Building Cross-Sectional Systematic Strategies By Learning to Rank

TL;DR

This paper introduces a novel framework that applies learning-to-rank algorithms to improve asset ranking in cross-sectional strategies, significantly boosting trading performance and Sharpe Ratios.

Contribution

It presents a new approach integrating learning-to-rank methods into portfolio construction, outperforming traditional heuristics and regression-based ranking techniques.

Findings

Approximately threefold increase in Sharpe Ratios.

Enhanced ranking accuracy through pairwise and listwise learning.

Significant performance improvements in cross-sectional momentum strategies.

Abstract

The success of a cross-sectional systematic strategy depends critically on accurately ranking assets prior to portfolio construction. Contemporary techniques perform this ranking step either with simple heuristics or by sorting outputs from standard regression or classification models, which have been demonstrated to be sub-optimal for ranking in other domains (e.g. information retrieval). To address this deficiency, we propose a framework to enhance cross-sectional portfolios by incorporating learning-to-rank algorithms, which lead to improvements of ranking accuracy by learning pairwise and listwise structures across instruments. Using cross-sectional momentum as a demonstrative case study, we show that the use of modern machine learning ranking algorithms can substantially improve the trading performance of cross-sectional strategies -- providing approximately threefold boosting of Sharpe Ratios compared to traditional approaches.