Grid Search for Adaptive RSI Periods and Trading Levels

Summary

The article develops an approach for selecting RSI periods and signal levels instead of relying on standard settings such as 70 and 30. It builds on an earlier method that centers signals on the indicator’s observed range and scales deviations by their average size. The new goal is to evaluate those levels and handle uncertainty about which RSI period to use.

The proposed workflow uses an MQL5 class to collect RSI readings across periods from 5 to 70 in steps of 5, then exports data for statistical analysis in Python. The article describes comparing alternative levels with traditional ones and using grid search to identify candidate inputs. It reports a revised test with profitable-trade accuracy changing from 57.38% to 56.47%, alongside a substantial increase in trades and risk, but the excerpt does not provide enough detail to assess the full test design or robustness. Results are specific to the presented setup and do not establish that selected parameters will generalize to other markets or periods.

Key ideas

• RSI values can behave very differently across indicator periods and markets.
• Centering signal thresholds on the observed RSI range can offer an alternative to fixed traditional levels.
• The proposed workflow gathers readings across multiple periods and analyzes them in Python.
• Grid search can help compare candidate RSI periods and levels without manually testing every setting.
• The reported test suggests a tradeoff between accuracy and risk, but does not establish out-of-sample robustness.

Tags

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