The Research Of VAR Risk Management Based On Price Range

U.S. financial crisis in 2008, quickly spread to global financial markets, so that people once again realize the importance of risk management. VAR as the latest developments in risk management can be integrated in a unified framework of financial institutions to measure the risks of different assets, many financial institutions use VAR model in risk management. VAR risk management systems in financial theory and practice have important significance, how to improve the effectiveness of risk management VAR later become the focus of academic research.In the existing method of calculation VAR, it is also not considered high and low price, just use the closing price; in the model of the highest price and lowest price, AV model and CARR model do not use the closing price. Using the closing price, high and low, in a unified framework, including price range and closing prices, to estimate the probability distribution of financial assets, in particular, describe the variance of price volatility. Based on this, the paper studies how to improve the validity of VAR.In our country, the application of GARCH model in calculation value at risk VAR, has achieved fruitful results from the probability distribution of returns subject to a variety of extensions to the GARCH model, but because of differences in sampling makes the existence of empirical conclusion is not consistent. For the current study, the model only do changes in the details, GARCH models use only the logarithm of the closing prices y, ignoring the other information included the highest price and the lowest. The price range between highest price and lowest price is the scope of price changes in trading days, including the formation of how the assets get closing price. According to GARCH models, Paper has established a range-GARCH model.In the range-GARCH model, Price range makes the conditional variance reflecting the risk timely, in turn, increases the effectiveness of VAR forecasts to reduce potential losses of financial assets.Paper uses the theoretical analysis and empirical method, the result of empirical analysis is to verify the reasonable of theoretical model. The part of theoretical analysis, based on VAR calculation principle and the core idea of GARCH model is to establish range-GARCH model and the corresponding VAR formula. After ARCH effect test, Empirical part uses the GARCH model and the EGARCH model to analyze volatility of asset, and calculates value at risk VAR, and finally to test the validity of VAR model by Kupiec likelihood ratio test.In order to expand the specific analysis and discussion, Thesis is divided into five parts:Part 1 is Introduction, describes the research background and significance, innovation of thesis, research methodology and framework structure.Part 2 is literature review, introduces background of VAR management and what research has done by Chinese scholars; Parameters in the calculation of VAR approach, many scholars have worked variety empirical analysis of GARCH models from the obedience of the probability distribution to the various derivatives of GARCH models; the research results of the price range.Part 3 is theoretical analysis about range-GARCH model and range-VAR, describes the meaning of value at risk, calculation principles, steps and the main method; established rang-GARCH model; under the conditional variance of range-GARCH model, derived VAR formula.Part 4 is empirical analysis of range-VAR, Describes the four selected samples, the data processing method; after the sample data station test and ARCH effect test, the four samples'empirical analysis of GARCH, EGARCH; Under the conditional variance of GARCH model and the range-GARCH model, calculate the risk value of the VAR, the results of the two methods were compared and the Kupiec likelihood ratio test. Part 5 is the research conclusions, Returns characteristics, the choice of GARCH model, the range-GARCH, VAR calculation and back measurements. Finally, the paper pointed out the inadequacies and the further research.Through the thesis research, we get the following conclusions:the returns of Shanghai and Shenzhen stock market do not exist significant differences; Price range shows that the volatility of the stock market in Shenzhen than Shanghai stock market; Returns approximately obey the random walk process. In most cases GARCH (1,1) model can be well fit the volatility of Returns. Price ranges can be used as an alternative conditional variance lagged variables to explain the sustainability issue of financial assets'volatility. Although the stock price ranges reflect the fluctuations in the trading day range, more prone to extremes with very small volume, making the price of two stocks on the conditional variance of range did not show uniform patterns. The price range contains important information about the price fluctuations, range-GARCH model can be more timely and sensitive reflection of the risk of financial assets, reducing the number failures of VAR, this improvement through the Kupiec likelihood ratio test, proved to be effective.