Analyzing Optimization Results by Removing Losing Time Intervals
Summary
The article presents a Frames Analyzer for reviewing strategy tester optimization results. It displays balance curves and ranks parameter sets by measures including profit, trade count, drawdown, and recovery factor. Its central feature identifies intervals containing losing trade sequences, removes a chosen number of those intervals from the historical results, and recalculates performance measures and balances. Users can compare original and adjusted results, save selected parameter sets and interval choices, and apply saved intervals to prevent an Expert Advisor from trading during them.
The tool stores optimization passes, parameter values, deal histories, and favorites in a database, and can read results from that database or an MQD file. The article illustrates the interface and workflow, but provides no independent evidence that interval selection will improve future results. Choosing intervals after inspecting the same historical data risks overfitting, so the proposed exclusions require out-of-sample validation before live use.
Key ideas
- The analyzer ranks optimization passes using profit, trade count, drawdown, and recovery factor criteria.
- It compares original balance curves with results after excluding selected losing periods.
- Saved time intervals can be used by an Expert Advisor to disable trading during those periods.
- Historical interval selection can overfit, so exclusions need validation on unseen data.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.