What you can validate
News sentiment & NLP
Transformer-scored news as a tradable signal.
News-sentiment strategies turn headlines and filings into signals via NLP scoring (FinBERT). We validate sentiment drift, overreaction reversals, and news-volume spikes with point-in-time timestamps and stale-news filters.
- US equities
20 example backtests
Each sample report is being prepared and will be downloadable here.
- 01
FinBERT headline-sentiment drift on US equities
Sample reportComing soon - 02
News-volume spike momentum study
Sample reportComing soon - 03
Negative-sentiment overreaction reversal
Sample reportComing soon - 04
Earnings-call tone vs subsequent drift
Sample reportComing soon - 05
Press-release sentiment continuation
Sample reportComing soon - 06
Sentiment dispersion as a risk filter
Sample reportComing soon - 07
Sector-level news-sentiment rotation
Sample reportComing soon - 08
Sentiment-confirmed breakout entries
Sample reportComing soon - 09
Macro-news sentiment overlay on index futures
Sample reportComing soon - 10
Same-day sentiment vs next-day return
Sample reportComing soon - 11
Sentiment-weighted small-cap momentum
Sample reportComing soon - 12
News-shock volatility expansion study
Sample reportComing soon - 13
Stale-news fade (already-priced) detector
Sample reportComing soon - 14
Analyst-language sentiment factor
Sample reportComing soon - 15
Sentiment trend (rolling 5-day) signal
Sample reportComing soon - 16
Headline-count anomaly screen
Sample reportComing soon - 17
Sentiment-conditioned mean reversion
Sample reportComing soon - 18
Cross-source sentiment agreement filter
Sample reportComing soon - 19
Sentiment-momentum top-decile basket
Sample reportComing soon - 20
Event-tagged sentiment reaction by category
Sample reportComing soon
Have a news & sentiment idea of your own?
Send us the rules in plain English. We confirm scope and feasibility, then validate it on the same institutional infrastructure.