Evaluating Backtests with Sharpe Ratio and Drawdown Measures

Summary

The article explains how to assess a strategy after an event-driven backtest by calculating portfolio-level statistics from its equity curve. It describes the annualized Sharpe ratio as mean periodic return divided by return variability, scaled by the square root of the periods-per-year factor. That factor should match the return sampling frequency; the example discusses daily, hourly, and minute data. The implementation assumes a zero benchmark return rather than subtracting a risk-free rate.

It also defines maximum drawdown as the largest peak-to-trough decline and drawdown duration as the number of trading bars spent below a high-water mark. These values are calculated from the equity curve and then combined with total return and Sharpe ratio in a portfolio summary. The article cautions that duration counts bars, so it does not directly express elapsed calendar time. This is a basic portfolio-level view: it omits trade-level analysis and other risk or reward measures, and the source gives no empirical performance results.

Key ideas

• Annualized Sharpe scales average periodic returns by their variability and sampling frequency.
• The example Sharpe calculation assumes a zero benchmark return.
• Maximum drawdown measures the largest peak-to-trough decline in the equity curve.
• Drawdown duration counts trading periods until the equity curve reaches a new high-water mark.
• Portfolio summary metrics can be extended, but these measures do not provide trade-level analysis.

Tags

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.