Compute The VaR Of 50ETF Based On GARCH Process And SV Model

50etf options has been listed on the Shanghai stock exchange officially on February 9,2015, opening the prelude to the development of options market in China.50etf options is the first listed options product in the capital market of our country, filling the gap of financial products in our country’s securities exchanges. However, as the subject matter of options, the importance of 50etf in the stock market has been further highlighted, which is because of that the biggest risk in the options price is its underlying asset price change risk(delta). If you can control the risk of 50etf, you will hedge the risks of the 50etf options and achieve steady investments, which has positive significance. Therefore, this thesis selects the accumulative unit net value of 50etf as the research object and calculates the rate of return series, with value at risk(VaR) as the index to measure the market risk.In this thesis, we firstly adopt historical simulation and monte carlo sim-ulation, which belong to full valuation method, to calculate the VaR of 50etf. Then we discussed the normal test, ACF test, stationary test and arch test, showing that the series of rate of return has the important formal characteris-tics such as heavy tails and volatility clustering existing in financial series. We use the GARCH model based on error of normal distribution and t distribu-tion respectively to fit heavy tails and volatility clustering. Through empirical analysis, we find that the GARCH model based on t distribution which has heavier tails fitting to return series relatively better. In order to reflect the influence on rate of return from the current environment and conditions more accurately, we put forward to stochastic volatility (SV) model to improve the accuracy. Considering the complexity of parameter estimation in SV model, this thesis chose the markov chain monte carlo(MCMC) method to implement the parameter estimates. Through the empirical analysis, we conclud that the SV model based on MCMC method has the best fitting effect on return se-ries of 50etf and control market risk more effectively. GARCH model and SV model based on MCMC method is the focus of this thesis.This thesis mainly consists of four chapters.In chapter 1, we introduce the research background and significance of value at risk of 50etf firstly; Secondly, we introduce the history and research status at home and abroad of risk measurements in financial market; Finally. we summarize the research contents and methods of this thesis.In chapter 2, we mainly introduce the measurement of market risk— —VaR, including the basic principle, calculation methods, test methods and the advantages and disadvantages of VaR systematically. In the part of meth-ods of VaR, we mainly introduce the historical simulation, monte carlo simu-lation and GARCH process.In chapter 3, we focus on the SV model based on MCMC method. Firstly we explain the basic idea of MCMC method and sampling method, such as the Metropolis-Hastings sampling and Gibbs sampling; Then this thesis tells us the basic theories of SV model and how to carry out the volatility forecast.In chapter 4. the empirical analysis, we firstly discuss a series of test such as the normal test and so on.Then we calculate the VaR by using historical simulation, monte carlo simulation, the GARCH model and SV model based on MCMC method. At last, by Kupiec failure frequency test on the calculation results, we come to the conclusion and summarize the inadequate of this thesis.