Don't fool yourself: backtesting hygiene
Pseudo-Mathematics and Financial Charlatanism: The Effects of Backtest Overfitting on Out-of-Sample Performance
A readable warning: try enough configurations and a high in-sample Sharpe is nearly guaranteed and means nothing.
The Statistics of Sharpe Ratios
Standard errors for Sharpe ratios, and why annualising by √12 is wrong when returns are autocorrelated.
Backtesting
Haircut Sharpe ratios for multiple testing. A practical rule for discounting backtests.
The Deflated Sharpe Ratio: Correcting for Selection Bias, Backtest Overfitting, and Non-Normality
Adjusts a backtest's Sharpe ratio for the number of variants tried, skew and fat tails. Tells you whether your best variant is just luck.
The probability of backtest overfitting
Combinatorially symmetric cross-validation estimates the probability that your in-sample winner underperforms out of sample.
A Reality Check for Data Snooping
A bootstrap test of whether the best of many strategies really beats a benchmark once you account for the search.
Does Academic Research Destroy Stock Return Predictability?
Anomaly returns are lower out of sample and lower still after publication. Expect alpha found in papers to decay.