(2007) examined if the movements of stock prices in a small open economy could be partly explained by the dynamics of the political environment. They researched the stock market of Greece by using the ASE index. In their paper they confirmed previous findings of the importance of political events in explaining the behavior of the stock exchange. They found evidence of both partisan and electoral effects on the prices of stocks, especially in the in period from 1988 till 2004(F. Siokis and P. Kapopoulos, 2007).

A.F. Herbst and C.W. Slinkman(1984) found strong support for a four-year political-economic cycle. Their results also showed the existence of a two-year cycle, although it did not peak during the election date and therefore is not labeled as a political cycle.

Based on multiple studies where political cycles are found the following hypothesis is formed:

The cumulative average abnormal returns in the period surrounding the election will be non-zero for the thirty top trading partners. This can be expressed as following:

H0: CAAR= 0

H1: CAAR ≠ 0

Niederhoffer et al. (1970) found evidence for the traditional Wall Street view that the market prefers republicans. Their results showed significant rise in the Dow Jones Index the day following a republican victory. After a republican victory the Dow Jones Index on average rose by 1,12 per cent, while the Dow Jones Index on average fell by 0,81 per cent following a democratic victory. This is also the case for the first month following the election(Niederhoffer, et al., 1970).

~ 8 ~

A more recent research of Huang(1985) contradict the finding of Niederhoffer et al.(1970). They found evidence against the myth that markets prefer a republican administration over a democratic administration. The results obtained in the research didn’t show any significant difference between the two administrations in four out of the six cases. Although for the two periods with a significant difference both showed higher returns for a democratic administration. Also, research of for example Allvine and O’Neill(1980) and from Santa-Clara and Valkanov(2003) have found that the election of a democratic administrations predicts positive stock market returns. Based on these finding the following hypothesis is formed:

The cumulative abnormal return in the election period will be higher when a democratic administration is elected. Which can be expressed as following:

H0: β2 = 0

H1: β2 > 0

In the research of Bialkowski, et al.(2006) is found that one of the factors that effects the stock market volatility is a change in political regime. According to their findings the volatility of stock prices rises when the political orientation of the elected administration changes. The model made by Pastor and Veronesi(2012) confirms the results found by Bialkowski, et al.(2006) that stock prices should fall when there is a change in government policy, which can be assumed when there is a change in political administrations. Following this knowledge, the following hypotheses is formed:

A change in the political orientation because of the election will result in lower cumulative abnormal returns.

H0: β3= 0

H1: β3 < 0

According to the research of Herbst and Slinkman(1984) the stock markets tend to rise more pre-election than post-election and therefore stock market are more favorable towards expectations than realization. Niederhoffer, et al.(1970) also found that stock prices tend to rise before an election, but only when a big victory is expected. Therefore, they assume that there is a relation between the change in the market index and the margin of victory. Bialkowski, et al.(2006) also found that one of the factors that increases volatility of the stock market is the margin of victory. And therefore, confirms that market indexes tend to fall when there is uncertainty about who will win the election. Following these findings the hypothesis formed is:

~ 9 ~

A large margin of victory will have a negative effect on the cumulative abnormal returns. Which can be expressed as following:

H0: β5 = 0

H1: β5 < 0

Di Giovanni and Levchenko(2009) discuss the relationship between openness to trade and the volatility of the domestic stock market in their paper. Although they couldn’t conclude that a higher trade openness makes countries more exposable to external shocks, because of the high difference between countries, they could conclude the following. First, the higher the trade in a certain sector, the higher the volatility in that sector(di Giovanni and Levchenko, 2009). Secondly, an economy with more trade has a higher specialization. Their results show that a change in trade openness is accompanied by an estimated rise in aggregate volatility. This volatility is about five times higher for developing countries than for developed countries. Milner and Rosendorff(1997) have examined the impact of elections on international cooperation. They found that elections make the endorsements of international agreements problematic, because of the uncertainty surrounding elections. Because elections can usually not be completely predicted, the negotiators must guess what kind of agreement will be acceptable for the future legislator(Milner and Rosendorff, 1997).

From the studies from di Giovanni and Levchenko(2009) and from Milner and Rosendorff(1997) it can be assumed that elections have a bigger impact on economies with international agreements and more trade with the country where the election takes place. Since the election brings more uncertainty for these countries concerning their export and economy the following hypothesis is formed:

A larger amount of export from a country will have a non-zero effect on the cumulative abnormal returns during the election period. This can be expressed as following:

H0: β1 = 0

~ 10 ~

Data and Methodology

Data

To perform an event study on the effect of the presidential elections in the United States on their trade partners, the current largest exporters to the U.S. will be selected together with their corresponding market indices. Market indices are used because of their intent to represent entire stock markets (https://www.investopedia.com/). The U.S. department of Commerce is used as a source for the most current data of total export by countries to the United States. The database of the U.S. department of commerce has historical data available till 1985. Therefore, the exports of 1985 as an estimate of the export of 1984.

The elections from before 1984 are not included in this study, because of the lack of data from market indices in that period. Also, the top of exporters to the United States changes over the years, which makes the period of 1984 till 2016 more representative. The data about the elections results is retrieved from the electoral college website(Historical election results, 2016).

