Transcript Analyst Question Uniqueness
In plain terms
Allee, Do & Do (2024): textual uniqueness of analyst Q&A questions (dissimilarity vs other analysts on the same call, the same analyst's prior calls, and the prepared remarks) reflects genuine private information-gathering, and the answers
How it works
Allee, Do & Do (2024): textual uniqueness of analyst Q&A questions (dissimilarity vs other analysts on the same call, the same analyst's prior calls, and the prepared remarks) reflects genuine private information-gathering, and the answers are more value-relevant. Compute a per-call uniqueness score (Jaccard/embedding distance vs same-call peers and prepared script), sort cross-sectionally: high-u
Data dependencies
- Earnings call transcripts (Q&A speaker-segmented)
Full earnings-call transcripts (prepared + Q&A), tokenised.
- Transcript finbert scores
A data feed this strategy reads, refreshed on its normal schedule.
- Tnic peers (optional)
Hoberg-Phillips text-based industry classification peer lists (annual).
Expected edge
- Tested over
- Modern daily equity data
info premium; decay risk 2/5.
Example tickers where this is likely to fire
Illustrative only, the signal fires based on the live data, not a fixed list.
Related families
Explore Transcript Analyst Question Uniqueness on alphactor.ai
See which tickers this family is currently firing on, with live signals and rankings.