Financial Instruments Business Operator Kanto Finance Bureau Director (Kinsho) No. 1960 / Membership Association Investment Trusts Association, Japan Membership No. 012-02323 Go to GogoJungle Home
icon List
M&K
09/20/2026 02:00
Release： 09/20/2026 02:00
Update： 09/20/2026 02:00
M&K
Investment style : Trading System
Investment career : More than 10 years
Investment amount : -
For over 10 years, I have been involved in market analysis for stocks and FX, and I develop EA using the know-how I have cultivated. What I value most is “how to avoid unnecessary losses and carefully capture only trends with an edge.” To realize this, I have built an in-house backtesting and automatic optimization system using Python, aiming for robust logic grounded in extensive data verification. To ensure long-term, reliable operation, I strive for sincere quality improvement and support. If you have any questions about setup methods, please feel free to contact me via message at any time.
icon
Series
USDJPY Ultimate Design Philosophy —Why did this EA choose the “royal road”? (All 6 parts)
USDJPY Ultimate Design Philosophy ―― Why this EA chose the “classic” approach? (Six parts in total) Automated trading (EA) may work well right after you start using it, but it can stop functioning as market conditions change. In many cases, the cause lies in “overfitting” to past data. In this series, using the USDJPY-specific trend-following EA “USDJPY Ultimate” as the subject, we will explain from the developer’s perspective why we chose the trend-following “classic” approach, and how its weaknesses are compensated for in the design, episode by episode. We will delve into not only the entry logic, but also money management concepts, the reasons for publishing parameters, and the intention behind narrowing currency pairs and timeframes, explaining “why it was designed that way.” ※This series does not constitute trading recommendations; it aims to explain the design philosophy of EAs. Past backtesting results do not guarantee future performance.
Article cover image

Final Episode — Therefore I chose the “Orthodox Path” — USDJPY Ultimate: A Summary of the Design Philosophy Series (All 6 Parts)

This series is also the final installment. For five installments, I have discussed the design philosophy of one EA called USDJPY Ultimate. In the final installment, I will summarize by looking back at what has been covered and why I ultimately arrived at the word “classic/orthodox.”

14-day free trial version, now available

You can verify the designs discussed in this series in your own environment.

Try it now →

Looking back on the five episodes

1

Why I chose the orthodox approach

To avoid overfitting to past data, I chose a trend-following approach that can be explained structurally as the foundation. Momentum determination using SRSI and narrowing with a four-filter system control its weakness (whipsaw in ranging markets).

2

Capital management design

I discussed a concept of managing risk across three levels: per trade, per day, and per account. The winning logic and the losing logic are designed as separate components.

3

Reasons for publishing parameters

Rather than sharing “winning numbers,” I discussed publishing the approach to design so that users can verify it themselves, leaving room for verification.

4

Judgment criteria when performance declines

I spoke about evaluating based on number of trades, market environment, safety mechanisms at the account level, and publicly available performance—making decisions based on verifiable material, not intuition.

5

Differences from discretionary trading

I discussed a pragmatic approach to AI that “eliminates execution variation” but does not remove the effort to understand; the real-world way to interact with EA.

Note that all of these logics were validated and designed specifically for the USDJPY pair on the H1 (1-hour) timeframe. Rather than broadening the scope, I prioritized operating within a range that has backing from verification results.

Something that remained consistent throughout the five episodes

A design that is open and verifiable

The core idea across topics—being orthodox, dividing capital management into layers, publishing parameters, and sharing results—has the same underlying philosophy. Instead of declaring the system a black box that “will win,” I explain why I think so and ensure users can verify for themselves. That is the essence of the “orthodox” approach for USDJPY Ultimate.

In fact, while writing this series, I have practiced the same approach. There were several occasions where expressions used in the articles diverged from the actual code behavior at fine detail, and I corrected these by revisiting the implementation. Some places were revised after third-party feedback comparing with EA code. This may be a less glamorous aspect of a series, but I will be transparent. It isn’t easy to balance easy-to-understand explanations with precise implementations, but prioritizing accuracy and correcting as needed has been my way of practicing the “do not hide” stance I aimed to convey in this series.

That’s why I chose the orthodox path

Using trend-following as a longstanding fundamental approach, suppressing weaknesses with filters, protecting capital through hierarchical structures, and openly revealing the mechanism and results. It may not be flashy, but this is the current content of USDJPY Ultimate and the conclusion of this series. I have filled the word “orthodox” with such substance.

Thank you for reading all five installments. It’s quicker to experience it than to read about it. You can try it free for 14 days.

Would you like to actually verify the designs introduced in this series?

You can try it free for 14 days

14 days free trial →

Details of the actual logic, parameters, and performance are published on the product page

View the “classic” trend-following EA for USDJPY Ultimate (USDJPY dedicated)

※This article does not recommend investment or trading in any specific financial product. It aims to introduce the design philosophy of the EA. Trading the EA can incur losses. Past backtests and forward tests do not guarantee future performance. When trading in reality, please use your own judgment and responsibility with full understanding of the risks.