Momentum strategies, information diffusion, and investor conservatism in Asian-Pacific region

This study performs an out-of-sample test on momentum effects in nine Asian-Pacific stock markets from 1990 to 2002. We find little evidence on the existence of intermediate stocks return momentum. Specifically, the unrestricted momentum trading strategies appear profitable only in major developed countries but not in the emerging countries in our samples. We further investigate momentum effects through two behavioural theories: Hong and Stein's (1999) gradual-information-diffusion model and Barberis et al.'s (1998) investor conservatism bias model. Our results support Barberis et al. (1998) model but reject the Hong and Stein's information diffusion hypothesis. It shows the profitability of momentum strategies is positively related to analyst coverage but negatively related to analysts' forecast dispersion. We argue that investors' psychological conservatism biases reflect slow updating of their beliefs and underrate the importance of new information. Interestingly, we also find that book-to-market is more important than dispersion and dispersion is more important than size in explaining momentum returns.