Do earnings call transcripts predict post-announcement returns?

Molinaro (2026) studies event-driven strategies in equities using a transformer. The abstract mentions an out-of-sample test but not a benchmark, transaction costs, multiple testing or the size of the universe.

“We construct a systematic machine learning pipeline that transforms raw transcript text into quantitative signals and evaluates their out-of-sample forecasting performance on the Russell 1000 universe.”

Also stated: Quarterly, Long-short, Fundamentals, Text and news, out-of-sample test

Missing from abstract: benchmark, costs, multiple testing, universe