5. Concluding remarks
The existing evidence on the relation between momentum and IV is mixed. Arena et al. (2008) document a positive relation between momentum and IV in the U.S. stock markets which supports both the underreaction and the overconfidence and self-attribution explanations of the momentum anomaly. However in a subsequent study McLean (2010) argues that there is no relation between momentum and IV and shows that Arena et al.’s (2008) results were obtained by excluding small size and low priced stocks thereby eliminating most of the high IV stocks in their sample. Apart from rejecting both the underreaction and the overconfidence and self-attribution stories, McLean’s (2010) results imply that idiosyncratic risk is not a limit to the arbitrage of momentum returns.
We verify the relation between momentum and IV in China and in selected Asian countries. Consistent with McLean (2010) we find at best, no relation between momentum and IV in China and in selected Asian markets supporting the view that idiosyncratic risk is not a significant arbitrage cost for momentum returns. While the absence of a positive relation between momentum returns and IV also rejects both the underreaction and overconfidence stories of momentum, we find support for the overconfidence and selfattribution story from our results on market dynamics and momentum. We find that when we condition momentum return on market dynamics, momentum returns are significantly higher when markets continue in the same state than when they transition to a different state, consistent with the prediction of Daniel et al.’s (1998) model but inconsistent with of the model of Hong and Stein (1999). We also find support for the suggestion that cross-country differences in momentum returns could be the result of different market dynamics rather than differences in levels of individualism as earlier suggested by Chui et al. (2010).