Table 1:

This table shows the results from the election of the United States from 1984 till 2016. The third column shows what the political orientation was of the elected president. The fourth column shows if there was a change of political administration because of the newly elected president. And the fifth column shows the margin of victory for the elected president in percentage points above 50 pro cent.

The historical daily prices of the market indices are retrieved from the Thomson Reuters DataStream database. Not all elections are included for the analysis in all thirty countries, because of the availability from the market indices that does not go back till 1984 in all cases. The historical Gross Domestic Products(GDP) of the countries in the study is retrieved from

Election dates Winner Administration Change Margin of victory 6-11-1984 Ronald Reagan Republican No 47,58% 9-11-1988 George Bush Republican No 29,18% 3-11-1992 William Clinton Democrat Yes 18,77% 5-11-1996 William Clinton Democrat No 20,45% 7-11-2000 George W. Bush Republican Yes 0,37% 2-11-2004 George W. Bush Republican No 3,16% 4-11-2008 Barack Obama Democrat Yes 17,84% 6-11-2012 Barack Obama Democrat No 11,71% 8-11-2016 Donald Trump Republican Yes 6,88%

~ 11 ~

the World Bank. There is no data of the GDP of Taiwan, and therefore Taiwan will not be included in the cross-section analysis. Both the daily prices of the indices and the GDPs are converted to US$ with the use of historical exchange rates.

To be able to calculate the expected return without the event the MSCI indices are used. The MSCI indices includes the most representable stocks from several countries in the particular market. Which MSCI index is used depends on the country and its region and can be seen in table 4 in the appendix.

Methodology

This study will use an empirical event study to test if the U.S. presidential elections have an effect on the U.S. their largest trade partners. The event study pioneered by Fama, Fisher, Jensen and Roll(1969) will be used for event study. First the period of the research will be discussed, followed by the methodology for abnormal returns.

According to MacKinlay(1997) both an estimation period and an event period have to be specified for the event study. The estimation period(T0,T1) will be one year before the event period and therefore consist of 261 trading days. The estimation window will be slightly smaller for the event of 1984 because of the start of many indices on January 1984. The event itself is election day, the event window(T1,T2) will be four days, one day before and two days after the event. This event window is chosen because of the different time zones in the studied countries and because of the weekend before the election day.

The daily return of the market indices in the chosen period will be calculated using the following formula. In this formula 𝑅_𝑖,𝑡 represents the daily return of the market index on day t. 𝑃𝑖,_𝑡 is the price index on day t and 𝑃_{𝑖,𝑡−1} is the prices index on the day before 𝑃𝑖,_𝑡.

𝑅_𝑖,𝑡 = ( 𝑃𝑖,𝑡

𝑃_{𝑖,𝑡−1}− 1) ∗ 100

The following formula is used to calculate the daily abnormal returns of market index i at day t. Where 𝑅_𝑖,𝑡 represents the observed daily return on the market index and 𝐸(𝑅_𝑖,𝑡) represents the daily expected return of the index without the event.

~ 12 ~

When the market model is used to calculate the expected return, the residuals can be seen as estimates of the abnormal return(Fama, et al., 1969). Rewriting the above formula gives the following:

𝐴𝑅_𝑖,𝑡 = 𝑅_𝑖,𝑡− 𝛼̂ − 𝛽_𝑖 ̂ ∗ 𝑅_{𝑖 𝑀𝑆𝐶𝐼} The OLS-estimators are estimated by the following OLS-regression:

𝑅_𝑖,𝑡 = 𝛼 + 𝛽 ∗ 𝑅_{𝑀𝑆𝐶𝐼}+ 𝜀_𝑡 ~𝑁(0, 𝜎2₎

The cumulative abnormal return is given by:

𝐶𝐴𝑅 = ∑ 𝐴𝑅_𝑖,𝑡

𝑇2 𝑡=𝑇1

The test is about more than one country and therefore we use an average of the cumulative abnormal returns. By testing this average, we can tell if there is a significant effect on all the trade partners instead of one of the trade partners.

𝐶𝐴𝐴𝑅 = 1

𝑁∑ 𝐶𝐴𝑅𝑖

𝑁 𝑖=1

The cumulative average abnormal returns(CAAR) will be tested for their significance from zero. A significant CAAR implies that the presidential election in the United States have an impact on the countries that have the largest export to the United States. The results will be tested for significance using their t-values. Since the actual standard deviation is unknow the following formula will be used to as an estimator of the standard deviation:

𝑆_𝑡 = √ 1

𝑁 − 1∑(𝐶𝐴𝑅𝑖,𝑡− 𝐶𝐴𝐴𝑅𝑡)2

𝑁 𝑖=1

The t-statistic would then be calculated with the formula:

𝑡 = √𝑁𝐶𝐴𝐴𝑅𝑡

𝑆_𝑡 ~ 𝑡𝑁−1

The study will also contain a cross-section analysis. This analysis will research which variables of the election have a significant influence on the CARs. The control variables are based on existing literature. The regression model used, is as followed:

~ 13 ~

𝐶𝐴𝑅 = 𝛽₀+ 𝛽₁𝐿𝑁(𝐸𝑥𝑝𝑜𝑟𝑡) + 𝛽₂𝐷𝑒𝑚𝑜 + 𝛽₃𝐶ℎ𝑎𝑛𝑔𝑒𝑅𝑒𝑔𝑖𝑚𝑒 + 𝛽₄𝐷𝑒𝑚𝑜 ∗ 𝐶ℎ𝑎𝑛𝑔𝑒𝑅𝑒𝑔𝑖𝑚𝑒 + 𝛽₅𝑀𝑎𝑟𝑔𝑖𝑛 + 𝜀

Where Export is the natural logarithm of the export as a percentage point of the Gross Domestic Product. Demo is a dummy variable which is 1 if the elected president is a democrat and 0 if the elected president is a republican. ChangeRegime is also a dummy variable which is 1 if there is a change in political administration because of the election and 0 if the administration stays the same. Demo*ChangeRegime is an interaction variable. And Margin is the percentage points above 50 per cent of the electoral votes of the elected president.

