Cross-Sectional Asset Retrieval via Future-Aligned Soft Contrastive Learning

TL;DR

This paper introduces FASCL, a novel representation learning framework that aligns asset retrieval with future return correlations, improving the selection of assets likely to exhibit similar future behaviors in quantitative finance.

Contribution

The paper proposes a new future-aligned soft contrastive learning method for asset retrieval, directly optimizing for future return correlation rather than historical similarity.

Findings

FASCL outperforms 13 baselines across all future-behavior metrics.

The evaluation protocol effectively measures future trajectory similarity.

Experiments on 4,229 US equities validate the method's effectiveness.

Abstract

Asset retrieval--finding similar assets in a financial universe--is central to quantitative investment decision-making. Existing approaches define similarity through historical price patterns or sector classifications, but such backward-looking criteria provide no guarantee about future behavior. We argue that effective asset retrieval should be future-aligned: the retrieved assets should be those most likely to exhibit correlated future returns. To this end, we propose Future-Aligned Soft Contrastive Learning (FASCL), a representation learning framework whose soft contrastive loss uses pairwise future return correlations as continuous supervision targets. We further introduce an evaluation protocol designed to directly assess whether retrieved assets share similar future trajectories. Experiments on 4,229 US equities demonstrate that FASCL consistently outperforms 13 baselines across all future-behavior metrics. The source code will be available soon.