Co-movement And Price Discovery In HS300ETF And Stock Index Future Markets

In recent years, financial innovations of index derivatives have been introduced into Chinese market, among which HS300stock index futures in2010and HS300ETF in2012are groundbreaking. As in developed countries, it now makes chance for civil scholar to study the relationship in closely related parallel markets (HS300Index, ETF and Stock Index Future). In this paper, HS300ETFs and Stock Index Future are chosen as target markets to research Co-movement and Price Discovery of Chinese spot and future markets.First, Mean and volatility spillover effects are examined. We apply Cointegration test, Vector Error Correction Model, Granger causality test and Quantile Regression as empirical methods, and find that HS300ETF, Index and Stock Index Future markets exit long-term equilibrium as well as short-term correction. It is no earlier than2minutes for ETF market to discover price, while the situation in future market is at least12minutes in advance. Granger causality test shows that there are bidirectional mean spillovers between spot and future markets; Quantile Regression coefficients are U-shaped, which means the interaction between spot and future markets is more intense at extreme cases. Volatility spillover effects are examined based on realized volatility method and Granger causality test. ETF and stock index future markets exit bidirectional volatility spillover effects; the overall volatility is overflowed from the future market to the spot market, while the continuous volatility from spot market to futures market; jump conduction between both does not exist.Then, we study price discovery ability through CFW model and IS model. The results show that, no mater the stock index or ETF funds being regarded as spot price, future market price discovery ability is stronger. Further more, the intraday CFW indicators are calculated based on1minute high-frequency data to measure time-varying price discovery ability, which reflects that price discovery ability is not stable. On this basis, through multiple regression model of liquidity and volatility as factors, we found that relative volume (as liquidity proxy variable) has no contribution on price formation process, while overall volatility and future continuous volatility are negative correlated with price discovery ability, as well as continuous volatility of spot market are positive correlated. Meanwhile, we also find that jump will not affect the price discovery process.