Look-ahead bias (peek bias) is any use of information in a backtest that the trader could not have known at the bar in question. The most common forms: using the bar's close to make a decision and then 'entering' at that same close, using the period-end value of an indicator that is only finalized after the period ends, or using fundamental data dated by reporting period rather than release date.
Look-ahead bias is subtle and catastrophic — it can turn a flat strategy into one that prints a perfect equity curve. The discipline is to lag every signal by one bar (decide on bar N, enter on bar N+1's open) and to use release-dated rather than period-dated fundamentals.
Even simple indicators can introduce look-ahead if computed sloppily — a rolling mean centered on the current bar rather than trailing behind it includes future bars. Always verify that indicators only depend on data up to and including the bar in question.
Cómo Noon Barbari usa Look-ahead bias
Cada concepto aquí está implementado en la plataforma. Abre la documentación o la herramienta correspondiente para verlo en acción.
How we prevent lookahead →Términos relacionados
- Backtesting
Survivorship bias
Testing only on assets that survived to the present, ignoring those that died.
- Backtesting
Backtest
Simulating a trading rule on historical data to estimate how it would have performed.
- Backtesting
Overfitting
Fitting a strategy so closely to past data that it captures noise, not signal.
- Backtesting
Slippage
Difference between the price a strategy assumed and the price it actually got.