Effect of Election Preferences on the Stock Prices

• Název práce v češtině: Efekt volebních preferencí na ceny akcií
• Název v anglickém jazyce: Effect of Election Preferences on the Stock Prices
• Klíčová slova: sentiment, volatilita, GARCH, lexicon
• Klíčová slova anglicky: sentiment, volatility, GARCH, lexicon
• Akademický rok vypsání: 2016/2017
• Typ práce: diplomová práce
• Jazyk práce: angličtina
• Ústav: Institut ekonomických studií (23-IES)
• Vedoucí / školitel: prof. Ing. Evžen Kočenda, M.A., Ph.D., DSc.
• Řešitel: skrytý - zadáno vedoucím/školitelem
• Datum přihlášení: 16.06.2017
• Datum zadání: 16.06.2017
• Datum a čas obhajoby: 16.01.2019 08:30
• Místo konání obhajoby: Opletalova - Opletalova 26, O206, Opletalova - místn. č. 206
• Datum odevzdání elektronické podoby: 30.12.2018
• Datum proběhlé obhajoby: 16.01.2019
• Oponenti: PhDr. Jiří Schwarz, Ph.D.

Seznam odborné literatury

1. Santa–Clara, P., & Valkanov, R. (2003). The presidential puzzle: Political cycles and the stock market. The Journal of Finance, 58(5), 1841-1872.
2. Adjei, F., & Adjei, M. (2017). Political Cycles, Investor Sentiment, and Stock Market Returns. Journal of Finance and Economics, 5(1), 1-10.
3. Bialkowski, J., Gottschalk, K., & Wisniewski, T. P. (2008). Stock market volatility around national elections. Journal of Banking & Finance, 32(9), 1941-1953.
4. Veronesi, P. (1999). Stock market overreactions to bad news in good times: a rational expectations equilibrium model. The Review of Financial Studies, 12(5), 975-1007.
5. Brooks, C. (2014). Introductory econometrics for finance. Cambridge university press.
6. Cutler, D. M., Poterba, J. M., & Summers, L. H. (1989). What moves stock prices? The Journal of Portfolio Management, 15(3), 4-12.
7. Durnev, A. (2010). The real effects of political uncertainty: Elections and investment sensitivity to stock prices.

Předběžná náplň práce v anglickém jazyce

Motivation:
It is common sense that politics and economics are deeply interconnected. In the modern world, every political act affects changes on the financial markets to varying degree. Stock markets are extremely volatile by its nature. Investors are aiming to determine any operation that might change the stock price and adjust their strategies accordingly. Stockholders are concerned about the elections since there is an uncertainty about how elected candidate would manage the economy.

Historical data provides us with information that presidential cycles are correlated with stock markets returns. For example, in past 120 years in USA Democrats have been better for stocks then Republicans, as more wars have started during the Republican presence in White House. This knowledge has an influence on the decision-making of investors and can be seen as an example of behavioral finance.

In the research we are going to focus on two election races: US and French President Elections in 2016 and 2017 respectively.

Various researches proves that the investors’ behaviour on stock market during the race period is changing due to some level of uncertainty about future of the economy. A lot of studies support the hypothesis that market reacts more to bad news rather then to good news. Rising support of less radical candidate is assumed as good news, while rising support for radical candidate is considered as a bad news because of higher level of uncertainty, investors are not sure how the elected candidate would manage the economy. We are going to test whether stock prices react differently to change in preferences of population for different candidates.

Hypotheses:
1. Hypothesis #1: Including candidates’ ratings into stock prices volatility modeling have positive effect on the predictions and provides us with more accurate outcomes of the model.
2. Hypothesis #2: Higher support for radical candidate does not imply stock price stability
3. Hypothesis #3: During election race period market does not tend to be declining.

Methodology:
First step of our Volatility analysis will be performed by gathering the timeseries data for the choosen indices, closed priced for selected indeces (S&P500, Dow Jones and etc.) from Yahoo Finance database.
Next, the surveys for the candidates support would be gathered and indexed according to their change in preferences. In order to perform this we will be usiing Naive Bayes algorithm – probabilistic classifier based on applying Bayes theorem. And Python programming language will be used to build the classification.
The volatility of the stock prices will be evaluated using GARCH family of models. We will use GARCH and EGARCH to estimate the volatility of stock prices These models are can estimate the variance of a series at a particular point in time and thus, they are the best estimators as volatility is simply described as ’the conditional variance of the underlying asset return’. Further we will use TGARCH modeling to find the support for our hypothesis that rising support of radical candidate have more effect on the stock prices volatility rather then less radical.

Expected Contribution:
The model will be tested on the latest dataset for selected indeces. We will construct two models for two elections races, what will give us the opportunity of comparison and improvement. Model will develop prediction for the stock price volatility and give more accurate results.

Outline:
1. Introuction
2. Literature review
3. Methodology (Machine learning, GARCH family of models)
4. Data Description
5. Model estimation and Empirical Results
6. Conclusion
7. Suggestions for Further Extensions