All regressions in the study will be conducted with robust standard errors to avoid heteroscedasticity.

~ 14 ~

Results

In this chapter the results of the event study and the cross-section analysis are presented and analyzed. The study is conducted as is described in the methodology section. The cumulative abnormal returns are an indicator for the abnormal returns per event. In addition, the cumulative average abnormal returns show the aggregate effect of the presidential election on a country level. Both variables are presented in the appendix with the corresponding t-statistic. In this section significant results of the cross-section analysis will be marked with asterisks. One asterisks is shown for a significance at a ten per cent level, two asterisks for significance at the five per cent level and one asterisks for significance at the one per cent level. Results will be discussed on both individual country level as well as all the countries, trade partners, combined.

Table 4 in the appendix shows the cumulative abnormal returns per country and per event. The cumulative abnormal returns are tested for their significance. Table 2 summarizes the cumulative average abnormal returns per country and their corresponding t-statistic. The results of the analyses show very different results in significance both per event and per country. The cumulative abnormal returns per event and country are not significant in almost all the cases. Therefore, there is no evidence of the individual effect of the event. Although when the cumulative average abnormal returns of the countries are tested for significance the results are different. The cumulative average abnormal returns show significance in thirteen out of the thirty countries. This suggest that the presidential election of the United States has an effect on some of the countries, although not all the trade partners. This suggestion is confirmed by a regression on all the cumulative average abnormal returns. This regression showed that there is a 18,8 per cent change that the found cumulative average abnormal returns are zero. There is not enough evidence to suggest that the presidential election in the United States has a significant effect on all the top trading partners of the U.S. Therefore, the first hypothesis can be rejected.

Table 2

Country CAAR t-test

China -1,552 -0,504

~ 15 ~ Canada 0,142 0,870 Japan 0,812 4,813 Germany -0,444 -15,405 South Korea 0,312 0,869 United Kingdom -0,067 -0,384 France 0,621 1,135 India 1,308 1,018 Ireland -0,350 -4,245 Italy 0,954 4,807 Vietnam -0,014 -0,024 Taiwan 1,666 3,048 Malaysia 0,755 2,516 Switzerland 0,435 2,100 Thailand 3,067 8,512 Brazil -1,631 -4,801 Israel -0,427 -11,899 Indonesia 1,748 10,233 Singapore -0,118 -0,170 Belgium 0,412 0,896 Saudi-Arabia 2,070 1,613 Netherlands -0,483 -0,930 Russia -2,394 -0,687 Colombia 0,381 0,161 Spain -0,302 -0,788 Venezuela 0,149 0,302 Austria 0,404 0,833 Philippines 2,679 4,590 Sweden 0,470 0,560

In the first column the countries are listed. Column two shows the cumulative average abnormal returns for the countries and column three lists the t-statistic of the corresponding CAAR.

In the cross-section analysis of this study the cumulative abnormal returns of all the countries and events are regressed on the following model:

𝐶𝐴𝑅 = 𝛽0+ 𝛽1𝐿𝑁(𝐸𝑥𝑝𝑜𝑟𝑡) + 𝛽2𝐷𝑒𝑚𝑜 + 𝛽3𝐶ℎ𝑎𝑛𝑔𝑒𝑅𝑒𝑔𝑖𝑚𝑒 + 𝛽4𝐷𝑒𝑚𝑜

~ 16 ~

In table 3 the results of this regression are shown. The first column of results shows the regression results when the model is regressed against all the cumulative abnormal returns of the countries. The second column shows the regression results when only the countries with significant cumulative average abnormal returns are included.

Table 3

(1) (2)

VARIABLES CAR CAR

lnExport 0.0346 0.722** (0.160) (0.291) Democrat 0.0200 0.239 (0.308) (0.962) Change 0.0945 0.933 (0.572) (1.088) DemocratChange 0.505 -0.555 (0.841) (1.503) Margin 0.00960 0.0145 (0.0133) (0.0388) Constant -0.689 -13.58** (3.060) (5.707) R-squared 0.010 0.079

CAR(1) gives the result for all countries. CAR(2) lists the results for all countries with a significant CAAR.

*= significant at the 10 per cent level **= significant at the 5 per cent level ***= significant at the 1 per cent level

When the above model is regressed on all the cumulative abnormal returns none of the variables are found to be significant. These results are to be expected since most of the observed cumulative abnormal returns were not significant. The output of the regression on only the countries with a significant cumulative average abnormal return is therefore more representable result and will be further discussed from here on.

