Building Cross-Sectional Systematic Strategies By Learning to Rank
This work offers a significant improvement in trading performance for quantitative finance practitioners by enhancing asset ranking accuracy in systematic strategies.
This paper addresses the sub-optimal asset ranking in cross-sectional systematic strategies by proposing a framework that incorporates learning-to-rank algorithms. Applying this to cross-sectional momentum, the authors demonstrate a threefold increase in Sharpe Ratios compared to traditional methods.
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.