Research On The Features Of Continuously Rising And Falling Stock Index Futures Based On Copula-TACD Model

In recent years,with the rapid development of China's securities financial market,stock index futures have become a relatively mature investment tool in the market with its advantages in hedging,price discovery and risk management.How to effectively avoid the risk in the stock index futures investment is a matter of great concern to every investor.This article uses the continuous rising(falling)characteristics of high-frequency data of stock index futures as the research object,and conducts research and analysis from the perspective of time and quantity respectively.This article starts with the one-minute high-frequency rate of return of the monthly continuous index of the Shanghai and Shenzhen 300 stock index futures,and analyzes the continuous rising(falling)minutes and continuous rising(falling)yields to study the characteristics of stock index futures.First,it summarizes the characteristics of weak autocorrelation and non-excessive dispersion,then uses survival analysis method to fit its empirical survival function,probability density function and hazard rate function respectively.Finally,it discusses the probability distribution and conditional probability distribution of the index price from rising to falling and from falling to rising.For the consecutive rising(falling)yields,the characteristics of leptokurtosis,excessive dispersion,strong autocorrelation and nonlinearity are summarized.Then the Copula-TACD model is used for specifying a joint distribution model.And two-stage maximum likelihood estimation method is used to estimate its marginal distribution and dependency structure.Specifically,the TACD model is first introduced to describe the time interval for the occurrence of financial events,and the marginal distributions of the consecutive rising(falling)losses are respectively fitted.Then the Copula function is used to fit the dependency structure.Finally,value-at-risk(VaR)and conditional value-at-risk(CVaR)are calculated based on the marginal distributions and joint distributions respectively.The results show that both VaR and CVaR could effectively measure the risk of falling losses,and that CVaR is more capable of capturing extreme risk than VaR.This article uses a variety of test methods to test the consecutive rising(falling)time and consecutive rising(falling)yields,and summarizes the different characteristics of continuous rising(falling)time and continuous rising(falling)yields of high frequency data,and uses two models to model them respectively.The survival analysis model gives the probability of price fluctuations from a time perspective.The Copula-ACD model provides a distribution of price fluctuations from a quantitative perspective,which can provide certain decision support for high-frequency trading investors.