基于ARMA-GARCH-t和Black-Litterman模型的资产投资组合研究

Research on Asset Portfolio Based on ARMA-GARCH-t and Black-Litterman Model

本文选取上海证券交易所不同行业3只市值热度较高的股票运用ARMA-GARCH-t模型对其日收益率与波动性进行预测,在推广的Black-Litterman(BL)模型的框架下,将投资者主观收益分布与资产的先验均衡分布相结合,计算资产的配置权重,并与传统的马科维兹均值-方差(MV)模型给出的组合权重进行对比,发现投资者对资产收益的信心水平越高,BL模型在投资组合中赋予相应资产的权重越高,且投资组合的收益也得到提高。最后,通过融资融券分析验证3只股票市场配置的有效性,说明BL模型给出的配置更符合投资者预期。该研究可以为资产管理者和投资者在资产配置方面提供更多借鉴。

In this paper,the ARMA-GARCH-t model is employed to forecast daily return and volatility of the three stocks with a high market value in different industries of Shanghai Stock Exchange.Under the framework of extended Black-Litterman(BL)model,investor’s subjective return distributions are combined with asset’s priori balanced distribution.Then the weights of the asset allocation is calculated and compared with the ones from traditional Markowitz mean-variance(MV)investment portfolio.The empirical results show that the BL model puts more weight to the asset with high investor’s confidence level and the return of portfolio is larger than the MV case.Finally,the analysis results of the margin trading are applied to verify the validity of the market allocation of the three stocks,which confirms the conclusion that the BL model is more in line with investor expectations.The study can provide more reference for asset managers and investors in asset allocation.