Test a strategy, written as one sentence
Load your own prices or practise on made-up ones, describe a rule in one line, and get the equity curve, the placebo sweep, the noise floor and the verdict, with the date it stopped working. Free, nothing to download.
The other four tools score numbers you bring. This one produces them: pick a market, build a rule from the menus below, and press Test it. It runs the rule, the fair benchmark, the placebo sweep, the settings next to yours and the noise floor, then says what happened and when. Nothing to download, nothing to install.
Watch: what goes in
What it tells you
The same seven conditions the ledger applies to every approach, applied to yours; the seventh, forward data, is one no backtest can meet. With the date it stopped working, if it did. Not whether the rule made money, since a third of arbitrary date windows are profitable, but whether it beat simply holding less, whether the settings next to yours agree, whether random entries with the same count and holding times do as well, and whether doubling the costs ends it.
The best this page can ever say is candidate. It is a backtest, and only forward data can promote one.
Made-up practice prices, so you can learn the tool before finding a file. Each behaves roughly like its kind of market, bear runs included, but none is a real market, and a result on one says nothing about any real one. To test a real market, pick My own file.
A CSV of daily prices whose first row names a date and a close column; open, high and low are used when present. Most charting sites and brokers export one, and stooq.com has a free download for nearly every market. Nothing you load leaves your browser.
- Sharpe
- –
- A year
- –
- Worst fall
- –
- Trades
- –
| Condition | Result | Measured |
|---|
How it is run: a signal on the close enters at the next day’s open; a stop that gaps fills at the open; a bar that touches both the stop and the target counts the stop; costs are charged per side. Buy-and-hold and the blend use the same closes as the rule, so dividends are in or out for all three alike. The twenty placebo runs are seeded, so the same prices and rule give the same twenty every time.
Where this number comes from
The free first module works through the arithmetic behind this calculator: where the bar comes from, why looking twice raises it, and what to do when your result sits below it. Subscribe for one email every two weeks about what I tested.
Your address is used for this and nothing else. Every email has a one-click unsubscribe.
If a tool here told you something uncomfortable
Why your backtest lied to you
The free first module explains the arithmetic behind these calculators in full: where the bar comes from, why looking twice raises it, and what to do when your result sits below it. No signup, ~14 minutes.
Read it freeThe full course is eight modules, nineteen data sources with the trap in each, twelve working download scripts and the searchable ledger of every approach I tested, one of which survived. See what is in it →