Selecting Among Trading Systems with Virtual Trades and Rolling Ratings

Summary

This article presents a method for comparing several trading systems through virtual trades, then allocating live trading to systems whose recent results meet a rating threshold. Each system receives a score based on its profit over a rolling window of completed trades, scaled against the best-performing system. The score can determine whether a system trades and how much of its configured lot size it uses. The example combines five EURUSD strategies, including moving-average and indicator-based rules, and discusses how to choose rating windows based on observed changes in each system’s balance curve.

The author reports historical tests with and without adaptation. In the supplied example, adaptation reduced the number of trades and drawdown, while also reducing total profit; the reported modeling method is explicitly described as crude and unsuitable for treating the results as reliable. The method depends on systems having discernible periods of gains and losses, and the author cautions against strategies with too few trades or irregular outcomes. A rolling profit rating cannot guarantee profitability or establish that past performance shifts will persist.

Key ideas

• Virtual trades let an Expert Advisor track candidate systems before assigning them live capital.
• Each system’s rolling profit is rated relative to the best system, and the rating can scale its trade size.
• The author recommends choosing rating windows based on observed shifts in a system’s profitability.
• The example’s adaptive test reduced trades and drawdown but also reduced total profit.
• The reported backtest used a crude model and does not demonstrate reliable future performance.

Tags

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