Optimal Beats Naive Diversification: Asset Allocation Using High-Frequency Data

Abstract

This article evaluates the usefulness of high-frequency data in optimal portfolio choice. The authors use a comprehensive list of major stock indexes and different frequencies of observations. Furthermore, they consider the impact of economic cycles, microstructure noise, and seasonality on performance. Their results show the ability of high-frequency data–based strategies to beat both monthly and daily based strategies and the benchmark equally weighted portfolio, even in presence of transaction costs. The authors also find that the outperformance arises from the reduction in the estimation error of the covariance matrix, which offsets the increase in transaction costs.