Each of the independent variables, except for the export, have been suggested to be an explanatory variable by academic literature for the cumulative abnormal returns created by the presidential election in the United States. The export as a percentage of the GDP has a significant positive effect on the cumulative abnormal returns at the 5% level. This result suggests that both the gross domestic product and the export to the United States of a country have an impact on the returns around an election. The beta of 0.722 suggest that a change of one percent in the export as a percentage of the GDP results, on average, in an 0.00722 percentage points increase in the returns of the index. These findings are in line with the suggestion made based on the literature of di Giovanni and Levchenko(2009) and Milner and

~ 17 ~

Rosendorff(1997). Therefore, the following hypothesis will not be rejected: A larger amount of

export from a country, as a percentage of their GDP, will have a negative effect on the cumulative abnormal returns during the election period.

A democratic administration has a positive effect on the cumulative abnormal returns, which is in line with the research from, among others, Allvine and O’Neill(1980), Santa-Clara and Valkanov(2003) and Niederhoffer et al. (1970). Although in this study there is no significant evidence that it matters which administration wins for the top-trade partners. Therefore, the following hypothesis can be rejected: The cumulative abnormal return in the

election period will be higher when a democratic administration is elected.

A change in administration, republic of democratic, has a large positive effect on the cumulative abnormal returns of the countries. This finding contradicts the expectation formed by the research of Pastor and Veronesi(2012) that stock prices should fall when there is a change in the political orientation. Although the coefficient in this study is relatively large, it is not significant and therefore there is no evidence that a change in political orientation affects the stock market of the top trade partners. This result causes a rejection of the following hypothesis:

A change in the political administration because of the election will result in lower cumulative abnormal returns.

The margin of victory in the election has no significant effect on the cumulative abnormal returns of the trade partners. This is in contradiction with the findings of Bialkowski, et al.(2006) and Niederhoffer, et al.(1970) that a big victory, and therefore assumed an expected victory, has a positive effect on the market indices. This finding causes a rejection of the following hypothesis: A large margin of victory will have a positive effect on the cumulative

abnormal returns.

The overall results show that all but one of the hypothesis of this study can be rejected. The presidential election is not found to be of significance on all the trade partners. Although when the election is found to be significant the amount of export as a percentage of the GDP partly explains the abnormal returns. The rest of the variables have no significant effect. The reason that many of the studied countries show no significance could be that most of the studied countries are not fully depended on trade with the United States. Because of the low dependence the volatility can be expected to be lower and therefore the abnormal returns as well.

~ 18 ~

Conclusion

The main focus of this study was to investigate if the presidential election of the United States has an effect on the top trading partners of the U.S.. This is investigated by conducting an event study. The expected returns of the indices per country where estimated by an OLS regression on the corresponding MSCI-index. The difference between the realized and the expected return in the four-day window surrounding the events is aggregated and tested for significance. In this research thirty countries and data dating back to 1984 is used to estimate the relation between the presidential election and the U.S. their current top trading partners.

The overall effect of the presidential election in the United States on their top trading partners is found to be insignificant. When limiting the analysis of explanatory variables to just the significant countries only the effect of export as a percentage of the GDP was found to be significant. Although the impact on the returns is low and only significant at a 5% level. The results suggest the presidential election in the U.S. could have a significant effect on a country, although the total amount of export to the U.S. is not the main explanatory variable.

One of the limitations in this research is the prediction of the abnormal returns by using the MSCI-index. This index contains the most important stocks from several countries in the market, but therefore it is also more sensitive to other events in the area. This could cause disruption in the predicted returns without the election event. Also, since there is a proven effect of the election in the domestic market this is already incorporated in the MSCI-world index which contains 54% of U.S. stocks.

In future studies the research could be conducted on countries were the export to the United States is measured as a total of the export of the country. It could also be studied further if the effect of the event can be explained by the type of specialization from the countries and therefore what kind of trade takes place, instead of total trade of goods and services.

~ 19 ~

References

Allvine, F. C., & O’Neill, D. E. (1980). Stock Market Returns and the Presidential Election Cycle: Implications for Market Efficiency. Financial Analysts Journal, 36(5), 49–56.

Bialkowski, J., Gottschalk, K., & Wisniewski, T. P. (2007). Political orientation of government and stock market returns. Applied Financial Economics Letters, 3(4), 269-273.

Białkowski, J., Gottschalk, K., & Wisniewski, T. P. (2008). Stock market volatility around national elections. Journal of Banking & Finance, 32(9), 1941-1953.

Binder, J. (1998). The event study methodology since 1969. Review of quantitative Finance and

Accounting, 11(2), 111-137.

Herbst, A. F., & Slinkman, C. W. (1984). Political-Economic Cycles in the U.S. Stock Market. Financial Analysts Journal, 40(2), 38–44.

Huang, R. D. (1985). Common Stock Returns and Presidential Elections. Financial Analysts Journal, 41(2), 58–61.

Imbert, F. (2017, September 12). S&P’s value has soared by 2 trillion since Donald Trump election. Retrieved from: https://www.cnbc.com/2017/09/12/sp-500s-value-has-soared-by-2-trillion-since-donald-trump-election.html

Kiersz, A. (2016, December 12). Stock markets after Trump election. Retrieved from:

https://www.businessinsider.nl/stock-markets-after-trump-election-2016-12/?international=true&r=US MacKinlay, A. C. (1997). Event Studies in Economics and Finance. Journal of Economic Literature,

35(1), 13–39.

