#27 Why I discarded PF9.46 and chose PF3.53 - Surviving food poisoning and the birth of the "Investment Trust Type EA" -
February 2, 2026: Note
A long time has passed since the last update.
“Did I hit a wall in development?”
“Could I really have been erased?”
Some of you may have wondered about that.
To cut to the chase, the EA was completed the day after I wrote the last article (January 16).
So, why have I remained silent until now…
The reason is extraordinarily human.
In the first week, as a reward for pushing hard for five months since August, I slowed down and rested like a zombie, binge-watching Netflix's Golden Kamuy.
And after that, it was hell. I got food poisoning.
I was tormented by a high fever and vomiting.
I didn’t have diarrhea, though…
However, there was not a single micron of energy left to open the computer and write an article (cries).
I’ve finally recovered, so I’m reporting this belatedly.
“Lowering performance” was easy
In the last article, I wrote this.
“If I publish PF9.46, FX brokers will notice and freeze the account entirely.
So I’ll deliberately lower the performance.”
Increasing performance is extremely hard, but lowering it was surprisingly simple.
I tightened the entry conditions and simply did not enter in any scene with even a hint of risk…
That’s all there is to it.
And thus the final form with its “fangs removed”… no, “fangs concealed” is here.
PF (Profit Factor) = 3.53
Generally speaking, this is still an outlier, a god-level figure.
Normally it might be at a level where you’d expect to be hunted (countermeasures by brokers).
But this time I believe it’s okay.
Because,I drastically reduced the number of trades.
Please think of it as not FX but a “mutual fund”
The concept of this EA is not a gambling machine aiming for 100% monthly returns by spinning rapidly.
It won’t waste shots, and it’s not afraid of going quiet for several days.
However, when it does shoot, it will certainly finish the job.
In other words,an “ultra-precise mutual fund”of an EA.
There’s no showiness, but your assets will increase before you notice it
A “battle you can’t lose” that has been automated successfully.
This is the backtest trade history.
Losing a little, winning big. Or breaking even.
There’s no flashiness, but you can see the balance steadily growing.
I won’t predict. I will only wait.
This is the idea from Mark Douglas’s Zone that I also mentioned on social media.
This forms the core logic of this time.
Most traders and EA developers tinker with indicators to predict the “future.”
“Will it go up next? Will it go down?”
I will make no predictions at all.
The high of the previous candlestick will always be surpassed, and the low will always be broken.
When the market moves, it inevitably passes through there.
So you should set up a net there and wait.
That’s enough.
Regarding the sealed “Fourth Work”
Thus, my maiden EA is complete.
This is already powerful, but I have more in mind for my concept.
I also plan to release a second and third work that take a bit more risk and aggressively attack while adjusting PF to avoid being hunted by brokers.
However,as the “Fourth Work,”I had envisioned a system where an individual trader can outrun Wall Street’s algorithms and legally siphon cash from megabanks like JPMorgan, UBS, and Deutsche Bank hidden behind brokers…
About this,I have stopped developmentand sealed it.
When I asked an AI to simulate it,
“If you release that, you’ll definitely be erased by the CIA.”
It warned me seriously (laughs).
Life comes first.
I’ll save fighting state power for another time.
Currently under review
So, the long development road that began in August has reached a provisional goal.
I have submitted to Gogojan and am waiting for review.
The sales page has not been published yet.
If it passes review, I’ll announce it here again.
Next:
I plan to complete the logic for the second EA and publish it in a separate serialization.
My EA is about making a computer do what humans would do manually.
After all, it’s hard to give a program “the human eye”… seriously! (˘•ω•˘)
※
The EA I’m developing is, as introduced in Episode 2, about automating the method
“Winning a lot even without scalping”.
This is the only paid article in this series so far.
150 yen!Inexpensive!... President, inexpensive! ٩(ˊᗜˋ*)و
× ![]()
