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...

Challenges to the empirical implementation of portfolio optimization abound, leading to a recent focus upon the naïve equally weighted portfolio as asset allocation benchmark. In this paper, we extend the performance analysis of the naïve allocation approach to encompass data both within and...

Extreme asset price movements appear to be more pronounced recently and have major consequences for an economy's financial stability and monetary policies. This paper investigates the extreme behaviour of equity market returns and quantifies the probabilities of these losses. Taking fourteen...

The evidence for international diversification as a means to curtail portfolio risk relies predominantly on short-run data. In this paper, we examine the extent to which the risk reduction benefits of international investment hold in the long-run. Employing a multi-horizon non-parametric filter,...

We present new international diversification indexes across equity, sovereign debt, and real estate. The indexes reveal a marked and near ubiquitous decline in diversification potential across asset classes and markets for the post-2000 period. Analysis of panel data suggests that the decline is...

We estimate trends in diversification for equity, debt, and real estate within and across countries. After 2000, we uncover a marked and near ubiquitous decline in diversification, which coincides with sharply higher levels of investment risk. This decline is associated with country economic...

We examine machine learning and factor-based portfolio optimization. We find that factors based on autoencoder neural networks exhibit a weaker relationship with commonly used characteristic-sorted portfolios than popular dimensionality reduction techniques. Machine learning methods also lead to...