Study On The Application Of Var Risk Measurement Methods In Chinese Stock Market

Followed by the developing of global economic, Chinese financial institution is now searching more opportunities in other countries. At the same time, foreign financial institutions are also arriving in china and will have a fierce competition with domestic institutions. Compared with foreign funded companies, domestics have less knowledge about risk management. Therefore, if we want to improve our ability of competition and avoid to become loser, we must enhance our capability of financial risk management and master some advanced knowledge about financial.During these years, the stock market of our country is developed a lot, but as well as some drawbacks. Currently, stock market is ongoing emerging and reforming status. The market is affected by policy seriously and the volatility of the market is ups and downs frequently. The correlation between the stock market and economic is high, the fragile capital market will lead to systemic risk. Consequently, in order to raise the standard of the security market, popularize risk management is essential.Firs, this paper recall some cases of financial risk management. Based on such knowledge, it provides an overview of the technology of risk management abroad and domestic. As a new method of risk management, VaR is used by experts widely.Second, it introduces the definition, formula and principle of the VaR as well as compared kinds of ways to compute VaR. This paper illustrates the way of historical simulation, monte carlo simulation and GARCH. Historical simulation and monte carlo simulation is full valuation and they consider all kinds of risk factors in the market, but historical simulation depends on historical data and mente carlo simulation is complicated during computing. The GARCH is one example of part valuation method, but it can include the fat tail behave.The stock index is used in this paper, using historical simulation and GARCH model to analyze. When utilizing the GARCH model, ARMA(1,1)-GARCH(1,1) model is used to compute VaR and the result is impartial. After using models to analyze, it uses back test to test the result. After the research, it finds that historical simulation is likely to overstate the risk and oppositely the GARCH method is likely to underestimate risk on our stock market.Finally, it summarizes this paper and discusses the prospects of application of VaR in china.