Milner, H. V., & Rosendorff, B. P. (1997). Democratic Politics and International Trade Negotiations: Elections and Divided Government as Constraints on Trade Liberalization. The Journal of Conflict Resolution, 41(1), 117–146.

~ 20 ~

Niederhoffer, V., Gibbs, S., & Bullock, J. (1970). Presidential Elections and the Stock Market. Financial Analysts Journal, 26(2), 111–113.

Pastor, Ľ., & Veronesi, P. (2012). Uncertainty about Government Policy and Stock Prices. The Journal of Finance, 67(4), 1219–1264.

Siokis, F., & Kapopoulos, P. (2007). Parties, elections and stock market volatility: evidence from a small open economy. Economics & Politics, 19(1), 123-134.

United States Department of Commerce. (2016). Export Data Trade US.Retrieved from: https://www.trade.gov/mas/ian/tradestatistics/index.asp

United States National archives and records administration. (2016). U. S. Electoral College: Historical Election Results. Retrieved from:

~ 21 ~

Appendix

Table 4: Summary of top exporting countries to the U.S. and used market indices

Ranking (export

to U.S.) Country Market Index Region

Global Stock Market Index(MSCI) Starting date Index (DD-MM-YY) Elections included

1 China SSE Asia MSCI Emerging Markets 2-1-1992 7

2 Mexico IPC-35 South-America MSCI Emerging Markets 4-1-1988 8

3 Canada S&P/TSX North-America MSCI World Index 1-1-1980 9

4 Japan Nikkei-225 Asia MSCI World Index 1-1-1980 9

5 Germany DAX30 Europe MSCI Europe 1-1-1980 9

6 South Korea KOSPI Asia MSCI Emerging Markets 1-1-1980 8

7 United Kingdom FTSE-100 Europe MSCI Europe 30-12-1983 9

8 France CAC-40 Europe MSCI Europe 9-7-1987 8

9 India NIFTY-500 Asia MSCI Emerging Markets 2-1-1991 7

10 Ireland ISEQ Europe MSCI Europe 5-1-1983 9

11 Italy FTSE MIB Europe MSCI Europe 31-12-1997 5

12 Vietnam HCMNVNE Asia MSCI Emerging Markets 28-7-2000 4

13 Taiwan TAIEX Asia MSCI Emerging Markets 1-1-1980 8

14 Malaysia FBM KLCI Asia MSCI Emerging Markets 2-1-1980 8

15 Switzerland SMI Europe MSCI Europe 30-6-1988 7

16 Thailand SET Asia MSCI Emerging Markets 1-1-1980 8

17 Brazil WIB South-America MSCI Emerging Markets 1-11-1994 6

18 Israel TA-125 Asia MSCI World Index 23-4-1987 8

19 Indonesia IDX Asia MSCI Emerging Markets 4-4-1983 8

20 Singapore STI Asia MSCI World Index 31-8-1999 5

21 Belgium BEL-20 Europe MSCI Europe 2-1-1990 7

22 Saudi Arabia TASI Asia MSCI Emerging Markets 19-10-1998 5

23 Netherlands AEX Europe MSCI Europe Index 3-1-1983 9

24 Russia MOEX Europe/Asia MSCI Emerging Markets 23-9-1997 5

25 Colombia IGBC South-America MSCI Emerging Markets 3-7-2001 4

26 Spain IBEX 35 Europe MSCI Europe 5-1-1987 8

27 Venezuela IBvC South-America MSCI Emerging Markets 1-4-1993 6

28 Austria ATX Europe MSCI Europe 7-1-1986 8

29 Philippines PSE Asia MSCI Emerging Markets 2-1-1986 8

~ 22 ~

Table 5: Summary of Returns(in %) per index

ReturnMSCI~g 7840 .0382048 1.117744 -9.511194 10.59764 ReturnMSCI~e 9927 .0325677 1.146931 -9.677408 11.29126 ReturnMSCI~d 9927 .0322514 .8751409 -9.844356 9.523239 ReturnSweden 8360 .0471098 1.623853 -9.79845 14.68407 ReturnPhil~s 8360 .0541737 1.798763 -14.29258 23.69713 ReturnAust~a 8357 .0365473 1.492725 -11.78219 13.4412 ReturnVene~a 6412 .1251763 7.435232 -100 100 ReturnSpain 8098 .0302653 1.510157 -14.16687 16.14648 ReturnColo~a 4317 .0649701 1.605956 -11.33333 15.625 ReturnRussia 5303 .0582311 2.890965 -38.63636 37.00787 ReturnNeth~s 9143 .0414005 1.393075 -11.18058 13.10647 ReturnArabia 5023 .042288 1.425151 -11.02938 17.60375 ReturnBelg~m 7317 .0240798 1.278953 -9.069081 11.00423 ReturnSing~e 4797 .0226974 1.231413 -8.307263 9.601815 ReturnIndo~a 9078 .0397341 2.144081 -32 47.51773 ReturnIsrael 8020 .0471383 1.590387 -10.61938 10.90782 ReturnBrazil 6057 .0456347 2.316802 -17.23295 22.68945 ReturnThai~d 9927 .0339075 1.625239 -20.25723 15.97444 ReturnSwit~d 7710 .036581 1.172561 -9.127943 10.53507 ReturnMala~a 9926 .027334 1.506319 -30.93367 26.37474 ReturnTaiwan 8623 .0490752 1.783584 -10.72228 14.8288 ReturnViet~m 4559 .2863514 7.656399 -50 100 ReturnItaly 5231 .0157085 1.684767 -14.29222 13.17998 ReturnIrel~d 9141 .0440557 1.343856 -14.0594 10.46135 ReturnIndia 7056 .0427542 1.653755 -13.09795 19.85012 ReturnFrance 7965 .0287878 1.459899 -11.07437 12.91153 ReturnUK 8884 .0302108 1.242371 -13.54755 12.99671 ReturnKorea 9927 .0455843 2.148051 -18.51852 31.57895 ReturnGerm~y 9927 .0442753 1.447732 -12.24141 13.16724 ReturnJapan 9927 .0312096 1.444179 -16.27702 13.39548 ReturnCanada 9927 .027847 1.124919 -12.88093 10.43448 ReturnMexico 7838 .068713 1.834295 -19.58374 16.51744 ReturnChina 6795 .0652107 2.640595 -32.15663 109.2412 Variable Obs Mean Std. Dev. Min Max

