backtested

Backtested refers to the evaluation of a trading strategy, model, or investment rule by applying it to historical market data to simulate how it would have performed in the past. A strategy is typically defined by its entry and exit rules, position sizing, risk controls, and assumptions about costs. When these rules are executed on a historical data series, the resulting results are known as a backtest or backtested performance.

The purpose of backtesting is to estimate potential profitability, risk characteristics, and robustness, and to compare competing ideas before deploying capital in live markets. It helps identify practical issues with the rules, such as how often trades occur, capital requirements, and sensitivity to market conditions.

The backtesting process generally involves selecting a data set and time frame, coding the strategy rules, simulating trades (including transaction costs and slippage), and computing performance metrics. Common metrics include cumulative and annualized returns, maximum drawdown, volatility, Sharpe ratio, Sortino ratio, win rate, and drawdown-related measures. Analysts may also examine risk factors such as exposure, turnover, and the probability of ruin.

Backtests come with limitations and potential biases. Look-ahead bias occurs when future information is inadvertently used in past data; survivorship bias ignores delisted or failed instruments; data snooping and overfitting tailor the model too closely to historical irregularities. Liquidity, market impact, and changing regimes are often underrepresented in simple backtests.

Best practices include separating in-sample and out-of-sample data, using walk-forward testing, performing sensitivity analyses, and ensuring transparent, reproducible methodology with clear assumptions and data sources.