KEY FINDINGSn Contemporary approaches (e.g., simple heuristics) used to score and rank assets in portfolio construction are sub optimal as they do not learn the broader pairwise and listwise relationships across instruments.n Learning to rank algorithms can be used to address this shortcoming, learning the broader links across assets, which consequently allow them to be ranked more accurately.n Using Cross-sectional Momentum as a demonstrative use-case, we show that more precise rankings produce long/short portfolios that significantly outperform traditional approaches across various financial and ranking-based measures.

No abstract

No abstract

Globally learned learning-to-rank (LTR) algorithms at the core of cross-sectional strategies ignore differences between the distribution of asset features over portfolio rebalances. This flaw produces inaccurate asset rankings that can happen over risk-off episodes and cause unwanted drawdowns.n The authors tackle this shortcoming using the idea of the local ranking context from information retrieval: that a query's top retrieved documents provide vital information about the query's own characteristics, which can then be used to refine the initial ranked list.n Using cross-sectional currency momentum as a case study, the authors encode the ranking context of instruments with a transformer-based architecture and demonstrate the superiority of strategies using context-aware LTR on various performance measures over both normal and risk-off market states.