~ 23 ~

Table 6: Cumulative abnormal returns per country

9618. 10/11/2016 9 1.49733 .3998414 8573. 08/11/2012 8 -1.494227 -.3990126 7528. 06/11/2008 7 1.582581 .4226064 6483. 04/11/2004 6 -1.684505 -.4498239 5443. 09/11/2000 5 2.061073 .5503811 4398. 07/11/1996 4 -.9246336 -.2469106 3353. 05/11/1992 3 -11.89829 -3.177275 Date event_id CARChina test 9618. 10/11/2016 9 -9.210472 -1.879185 8573. 08/11/2012 8 -2.782416 -.5676878 7528. 06/11/2008 7 -3.252977 -.663695 6483. 04/11/2004 6 1.524387 .311016 5443. 09/11/2000 5 -1.814701 -.370248 4398. 07/11/1996 4 .1884821 .0384554 3353. 05/11/1992 3 -.2781067 -.0567413 2314. 11/11/1988 2 2.988104 .6096538 Date event_id CARMexico testMex~o 9618. 10/11/2016 9 -2.033238 -1.297969 8573. 08/11/2012 8 .944016 .6026364 7528. 06/11/2008 7 4.861448 3.103428 6483. 04/11/2004 6 -1.697218 -1.083462 5443. 09/11/2000 5 -1.899286 -1.212457 4398. 07/11/1996 4 1.005332 .6417789 3353. 05/11/1992 3 -.2896007 -.1848739 2314. 11/11/1988 2 -1.453463 -.927855 1268. 08/11/1984 1 1.843869 1.17708 Date event_id CARCanada testCan~a 9618. 10/11/2016 9 -1.630692 -.414925 8573. 08/11/2012 8 -.9371667 -.2384595 7528. 06/11/2008 7 5.982233 1.522163 6483. 04/11/2004 6 -1.349802 -.3434534 5443. 09/11/2000 5 2.332216 .593426 4398. 07/11/1996 4 .1408995 .0358515 3353. 05/11/1992 3 2.158833 .5493091 2314. 11/11/1988 2 1.730152 .4402325 1268. 08/11/1984 1 -1.118562 -.284615 Date event_id CARJapan testJapan

~ 24 ~ 9618. 10/11/2016 9 .3911467 .2768309 8573. 08/11/2012 8 -.2467786 -.1746555 7528. 06/11/2008 7 -.4242682 -.3002724 6483. 04/11/2004 6 -.4386568 -.3104558 5443. 09/11/2000 5 -1.045108 -.7396666 4398. 07/11/1996 4 .5767609 .4081978 3353. 05/11/1992 3 -2.543337 -1.800026 2314. 11/11/1988 2 -.5117904 -.3622155 1268. 08/11/1984 1 .2457191 .1739057 Date event_id CARGerm~y testGer~y 9618. 10/11/2016 9 1.210034 .3832937 8573. 08/11/2012 8 1.063894 .337002 7528. 06/11/2008 7 -1.82107 -.5768475 6483. 04/11/2004 6 -.6301411 -.1996053 5443. 09/11/2000 5 1.344775 .4259749 4398. 07/11/1996 4 -2.910065 -.9218004 3353. 05/11/1992 3 3.224632 1.021443 2314. 11/11/1988 2 1.011487 .3204017 Date event_id CARKorea testKorea 9618. 10/11/2016 9 .6132777 .6409754 8573. 08/11/2012 8 -.0052128 -.0054482 7528. 06/11/2008 7 -1.68015 -1.756031 6483. 04/11/2004 6 .0024532 .002564 5443. 09/11/2000 5 .8021134 .8383394 4398. 07/11/1996 4 -1.945161 -2.03301 3353. 05/11/1992 3 1.38712 1.449766 2314. 11/11/1988 2 -.3254659 -.340165 1268. 08/11/1984 1 .5446141 .5692106 Date event_id CARUK testUK 9618. 10/11/2016 9 .3489873 .2985774 8573. 08/11/2012 8 -.1822605 -.1559337 7528. 06/11/2008 7 .3689447 .315652 6483. 04/11/2004 6 -.6406877 -.5481429 5443. 09/11/2000 5 -.7061312 -.6041333 4398. 07/11/1996 4 2.355015 2.014842 3353. 05/11/1992 3 2.233903 1.911224 2314. 11/11/1988 2 1.187602 1.016057 Date event_id CARFrance testFra~e

