Overfitting: when a backtest is too good to be true

Overfitting, also called curve fitting, happens when a strategy is tuned so closely to past data that it captures random noise instead of a real, repeatable pattern. It looks excellent in the backtest and then fails on new data.

Most strategies have parameters: adjustable numbers such as "use a 20-day average" or "exit after a 7% drop". Changing them changes the results. The trap is testing many combinations and keeping the best one.