Most quant tools tell you what to buy.
We tested whether ours could — and published the answer.
Our own leak-free, survivor-free probes show factor similarity does not forecast returns. What it does forecast is risk — a stock's factor analogues cluster its forward volatility (IC +0.075, t +10.2). Type a ticker: see its factor state and its closest historical twins — the setups the risk signal is built on. See the proof →
factor state
as ofRisk twins — the factor analogues the risk signal reads
Historical setups whose factor profile is closest to this one (cosine similarity). Our research shows analogues like these cluster forward realised volatility — the one signal that survived leak-free, survivor-free testing. They do not predict returns, and we’d rather show you the proof than pretend otherwise.
| Ticker | Date | Similarity | RSI | Momentum | 20d return | Composite |
|---|---|---|---|---|---|---|
No precomputed look-alikes for this ticker.
Want the whole dataset?
A free account adds all 28 factors, 252 days of point-in-time history, four similarity methods, forward-return labels, the REST API and the MCP server.
Create a free account250 calls/day, no card. See the research note for what the data does and doesn't do.