Essays on financial market anomalies

Abstract

This dissertation consists of three essays on financial market anomalies. The first essay examines whether asset pricing anomalies are evidence of a failure of efficient markets and equilibrium asset pricing models based on stochastic dominance test. I document that stocks with low anomalous characteristics stochastically dominate stocks with high anomalous characteristics in second- and third-order stochastic dominance over the period from 1975 to 2010. I conclude that most of asset pricing anomalies are more likely evidence of market inefficiency, rather than result of omitted risk factors and suggest the possibility of disappearance of market inefficiency over time. This second essay investigates style momentum phenomenon in anomalies. I define new styles using anomalous characteristics in recent anomalies (accruals, asset growth, composite stock index, distress risk, idiosyncratic volatility, corporate investment, and net stock issue) and find that style momentum exists. I study all NYSE, Amex, and NASDAQ stocks since 1975. Price and earing momentum strategies that buy winner style portfolio in each anomalous characteristics and that sell loser style portfolio perform well. Style momentum based on each anomalous characteristic is distinct from individual stock return and earnings momentum. Strategy using all anomalous portfolios also performs well. From this empirical phenomenon, I suggest the momentum strategy which selects the best portfolio among 7 anomalies. The third essay examines systemic risk is priced in the equity market. I focus on the CoVaR measure and find strong evidence that stocks with low systemic risk perform better than stocks with high systemic risk. This finding is different from existing anomaly about low beta and size effect. There is no evidence that higher returns of the lowest portfolio are compensation of distributions and risks. I conclude these return patterns are new anomaly and the equity market does not reflect systemic risk.