Essays on financial markets

Abstract

This thesis comprises three empirical studies, which investigate the influential factors of financial markets and their participants’ behaviour. These studies can be read independently. Using a sample of European banks, the first study, “Corruption culture and bank short-termism”, provides evidence that country-level corruption is strongly associated with short-termism (a behaviour that focuses on short-term benefits at the expense of long-term shareholders’ wealth growth). To measure short-termism, I construct a multi-dimensional index which combines earnings management, tail risk, and short-term debt ratio. I show that banks headquartered in countries that are more corrupt behave more short-sightedly than banks headquartered in countries that are less corrupt. I further demonstrate that foreign shareholders act as a channel through which corruption is imported. For banks located in countries with a lower than average corruption level, having more shareholdings by investors domiciled in countries that are more corrupt leads to more short-termism. This study highlights the link between bank short-termism and corruption of both headquartered countries and foreign shareholders. The second study, “Macroeconomic and political uncertainty and cross sectional return dispersion around the world”, examines whether return dispersion (the cross sectional variation in stock returns) could be used to measure the macroeconomic and political uncertainty of international financial markets. I show that return dispersion is able to capture uncertainties including local and global business cycles, international political instability, market general uncertainties, international country risk, and economic policy uncertainty. Stocks that are more sensitive to return dispersion have higher returns. Moreover, I compare return dispersion with another commonly used uncertainty measure: implied volatility. I find that they capture different aspects of uncertainty. This study aims to provide a simple and easy-to-use measure of uncertainty for both academic purposes and professional practice. The third study, “The performance of asset allocation strategies across datasets and over time”, evaluates the ex-ante performance of various commonly used asset allocation strategies including equal weighting, mean variance weighting, risk parity weighting, minimum variance weighting, equal risk contribution weighting, and maximum diversification weighting. The results show that there are no statistically significant differences between asset allocation strategies if the portfolios are based on country or industry indices. If the portfolios are formed of stocks or multi-asset classes, then the differences between strategies are large; however, none of the strategies can consistently outperform the others over time. I also identify that the implementation of the mean variance rule leads to wide fluctuation in risk shifting, which is undesirable for investors. Last but not least, I illustrate how all of the strategies are far from ex-ante optimal.