Covariance matrix filtering and portfolio optimisation: the Average Oracle vs Non-Linear Shrinkage and all the variants of DCC-NLS

TL;DR

The paper demonstrates that the simple Average Oracle covariance filtering method outperforms complex DCC+NLS variants in portfolio optimization, achieving higher Sharpe ratios due to lower leverage and turnover.

Contribution

It provides empirical evidence that the Average Oracle surpasses state-of-the-art covariance filtering methods in large-scale experiments, with analytical insights into its advantages.

Findings

Average Oracle yields higher Sharpe ratios than DCC+NLS variants.

DCC+NLS variants have higher leverage and turnover, reducing returns.

Simple covariance filtering can outperform complex models in dynamic markets.

Abstract

The Average Oracle, a simple and very fast covariance filtering method, is shown to yield superior Sharpe ratios than the current state-of-the-art (and complex) methods, Dynamic Conditional Covariance coupled to Non-Linear Shrinkage (DCC+NLS). We pit all the known variants of DCC+NLS (quadratic shrinkage, gross-leverage or turnover limitations, and factor-augmented NLS) against the Average Oracle in large-scale randomized experiments. We find generically that while some variants of DCC+NLS sometimes yield the lowest average realized volatility, albeit with a small improvement, their excessive gross leverage and investment concentration, and their 10-time larger turnover contribute to smaller average portfolio returns, which mechanically result in smaller realized Sharpe ratios than the Average Oracle. We also provide simple analytical arguments about the origin of the advantage of the Average Oracle over NLS in a changing world.