VaR Empirical Research On China's Stock Market

VaR (Value at Risk) is a risk measuring technology that has just been developed in recent years. As it is straightforward, comprehensive and applicable, VaR has been welcomed by a variety of organizations, including the governmental institutions of the Basel Committee, Bank for International Settlements, and kinds of banking and financial institutions (such as insurance companies, security agencies, especially institutional investors), since J.P. Morgan first announced its VaR system in 1994. At present it has become the mainstream method for financial market risk management.Research on VaR in other countries is mature and there have been a lot of technological methods proposed since its introduction. In China the research on it is left relatively behind. With the deepening of Chinese financial sector reform, the domestic financial institutions are to establish risk management systems based on VaR risk measurement as have been done by international institutions. So it is quite necessary and realistic for us now to study VaR and its computation models thoroughly, and compare their different characteristics, while trying to master risk measuring technology.In this paper, we introduce VaR model systematically taking SSE 180 index in decreasing trend as the empirical research subject. We compare different models for VaR and point out their advantages and disadvantages. There are altogether six chapters. Chapter 1 is introduction. In Chapter 2, security market risk definition, meaning, source and its characteristics are briefly discussed, and security market risks are categorized. In Chapter 3, financial market measuring technology framework is introduced first, and then sensitivity analysis, volatility method, VaR, stress testing, and extreme value theory are discussed sequentially in the following sections. Chapter 4 is on the essential theories in this paper. Here we introduce models for VaR computation, especially historical simulation method, analytical method, and Monte Carlo simulation method, and then compare the three in detail. Chapter 5 is on VaR empirical research on China's stock market. Section 1 explains sample selection and manipulation.. Section 2 computes SSE 180 index daily geometric return rate VaR with the three methods of historical simulation, analysis and Monte Carlo simulation and this has been realized by a QBASIC program.Section 3 uses failure probability test to posterior testify and evaluate VaR efficiency. Chapter 6 concludes the paper and makes proposal on Chinese stock market risk management based on previous discussions.