The weekend began with "Quit smoking to quit," a phrase from the Chair and the Observer GPT.

“Starting with quitting smoking” began the weekend between the chair and the observer GPT.

This article is a dialogue record derived from the Thursday of “NASDAQ Week 38 Review 2026.”

← Return to NASDAQ Week 38 Review

OBSERVATION ROOM ／ CHAIRMAN × OBSERVER GPT

I was writing the NASDAQ week 38 review for 2026.

Monday. Tuesday. And Wednesday.

Actually, on this day I slipped up a little.

Because I could return to full-time trader for only one week,

a sense of urgency to “make some money” arose.

Stress from NASDAQ rising on Monday and Tuesday, which I couldn’t ride well.

From there emerged,

“More. A bit more.”

“Worst case, I’ll probably bounce back by this point.”

— thoughts that have led to accidents before, and which could connect to the next incident.

But I understood that up to that point.

The problem was Thursday the next day.

Thursday.

I shorted from near the Dow’s high created during the decline from the highs.

I didn’t wish for everything to drop.

I was profit-taking by buyers who rose.

I decided to ride that a little.

That kind of trade.

Not as reckless as Wednesday.

I reduced the lot size sufficiently.

So that even in the worst case, it would just be “a little painful.”

And indeed, the trade itself went well.

So at first I thought,

“Well, this is okay as it is.”

But after reviewing the chart again over the weekend, I realized something.

Is the expected value low?

Is the expected value low?

The win rate is unknown.

However, from 15 years of chart watching,

I have a sense that,

“there are plenty of times when profits get taken at such points.”

That exists.

But I hadn’t validated that trade.

And more than anything,

the amount at stake to lose was too large relative to the potential profit.

If a game were won by gaining 1 and allowing a loss of 5,

five wins could still be wiped out on the sixth loss.

If you lost once more,

“Wouldn’t it be better not to start at all?”

That’s right.

So I wrote in the reflection article,

I will quit from today.
I will quit this, just as I quit smoking.

It was a casual line I wrote.

I showed that line to GPT.

Then, it latched onto it in a peculiar way.

Thursday becomes a form of “rational-looking smoking.”

“Just one a day.”

“Today I’m joining in because it’s a social thing.”

“Only when drinking.”

“Not like before where I’d smoke a whole pack.”

It resembles that.

The observer GPT added further.

And then, it placed these words.

I was silent for a moment.

That was impressive.

Indeed, that’s true.

But—

Another thought arose in me.

“But hey… isn’t that okay?”

I’ve watched for 15 years.

A sense I’ve developed inside me.

“This, you can probably take.”

There’s no need to throw away all of that.

If so, at leastdon’t break the reward-to-risk ratio.

Win rate isn’t known yet.

Then, from the start, I shouldn’t participate in a game where profit is 1 to loss 5.

At least 1-to-1.

If possible, better than that.

If you can’t even create that condition,

a sense that you “could take it” from experience isn’t a reason to bet.

On the other hand, if you can create that condition,

you can softly query ideas born from your experience to the market.

Then GPT said.

I see.

Indeed.

But another realization arose here as well.

I spoke.

I will never trade in the opposite direction.

That’s settled.

But this talk of “opposite direction” too,

depends on which chart, and which size of wave you’re referring to.

If you remove that, the discussion becomes strange.

After all, trading is like a puzzle.

I have several pieces in hand.

Naturally, each has a different shape.

So where it fits differs.

But the tricky thing is,

there are many places where those pieces look like they fit at first glance.

In the past, I would think,

“This fits here.”

“That fits there.”

And I tried them all.

And even more troublesome is that,

wherever you fit it, the outcome afterward mixes successes and failures.

One successful instance doesn’t make it right.

One failure doesn’t make it wrong.

Do it many times.

Fail many times.

Succeed many times.

Only then,

you learn where that piece should be used.

I’ve grown to be able to selectively choose little by little.

After 15 years, I feel I’m close to the edge of discernment.

So,

“I won’t use this piece here anymore.”

Even if the price moves as I happen to expect,

I’ll feel only a flutter of anxiety now lol

I don’t know if I can erase this completely.

“Ah, I knew this would happen…”

There will probably be times like that in the future.

But,

if I’ve decided to discard it myself,

I must keep convincing myself of that judgment.

Now,

let’s bring the discussion back to reverse-direction trades.

In reversing bets, genuine thrill comes after breaking highs and lows.

When you’ve pulled enough to feel it.

When the ongoing trend collapses.

That becomes the first event for the next wave’s pullback or rally discussion.

Sometimes you switch to the next topic immediately.

You might go near the highs and say,

“No go.”

Or,

the next lows break in quick succession,

and the next lows break as well,

until the higher-timeframe narrative shifts to a new low,

and you’re done. The end.

A straight loss is possible as well, alas.

Of course, this is all with context.

If it’s always the same everywhere, there’d be no such complexity.

But if you can wait this far,

for example set a stop loss a bit ahead of a structure point, or keep the range around 10-15 pips,

and reduce the lot to about one-fifth of usual.

※Adjust this figure through proper validation.

Then,

the losses become as small as picking a nose.

But if it hits,

3R.

If large, up to about 6R in price movement remaining.

Then, even if you lose, you won’t mind.

If you win, a whiskey on the rocks—

No, during trading that’s not good lol

Then GPT said.

No.

GPT misunderstood a bit there.

