Indicator guide
RSI Backtesting Guide
RSI is a momentum oscillator bounded from 0 to 100. A valid backtest converts its readings into explicit, time-aware rules and evaluates them after costs without assuming profitability.
How RSI is commonly used
Researchers use RSI for oversold/overbought entries, centerline filters, and divergence-inspired rules. The indicator should be calculated only from information available at each simulated decision time.
Parameters to define
State the lookback length and threshold levels. Parameter choices affect signal frequency and lag, so compare a limited, predefined range rather than selecting one value after reviewing the full history.
How to evaluate a test
Review net return, maximum drawdown, volatility, trade count, win/loss size, exposure, and behavior across subperiods. Compare with a simple benchmark and test unseen data.
Research checklist
- lookback length and threshold levels
- Signal timing and execution price
- Fees, spread, and slippage
- Out-of-sample stability
Questions and answers
How do I backtest RSI?
Define the RSI calculation and signal rules, prevent future data from entering each decision, apply realistic execution costs, and validate on a separate period.