Aggregating Information for Optimal Portfolio Weights

I attempt to address an important issue of the portfolio allocation literature -- none of the allocation rules developed in prior literature seems to consistently deliver good performance across different asset samples. For this purpose, I develop an approach that aggregates information from multiple sources for optimal portfolio weights. In particular, my approach uses the weights implied by extant allocation rules as instruments and decides the relative contribution from each rule through Elastic Net, a machine learning technique. Out-of-sample tests suggest that, by aggregating information from twelve allocation rules, my approach consistently achieves good performance across a variety of asset samples whereas none of the twelve rules can match the consistency. My paper also emphasizes the relevancy of the mean-variance framework and the allocation rules developed in prior studies. Even though these rules might not deliver satisfactory performance individually, their weights still contain valuable information and serve well as instruments.