Calibrating a trading rule using a historical simulation (also called backtest) contributes to backtest overfitting, which in turn leads to underperformance. We propose a procedure for determining the optimal trading rule (OTR) without running alternative model configurations through a backtest...