Empirical Study On Systematic Risk In Stock Market

The beta coefficient is an important indicator for measuring systemic risk and plays an important role in risk management and investment analysis.The traditional CAPM capital asset pricing model uses the variance of the yield to measure the risk when studying the beta coefficient,which is defined as the future uncertainty,that is,there may be positive income or negative income,and the rational investors hold the same attitude for both results.In the actual investment process,investors tend to be more averse to downside risk.At the same time,China's stock market has also been in a downturn recently.Therefore,the article focuses on the systematic risk in the downside risk of the stock market,that is,the estimation of the downside beta coefficient.There are many ways to measure downside risk.The semi-variance can satisfy the investors' real thoughts about risk.The value at risk(Va R,CVa R)can directly obtain the corresponding maximum loss under different probabilities.The extreme value theory(EVT)can estimate extreme tail risk by describing the distribution of the tail.The above methods focus on the overall distribution of downside risk,while ignoring specific systemic risk factors.In this article,the traditional beta coefficient is combined with the theory of semi-variance and value at risk.60 stocks are randomly selected from the Shanghai and Shenzhen 300 Index as samples,and the corresponding beta coefficients of different methods are calculated.Among them,when estimating the downside beta coefficient using Va R,tail index is described by the important Hill estimator in the POT model,and the systemic risk corresponding to different downside risk levels is calculated by relaxing the selectionrange of the k.The empirical results show that the downside beta coefficient of most stocks is higher than the traditional beta coefficient.The downside beta coefficient obtained by the semi-variance model changes with the value of the target rate of return.When the target rate of return is the value of the previous period,the difference between the result and the value of the downside beta coefficient obtained based on the Va R-Hill model(parameter k/n=10%)is small.When the k/n parameter value keeps shrinking,a small number of stocks will have a downside beta coefficient of 0 due to the characteristics of the stock itself or the small amount of data at the end of sample.Some stocks will have a large increase in the value of the downside beta coefficient due to the high loss rate in extreme cases.That is,when the parameter value in the Va R-Hill model is too small,the result of the estimation of the downside beta coefficient will be distorted and cannot reflect systemic risks under normal downside risks.Compared with the traditional beta coefficient,the semi-variance model and the Va R-Hill model can select parameters according to the real risk condition of the stock market,such as target rate of return,risk probability and threshold,and analyze the linkage between individual stocks and the overall market,determine the corresponding systemic risk.In addition,in the robustness test of beta coefficient,the transfer matrix results show that the downside beta coefficient maintains stability in the short-term,and when the lag period is more than one year or longer,the downside beta coefficient of some stocks will have more fluctuations.Therefore,downside beta coefficient can be estimated by using historical data and the short-term systemic risk of individual stocks can be predicted in the future.