Free while in beta
Your strategy looks great on paper. Let’s try to break it.
Backte.st runs a transparent backtest of your trading rules on real market data, then stress-tests the result with unseen data, higher costs, delayed entries and more, so you find the weak spots before the market does.
- Up to 10 markets per run
- 6 robustness checks
How it works
- 01
Build your strategy
Pick indicators such as RSI, MACD, Bollinger Bands or Supertrend, set entry and exit rules, and add stops and targets. No code.
- 02
Choose markets
Stocks, ETFs, indices, forex or crypto. Up to 10 at once, from 1-minute to monthly candles.
- 03
Run & share
Get every trade, the win rate, average R, profit factor, drawdown and equity curve in seconds, then share it as a link.
Strategy lab
Test a strategy in three steps
Step 1: Build your strategy
Pick indicators, set entry and exit rules, and add stops and targets. No code needed. Or start from an example.
Buy when
Sell when
Step 2: Choose markets
Pick up to 10 symbols. Each market is tested separately with the same rules.
Stocks, ETFs, indices, forex and crypto.
Step 3: Set costs & run
Set the position size and realistic trading costs, then run the backtest.
How the test works
Each market starts with 10,000 in its own currency. Signals are read from completed candles and orders fill at the next open. Stops and targets are checked against each candle’s high and low; if one candle touches both, the stop counts first, and a gap past either fills at the open. Every fill pays half the spread plus commission, and market and stop fills also slip. One position at a time, no leverage; anything still open is closed at the end. The last 30% of the history is also scored on its own as an unseen period.
Built for skeptics
One good equity curve proves little. So we attack it.
You approve explicit rules before a deterministic engine runs the numbers. Then every run is re-tested under harsher conditions:
Unseen period
The last 30% of history is scored on its own, as if you’d never seen it.
How this check worksDoubled costs
Commission, spread and slippage doubled on every trade. Thin edges disappear fast.
How this check worksNearby parameters
Every indicator period shifted ±10% and ±20%. Curve-fitted rules fall apart.
How this check worksDelayed entry
Every order filled one bar late. Shows how much timing luck you’re relying on.
How this check worksBest trades removed
Drops your five biggest winners. Is there an edge, or just a few lucky trades?
How this check worksTrade-order shuffle
Reshuffles the trade order 200 times to show how deep drawdowns could get.
How this check works
From the blog
All postsHow to read a backtest report without fooling yourself
Net return, drawdown, trade count, buy and hold, and robustness checks: what each number in a backtest report tells you, and what it can’t.
Overfitting in backtesting: why great results often fail live
How testing many variations makes noise look like an edge, the warning signs in a report, and simple habits that reduce overfitting.
Out-of-sample testing: how to use a holdout period honestly
Why scoring the last part of your history on its own catches overfitting, how Backte.st’s unseen-period check works, and how to avoid using it up.