Financial Market Risk Measurement Based On Self-Excited Hawkes Process And Extreme Value Theory

Under the current reform of the supply-side structure,the inherent characteristics of China's financial market have forced it to experience cyclical crises,while the uncertainty of the domestic and international economic environment has increased,and the successive negative developments in the development of global financial integration.The impact made the stock market show huge volatility and vulnerability,which led to the volatility of the Shanghai and Shenzhen stock markets,resulting in a negative external spillover effect of the financial market on the real economy.Therefore,in the development of China's financial market,it is indispensable to prevent and resolve financial market risks.Effectively measuring financial market risks is particularly important for the healthy and stable development of China and the global economy.In the method of measuring financial market risk,the VaR method can accurately calculate the potential losses of various financial assets under different confidence intervals in the future,which is in line with the needs of modern financial institutions,so VaR measurement has become the main measure of financial market risk.means.Before measuring the risk value of financial market risk,it is necessary to choose the estimation method of the financial asset income sequence.The estimation methods of the yield series are divided into non-parametric method,semi-parametric method and parametric method.The nonparametric method is based on the simulation of historical data ranges.There are logical inconsistencies pointed out by Manganelli and Engle,that is,the nonparametric method improves accuracy through a long time window,while the financial asset return rate sequence requires shorter opening.The time window satisfies its time-varying characteristics.The nonparametric method also cannot explain the phenomenon of fluctuations in the financial assets' yield series,autocorrelation,peak and thick tail.Therefore,the nonparametric method has certain theoretical flaws.The semiparametric method is based on different probability distribution assumptions for different sample grouping data.It mainly has conditional autoregressive value(CAViaR),extreme value theory and GARCH model of quasi-maximum likelihood estimation.The extreme value theory is widely used.The extreme value theory focuses on the tail model of the distribution of financial assets' yields to effectively highlight the thick tail characteristics of financial assets' return.The parameter method is mainly developed on the basis of the GARCH series model,which can significantly characterize the phenomenon of fluctuation of financial assets' return rate and long memory.In view of the shortcomings of the nonparametric method,this paper will mainly discuss the VaR model of the parametric method and the semiparametric method.The empirical analysis part of this paper will be divided into the following three parts: the GARCH model,the POT model in extreme value theory,and the self-excited Hawkes-POT model to measure the financial market risk in China.The calculation of VaR under the assumption of different income distribution shows that the financial market risk measured by the self-excited Hawkes-POT model and POT model is greater than that of the GARCH model.Then,through the backtest results of VaR under different income distribution assumptions,the results of the self-excited Hawkes-POT model and the POT model are significant under the three probability levels of 5%,1% and 0.1% required by Basel III.Better than the GARCH model.Therefore,through the self-excited Hawkes-POT model for financial market risk measurement,it can not only effectively describe the peak and thick tail and volatility clustering phenomenon of financial assets return rate,but also improve the risk warning ability,and establish the early warning mechanism of financial market risk in China.Perfection is of great significance.The main innovation of this paper is to introduce the Hawkes class point process into the financial market risk measurement method in China,and make up for the defects of the POT model.The essence of the Hawkes model belongs to the linear self-excited point process,which is characterized by the fact that the events that have occurred will stimulate the occurrence of new events.Using this model to measure the market risk of China's stock market can effectively solve the weaker ability of extreme value theory in the distribution characteristics of the yield..Therefore,it is of great significance to improve the method of risk measurement in China's financial market,and improve the risk management ability of financial institutions.