What Makes An Asset Useful?

TL;DR

This paper defines criteria for assessing the usefulness of a new asset in investment portfolios, focusing on diversification, predictability, tail risk mitigation, and passive investment suitability, with scalable tests for each.

Contribution

It introduces a formal framework and scalable algorithms to evaluate asset usefulness based on four key features relative to a reference universe.

Findings

Proposes quantitative measures for incremental diversification and predictability.

Develops scalable algorithms to test asset usefulness criteria.

Provides a practical framework for asset evaluation in investment management.

Abstract

Given a new candidate asset represented as a time series of returns, how should a quantitative investment manager be thinking about assessing its usefulness? This is a key qualitative question inherent to the investment process which we aim to make precise. We argue that the usefulness of an asset can only be determined relative to a reference universe of assets and/or benchmarks the investment manager already has access to or would like to diversify away from, for instance, standard risk factors, common trading styles and other assets. We identify four features that the time series of returns of an asset should exhibit for the asset to be useful to an investment manager, two primary and two secondary. As primary criteria, we propose that the new asset should provide sufficient incremental diversification to the reference universe of assets/benchmarks, and its returns time series should be sufficiently predictable. As secondary criteria, we propose that the new asset should mitigate tail risk, and the new asset should be suitable for passive investment (e.g. buy-and-hold or short-and-hold). We discuss how to quantify incremental diversification, returns predictability, impact on tail risk, and suitability for passive investment, and for each criterion, we provide a scalable algorithmic test of usefulness.