Backtesting

Use this category to evaluate a strategy on historical candles before connecting it to an order workflow. FYERS provides the historical data; the strategy logic, cost model and report run in your backtest code.

A useful sequence is: fetch candles → build a time-ordered signal → apply the signal without looking ahead → include brokerage, statutory charges and slippage → report returns and drawdown → test on a separate period. The trading skill provides example_strategy.py and a QuantStats report helper for this workflow.

Intraday timestamps need Indian session handling. An incomplete current candle should not be treated as a completed signal, and open positions at the end of available data should not be reported as closed trades. Keep the tested strategy separate from any live order path.