Risk Measures And Portfolio Based On VaR

VaR (Value at Risk) is currently the mainstream of risk measurement methods in financial markets, and is widely used in the portfolio, financial derivatives, market risk and credit risk analysis. This article compares the systematic study of many ways to calculate the portfolio based on VaR(Value at Risk), as well as constraints on the portfolio VaR model of the empirical analysis. Core content includes the following aspects:(1) It shows the definition of risk measures tools VaR and CVaR, compares the similarities and differences between the two, tells about the correspondence between VaR and CVaR from a new perspective, and analyse the VaR-based portfolio model combined with graph.(2) Combined with the historical simulation method, Monte Carlo simulation and the delta-gamma method, the article gives the portfolio values based on VaR of 10 options, and the comparative analysis of the three kinds of methods on the basis of simulation results.(3) In the framework of state-space model, we reconstruct the classical CAPM model, introduce random walk (RW) model for the time-varying behavior ofβcoefficient of 10 sectors on Shanghai stock market trades, bring in unknown parameters before the regression variables of the random walk model, estimate the new changed model and predict the risk coefficient with Kalman filter, and combining the time-varyingβcoefficients with Sharp--diagonal model, we can calculate the value of the industry portfolio VaR(4) We use a new intelligent approach -- genetic algorithms to solve M-V portfolio model with constraint of VaR, give the specific steps to solve the model, and show the empirical analysis of 10 stocks.