premise

Every year, at least one stock 10x's.

The job is to recognize the setup before the run. This is a working library of every documented 3x, 5x, and 10x move inside a ~12-month window in US equities since 2016 — with the catalysts, dates, technicals, and patterns that preceded each one.

Density per year

how many 3x / 5x / 10x'ers each year produced

why this exists

Pattern, not prediction

We're not picking 2026's winner. We're building the muscle to see the setup forming — tiny float + narrative, binary catalyst, short squeeze, sector pivot — so when one shows up, we recognize it on day three instead of day thirty.

the gradient

3x / 5x / 10x

The 10x'ers are rare and famous. The 5x'ers are denser and equally instructive. The 3x'ers are common enough to calibrate "false positive" base rates. Studying all three tells you what tips a setup from "ran" to "ran hard."

what's in each entry

Date, catalyst, float, setup

For every stock: the low date, the peak date, the headline that lit the fuse, the float and short interest entering the run, and the chart pattern. So you can backtest your own pattern-recognition.

All stocks

Year	Ticker	Company	Low → Peak	Multiple	Window	Catalyst	Patterns	Sector

Method

How to use this site

Start in Years for a high-level view of density per year (some years produce 7+ 10x'ers, others produce 0). Click any year card to drill into its stocks. Use All Stocks to filter by tier (3x/5x/10x), pattern, sector, or year. Click any row to see the timeline, catalysts, and float/short-interest data behind the move. Patterns shows what 3x, 5x, and 10x'ers have in common — and what distinguishes each tier. 2026 Candidates is the forward-looking watchlist with specific tickers.

Threshold

Strict multiples (3x / 5x / 10x) on a ~12-month low-to-peak basis. Intraday spikes are counted but flagged in the entry's notes — a 10x print you couldn't realistically have exited at is a different thing than a sustained run.

Inclusion

Documented, sourced names from US-listed equities (NYSE/Nasdaq, including ADRs). Micro-cap pump-and-dumps with no press coverage are excluded — not because they don't happen, but because there's no signal we could have caught.

Survivorship

This list is the winners. To make pattern-matching honest, we also need the stocks that looked like the setup but didn't run. That's a planned future addition (see "false positives" in the patterns section).

Confidence labels

high — multiple sources confirm. med — single source or imprecise price points. low — screener data that may be capped/placeholder. Verify before acting on.

Data approximation

Low/peak prices are rounded to the cleanest documented values from public reporting. Multiples are calculated trough-to-peak. Intraday spikes (TLRY, HKD, KOSS) are counted but flagged — these are different from sustained year-long runs (AXSM, OKLO). For position-sizing decisions, distinguish "10x print on the tape" from "10x you could realistically have exited at."

Backstory: pgintel

10X is a side library of pgintel.dev. The thesis: pattern recognition beats prediction. We're not picking next year's winner — we're building the muscle to see the setup forming so when one shows up, we recognize it on day three instead of day thirty.