Loading the page…
Loading the page…
Before an FDA decision, a drug trial result or an earnings report, some traders pile into options that only pay off if the stock jumps, or only if it drops.
Foreshadow rates how unusual that betting was, from 0 to 100, using trade data from Unusual Whales. Then it checks what the stock actually did.
KOD Endpoint Met: KSI-501, Sep 28, 2026
Watch the bets build up, day by day.
FDA decisions, drug trial results and company earnings reports, from the Unusual Whales calendars. Each one has a date on the calendar, so there is a clear “before” to look at.
In the five trading days before the news: how much money went into options bets, how lopsided it was, and how that compares with the same stock's normal 60 days.
After the news, did the stock move more than options traders expected, and in the direction the bets pointed? If yes, the bets were right.
The score is plain arithmetic you can check. Try it with the sliders.
Normal period
60 trading days
The 5 days before
what we measure
The news
then we check the move
In the last 90 days alone, Foreshadow checked 640 past events, and the betting was unusual before 37 of them. Whether unusual betting tends to be right, compared with ordinary betting, is just below, over all 1,657 finished cases.
Foreshadow grades itself in public. Across 1,657 finished cases, here is how often unusual betting turned out right compared with ordinary betting.
FDA
Before FDA news, when the betting leaned one way, unusual betting was right 26% of the time (6 of 23), about 3.5 times as often as ordinary betting (7%). The edge is in the size of the move: the stock moved more than options expected 27% of the time, against 14%. Which way it moved was close to a coin flip (61% the way the bets leaned).
Earnings
Before earnings, unusual betting hasn't clearly beaten ordinary betting so far: 15% right against 14%, counting cases where the bets leaned.
Foreshadow is also an MCP server, so Claude can use it as a tool. Add it with one line, then ask in your own words. Claude looks up the score, the biggest bets, similar past cases and the scorecard, and answers in plain English.
In Claude on the web or Desktop, add it under Settings → Connectors. It also works in Cursor. Setup and a live tool tester
Add it to Claude Code
claude mcp add --transport http foreshadow https://foreshadow-six.vercel.app/api/mcpThen ask, for example
Or pick one of its ready-made prompts: investigate_case, week_ahead, stock_track_record.
Big bets before news can come from good research, hedging or luck. Foreshadow shows what traded, never who placed it or why.
The same inputs always give the same number, and every trade behind it is listed on the case page. Claude writes the short summaries and never sets a score.
A high score says something unusual happened before the news. It says nothing about what a stock will do next.
Want the details? Read how the score is calculated.
Reading the small chart: each bar is a day of large options bets. Bets on a rise (calls) go up, bets on a fall (puts) go down, and the yellow part is the five days just before the news. “Expected” is the move options traders were pricing in. “Actual” is what the stock did.