Publication Type

Master Thesis


I propose that various measures of mutual funds’ performance are more consistent with their investment capability when mutual funds present low idiosyncratic risks. This paper finds conditional predictor for funds’ returns: alpha predicts returns positively for low idiosyncratic risk funds. It suggests that mutual funds which showed high alpha and low idiosyncratic risk in the past may be capable in investment. Their performance is consistently higher than funds with low idiosyncratic risk and low alpha. On the other hand, the performance of high idiosyncratic risk funds is more likely to reverse in the future: expected returns are low for high alpha funds, and low alpha funds’ expected returns are high. I split the sample into 3 categories: funds with high idiosyncratic risk, low idiosyncratic risk and low alpha, low idiosyncratic risk and high alpha. Following Barras, Scaillet and Wermer(2010)’s method, I find out that the proportion of zero-alpha fund is highest within high idiosyncratic risk funds, and low alpha low idiosyncratic risk funds include the most unskilled funds. This paper also studies the predictive power of a variety of fund characters: alpha, idiosyncratic risk exposure, information ratio, and so on. However, none of them shows clear predictive pattern for expected returns. My observation reveals that information ratio does not predict returns in the full sample, but it indeed has strong predictive power for funds which keep long term growth, or growth and income investment objective.

Year Dissertation/Thesis Completed



mutual funds, idiosyncratic risk, risk taking, investment skills


Finance and Financial Management | Portfolio and Security Analysis

Degree Awarded

Master of Science in Finance


Joe Zhang


Lee Kong Chian School of Business