I. VaR risk measurement methods There are two main types of risk measurement models: parametric models and nonparametric models. Parametric model includes models of various analysis methods, which simplifies VaR by using the characteristics of sensitivity and statistical distribution. However, due to the assumption of distribution form and the local characteristics of sensitivity, it is difficult for analytical methods to effectively deal with the nonlinear problems of actual financial markets, which leads to measurement errors and model risks. Nonparametric methods include historical simulation method and Monte Carlo simulation method. Compared with analytical method, simulation method can deal with non-normal problems well, is a complete estimation, and can effectively deal with nonlinear problems. 1. parameter analysis is the most commonly used method in VaR calculation, which simplifies the calculation of VaR by using the approximate relationship between the value function of securities portfolio and market factors and the statistical distribution of market factors (variance-covariance matrix). According to the different forms of portfolio value function, analysis methods can be divided into two categories: Delta model and Gamma model. Among them, the Delta model can identify linear risks, and the Gamma model can identify convex risks, such as derivatives with options in the portfolio. This paper will use Delta- normal model and Delta- GARCH model in Delta-class model for analysis. 2. Non-parametric method (1) Historical simulation method The simplest and most intuitive method is historical simulation method, the core of which is to simulate the future profit and loss distribution of securities portfolio according to the historical sample changes of market factors, and to represent the future changes of market factors by the changes of market factors observed in a given historical period. Then, re-evaluate the position according to the future price level of market factors and calculate the change of the value profit and loss of the position. Finally, in the historical simulation method, the profit and loss of the combination are sorted from small to large, and the profit and loss distribution is obtained, and the VaR is obtained through the quantile under the given confidence. (2) Monte Carlo simulation method makes use of statistical distribution characteristics. If there is a nonlinear problem of heavy tail and large fluctuation in the market, the deviation of risk measurement will be relatively large. Monte Carlo simulation is a stochastic process of repeatedly simulating and determining the price of financial instruments. Every simulation can get a possible value of the portfolio at the end of the holding period, and then a large number of simulations are carried out, and then the simulated distribution of the portfolio value will converge to the real distribution of the portfolio, and then the VaR will be obtained according to the confidence.