Current ACE Gold is

From 2018 to 2026, about 8 years and a half

backtesting has been conducted.

But,

we did not run for 8 years and a half from the very start.

The first validation period is

July 2024 to August 2026, about 2 years and 1 month.

Based on that result,

I decided that ACE was completed once and proceeded to sell.

At that time,

I did not think I would need to remake the completed ACE again.

The trigger was a problem at the time of sale

What happened first was,

it was not an issue of ACE's logic.

It was about the handling of overseas broker names that were used.

So, for a while,

I redid a backtest for about 3 years using a domestic environment.

Further, during that process,

we decided to conduct a longer-term validation

.

Then,

let's do as much as possible.

At first, the aim was

10 years.

However, when looking back into the past,

data around June 2017 caused a data issue.

So the starting point was changed to January 2018.

January 2018 to August 2026.

Backtesting for about 8 years and a half began.

Not just extending the period...

ACE, which had been running for 2 years and 1 month,

ran for 8 years and a half.

At first,

I simply thought the validation period would just be longer.

But,

it was actually different.

A market I would not have encountered in 2 years.

A period when profits could not be made for a long time.

A period when losses accumulated.

By extending the period,

weaknesses of ACE that had not been seen until now

began to surface one after another.

And,

the ACE that was supposed to be complete

could not run for 8 years and a half.

funds ran out.

ACE developed a “defense”

From here,

ACE's development changed dramatically.

Until then,

we looked at how much profit we could make

but in 8 years and a half,

there were things that had to be confirmed before that.

Could it survive to the end?

After seeing the initial zero drawdown,

we investigated the cause,

and added defenses to ACE.

And,

set off for another 8 years and a half.

This time, we believed we could finish.

We thought so.

again, funds ran out.

The second zero drawdown.

Even guarding wasn’t enough

After the first failure, we added defenses.

Still,

8 years and a half could not be completed.

So again,

we investigated where ACE collapsed.

Dangerous entries.

Scenarios where losses piled up.

Situations where the number of positions increased.

Things that in short periods did not become big problems,

in a long time,

gradually pressed ACE into a corner.

Each time, we investigated the causes,

and added defenses we thought were necessary.

And,

we ran the backtest again.

Third time, finally completed 8 years and a half

And on the third time,

finally ran the full 8 years and a half.

ACE that had been completed in 2 years and 1 month

was supposed to be extended to 8 years and a half,

but the first zero drawdown.

With added defenses,

second zero drawdown.

Further investigate the causes,

and remake once again.

And,

third time, first complete run.

Here,

the approach to long-term backtesting changed.

Long-term backtesting is not simply about extending the period to look at profits.

It is also about finding weaknesses that did not appear in a short period,

and testing to see where ACE might break

It was also a test to explore where ACE could fail.

Because the place of failure became visible,

a concept of “defense” was added to ACE.

This should have been the completion...

We ran the 8 years and a half.

This time for sure,

complete.

We thought so.

However,

Until now I had

kept ACE alive by

adding defenses.

But,

as I watched ACE that could finish running,

a reverse question emerged.

Are all of these defenses really necessary?

by over-protecting, are we also stopping ACE's original profits?

ACE development proceeds in the opposite direction.

The defenses added

will now be questioned one by one.

Next Episode Preview

Why were the defenses built to protect removed?

After two zero drawdowns,

ACE finally completed the full 8 years and a half on the third attempt.

However, when we tested the defenses,

some defenses, when removed, improved performance.

Next, Episode 12.

ACE moves from “additive development” to “subtractive development.”