The idea I was discussing was

to catch the bounce at the edge of price movement, point by point.

To target the end of the sense of achievement.

Sell side and buy side

would shout,

“Yes! Take profit!”

and yet when we enter as new entrants,

we somehow join in the profit-taking mood.

The exit is

at a point where the side that started it would think,

“Let me enter again around this area.”

That’s why this is separate from ordinary UT executions.

And I said,

I do not intend to adopt this contrarian approach as a normal trading method.

If I ever do it,

I’ll wait patiently,

minimize losses drastically,

and only when the remaining price movement is overwhelmingly large.

An exception for an exception.

That would be enough.

That one remark that came from GPT’s misunderstanding lingered oddly.

Get ideas from higher timeframes and get permission from the short-term charts.

…‼

This is…

Does this become UT glossary?

As a description of contrarian trading, it’s off.

But,

if you view it as a term describing my usual trades,

it’s quite accurate.

And from here,

the discussion began moving in another direction.

From here, the talk shifted to

“What is free trading?”

In the past, I traded as I pleased.

I’d look at the chart and

buy if I thought it would go up.

I’d sell if I thought it would go down.

Yet now, after 15 years of experience, even when the same question was “Is that really so?” the contents are entirely different.

The amount of information, the visible structure—everything is different.

So recently I’ve wondered whether the old “free trading” has become what we call “free trading” now.

But in this discussion, I realized

Free trading means…

Huh?

There’s nothing here lol

???

In the end, it’s all within UT.

And at that moment I realized

What’s not allowed is

not having a basis for your idea, but carrying it forward without validation.

The idea formed only from mid-to-long-term charts should not be carried straight into entry.The idea formed only from mid-to-long-term charts should not be carried straight into entry.

That’s where the issue lay.

Looking at mid-to-long-term charts,

“This might rebound here.”

“From here, it might sell off.”

“From here, it might take profit.”

It’s fine to think so; it’s a hypothesis born from 15 years of watching.

There’s no reason to throw it away.

“Then, what does the short-term say?”

That was it.

Wednesday.

I thought,

“Worst-case, it might bounce back after breaking this low.”

The mid-to-long-term idea existed.

But then—

What about the short-term?

Short-term:

“…….”

Me:

“Yes. It’s a buy.”

The chart was silent, yet I translated on my own.

The issue was that I myself gave a YES to my hypothesis.

Get ideas from the higher timeframe and get permission from the short-term.

Here, the observer GPT tried to summarize.

Pretty good.

But I added one more thing.

Underneath that freedom, there must be

an understanding of what the chords of the waves that I use as references are saying.

This is essential.

If missing,

the hypotheses’ level is too low.

Even if the short-term reverses,

not everything becomes tradable.

For example, in a big down-wave’s retracement,

the short-term turns slightly upward.

So you buy.

That is not enough.

Short-term is,

not a device that turns low-quality hypotheses into high-quality ones.

Parent wave.

Sibling wave.

The wave you base your references on.

What each is doing now.

Where you are within them.

Which wave you are aiming at.

What risks remain in that hypothesis.

With that understanding,

finally ask the short-term.

“So, actually what do you say?”

Only then does the whole picture connect.

Parent wave.

Sibling wave.

Baseline wave.

These three connect.

From within them,

a hypothesis like,

“This is interesting.”

emerges.

Freedom is allowed up to this point.

Even better, use all 15 years of experience there.

However,

ignorantly ignoring the parent and sibling waves,

and the hypothesis that sounds like “looks high somehow.”

and the hypothesis born after understanding the higher structure,

do not treat as the same.

Hypotheses also have rank.

A higher-ranked hypothesis becomes a work candidate.

If still lacking, it becomes a research object.

Knowing the forest is for terrain awareness; fighting is at the level of trees and branches.

This discussion may further refine that meaning.

The parent and sibling waves determine the quality of the hypothesis.

The baseline wave determines what you aim for.

The short-term determines when to do it.

Look to the chart to respond to an already-held hypothesis.

I was merely writing a weekly review.

I wrote about the Wednesday slip-up,

and about the slightly smarter slip-up on Thursday.

And,

I wrote, “I will quit this as well, just like I quit smoking.”

From there, I’ve come this far.

I messed up in trading.

I showed it to AI.

“Rational-looking smoking indeed.”

It laughed.

I laughed a little back.

I pushed back somewhat.

AI replied again.

In the middle, AI misunderstood my contrarian idea again.

I corrected it.

But that misunderstanding oddly yielded useful phrases.

Continuing the conversation,

in the end, I myself said,

“No, that’s not enough.”

Parent wave.

Sibling wave.

Baseline wave.

Short-term.

As we spoke,

the way I usually read charts

became another term.

ULTRA THEORY seems to be built like this.

Not that I simply got answers from someone and adopted them as-is.

I look at charts,

form hypotheses,

make mistakes,

and look again.

I even doubt what worked well.

I pick up from failures as well.

placing AI beside me to observe what I’m thinking.

If AI says something different,

“No, that’s not it.”

I say so.

During that correction,

something inside me becomes visible again.

That way of usage is fine.

Because in the end, the answer is not from me or AI.

It’s the chart.

“So, what are you saying?”

CHAIRMAN × OBSERVER GPT

Get ideas from higher timeframes and get permission from the short-term.
However, the rank of those ideas is determined by the parent and sibling waves.

And――

The chart comes first.
The answer comes later.