Every match is explained and traceable to source.
| Company | Analogues did | Market did | Edge vs market | Since you added |
|---|---|---|---|---|
| MSFTMicrosoft | +7.0% | −2.6% | +9.6% | +25.7% |
| AAPLApple | +8.0% | +2.1% | +5.9% | +22.8% |
| JPMJPMorgan Chase | +7.3% | +0.5% | +6.8% | +19.3% |
| NVDANVIDIA | +4.8% | +0.0% | +4.8% | +19.2% |
| LLYEli Lilly | +4.9% | +1.8% | +3.1% | +14.0% |
| COSTCostco | +3.7% | −0.8% | +4.4% | −6.3% |
4,515 US equities · SEC EDGAR filings, point-in-time · no look-ahead, enforced in tests · validated across 922 historical periods
Research tool · not investment advice · anchor dates from 2019
Find today's opportunities that resemble historical winners.
Nvidia, 2016 — which US stocks look like that today? Matched on fundamentals and price behaviour, not sector.
Run a searchStress-test your investment thesis against history.
Microsoft, today — when has the market looked like this before, and what followed over the next six months?
Test a thesisStress-test your whole portfolio, every trading night.
Microsoft, Apple, everything you hold — that same question, answered again every night while you sleep.
Open your watchlistWe ran 922 historical searches on periods where the reference company later gained 25% or more — chosen by what it did, never by what the matches did. Random picks over the same dates returned −0.5%. The matches beat them, and the margin grew with the score.
| Match score | Candidates | Median 6-month return | vs. randomrandom: −0.5% |
|---|---|---|---|
| 90%+ | 942 | +2.3% | +2.9 pts |
| 85–89% | 16,951 | +1.2% | +1.7 pts |
| 80–84% | 79,918 | +0.7% | +1.2 pts |
| 75–79% | 183,293 | −0.2% | +0.4 pts |
| below 75% | 3,616,694 | −0.4% | +0.2 pts |
Same engine, three ways in:
Any US public company — then, depending on the question:
Fourteen fundamental and price metrics, each ranked against the companies of that same era — 4,515 candidates at one date, or every month back to 2016.
Every match opens metric by metric, down to the filed figure and the filing date it came from.
Everyone else gives you a score. We show you the evidence: which metrics drove it, where the candidate diverges from the anchor, and a confidence grade reflecting how complete the underlying filing data actually is.
Three real Discovery searches — a company at a moment in its past, and the stocks that most resemble it today. Each card shows that search’s top four matches, unedited and with losses included. The sector column is the point: these are companies a sector screen would never put next to each other.
Built and backtested for 6–12 month pattern research. When a company’s filing data is incomplete, we show it — a low grade or an N/A, never a smoothed-over guess.
This is also an early MVP, not a finished product: anchor dates currently run from 2019 across 4,515 US-listed stocks. The limit isn’t price data — we hold far more — it’s how far back enough companies have resolvable fundamentals to rank a match against. We’re gauging interest before building out deeper history, more markets, and more depth.