01 What is lookahead bias?

Testing on too much dataUsing information in a test that was not available at the time of the decisionTrading in the wrong sessionIgnoring commissions

02 Which of these is lookahead bias?

Using the previous day's highUsing a candle's close to decide a trade during that same candleUsing a fixed stopTesting on minute data

03 What is overfitting?

Using too few rulesTuning a strategy so closely to past data that it captures noiseTrading too large a positionTesting on recent data

04 What is out-of-sample data?

Data from another marketData that played no part in building or tuning the rulesData with missing barsThe most recent week only

05 Why can assuming a limit order fills on a touch overstate results?

Limit orders cost morePrice can touch a level without the order filling, because of the queueLimit orders are not allowed in futuresIt ignores margin

06 Which cost matters especially in forex?

DividendsThe bid-ask spreadStamp dutyContract roll

07 A strategy has 9 trades in its backtest. What is the main concern?

Too many tradesThe sample is too small to say muchThe drawdown is too lowNothing

08 What does maximum drawdown show?

The best tradeThe largest fall from a peak in the resultsThe average trade lengthThe commission paid

09 Small changes to settings make a backtest's result swing wildly. What does that suggest?

The strategy is robustThe result is fragile and possibly overfittedCosts are too highThe data is too recent

10 Do backtest results guarantee future performance?

Yes, if out-of-sampleYes, with enough tradesNoOnly for futures

Quiz: Backtesting Without Fooling Yourself