An empirical analysis of idiosyncratic volatility and extreme returns in the Japanese Stock Market

Abstract

Traditional finance theory assumes that systematic risks cannot be diversified in the market and need to be priced, while idiosyncratic risks should be diversified away in a rational investor’s portfolio and should not be compensated by risk premiums. However, due to various reasons such as incomplete information or market failures, investors in reality often do not hold perfectly diversified portfolios and thus idiosyncratic volatility is priced. Investors who hold stocks with higher idiosyncratic volatility expect a higher return. This study examines idiosyncratic risk behaviour from different perspectives. The objective of this research is to determine the effects of idiosyncratic volatility and extreme returns on the Japanese stock market. This research will determine whether there is a trend in the Japanese stock market using the t-dan test introduced by Bunzel and Vogelsang (2005). If the trend exists, we test on the factors that lead the trend behaviour. This study also examines if idiosyncratic volatility and extreme returns are priced in the Japanese stock market. The Fama and French three factor model, cross-sectional Fama Macbeth (1973) regression and double sorting stocks into portfolios based on variables of interest will be employed to test this effect. The study results identified four major findings on idiosyncratic volatility and extreme returns in the Japanese stock market. First, the result showed a negative and robust trend in idiosyncratic and market volatility in the Japanese stock market between 1980 and 2007. Second, the research findings confirm that volatilities, whether equal weighted, value weighted idiosyncratic volatility, market volatility or maximum daily returns are unable to forecast one month ahead excess market returns. However, the research results confirm with Brockman and Yan’s (2006) finding, who found no evidence of forecasting ability during their research sample from January 1926 to June 1962 in the US stock market. Third, the result showed a negative relationship between idiosyncratic volatility and expected stock return and the reverse between maximum daily return and expected stock return. Both the idiosyncratic volatility and extreme return finding is consistent with Ang, Hodrick, Xing and Zhang (2006), Brockman and Yan (2006) and Bali et al.’s (2010) findings for the U.S market. Finally, the result showed a highly significant inverse relation between idiosyncratic volatility, maximum daily return and cross-sectional returns.