A Study Of Stock And Bond Portfolio Strategy Based On Copula Model

Financial markets fluctuate fiercely,from 2008 to now China has experienced two relatively large economic turmoil,but also it reflects the securities market are full of systemic risk and lack of monit oring indicators.For all Participants,how to avoid risks in such a market environment become more and more important,for investors,rate of return is the most concerned problem,in China's stock market,market segmentation is more obvious,but at the same time the two markets are two pillars of the capital market,The stock market and bond market attract a lot of investors because of their huge capacity and different risk,so the study of the relationship between the two markets is important,and from the perspective of the portfolio,it is significant for investors to make recommendations.With the deepening research of the correlation,we can find that the relationship of the two markets is not a simple Granger causality or linear correlation,their rate distributions have the character of "non-normality","asymmetry","spikes" and "volatility",Copula function is prefer because of it is not required for the distribution of the rate,it can be directly to model the edge distribution and also have other advantages,so it plays a significant role in the research of the time series correlation.In the aspect of theory,five constant Copula functions,a mixed Copula function,three vine Copula functions and three dynamic Copula functions are described.and introduces the theory of risk measure.In the aspect of empirical,this paper chooses the Shanghai Composite Index,Shenzhen Stock Index and CSI Bond Index as the research object.The Shanghai Composite Index and Shenzhen Stock Index represent trend and fluctuation of the stock market.The CSI index is an important index to measure the trend and fluctuation of the bond market.Due to the impact of the Hysteresis effect and to ensure the continuity of the data,this paper does not delete the data non-synchronous,First of all,the paper describes the development and fluctuation of the two markets in 2008,then GARCH(0,4)-t model GARCH(3,0)-t model are used to describe the edge distribution of the Shanghai Composite Index and the CSI bond index,and the GARCH(1,1)-t model to analyze the marginal distribution of the Shenzhen Stock Index.In the study of the properties of dynamic models,We choose the Shanghai Composite index and Shenzhen stock index to construct three variable model and a static mixing Copula model were compared.analysis to choose the best fitting effect model,finally to realize the combination of Copula and value at risk,according to the historical index,randomly generated by monte carlo method to obey GARCH-t marginal distribution and copula connect joint distribution of the random sequence,calculating the value of such as proportion of Va R and CVa R,through the test of failure rate method to examine the effects of fitting,and the portfolio with the 10000 times of simulation optimization,choose the optimal portfolio.It is found that in the constant Copula,the Copula function of the Archimedes Copies is more suitable for fitting the "asymmetry" and "spike tail" characteristics of the yield series than in the Copula function of the ellipsoid.The Copula function fits the effect better than any single static model.After the quasi-static Copula model and the Time-Varying SJC Copula model,it is found that the Time-Varying SJC Copula model can measure the ever-changing dependency between time series.Based on the Copula model,the optimal ratio of stock and bond market is studied,and the Copula-mean-Variance model and mean-CVa R model are compared to verify the superiority of CVa R in measuring risk.At the same time,we can see that as the expected yields rise,the optimal portfolio will shift towards the stock market,and the proportion of stocks with thick tails will increase.