The S&P is still king, the rest of the world is already correcting

The Liquidity Map · Stocks report

The S&P is still king, the rest of the world is already correcting

Nikkei, Hang Seng and Nifty are already 9-11% below their 2026 highs.

The S&P is 2% below its own.

We're entering the autumn seasonal window, and I have a number in mind for it: about 5%, not more.

And the Dow already did what I was waiting for in March, it took out the 2025 highs.

September 3, 2026 · 6 min read

Key takeaways

The S&P 500 sits at 7,667, 1.7% below the 13 August high and up 12% on the year. The Nikkei is 11.3% off its June high, the Hang Seng 9.8%, the Nifty 9.2% and negative on the year.
Romania's BET, the best index of the year at +41% since 1 January, has entered a correction too: -5.9% from its 19 August all-time high, with a -3.6% day on 31 August.
Since 1993, the September-October window has brought a median drawdown of 2.5% from the 31 August close, and it exceeded 5% in 11 of 33 years.
On the big US indices I don't see more than about 5% in this window. The leveraged ETF short from 27 August stays a hedge for it, not a position to add to.
The Dow Jones sits at 53,062, 8.9% above its 2025 high. The March scenario played out: a test of the highs, then continuation.
The big discount of 30% or more on the major indices is one I see in the Shmita year, 20 September 2028 to 9 September 2029. 2027 stays a growth year.
Nothing changed on allocation. SpaceX still at 140, MicroStrategy at 123, the 86-94 zone is still the one I'm watching.

This week I looked outside America.

Nikkei, Hang Seng, Nifty. All three have been correcting for months.

The S&P hasn't. It sits 2% below its high and behaves like a king.

But September has started, and September has a track record.

Definition

The seasonal window

The September-October stretch, historically the S&P 500's weakest returns and its most frequent shakeouts of the year. It isn't a law, it's a statistic: since 1993, September is the only month with a negative average return, and the maximum drawdown inside the window versus 31 August has been 2.5% in the median year. It's the window where a hedge makes sense, not a crash prediction.

The king

The S&P 2% below its high, Asia 10%

The Nikkei topped on 25 June at 72,366 and sits at 64,214 today. Minus 11.3%.

The Hang Seng topped in January and is down 9.8%. On the year, it's negative.

The Nifty is the weakest of them all: minus 9.2% from its 2 January high and minus 8.5% on the year. India hasn't recovered since February.

Meanwhile the S&P sits at 7,667, 1.7% below the 13 August high and up 12% on the year. The Dow, up 10%.

Romania is no different. The BET is the best index of the year, up 41% since 1 January, and it set its all-time high on 19 August.

Since then, minus 5.9%. With a single day of minus 3.6%, on 31 August.

That means two things.

One, the global correction has already started, just not in America.

Two, when it gets here, it lands on an index that hasn't given anything back yet. That's why a hedge is worth having.

-11.3%How far the Nikkei sits below its 25 June 2026 high. The Hang Seng is at -9.8%, the Nifty at -9.2%, the BET at -5.9%, the S&P at -1.7%.

S&P 500, Nikkei, Hang Seng and Nifty, indexed to 100 at the start of 2026. The S&P is the only one near its high.

How far each index sits below its 2026 high, BET included. Red, more than 5%.

The BET, last month of closes. All-time high on 19 August, then a 5.9% correction.

And the month with the worst track record of the year has just started.

The window

Sept-Oct: about 5%, not more

I took the S&P day by day since 1993 and measured one thing: how far it fell, at some point between 1 September and 31 October, versus the 31 August close.

The median year: minus 2.5%.

In 11 of 33 years, the drop went past 5%. One year in three.

On the big US indices I don't see more than about 5% in this window. Turbulence, yes. A crash, no.

It isn't a crash prediction, it's the opposite.

I went short on the S&P on 27 August as a hedge for this window. It stays a hedge. I'm not adding to it.

Last week's checklist held up: the S&P kept 7,400-7,700 and closed at 7,667. SpaceX is still glued to 140, at 140.71.

After October, if the window passes the way I think, a new cycle may start on these big indices too.

11 of 33Years since 1993 in which the S&P fell more than 5% in the September-October window versus 31 August.

The S&P's maximum drawdown inside the Sept-Oct window, year by year since 1993. Red, the years above 5%.

And on the Dow, something I'd been waiting for since March has already happened.

The Dow

The 2025 highs are taken, the big discount moves

In March, with the Dow at 46,900, I wrote the scenario down.

If yields don't collapse this year, we only see a test of the 2025 highs, then continuation up into the next Shmita year.

Today the Dow sits at 53,062. That's 8.9% above the 2025 high of 48,731.

The test happened, the continuation came. So the big discount, 30% or more, isn't for this year.

I see it in the next Shmita year, which starts on 20 September 2028 and runs to 9 September 2029. Possibly with a retest of the 2026 highs first.

2027 stays, in my framework, a year of massive growth.

On Shmita years and why they matter for liquidity, I wrote on 19 August, I won't repeat it.

It stays a scenario I'm tracking, not a target. If yields do collapse this year after all, the calendar changes.

+8.9%How far the Dow Jones sits above its 2025 high of 48,731, set on 24 December.

Dow Jones since 2014, log scale. The dotted line is the 2025 high. Past Shmita years shaded, the next one marked on the right.

Concretely, for this week, nothing changes.

How we're positioned

I changed nothing on allocation this week.

The leveraged ETF short stays a hedge for Sept-Oct, at the same size.

Toward 2028 the plan is to step out of the market gradually, not to argue with everyone about whether it's the top.

SpaceX at 140, MicroStrategy at 123. The 86-94 zone on MicroStrategy stays the one I'm watching.

Conclusions

The S&P is 2% below its high, Asia is 10% below. The global correction is already under way, it just hasn't reached America.

The Sept-Oct window has a clear record: a 2.5% median, more than 5% one year in three. My frame for the big US indices stays about 5%.

The Dow took out the 2025 highs, so the March scenario played out, and the big discount moves toward the 2028-2029 Shmita year.

2027 stays a growth year. The rest is patience and a hedge kept where it belongs.

What to watch this week

Does the S&P hold above 7,400 through the end of October, or does it step into the 5%?
Do the Nikkei and the Nifty keep making new lows, or do they stop where they are?
Does the Dow stay above its 2025 high of 48,731 if the shakeout comes?

S&P is still king, while the rest of the world is already correcting.

Keep reading

2027 is THE YEAR

The consensus wants an October bottom. A green August says otherwise

The math for a rate cut