Essays on international financial markets

This dissertation focuses on three issues on international financial markets. In the first essay, we examine an investor visibility event, a high volume shock, across countries to determine whether the event is pervasive and whether systematic differences across countries in their demographics, market characteristics and investor confidence affect the magnitude of the resulting return premium. We find that the high volume return premium is a persistent phenomenon found in both developed and emerging markets. Using Merton’s (1987) investor recognition hypothesis as a guide, we find the magnitude of the premium is associated with country characteristics hypothesized to affect returns subsequent to a change in a stock’s visibility. In the second essay, we investigate whether the liquidity risk documented by Pastor and Stambaugh (2003) for the U.S. market is a common risk factor by using data from 19 international stock markets. For many international markets, we find strong evidence that on the cross-section expected stock returns are associated with the sensitivities of stock returns to innovations in market liquidity. Stocks with higher sensitivities to market liquidity have significantly higher expected returns, in magnitudes comparable to, in some cases higher than, that of the U.S. market, even after controlling for exposures to the market return, size, value and momentum factors. Therefore, our research lends substantial support to the market liquidity as a priced systematic risk. In the third essay, we use a “natural experiment": a stock is traded in two marketplaces, where one is subject to price limits while the other is not, to identify the effect of price limits. We use the information on the U.S. market to estimate the price differential between the two claims and then predict what the price would have been on the Tokyo Stock Exchange if there were no price limits. We then predict the moments for the days that price limits were hit and use the predicted moments to compare with the observed stock moments. We conclude that price limits do not have a significant effect on the means and variances, so the proposed intent of having price limits is not supported by our analysis.