Empirical Study On Cross-District And Cross-Industry Asset Allocation Model Based On The Black-Litterman

The traditional mean variance model has strict presuppositions for the expected return and variance.The model assumes that the returns of the securities are subject to the normal distribution.Investors,by analyzing and calculating the historical return distribution,use the expected return to measure the future real return Uncertainty about the future rate of return measured by the variance of the historical rate of return.However,the traditional mean variance model has many deficiencies in the actual investment portfolio.Many scholars have proposed some new models based on the mean variance model.Fisher Black and Robert Litterman(1992)based on Markowitz theory,combined with Capital Asset Pricing Model(CAPM),Sharp inverse optimization theory and Bayesian theory to establish the Black-Litterman model.The core of this model is to use Bayes theorem Investors view the income and the balanced income combined to derive the BL model yield distribution,and finally use Markowitz optimization method to solve the return on the BL model based on the distribution of asset allocation weights.The BL model relaxes the strict qualification of the traditional Markowitz model and also adds the investor viewpoint to the investment theory model,opening up a new road for the application of investment theory in practice.In this paper,we use the GJR-GARCH-M model to fit the indices of the Shanghai-Shenzhen 300 Index and the six regions and 10 industry sub-indices included in the index,and predict the next period's returns and volatilities.Fluctuation,respectively,as the BL model for investors on the various indices of view of the rate of return and point of view earnings error.Build a BL-based asset allocation model,and on this basis,cross-regional and cross-industry asset allocation,and then use the BL model for asset allocation results for detailed analysis.In order to verify the robustness of the asset allocation method based on the BL model and to determine whether the asset allocation result is affected by the asset allocation weight adjustment cycle or the data calculation cycle,this paper tests the robustness of the asset allocation method based on the BL model.The results show that the BL model based on BL model with different scalars can be used to allocate assets across regions and industries,calculate the cumulative returns,volatility and Sharpe ratio under different BL models.The results show that compared with Shanghai and Shenzhen,300 index and mean variance model have great advantages.For the selection of scalar value in BL model,the optimal value can not be determined through the comparative analysis of the results.The selection of optimal scalar value is different under different background conditions.Robustness test results show that the BL model has a strong advantage in asset allocation,and the advantage will be greater in the medium and long term periods.