Generative AI for Stock Selection

TL;DR

This paper demonstrates that generative AI, combined with retrieval-augmented generation, can automate feature discovery in stock prediction, yielding competitive and interpretable signals that enhance forecasting accuracy.

Contribution

It introduces a novel approach using large language models with retrieval-augmented generation for automated feature synthesis in equity markets.

Findings

AI-generated features are consistently competitive with traditional baselines.

Better retrieval knowledge bases significantly improve forecasting outcomes.

Generated signals are weakly correlated with traditional features, enabling effective combination.

Abstract

We study whether generative AI can automate feature discovery in U.S. equities. Using large language models with retrieval-augmented generation and structured/programmatic prompting, we synthesize economically motivated features from analyst, options, and price-volume data. These features are then used as inputs to a tabular machine-learning model to forecast short-horizon returns. Across multiple datasets, AI-generated features are consistently competitive with baselines, with Sharpe improvements ranging from 14% to 91% depending on dataset and configuration. Retrieval quality is pivotal: better knowledge bases materially improve outcomes. The AI-generated signals are weakly correlated with traditional features, supporting combination. Overall, generative AI can meaningfully augment feature discovery when retrieval quality is controlled, producing interpretable signals while reducing manual engineering effort.