EA developers should invest their own funds to conduct forward testing

On April 29 and May 2, there was a double currency intervention, making the market extremely volatile. How is everyone doing?

As for my trend-following EA, in the first intervention, long positions were wiped out, and most of the profits were blown away...

In the second intervention today, the trend-following EA held short positions, so the intervention led to a surge in unrealized gains followed by a half retracement, halving the profits. Compared to last week, assets have increased a little.

This time, I plan to write an article titled “FXEA developers should fund forward testing with their own capital.”

In the world of stocks, it is common for fund founders to invest their own money in the fund.

Because investors and fund founders share profits and losses, healthy fund management can be achieved.

So what about the relationship between EA developers (providers) and buyers (users) in FX?

Forward results of an EA cannot be predicted even by the EA developers themselves.

Even if you guess what might happen, since no one can predict future markets, such guesses are unreliable.

Still, I publish with confidence on GoGoJungle, and for EA with strong confidence, I believe its real-account performance should be disclosed.

If there are many, it becomes difficult to run all EAs, so I think we should prioritize the best-selling, high-performing EAs.

In the case of GoGoJungle, since it can be linked to “REAL TRADE,” once a certain number of copies are sold, users register the EA in REAL TRADE to take forward testing for themselves.

But this becomes dependent on the users.

If you want to be successful as an EA creator, you should operate in a real account—even with small lots.

In my case, I use FXTF accounts as my main trading account for EA operations.

FXTF is my primary account.

Deposited 300,000 yen in June 2022; as of September 2024, 2,500,000 yen

Since I began automated FX trading in 2021, I have not lost money on an annual basis in FX overall.

From the purchaser’s perspective, you wouldn’t want to buy an EA from a developer who is losing money in FX, right?

Demonstrating steady wins on a real account and sharing forward results from that account to buyers is necessary to earn trust.

GoGoJungle’s forward testing is done on a demo account, and for swing-type EAs that trade relatively slowly, the gap between demo and real accounts is small.

Therefore, some think forward testing should be handled by GoGoJungle’s official forward tests, but

EAs that diverge from the demo account include morning scalping EAs, high-frequency trading types, and some indicator-based EAs.

At midnight server time, spreads widen, and spreads also widen during macro releases and events.

Also, EAs with an extremely high number of trades (over 1,000 per year) are affected by slippage and other factors.

Therefore, on the sales page, for EAs with comments that they diverge from the demo, their “REAL TRADE” results must be cross-checked with their live-trade performance.