Contrast Study And Empirical Analysis Of VaR And CVaR

Since China has joined WTO, its financial industry will have to be faced with the challenges from the world. So it is an important and urgent task to search for fitful methods and systems to manage risk. Value-at-Risk (VaR) approach has become the standard method for risk management organizationVaR is very popular recently. But it has various theoretical deficiencies as a measure of market risk. Conditional VaR (CVaR) is an alternative risk measure to the quantile which overcomes the theoretical deficiencies of VaR. In particular, this risk measure gives some information about the size of the potential losses given that a loss bigger than VaR has occurred. This paper estimates and assesses tail-related risk using VaR and CVaR together.Though VaR and CVaR have many computing methods, they have limitations. Because almost all of the traditional methods estimating tail-related risk VaR and CVaR focus on the central observations or, in other words, on returns under normal market conditions.However, VaR and CVaR are risk measures that relates solely to the tails of the distribution. The extreme values which lies in the tail are some rarely happened events that have significant influence. Extreme Value Theory (EVT) is the statistical model to study the behavior of extreme values. This paper introduces the basic knowledge of EVT and estimates VaR and CVaR using EVT. The application methods of EVT have block maximum method and peak over threshold method according to related ways of identifying extremes in real data. This paper gains the formulas computing VaR and CVaR using peak over threshold method and block maximum method. Then makes empirical analysis for Shanghai index using peak over threshold method. Based on the stock data of S&P 500 index. this paper makes an empirical analysis of VaR and CVaR estimation using block maximum method of EVT. Empirical findings conclude that block maximum method can well approximate the tail of financial return distribution. Backtesting indicates block maximum method can accurately estimate VaR and CVaR for high probability excess 95%.In the last part, study of VaR estimation based on GARCH model is used to analyze the volatility clustering phenomenon in the return ratio series. As for the random component in the model, normal distribution, t-distribution and generalized Pareto distribution assumptions are applied, respectively. Best result is obtained based on the generalized Pareto distribution assumption for the random component in the GARCH model.