Profitable EA: Avoiding Over-Optimization of Parameters in Pseudo-Forward Testing; Thorough Explanation of the EA "Sanshi Suimei"

How to Avoid Curve Fitting with Pseudo-Forward Testing

A Thorough Explanation of EA “Sanzu Suimei” 

■ Overview

EA “Sanzu Suimei” and “Senzi Ban” perform pseudo-forward testing added in MT5 to prevent excessive parameter optimization.

They avoid overfitting of parameters through backtesting and forward testing.

(Sales site)

“Senzi Ban” https://fx-on.com/systemtrade/detail/?id=14944

■ What is Pseudo-Forward Testing

As shown in the figure below, historical price data is split into backward testing and forward testing for backtest and forward test respectively,

where parameters are optimized in the backtest, and in the forward test we verify that they have not been excessively optimized.

This confirms the absence of overfitting.

■ Example

Which would you choose?

Case A Backtest

Case A Forward Test

Case B Backtest

Case B Forward Test

Compared with Case A, Case B’s backtest shows PF over 4 and looks favorable, but the forward test results indicate Case B’s performance declines considerably.

As a result, Case A was adopted as “Sanzu Suimei.”

A similar verification is performed for “Senzi Ban.”

■ Sales Site

“Senzi Ban” https://fx-on.com/systemtrade/detail/?id=14944

“Sanzu Suimei” https://fx-on.com/systemtrade/detail/?id=15576