~ 25 ~ 9618. 10/11/2016 9 1.423103 .4292083 8573. 08/11/2012 8 .3479191 .1049325 7528. 06/11/2008 7 7.407605 2.234136 6483. 04/11/2004 6 .901568 .2719131 5443. 09/11/2000 5 2.558402 .7716148 4398. 07/11/1996 4 -2.289927 -.6906425 3353. 05/11/1992 3 -1.193083 -.3598341 Date event_id CARIndia testIndia 9618. 10/11/2016 9 2.068431 1.067748 8573. 08/11/2012 8 .5042074 .2602776 7528. 06/11/2008 7 -3.793458 -1.958226 6483. 04/11/2004 6 .9165719 .4731449 5443. 09/11/2000 5 .7224272 .3729252 4398. 07/11/1996 4 -.8068684 -.4165146 3353. 05/11/1992 3 -2.694288 -1.390822 2314. 11/11/1988 2 -.777379 -.4012919 1268. 08/11/1984 1 .7082189 .3655907 Date event_id CARIrel~d testIre~d 9618. 10/11/2016 9 -.136473 -.0860232 8573. 08/11/2012 8 -.7721772 -.4867272 7528. 06/11/2008 7 4.520136 2.849182 6483. 04/11/2004 6 .1542705 .0972415 5443. 09/11/2000 5 1.004787 .6333486 Date event_id CARItaly testItaly 9618. 10/11/2016 9 -1.407926 -1.529975 8573. 08/11/2012 8 0 0 7528. 06/11/2008 7 1.366464 1.484919 6483. 04/11/2004 6 -2.549923 -2.770969 Date event_id CARViet~m testVie~m

~ 26 ~ = 9618. 10/11/2016 9 .8572999 .2541179 8573. 08/11/2012 8 1.692111 .5015698 7528. 06/11/2008 7 -1.491298 -.4420455 6483. 04/11/2004 6 .9339885 .2768497 5443. 09/11/2000 5 6.333215 1.87727 4398. 07/11/1996 4 2.136784 .6333783 3353. 05/11/1992 3 -.922604 -.2734751 2314. 11/11/1988 2 3.786434 1.122362 Date event_id CARTaiwan testTai~n 9618. 10/11/2016 9 -1.153566 -.5863563 8573. 08/11/2012 8 -1.050171 -.5338005 7528. 06/11/2008 7 5.13036 2.607755 6483. 04/11/2004 6 .2730775 .1388049 5443. 09/11/2000 5 -.8644455 -.4393965 4398. 07/11/1996 4 .0374732 .0190476 3353. 05/11/1992 3 2.934988 1.49185 2314. 11/11/1988 2 .7290219 .3705609 Date event_id CARMala~a testMal~a 9618. 10/11/2016 9 1.884329 1.350895 8573. 08/11/2012 8 1.04188 .746935 7528. 06/11/2008 7 -2.675784 -1.918298 6483. 04/11/2004 6 1.825182 1.308492 5443. 09/11/2000 5 .2908417 .2085075 4398. 07/11/1996 4 .9951128 .7134067 3353. 05/11/1992 3 -.3136378 -.2248502 Date event_id CARSwit~d testSwi~d

~ 27 ~ 9618. 10/11/2016 9 -4.701962 -.9311698 8573. 08/11/2012 8 -.84524 -.1673901 7528. 06/11/2008 7 -.9529824 -.1887273 6483. 04/11/2004 6 -.1407084 -.0278657 5443. 09/11/2000 5 -1.597562 -.3163788 4398. 07/11/1996 4 -1.549008 -.3067632 Date event_id CARBrazil testBra~l 9618. 10/11/2016 9 -.5324584 -.2202861 8573. 08/11/2012 8 .9270616 .3835395 7528. 06/11/2008 7 -1.635873 -.6767856 6483. 04/11/2004 6 .487504 .2016878 5443. 09/11/2000 5 -.9045703 -.3742345 4398. 07/11/1996 4 2.176657 .9005163 3353. 05/11/1992 3 -5.90968 -2.444925 2314. 11/11/1988 2 1.978018 .8183361 Date event_id CARIsrael testIsr~l 9618. 10/11/2016 9 1.240868 .3826743 8573. 08/11/2012 8 .1102757 .0340082 7528. 06/11/2008 7 6.563006 2.023981 6483. 04/11/2004 6 .5874755 .181173 5443. 09/11/2000 5 1.198105 .3694866 4398. 07/11/1996 4 3.355864 1.034923 3353. 05/11/1992 3 -1.624089 -.5008565 2314. 11/11/1988 2 2.553819 .7875783 Date event_id CARIndo~a testI~sia 9618. 10/11/2016 9 1.730504 .5128901 8573. 08/11/2012 8 -.7778048 -.2305273 7528. 06/11/2008 7 12.69308 3.762001 6483. 04/11/2004 6 -.1066328 -.031604 5443. 09/11/2000 5 2.60257 .7713547 4398. 07/11/1996 4 4.611129 1.366655 3353. 05/11/1992 3 1.810474 .5365919 2314. 11/11/1988 2 1.969776 .583806 Date event_id CARThai~d testTha~d

