|
|
Seasonal Effects on Stock Markets in Europe
Rosol, Jaroslav ; Kukačka, Jiří (advisor) ; Čornanič, Aleš (referee)
This thesis researches the problem of stock market efficiency and market anomalies. Specifically, we look on European stock markets and possible presence of four seasonal effects - January, Halloween, Turn-of-the-month and Monday effects. These seasonal anomalies imply that returns for specific period are unusually higher or lower than returns for the rest of the time, which presents a challenge for the Efficient Market Hypothesis. The empirical side of this problem is the possible opportunity for excessive profit from trading on stock markets that could be based on the seasonal anomalies. Firstly, we summarize previous research in the field and attempts of explanation of individual effects. Further, we present the tools needed for our analysis - Ordinary Least Squares regression with dummy variables and few extensions. Data used for the analysis consists of 32 European stock indices. The actual analysis is performed as a comparison of returns on stock for certain specified periods. The evidence on January and Monday effects is found not strong enough to confirm the presence of such anomalies. On the other side, there is enough significant evidence on the presence of Halloween and Turn-of-the-month effects. Moreover, we are unable to explain the Halloween effect as manifestation of January effect. Powered...
|
|
|
Impact of investors' mood on European stock markets
Rosol, Jaroslav ; Kočenda, Evžen (advisor) ; Šťastná, Lenka (referee)
This thesis investigates into relationship of mood and stock markets. As proxies for mood we selected Daylight Savings Time changes, Lunar cycles and football results. Any abnormal returns on related days would present a challenge for Efficient Markets Hypothesis. We examine stock returns on corresponding days using up-to-date daily data of European All Share indices and analyse them us- ing ARMA-GARCH model. Our results for Daylight Savings Time change are neither uniform nor statistically significant with one exception. Even though results for Lunar cycles mostly follow expected positive relationship, they also lack statistical significance. For France and Croatia we find statistically signif- icant negative relationship between wins of national teams and stock returns, which goes against our expectations and previous research. Statistically sig- nificant negative relationship is expected for days after losses and potential elimination of national teams from competition as we confirmed for Croatia and Italy. JEL Classification G14, G41, D53, D91 Keywords Finance, Shares, Mood, Seasonals Author's e-mail 23060835@fsv.cuni.cz Supervisor's e-mail evzen.kocenda@fsv.cuni.cz
|
|
|
Seasonal Effects on Stock Markets in Europe
Rosol, Jaroslav ; Kukačka, Jiří (advisor) ; Čornanič, Aleš (referee)
This thesis researches the problem of stock market efficiency and market anomalies. Specifically, we look on European stock markets and possible presence of four seasonal effects - January, Halloween, Turn-of-the-month and Monday effects. These seasonal anomalies imply that returns for specific period are unusually higher or lower than returns for the rest of the time, which presents a challenge for the Efficient Market Hypothesis. The empirical side of this problem is the possible opportunity for excessive profit from trading on stock markets that could be based on the seasonal anomalies. Firstly, we summarize previous research in the field and attempts of explanation of individual effects. Further, we present the tools needed for our analysis - Ordinary Least Squares regression with dummy variables and few extensions. Data used for the analysis consists of 32 European stock indices. The actual analysis is performed as a comparison of returns on stock for certain specified periods. The evidence on January and Monday effects is found not strong enough to confirm the presence of such anomalies. On the other side, there is enough significant evidence on the presence of Halloween and Turn-of-the-month effects. Moreover, we are unable to explain the Halloween effect as manifestation of January effect. Powered...
|
guest ::
RSS feed.