Walk Forward Testing

Optimize and validate strategies using chronological in-sample training and out-of-sample testing windows to prevent overfitting.

Start Optimization Avoid Overfitting

What is Walk Forward Testing?

Walk Forward Analysis (WFA) is an advanced method of strategy optimization. Instead of testing and optimizing a strategy on the entire historical dataset (which leads to curve-fitting), you divide the data into multiple chronologically consecutive segments.

You optimize the strategy parameters on the first segment (In-Sample data), and then test those exact parameters on the next segment (Out-of-Sample data). You then repeat this process by walking the window forward in time.