Abstract: The paper studies the changes of comovement between international crude oil market and China and U.S.stock market before and after the U.S.sub-prime crisis based on dynamic conditional correlation multivariate GARCH model (DCC-MVGARCH).The empirical results demonstrate that the comovement increases after the sub-prime crisis while volatility between different markets has significant conduction.The comovement volatility between international oil market and U.S.stock market is stronger than China stock market,indicating the shock has more violent impact on international oil market and U.S.stock market and their response is more sensitive.In addition,we select some factors affecting international oil market and China and U.S.stock market to make further explanations on comovement and employ partial least squares (PLS) method to analyze the explanation power changes before and after U.S.sub-prime crisis.Results show that the sub-prime crisis significantly affected the explanations power of those factors.

Key words: comovement, dynamic conditional correlation, GARCH, PLS, sub-prime crisis