The Spillover Effect Between "China Concept" Shares, Chinese And American Securities Market

Under the background of increasingly closely tied of global economic, the transaction costs are lower and lower, and the speed of information transmission is faster, caused by the great development of information technology which strengthens the relevance of the changes between the domestic and international stock markets. It is found that under perfect market integration, the pricing of cross-listed assets should be predominantly driven by fundamental information, and that under partial market segmentation, non-fundamental factors such as the locations of their businesses may also have significant impacts on cross-listed asset returns. It is found from examining the price discrepancies between ADRs and their underlying shares that ADR returns were driven by US market sentiment.As a continuation of the existing literature, our study concentrated on the"China Concept"shares affected by the shaking of American and Chinese security markets. The analysis focused on the concept of spillover effect, the mechanism of spillover effect, and the main factors influencing the spillover effect. To examine the return spillover effect, Granger causality tests were performed. We also used Impulse Response Functions to examine the response of"China Concept"shares, Chinese and American security market returns over time. Variance Decompositions were employed to examine explanatory relationships. After comparing the advantages and disadvantages of models for analyzing the volatility spillover effect between stock markets, including univariate GARCH models,multivariate GARCH models and vector SV model, we chose Dynamic Conditional Correlation Multivariate GARCH (DCC-MVGARCH) to analyze the volatility spillover effect of"China Concept"shares, Chinese and American security markets. DCC-MVGARCH model is widely used not only because it takes the relevance of time-varying conditions into consideration but its explanatoriness is better than other models. To examine lead-lag relationships, a Vector Autoregression (VAR) with two lags was estimated. At last, we examined the descriptive statistics of intraday returns. Granger causality tests, Impulse Response Function and Variance Decompositions were examined by Eviews, and Winrats was used to make the estimation of DCC multivariate model and characterize the time-varying correlation coefficient.This research can not only investigate the spillover effect of Chinese security market and the'China Concept'shares, but also evaluate the degree of correlation between the stock markets. Based on the return and volatility spillover effects, the conclusion of the study is valuable, which can help the investors determine their investments.