Asset allocation with correlation: A composite trade-off

Abstract

We assess the ability of minimum-variance portfolio allocation strategies accounting for time-varying correlation between assets to provide performance benefits relative to an equally-weighted portfolio. Prior to transaction costs correlation-based strategies emphatically outperform the equally-weighted benchmark. This finding is strongest for short horizon correlation forecasts and attributed to dynamic correlation as opposed to variance forecasts. Thus, estimation error is not found to be the primary obstacle to successful portfolio optimization. Rather, frequent rebalancing and associated transaction costs pose a significant challenge. Limiting portfolio turnover through short-selling restrictions and greater rebalancing error tolerance results in regular outperformance of the correlation based strategies even for large transaction costs. Taken together, these findings provide evidence of a trade-off between optimal portfolio performance, forecasting horizon, rebalancing frequency and transaction costs.