Do I need minute data to backtest, or is daily data enough?

Daily bars are fine for strategies that only act at opens and closes with market orders. Minute data becomes necessary the moment intraday mechanics matter: resting limit orders (did price actually touch your +7% target on Tuesday morning?), stop-losses (was the stop hit intraday before the close recovered?), time-of-day rules ('exit at 2pm if losing'), and realistic fill prices at the open. Daily-bar simulations of such strategies must guess what happened inside the bar, and the guesses are systematically optimistic. Minute resolution removes the guessing while remaining fast to compute — the backtests on this site run 5.5 years of minute data in under a second per strategy.