Research On The Jump Behavior And Jump Risk Of Chinese Stock Market

Study on financial asset return and volatility is important for asset pricing,portfolio selection and derivative product design.Usually,stock market returns change in a smooth and persistent way.However,in some cases asset returns may experience large discrete changes in a short time.In financial econometrics theory,this phenomenon is called jump.The probability of a jump occur is very small,but once occurred it will have a tremendous impact on the market.Multiple stock market crash in history started with a giant downward jump in the stock price.Chinese stock market as a developing market,it lacks a sound operational mechanism.The announcements of political policies,dramatic changes of macro economy and important international events may have a huge impact on Chinese stock market,causing drastic fluctuations in stock price.The fluctuation become more pronounced after the shock of financial crisis in 2007,snow storms and earthquake in 2008,and European debt crisis in 2012.From a micro perspective,Frequent jumps in volatility may increase investors difficulty in making investment decisions.While from a macro perspective,Frequent jumps in volatility intensifies market risk of the whole financial system,causing regulatory control difficulties and may even threaten the financial security of our country.Therefore,using an appropriate model to study the volatility and jump behavior of Chinese stock market becomes an urgent topic.An on appropriate jump model can provide effective risk management tools for investors and supervision departments and ensure stock market stability and healthy development of our country.In addition,with the contact between virtual economy and real economy becomes more and more closely,the impact of changes in macroeconomic policy on stock market volatility increase.Identify the relations between stock market and macro economy plays an important role in guarantee the stability of stock market development.As asset prices fluctuate frequently is one of the distinctive features of the financial market,financial risk that comes from financial asset price fluctuation is inevitable and unavoidable.The negative effects of financial risk increased with the development of economy and society.Therefore,effective risk management is widely recognized as the key to the sustainable development ability and competition ability of the financial industry in a country.As the basis of risk management,how to improve the accuracy of risk measurement under jump behavior is of great significance.At the same time,risk-return relationship is the core of the financial theory.Examine whether there is a relationship between them and to what extent directly relate to the applicability of different asset pricing model and the rationality of asset allocation.Thus the risk-return relationship is often referred to as "the first basic principle of finance".Based on these purpose,this research firstly adopt the ARJI-GARCH-Jump model proposed by Maheu and McCrudy(2004)to construct the random process of the jump intensity under a autoregressive structure and use a feedback function considering the feedback effect of jump on normal volatility.Empirical research shows that jump of Chinese stock market display significant time-varying and cluster effects.The feedback effect of jump on diffusive volatility is significant and asymmetric.The trend of jump intensity and ex post jump probability is basically the same,indicating that the ex ante forecast of jump basically verified after the event.A high ratio in conditional variance is caused by jump disturbance,and the GARCH variance component mainly explains the normal and smooth change of total variance while jump disturbance variance component explains the abnormal change of total variance.When studying the reason of jumps occur in stock market,we extend the autoregressive conditional jump intensity model by adding the current and lag variables of monetary policy and macroeconomic news announcements.Empirical study indicates that monetary policy and macroeconomic variables do have a significant effect on stock market jump behavior.Current variable of issuing Adjuste the deposit reserve ratio policy and lag variable of adjusting the deposit reserve rate have a significant impact on jump intensity.Interest rate adjustment policy do not have a significant impact on jump intensity.The current and lag variables of announcements of GDP,FI,RS,LOAN,M2 and the current variable of announcements of IAV significantly affect jump intensity.The current and lag variables of unexpected shock of GDP,TB,LOAN,M2,CPI,PPI,the current unexpected shock variable of RS and the lag unexpected shock variable of FI significantly affect jump intensity.When studying the risk measurement under extreme events,this paper estimates VaR via a two-stage process.The procedure starts with the ARJI-GARCH-Jump model to estimate the conditional mean and volatility of the entire distribution.In the second stage,the POT method of EVT is used to model the distribution of the residual.The empirical results show that:ARJI-GARCH-GPD model and GARCH-GPD model outperform simple GARCH class model and simple extreme value theory model.Under lower confidence level(90%)GARCH-GPD model outperforms ARJI-GARCH-GPD model while under higher confidence level(95%,99%,99.5%)the latter outperforms the former.Based on the results of a variety of back tests,ARJI-GARCH-GPD performs best on the whole while simple extreme value theory model performs worst.When examining the risk-return relationship in Chinese stock market,we integrate the autoregressive conditional jump intensity model into the GARCH-M model.In addition,we also study whether different risks are priced in the market and the effect of macro new announcements on the risk-return trade-off of stock market.Empirical study shows that the risk premium parameter of GARCH-M model is positive but insignificant while the risk premium parameter of jump integrated GARCH-M model is positive and statistically significant,suggesting that only when the jumps are accounted for can the GARCH-M model detect the risk-return relationship in stock market.Different source of risk is priced in the market level and different macro news announcement have different impact on the risk-return relationship.CPI and PPI have a positive impact on stock market risk-return relationship while GDP,IAV,FI,RS,TB,LOAN and M2 have a negative effect.