A Comparative Study Of Common Methods In Portfolio

More and more people pay attention to the depreciation of RMB,inflation,interest rate reduction and the existence of various uncertain factors.The study of optimal portfolios is also an important issue which has been widely concerned by scholars.As the market continues to fluctuate,the majority of investors begin to treat investment rationally.How to get profits in the continuous financial market is a very valuable and meaningful topic.More and more computers,statistics and optimization theories have been applied to the determination of optimal portfolios.The analysis and research of financial market has gradually entered into the more data oriented benign development of market,data,theory,model and analysis.If we can get some information from the data and apply these useful information to the establishment of the model,it is bound to guide the investors better,so as to better guide the development of the financial market.With the fluctuation of financial market,we think the covariance matrix is not constant.Therefore,on the basis of the minimum variance model,we replace the covariance matrix into a time-varying dynamic covariance matrix,and propose a minimum variance portfolio model(TVMV)for the time-varying covariance matrix.This paper uses the Fama-French database enterprise data to compare the mean variance model,the global minimization variance model,the minimized variance model,the NAIVE,the CVaR model,the COPCVaR model and the TVMV model,and compare the five indexes,such as the out of sample income,the outside sample variance,the sample out of sample SHARP ratio,the turnover rate and the turnover rate after adjusting the SHARP ratio.Its advantages and disadvantages to the choice of portfolio.The following conclusions are drawn: the average variance obtained by the mean variance model is the largest,the external variance of the sample obtained by CVaR is the smallest,and the SHARP ratio is the largest in CVaR,and the turnover rate of the NAIVE is the smallest.The turnover rate of the minimum variance investment combination model of the time-varying covariance matrix is adjusted after the turnover of the sample.Poubi is the largest.Although the minimum variance portfolio model of the time variant covariance matrix is in the five index,the best performance of the SHARP ratio is achieved only after the turnover rate is adjusted.But considering the consideration of the five index,we think the minimum variance portfolio model of the time variant covariance matrix is the most stable.