Headline Sentiment Lab
Paste a market headline. See how an off-the-shelf sentiment model scores it, and how the tape actually reads it. They disagree more often than you would like.
Try one of these
› Acme Bio announces pricing of upsized $150 million underwritten public offering› Nexon Corp beats Q3 estimates on record revenue; lowers full-year guidance› Orbis Therapeutics receives Complete Response Letter from FDA› Cardin Industries agrees to pay $45 million to resolve SEC charges› Halcyon Mining halted, news pending› Delta Systems to join the S&P MidCap 400 effective prior to openWhy generic sentiment fails on market headlines
General-purpose sentiment models score words. Markets price events. So "announces pricing of upsized $150 million offering" — six positive-sounding words — is one of the most reliably bearish headlines on the tape, because it means dilution. "Agrees to pay $45 million to resolve SEC charges" reads negative and is usually bought, because a capped, known number removes the overhang that kept institutions out.
The engine here matches event phrases first and only falls back to word polarity when nothing matches. That ordering is the whole trick, and it is also why a sentiment tag is only as good as the feed underneath it: scoring a headline you receive ninety seconds late is arithmetic, not an edge.
Try the live feed free at toptiernewswire.com
Top Tier Newswire aggregates 40+ sources into one feed with sentiment tagging and sub-2-second latency. Free tier is delayed; Pro is real-time. Not affiliated with this tool — it is just what I check the fast headlines against.
The 12 headline patterns worth knowing cold
| Pattern | Typical direction |
|---|---|
| "Prices $X million offering" | Bearish |
| "Beats on EPS and revenue" + guidance cut | Bearish |
| "Halted, news pending" (T1 / LUDP) | Direction unknown, volatility certain |
| "Substantial doubt about ability to continue as a going concern" | Strongly bearish |
| FDA milestones: accepts vs approves vs CRL | Depends entirely on which word it is |
| "Exploring strategic alternatives" | Bullish on the print, often bearish six months out |
| "Announces reverse stock split" | Bearish |
| "To join the S&P 500 / Russell rebalance" | Bullish, and mechanically so |
| Activist short report published | Bearish on impact, mean-reverting after |
| "CEO steps down, effective immediately" | Bearish, and the modifier does the work |
| "Authorizes $X billion share repurchase" | Mildly bullish, frequently overrated |
| "Reaches settlement with the SEC" | Usually bullish |
Educational tool, not investment advice. Scores are a rules-based read of language, not a prediction.