~ 28 ~ 9618. 10/11/2016 9 -1.867679 -.4070933 8573. 08/11/2012 8 .3905246 .0851217 7528. 06/11/2008 7 3.925981 .8557364 6483. 04/11/2004 6 -.3665482 -.0798956 5443. 09/11/2000 5 -2.671928 -.5823937 Date event_id CARSing~e testSin~e 9618. 10/11/2016 9 -1.552751 -.7816679 8573. 08/11/2012 8 .0576875 .0290404 7528. 06/11/2008 7 4.82954 2.431231 6483. 04/11/2004 6 -.3422451 -.172289 5443. 09/11/2000 5 .6293262 .3168081 4398. 07/11/1996 4 .8750057 .4404852 3353. 05/11/1992 3 -1.609482 -.810227 Date event_id CARBelg~m testBel~m 9618. 10/11/2016 9 7.198056 3.369955 8573. 08/11/2012 8 1.749487 .8190674 7528. 06/11/2008 7 -.2114086 -.0989764 6483. 04/11/2004 6 1.134239 .5310231 5443. 09/11/2000 5 .4789042 .2242114 Date event_id CARArabia testAra~a 9618. 10/11/2016 9 -.9541386 -.6954516 8573. 08/11/2012 8 .1828566 .1332803 7528. 06/11/2008 7 1.14937 .8377517 6483. 04/11/2004 6 -1.135189 -.8274152 5443. 09/11/2000 5 .302281 .2203263 4398. 07/11/1996 4 .6238648 .4547219 3353. 05/11/1992 3 -.8315554 -.6061032 2314. 11/11/1988 2 -2.173188 -1.583991 1268. 08/11/1984 1 -1.508929 -1.099826 Date event_id CARNeth~s testNet~s 9618. 10/11/2016 9 3.78775 .8743208 8573. 08/11/2012 8 -1.231488 -.2842625 7528. 06/11/2008 7 -.9222625 -.2128845 6483. 04/11/2004 6 -4.019522 -.9278203 5443. 09/11/2000 5 -9.58439 -2.212351 Date event_id CARRussia testRus~a

~ 29 ~ 9618. 10/11/2016 9 -3.390077 -.7546313 8573. 08/11/2012 8 .7855016 .1748527 7528. 06/11/2008 7 -1.48632 -.3308549 6483. 04/11/2004 6 5.614052 1.249688 Date event_id CARColo~a testCol~a 9618. 10/11/2016 9 -3.534301 -2.043291 8573. 08/11/2012 8 -1.688595 -.9762303 7528. 06/11/2008 7 3.54742 2.050876 6483. 04/11/2004 6 -.8422398 -.4869255 5443. 09/11/2000 5 -1.163984 -.6729363 4398. 07/11/1996 4 2.047761 1.183876 3353. 05/11/1992 3 -.9997268 -.5779737 2314. 11/11/1988 2 .218179 .1261362 Date event_id CARSpain testSpain 9618. 10/11/2016 9 2.268835 3.074739 8573. 08/11/2012 8 -1.821409 -2.468384 7528. 06/11/2008 7 0 0 6483. 04/11/2004 6 0 0 5443. 09/11/2000 5 0 0 4398. 07/11/1996 4 1.501542 2.034899 Date event_id CARVene~a testVen~a 9618. 10/11/2016 9 .3922852 .1773293 8573. 08/11/2012 8 -.1964149 -.0887877 7528. 06/11/2008 7 4.281562 1.935445 6483. 04/11/2004 6 1.239629 .5603641 5443. 09/11/2000 5 -.199042 -.0899753 4398. 07/11/1996 4 -.2489194 -.1125219 3353. 05/11/1992 3 -2.968912 -1.342072 2314. 11/11/1988 2 .9346389 .4224957 Date event_id CARAust~a testAus~a

~ 30 ~

Table 6: Regression analysis of the cumulative average abnormal returns

9618. 10/11/2016 9 -.6228718 -.0699623 8573. 08/11/2012 8 .8529559 .0958058 7528. 06/11/2008 7 2.041605 .2293174 6483. 04/11/2004 6 -1.186211 -.1332377 5443. 09/11/2000 5 20.28384 2.278324 4398. 07/11/1996 4 1.816086 .2039867 3353. 05/11/1992 3 -1.394618 -.1566464 2314. 11/11/1988 2 -.3606672 -.0405109 Date event_id CARPhil~s testPhi~s 9618. 10/11/2016 9 1.215144 .5734557 8573. 08/11/2012 8 1.344564 .6345324 7528. 06/11/2008 7 5.266338 2.485312 6483. 04/11/2004 6 -.0428292 -.0202121 5443. 09/11/2000 5 -2.972742 -1.402909 4398. 07/11/1996 4 -.9771639 -.4611473 3353. 05/11/1992 3 1.284742 .6063008 2314. 11/11/1988 2 -1.355629 -.6397542 Date event_id CARSweden testSwe~n

_cons .3007951 .2233096 1.35 0.188 -.1559243 .7575145 CAAR Coef. Std. Err. t P>|t| [95% Conf. Interval] Robust Root MSE = 1.2231 R-squared = 0.0000 Prob > F = . F( 0, 29) = 0.00 Linear regression Number of obs = 30