Name: Tetsushi O-nishi
Forex system trader
MQL5 programmer / EA (automated trading system) developer
Background
2018: Starting with discretionary trading
In 2018, I began discretionary Forex trading. At first I relied on indicators and intuition, and went through a long period where I simply could not win consistently.
Feeling that “I can’t go on like this,” I started visiting bookstores and devouring trading books. During this time I came across Mark Douglas’s book Trading in the Zone and realized that:
- The biggest enemy is not the market, but my own emotions.
- You cannot win over the long term without disciplined and consistent trading.
From there, I began to think that a “system” that allows rules to be executed in a testable, emotion-free way might suit me better. This led me to take an interest in Forex automated trading using MT5 EAs.
At first I did not build my own systems. Instead I used purchased EAs and copy-trading services. I searched MQL5 Market and Myfxbook for systems with good performance, and whenever I saw a beautiful upward equity curve I jumped in and tried that EA.
2018–2021: Heavy losses from high-risk EAs
During this period I experienced:
- Account blow-ups caused by grid / martingale / averaging-down EAs.
- Large drawdowns from copy-trading services.
- Buying expensive EAs and then losing the remaining account balance on top of that.
It was a repeated pattern of “equity increases for a while, then ends with a big loss.”
Over roughly three years from 2018 to 2021, the combined cost of tools and trading losses exceeded 2 million JPY. I became deeply disappointed with FX as a whole and strongly felt that:
“Many Forex-related tools and EAs focus only on short-term performance and involve substantial hidden risk.”
2021: Starting to build my own EAs and research trading logic
Even so, I couldn’t give up. In 2021 I started seriously studying MQL4 and MQL5, and began challenging myself to design and implement my own trading systems (EAs).
At the same time I studied intensively:
- The structure of trading logic.
- Market characteristics.
- The importance of backtesting and forward testing.
I spent countless hours developing and testing, searching for what might be called a “holy grail” trading logic.
2023: Realizing there is no holy grail
After developing many EAs and running extensive MT5 backtests, I eventually reached the conclusion that:
“There is no ‘holy grail system’ that can win in every type of market.”
In particular, the Forex market:
- Shows highly random price movements.
- Is heavily affected by trading costs such as spreads, commissions, and slippage.
Through the data I realized that earning stable profits is far from easy.
At the same time, by designing systems with a strong focus on:
- Expectancy
- Robustness (stability and reproducibility)
I began to see hope that it is possible to build systems with a genuine edge, even in a complex market like Forex.
Such systems:
- Do not necessarily have a high win rate.
- Experience drawdowns.
- Do not always show a perfectly smooth equity curve.
On the surface these logics may look plain and unimpressive.
However, by combining:
- Favorable risk–reward (cutting losses small and letting profits run),
- Proper money management (e.g. variable position sizing and compounding),
- And a design intended for long-term operation over many years or even decades,
I see the potential to “stay in the game and grow the account balance over time,” even starting from a small amount of capital.
(This is only a possibility and does not guarantee future profits.)
From 2025: Real trading only with EAs I trust enough to use my own money
After repeated testing, I began live trading—starting only with EAs I judged to be worth risking my own capital on.
As of December 2025, 14 EAs are running on my real accounts (the number of EAs in operation may increase or decrease in the future).
Purpose of Forex Robot Lab
1. Sharing accurate knowledge about system trading
Because I have personally experienced:
- Failures in discretionary trading,
- Account blow-ups from high-risk EAs and copy trading,
- And losses from buying tools based only on marketing claims,
I want to help reduce the number of people who repeat the same mistakes.
On this site I share:
- Reviews and cautions regarding third-party EAs,
- Trading knowledge based on probability thinking, such as expectancy, risk–reward and drawdown,
- Methods for running backtests and forward tests,
with the goal of improving literacy around system trading and EA operation.
2. Introducing my own EAs (trading systems) and publishing tests
For the trading systems (EAs) I develop, I publish, as far as reasonably possible:
- Long-term backtest results,
- Live account performance,
- Expected risks and potential drawdowns.
These EAs are designed with the policy of not depending on:
- Martingale,
- Overly aggressive grid / multiplier averaging,
- Extreme scalping that causes abnormal position build-up,
which I consider very dangerous from a risk management standpoint.
You can purchase and use these EAs via the MQL5 Market based on your own judgment, but please note:
They do not guarantee profits, and there is always a possibility of losses. All trading and EA usage is strictly at your own risk.
Disclaimer (Summary)
The content on this site is based on my personal experience, opinions and research and is provided for information purposes only. It does not constitute investment advice or a solicitation to buy or sell any specific financial instrument.
Information about EAs and trading performance does not guarantee future results.
Neither this site nor its operator can be held responsible for any losses arising from the use of this information or any EA introduced here. Please make all trading decisions at your own discretion and responsibility.
External links & activities
I also share information about EAs and trading on the following platforms:
- Myfxbook (live and backtest performance)
- MQL5 (EA sales, signals and performance statistics)
- X (Twitter)
- YouTube (content to be expanded in the future)
*Links and profiles will be updated on this page as needed.