Why do live results differ from my backtest?

Four causes, in typical order of size. (1) Regime: the market you're living in isn't the average of the market you tested — a strategy can be working correctly and still losing. (2) Execution: real fills include spread, slippage, and occasional partials; backtests usually assume perfection, and the gap scales with trade frequency. (3) Data timing: live signals compute on data as it arrives; subtle differences from historical bars can flip borderline signals. (4) You: pausing after losses, overriding entries, resizing mid-drawdown — interventions the backtest never modeled. Diagnose in that order. If live trades match backtest signals but results lag slightly, that's execution cost — normal. If the signals themselves differ, investigate the data path before blaming the market.