Research On Risk Spillover Between Stock Market Sectors And Portfolio Risk Measurement Based On Copula Function From Time-frequency Perspective

At present,with the continuous development of the multi-level capital market system,the connection between different sectors of the stock market is becoming closer and closer,the circulation of resources is becoming more and more rapid,and the efficiency of information transmission has also been greatly improved.However,once economic fluctuations occur in a certain market segment and then trigger financial risks,they will soon spread,thus triggering risk spillovers between different segments of the stock market.Therefore,from the time-frequency perspective,this thesis studies the risk spillover effect among the main board,GEM and New Third Board markets in the stock market based on the Copula function,and measures the time-frequency risk of the portfolio based on the main board,GEM and New Third Board markets.Based on the theory of maximum overlap discrete wavelet transform,Copula function,CoVaR and Monte Carlo simulation,this thesis studies the risk spillover effect between stock market sectors and the measurement of portfolio risk.Firstly,this thesis uses MODWT to decompose the original return series of each board at multiple scales,and conducts Granger causality test on the decomposed scale series to determine the direction of risk spillover.Secondly,the dynamic Copula function is used to fit each scale sequence after the probability integral transformation,and the risk spillover intensity indicators such as dynamic CoVaR are calculated to determine the intensity of risk spillover between plates at different scales.Finally,from the time-frequency perspective,the portfolio risk between the stock market sectors is measured based on Copula function.In this study,Monte Carlo simulation is used to measure the portfolio risk at various scales based on static Copula function,and Monte Carlo simulation is combined with rolling window method based on dynamic Copula function.The portfolio risk is dynamically measured at each scale.The results show that:(1)short-period trading is the main factor causing the volatility of each market sector;The spillover direction at different scales is different.The spillover direction between the main board and GEM first shows the one-way spillover from the main board to GEM,and with the extension of the trading period,the one-way spillover direction changes to the two-way spillover,while the spillover direction between the main board and the New Third Board,GEM and the New Third board mainly shows the one-way spillover from the main board to the new Third board and GEM to the New Third board.(2)From the perspective of dynamic correlation coefficient,the mean value of the main board and GEM is the largest and the standard deviation is the smallest,indicating that there is a strong positive correlation between the two boards and the fluctuation is relatively stable,while the main board and the New Third Board,GEM and the New Third board are weak positive correlation and the fluctuation degree is high;From the perspective of risk spillover intensity indicators such as dynamic CoVaR,the spillover intensity from GEM to the main board is larger than that from the main board to GEM at all scales,and the spillover intensity from the main board to the New Third board and GEM to the New Third board are both larger;With the increase of scale,the intensity of risk spillover between plates decreases gradually.The intensity of risk spillover measured by considering multiple markets at the same time based on multivariate Copula is larger than that measured by considering only two markets.(3)With the increase of scale,the portfolio risk among different sectors decreases;The portfolio risk measured by multivariate dynamic Copula model is larger than that measured by binary static Copula model,so it is necessary to measure portfolio risk under the comprehensive effect of multiple markets.This thesis has certain practical guiding significance,which can provide reference for investors to make better choices in investment decision and risk control,and provide empirical basis for regulators to strengthen the joint supervision of stock market sectors.