Backtesting
Backtesting is running a trading strategy on historical market data to see how it would have behaved: which trades it would have taken, where they would have closed, and what the result would have been after costs. It is the first test of any rule-based strategy, and the cheapest place to find out that an idea does not work.
What a good backtest includes
- Exact rules for entries, exits, stop loss and size.
- Real historical candles for the market being tested.
- Costs: exchange fees, slippage and, on futures, funding.
- No look-ahead: signals use only closed candles and fill afterwards.
- Futures mechanics: leverage, margin and liquidation.
- Every trade visible, with its prices, costs and reason for closing.
Its limits
A backtest describes the past. Live trading has real fills, delays and markets that change. A strategy tuned until its backtest looks perfect is often overfitted and does worse on new data. That is why a backtest is followed by paper trading on live prices before real money is used.
On Algonney
Backtests run on real historical candles from Binance's public data and can test as any supported exchange with its fees and trading rules. Fees and slippage cannot go below a floor, signals on a closed bar fill at the next bar open, and on futures funding and liquidation are applied. Each result shows its trades, equity curve, drawdown and a proof box with the data source, rules and engine version.
Related: slippage, drawdown, paper trading and Backtesting. Step by step: How to backtest a crypto trading strategy.