An Empirical Study On ETF For Price Discovery And Information Spillover In China's Stock Exchange Market

Based on CCF(cross-correlation function),we apply the S-statistic made by Cheung and Ng(1996) and the Q(M) -statistic proposed by Hong(2001) to test the price discovery ability and information spillover effect of the five ETFs in China's stock exchange market.The empirical results show that:1) the 50ETF,180ETF,HL ETF and SZ100ETF Granger in mean cause the corresponding indices.Meanwhile,the above four ETFs also have significant information spillover effects in mean on the composite indices for each one's market.On the opposite direction,there are no significant evidence about information spillover from indices to ETFs.These results imply that information spillover is mainly from ETFs to indices.On the other hand, ETFs have stronger price discovery ability than the corresponding spot indices in China.2) As to the above four ETFs,we don't find significant evidence on volatility spillover between ETFs and the corresponding indices.However,although the ZXETF doesn't have significant Granger causality in mean with the corresponding indices,there is significant evidence on volatility spillover from ZXETF to the indices Generally,we believe that the introduction of ETFs to China's stock exchange market increases the market efficiency;however,more improvement should be made to the arbitrage rules of ETFs.Additionally,our study also shows that compared with Cheung and Ng's(1996)S-statistic,Hong's(2001)Q(M)-statistic has more advantage in capturing correlation information when a relatively long lag period is considered.