Ah Dual Listings Of Asynchronous Transaction Earnings Predictability Research

Market efficiency has been a concern in academic. According to the Law of One Price, A, H shares have the same subject, both should have the same price and similar returns. Yet, researches on Behavioral Finance shows that the market is not efficient, people can judge the movement of stocks based on historical information and public information. In this paper, We set1999-2010dual-listed shares in A, H shares market as samples to study the return predictability.In this paper, we regress the H shares’returns during the holiday on A shares’ returns after the holiday, and find that there is significant positive correlation between them. Subsequently, referencing Cohen and Frazzini (2008)’s method, we design a zero-cost strategy:buying H shares whose returns during the holiday ranked top1/3, then selling H shares whose returns during the holiday ranked bottom1/3. Our results show that this strategy can achieve significant positive abnormal return. This results shows that if the cumulative return of H-share during the holiday is high (low), the return of corresponding A-share after the holiday will be high (low)we carry on empirical testing to make sure of the robustness of the results. We conduct Granger causality test, and if there is Granger causality in time series, we will excluded these samples. Then we carry on empirical testing using the remaining samples, and find that the result is the same as the former. In another robustness test, we combine62dual-listing AH stocks, calculating the average of62H shares’ cumulative return during the holiday, then use the average returns to predict the average cumulative abnormal return of A shares after the holiday. After the combination, we can see that the return predictability does not exist. All of the empirical results demonstrate the return predictability of A, H shares does not come from the linkage in time series, but comes form cross-section which caused by "asynchronous".