NLPAI

LLM-Extracted Earnings Sentiment as an Alpha Factor

Rodriguez, M. · Kim, S.

May 2026

Abstract

We evaluate frontier LLMs as extractors of forward-looking sentiment from earnings call transcripts. Our factor, constructed from semantic delta across consecutive calls, achieves a 3-day alpha of 1.8% with Sharpe 1.72 OOS across the S&P 500.

Method

We prompt frontier LLMs to score forward-looking sentiment in earnings call transcripts, then construct a factor from the semantic delta between consecutive quarterly calls for each issuer.

Results

The factor achieves a 3-day alpha of 1.8% with an out-of-sample Sharpe of 1.72 across the S&P 500, robust to transaction cost assumptions up to 15bps.

BibTeX

@article{quanthq2026llm,
  title={LLM-Extracted Earnings Sentiment as an Alpha Factor},
  author={Rodriguez, M. and Kim, S.},
  year={2026}
}