Research On The Dynamic Correlation Between China Stock Market And Bond Market And Influencing Factors Based On MS-VAR And DCC-GARCH Models

The correlation between the stock market and the bond market will affect the profitability of investors,market integration,etc.It is an important topic in the capital market.In recent years,there have been frequent occurrences of domestic and foreign economic policy uncertain events.Such economic policy uncertainty will affect investor behavior and even the phenomenon of "flight to quality",which in turn affects the relationship between the stock market and the bond market.At the same time,the existing literature seldom involves the research on the influence of economic policy uncertainty,interest rate,exchange rate and other factors on the dynamic relationship between the stock market and the bond market.Based on this,this essay chooses the CSI 300 Index and Small and Medium Enterprise Board Composite Index from June 1,2007 to December 21,2020 to represent the stock market,and China Bond Composite Wealth Index to refer the bond market.The detailed research approaches and ideas are as follows.This article firstly analyzes the dynamic correlation between the stock and bond markets by constructing the VAR model and the DCC-GARCH model of the stock and bond markets.Subsequently,this paper studies the impact of economic policy uncertainty,interest rates,exchange rates,and inflation rates on the dynamic correlation of stock and bond markets.Since the test found that their impact on the correlation of stocks and bonds is non-linear,this paper chooses the MS-VAR model that can distinguish different economic states for empirical research,and finally determines it as the MSIH(2)-VAR(2)form.Next,this paper adopts the impulse response function method to test the influence of these external factors on the dynamic correlation of stock bond yields on the basis of the above-mentioned partition system.Further,consider the effects of four segmented China's economic policy uncertainties on the dynamic relevance of stock and bond separately.This essay selects the dynamic correlation coefficients of the Shanghai and Shenzhen 300 Index and the China Bond Comprehensive Wealth Index,and the monthly data of the four sub-indices subdivided by economic policy uncertainty,and constructs their respective MS-VAR models to analyze two different stocks and bonds.Under the state of relevance,the effect of each specific economic policy uncertainty on the relevance of changes in the two markets over time.Finally,this paper finally applies practical strategies to the monthly dynamic correlation coefficients of the stock and bond markets obtained by the empirical study to reduce portfolio risk,which proves the practical significance of studying the correlation between stocks and bonds.This paper uses the method of nonlinear programming to adjust the correlation between stock and bond yields to adjust the weight of each asset in the portfolio to achieve risk hedging.Through VAR model,DCC-GARCH model and MS-VAR model to research and parse the dynamic correlation between them,this paper draws the following conclusions.First,on the overall level,the correlation between China's stock market and bond market is relatively weak,rendering negative correlation in most periods,and the bond market is more sensitive to the impact of new information;Second,most of the samples are in the state of low correlation of stocks and bonds.The high correlation stage started from 2016-2017 approximately,which is represented in the conversion of the two zoning states and the correlation reflected by the smooth curve has undergone a structural change.Third,the probability of the degree of correlation between the two markets index returns from a high-correlation state to a lowcorrelation state is extremely small,while the probability of a transition from a lowcorrelation state to a high-correlation state is 0.9896.This paper uses the MS-VAR model to analyze the driving factors of the dynamic correlation of stocks and bonds and the differential effect of the segmented economic policy uncertainty,and draws two conclusions.First,the correlation between equity and debt is in high correlation and low correlation states,The inflation rate is negatively correlated with the correlation between stocks and bonds;interest rates are positively correlated with the correlation between stocks and bonds,that is,an increase in interest rates will lead to an increase in the correlation of stocks and bonds.Under the two correlation states,the influence of exchange rate on the correlation between the two markets is very weak.The increase in economic policy uncertainty has led to a downward trend in the relevance of stocks and bonds,The decline in economic policy uncertainty will have two effects with different directions in the relationship between the two.Second,in a state of high and low correlations between stocks and bonds,the increase in monetary policy uncertainty will increase the correlation between the stock and bond markets in the short term,and the degree of enhancement will be greater under high correlation.The increase in the uncertainty of fiscal policy,exchange rate and capital account policy will weaken the correlation of stocks and bonds in the short term,and the degree of weakening will be greater under the state of low correlation.Increased trade policy uncertainty will increase the correlation of stocks and bonds in the short term,and the degree of enhancement will be even greater in the state of low correlation.Finally,this article uses strategy construction and back testing,it is found that there are obvious differences in the proportion of each asset allocation in the optimal solution of the investment portfolio in different economic cycle stages.Using the dynamic correlation coefficient of stock and bond to adjust the proportion of each asset allocation can reduce the portfolio variance during the economic overheating and stagflation period.The analysis and research of this paper have the following three innovations.(1)In terms of theory,the article inducts an investigation of the effect of economic policy uncertainty on the relevance of stocks and bonds,and designs the use of the monthly dynamics of the stock market and the bond market.The coefficient adjusts the strategy of asset allocation in hybrid funds.(2)In terms of research methods,the MS-VAR model is introduced to divide the correlation of stocks and bonds into two states: high correlation and low correlation,and investigate the effect of economic policy uncertainty index,interest rate,etc.,and four segmented economic policy uncertainties on the relevance of stocks and bonds under the two states.(3)In terms of practical significance,the study found that the increase in economic policy uncertainty will reduce the relevance of equity and debt.The uncertainties of the four sub-economic policies have different impacts on the relevance of equity and debt,which is of great significance to the formulation of economic policies.