Research On VaR Of Chinese Stock Market Based On Quantile Regression

Financial innovation and financial globalization make modern financial institutions face greater market risks.The continuous reform of China’s financial market,on the one hand,accelerates the opening up,on the other hand,it also brings huge challenges to financial management.Therefore,it is imperative to strengthen financial risk control.The key to risk management is risk measurement.There are many methods of risk measurement,among which VaR is the most commonly used one.Since the essence of VaR is a quantile and it is in good agreement with the quantile regression method in theory,this paper combines the theory of quantile regression with VaR theory to conduct research and explore a model suitable for China’s national conditions,which is of great significance for the development of China’s financial market.The main content of this paper is the VaR research of Chinese securities market based on quantile regression.Firstly introduces the VaR and quantile related theory knowledge,then combination model,adopting from Shanghai 50 index and composite index data,small and medium-sized plate and its logarithm,respectively in four assumptions: the confidence level,different GARCH model,the residual distribution form and the different holding period under quantile regression,and calculate the VaR value,in order to analysis the most excellent of assumptions in the calculation results of hypothesis.Finally,the effectiveness of the model evaluation under different hypotheses was evaluated by the kuidization likelihood ratio test method.Through empirical analysis,it is found that the VaR model of quantile regression based on the basic idea of quantile regression is very suitable for studying the financial market of China.There is no obvious difference in the VaR model of the quantile regression of sse 50 index under the assumptions of different residual models and different residual distribution forms.However,there is a significant difference between the calculated results of sse 50 composite index in different residual distribution forms and the VaR model of the quantile regression of sse 50 index.In addition,if the holding period is longer,the absolute value of VaR will be larger,and the estimation accuracy of the quantile regression VaR model will be higher.In this paper,the quantile regression method is combined with VaR to establish a risk measurement model based on the characteristics of China’s financial market,which has certain reference value for financial management institutions.