Three Essays on Momentum Returns

Abstract

This dissertation consists of three essays on momentum returns. The first essay is entitled ‘Momentum Returns, Market States and the Global Financial Crisis.’ This essay investigates the profitability of the momentum trading strategy in the stock exchanges of Shanghai, Shenzhen and Hong Kong over the period 1994 to 2010. My results show that there are significantly large momentum returns for Shanghai and Hong Kong but small and insignificant momentum returns for Shenzhen. Momentum returns remain almost the same for all datasets even after adjusting for the Fama-French risk factors. However, the momentum trading strategy generates negative returns for all markets during the Global financial crisis; it appears that the momentum trading strategy fails during a financial or stock market crisis and especially in the months when the market conditions improve. I find no significant relationship between momentum returns and market states, which contradicts the results of an earlier study conducted in the U.S. market. Instead of market state it appears that it is economic activity that explains momentum returns. In the second essay, entitled ‘Momentum Returns and Information Uncertainty’, I study the impact of information uncertainty on the profitability of the momentum trading strategy. Prior literature attributes returns of the momentum trading strategy to investor behavioural biases, such as under reaction to new information, and provides evidence that firms with higher information uncertainty earn lower future returns and higher momentum returns. This study investigates the relationship between information uncertainty, future returns and the momentum trading strategy in China over 1994 to 2010. Using age, volatility, volume and firm size as information uncertainty proxies, I find that firms with higher information uncertainty earn higher average returns over the succeeding 6-month period when information uncertainty is defined in terms of firm size and volume, consistent with the traditional risk based theories. However, firms with high information uncertainty earn lower returns if information uncertainty is defined in terms of firm age; however, firm age might not be an important factor in explaining future returns as the age difference among young and old firms is quite small and it might be the other firm-specific characteristics that drive lower future returns for the younger firms instead of their age. Therefore, my study rejects the findings of Jiang, Lee and Zhang (2005) that stocks with higher information uncertainty earn lower returns over the succeeding 6-month period. Using portfolio-level analysis and firm-level Fama and Macbeth regressions, I find that no robust significant relationship exists between information uncertainty and momentum returns. My findings suggest that it might be the activities of retail investors that drive momentum returns since China’s market is dominated by retail investors, whereas the momentum effect is weak among the U.S. stocks with higher institutional holdings.In the third and final essay, entitled ‘Momentum Returns, Long-Term Reversal and Idiosyncratic Volatility’, I study the impact of idiosyncratic volatility (IV) on the profitability of momentum and long-term reversal trading strategies in China over the period 1994 to 2010. The literature suggests that IV deters arbitrage, which results in higher momentum and reversal returns for stocks with high idiosyncratic volatility. However, my results indicate that the choice of the proxy used for IV, sorting method and number of portfolios (tercile versus quintile) play critical roles in determining the existence and significance of a relationship between IV and the profitability of momentum or reversal trading strategies. Both portfolio-level analysis and firm-level Fama and Macbeth regressions indicate that no robust relation exists between IV and momentum or reversal, which contradicts the results of earlier studies in the U.S. market. However, my findings suggest that reversal returns might be related to transaction costs since I find greater reversal returns among the stocks with high transaction costs. My findings also do not support the suggestion that momentum and reversal are part of same phenomenon.