Evaluating Trading Equity Curve Smoothness with R-Squared
Summary
The document proposes using the coefficient of determination to assess how closely a strategy’s balance or equity curve follows a linear trend. It contrasts this measure with common tester statistics such as net profit, profit factor, expected payoff, and trade count, arguing that those figures can hide path quality, depend on sample size, or be sensitive to the test interval and position sizing. R-squared is presented as an additional custom optimization criterion for selecting strategies with steadier growth.
The article discusses linear regression and correlation, explains collecting floating equity data, and describes implementing the statistic in MetaTrader 5. It illustrates that strategies with similar ending profits can have very different equity paths, and presents R-squared as a way to distinguish their trend smoothness. The measure is not a standalone test of strategy quality: it does not replace profitability, trade-count, or execution analysis, and the article acknowledges limitations. Its reported range and interpretation are given by the source, but a high value alone does not establish robustness or future profitability.
Key ideas
- Ending net profit alone does not describe the path a strategy took to reach it.
- R-squared can summarize how closely balance or equity observations fit a linear trend.
- Floating equity data are needed when the goal is to evaluate the strategy’s equity curve.
- R-squared can be added as a custom Strategy Tester criterion alongside conventional performance measures.
- The statistic has limitations and should not be treated as proof of robustness or future profitability.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.