Ask whether a strategy actually worked
Type it the way you would say it. We look the answer up in 660,005 backtests we already ran and published — the same method, the same costs, the same held-out sample for every one. No new backtest is run for your question, which is exactly why the answer can be trusted: it existed before you asked.
One of three verdicts, and two of them are negative. BEAT BUY-AND-HOLD means the best variant beat simply holding the asset on data it never saw. DID NOT BEAT BUY-AND-HOLD is the most common answer. INSUFFICIENT EVIDENCE means the sample was too thin to say either way — and that verdict fires on plenty of results that look spectacular, because a 27% edge over three trades is not an edge. Each answer carries how many variants of that indicator beat buy-and-hold on that asset, and the family's own base rate across every asset we test.
Questions people ask
Is this running a new backtest?
No — and that is the point. All 660,005 backtests were run in advance under one method, with the same costs and the same out-of-sample split, and published. This looks the answer up. A result that existed before you asked the question cannot have been shaped by the question.
Why does it sometimes refuse to give me a number?
Because below about 10 out-of-sample trades a Sharpe ratio is noise wearing a decimal point. When the sample is too thin the verdict is INSUFFICIENT_EVIDENCE and the number stays behind it. Roughly one query in five lands there, usually on the combinations that look most exciting.
Why show me every variant instead of the best one?
Because "does RSI work" has no single answer — we test 18 distinct RSI rules and they disagree. Showing you the best one and calling it RSI would be the most flattering presentation of the data rather than the most honest one, so you get the spread and the count that beat buy-and-hold.
Does it cost anything?
No. There is no key, no account and no per-call charge, because a lookup costs us nothing to serve. It is also available to AI assistants as the ie_test_strategy tool on our free MCP server.
Go deeper
Get the weekly edge report
The best-performing indicator per asset, what changed this week, and the honest caveats — straight to your inbox.