Modeling The Investment Strategy Of Commodity Markets And Empirical Evidences

In recent years,global institutional investors have paid more and m ore attention to the key role of commodities in asset allocation,and they have gradually increased the proportion of commodity futures in their portfolios.commodity futures have become one of the most attractive asset classes in the eyes of institutional asset managers and private investors.From the perspective of hedging and speculation,investment in commodity futures provides an effective way to diversify the risks of portfolios.Quantitative asset allocation methods have become the main analysis methods in mature capital markets.Mean-variance portfolio theory and Black-Litterman model are two popular asset allocation models.The mean-variance model starts with the expected return on assets.It is universally acknowledged that both linear Vector Autor egressive(VAR)models and nonlinear econometric frameworks,such as Markov Regime Switching(MRS)model,can accurately estimate the expected return on assets.Therefore,this paper studies whether simple VAR models can produce commodity portfolios whose performance is superior or similar to that obtained under MRS model,and also discusses the impact of investors’ different risk preferences on commodity portfolio performance under the mean-variance model.The empirical results indicate that,first,VAR mo dels can produce portfolios with better performance than MRS model in the case of a long sample.Second,an increasing risk aversion of investors may significantly reduce the portfolio performance.Among them,investors who prefer profit-making to risk-resisting and who pay same attention to returns and risks are more likely to obtain portfolios with higher returns and lower risks.Finally,WTI is an appropriate investment target for investors who only focus on the maximization of returns,while for investors with other risk preferences,gold proves an ideal investment target.However,subsequent studies have sho wn that the mean variance model has problems in practice such as the sensitivity of the input variables and the high concentration of the optimal solution.The Black-Litterman model is an asset allocation method that can overcome the above disadvantages.In order to explore whether the improved mean-variance model previously mentioned can overcome the shortcomings of the model and generate a better portfolio than Black-Litterman model,this paper studies the portfolio performance under the Black-Litterman model using the GARCH volatility estimation as the subjective views of investors.The empirical results show that adding the subjective views of inve stors can improve the portfolio returns,and Black-Litterman model with GARCH volatility estimation can produce better portfolio than the improved mean-variance model.Theoretically,this paper studies the impact of investors’ different risk attitudes on investment decisions in commodity markets,confirming the Expectancy theory.In addition,scholars have discussed the advantages and di sadvantages of the mean-variance model and the Black-Litterman model in investment practice.The empirical results of this article answer this question.From a practical perspective,the conclusions of this paper can provide investment decision-making reference for commodity investors.For example,investors who prefer profit-making to risk-resisting and those who pay the sam e attention to investment returns and risks are more likely to obtain favorable portfolios with higher returns but lower risks,and if the investors are extremely risk-averse and only want to minimize risk,it is better for them to invest in a single reliable commodity.