🎩 Hedge Funds Got Crushed by the AI Trade. We Had One Warning: Price Still Matters.

Goldman’s Hedge Fund VIP Basket Recently Suffered Its Worst Relative Month in More Than 20 Years

Seven weeks ago, we warned that semiconductor stocks were priced for exceptional execution. Then crowding, momentum and leverage met gravity.

"The lesson isn't that AI was wrong. It's that crowded trades, momentum and expensive valuations eventually meet gravity."

Sometimes it's worth revisiting things you've written.

Not because you predicted the future.

You didn't.

But because investing principles have an annoying habit of surviving whatever happens next.

Back on July 9, 2026, we wrote about hedge funds enjoying their best start in years while simultaneously behaving rather cautiously.

And we highlighted two things.

The U.S. stock market was expensive.

And:

Semiconductor stocks were priced for exceptional execution.

Then July happened.

Ouch. 🤕

Goldman Sachs' closely watched Hedge Fund VIP basket — the stocks appearing most frequently among hedge funds' largest long positions — suffered its worst one-month underperformance versus the S&P 500 in more than 20 years.

July also produced one of the sharpest hedge-fund de-grossing episodes of the past decade as managers rapidly reduced exposure.

Translation?

Everybody loved the party.

Then somebody yelled FIRE! 🔥

And suddenly everybody discovered there was only one door.

🤖 The AI Trade Meets Gravity

For much of 2026, AI had been one of Wall Street's greatest money machines.

Semiconductors.

Memory.

AI infrastructure.

Mega-cap technology.

The thesis wasn't imaginary. The earnings weren't imaginary. The capital spending wasn't imaginary.

But neither was the crowding.

Hedge funds had piled into many of the same winners. When momentum reversed, managers began reducing their longs while simultaneously covering shorts.

The trade that had worked beautifully going up became considerably less beautiful when everybody wanted out at once.

Some popular momentum names suffered enormous drawdowns from their peaks.

And one fund demonstrated just how dangerous the combination of conviction + concentration + leverage can become.

💥 Situational Awareness: When 67% Disappears in a Month

Leopold Aschenbrenner knows something about AI.

The former OpenAI researcher founded Situational Awareness, an investment fund built substantially around his conviction that artificial intelligence would transform the global economy.

And for a while?

Boy, did that work.

The fund reportedly grew to approximately $45 billion at its early-July peak.

Then came the unwind.

Situational Awareness's portfolio value fell approximately 67% in July as its AI-related positions collapsed and leverage magnified the damage.

The fund ultimately sold most of its public-equity portfolio, including a substantial portion to Ken Griffin's Citadel, while eliminating leverage.

Aschenbrenner himself acknowledged how dangerous the situation had become.

This is worth emphasizing because the lesson isn't:

Aschenbrenner was stupid.

Quite the opposite.

You can understand an industry extraordinarily well.

You can correctly identify one of the defining technological revolutions of your lifetime.

You can even be right about the companies.

And still construct a portfolio capable of blowing itself up.

Being right about the future and surviving the journey to that future are two entirely different investment skills.

🍽️ We Wrote This Seven Weeks Ago

Here's what we wrote on July 9:

“Good performance doesn't eliminate risk. Sometimes it simply changes where the risk resides.”

And regarding semiconductors:

“The semiconductor industry isn't necessarily overpriced. It's priced for exceptional execution.”

That distinction mattered then.

It matters considerably more now.

Our argument wasn't that artificial intelligence was a bubble.

Our argument wasn't that semiconductor companies were bad businesses.

Some are extraordinary businesses.

The warning was simpler:

Expectations have a price.

When investors pay increasingly large multiples for increasingly popular companies, progressively more good news becomes embedded in the stock price.

Eventually, being good isn't enough.

You have to be better than everybody already expects you to be.

And that's a very high bar.

🧠 Three Evergreen Lessons

1. 🚪 Beware the Crowded Trade

Crowds aren't necessarily wrong.

Sometimes they're crowded into an investment precisely because the investment thesis is excellent.

The problem arrives when everyone owns the same thing and then needs liquidity simultaneously.

The exit doesn't get wider because more sophisticated investors are trying to use it.

2. 🚀 Momentum Is Wonderful. Complacency Isn't.

Momentum can persist far longer than skeptics imagine.

That's precisely why it becomes dangerous.

Success creates confidence.

Confidence creates larger positions.

Larger positions create leverage.

And eventually confidence can quietly mutate into an assumption:

This can't go down.

It can.

Everything can.

3. 🌎 Valuation Always Matters

This may be the least fashionable sentence in a bull market.

It's also one of the oldest truths in investing.

An extraordinary company can be a terrible investment at the wrong price.

A mediocre company can occasionally become a terrific investment at the right one.

Price determines how much future success you're already paying for.

And the higher the valuation, the smaller the margin for disappointment.

On Earth, gravity always wins.

Wall Street occasionally forgets that stocks live here too.

🎯 The Irony

Here's the interesting part.

After all that carnage, U.S. equity long-short hedge funds were still up roughly 10% for the year through mid-August.

And hedge funds haven't abandoned AI. Recent Goldman data suggests exposure remains substantial even after July's cleanup.

That's important.

This isn't necessarily the end of the AI investment story.

It may simply have been the end of one particularly complacent chapter.

And that brings us back to July 9.

We concluded our semiconductor discussion with this:

“Great businesses don't always produce great investments. Sometimes investors simply pay too much for them.”

We wouldn't change a word.

Except perhaps add three more:

Especially with leverage.

Invest carefully.

And Carpe Diem!