Abstract
VaR means that at some confidence level, the estimated greatest loss of the financial assets may be suffered within given period of time in the future. In the paper we use the Jarque-Bera method to testify Chinese Shanghai Index, Chinese Hong Kong hang seng index, America Dow Jones Index, French CAC40 Index, Germany DAX Index, UK FTSE100 Index and Japanese Nikkei 225 Index, the result is that the distribution of index returns of every stock market present sharp peak and heavy tail. We use BMM model to measure VaR of stock market, this method can solve the problem of underestimating the tail risk effectively. We choose the time scope from Jan-2 1992 to Feb-28 2011, and use BMM model to analyze the returns of the seven indices of stock markets. The result is that, the VaR of Chinese stock market is at the highest level about 24.408%, while the VaR of other six stock markets arc at relatively lower level. Stepping into 21(st) century, we can find that the gaps of VaR between Chinese stock market and advanced stock markets are reducing gradually. The relatives between Chinese stock market and the advanced stock markets are become more closely. At the end, we propose several measures about risk control in Chinese stock market.