To The Moon

Everyone opens the latest 13F filings looking for the same thing:

What did smart money buy?

I think that misses the more interesting question.

What if smart money itself cannot agree? 🤔

The latest Q2 2026 filings show exactly that.

Across more than 6,000 institutional filings reviewed by Reuters, there was surprisingly little agreement about where the next big returns in technology will come from.

Nearly 44% of institutions reduced exposure to the Magnificent Seven, while about 42% increased or initiated positions.

That is almost a coin flip.

For me, that might be more useful than any individual billionaire’s purchase.

🧠 Signal #1: The AI trade is splitting into different camps

AI is no longer one trade.

A year or two ago, the strategy could almost be simplified to:

AI spending rises ➡️ chip demand rises ➡️ buy the obvious winners.

Now institutions seem to be asking a harder question:

Who actually captures the next dollar of AI spending?

That creates completely different answers depending on the investor.

Tiger Global, for example, reduced several huge technology positions during Q2, including $NVIDIA(NVDA)$, Microsoft and Meta, while cutting Alphabet exposure by more than 45%.

At the same time, Tiger Global increased its Intel position.

That caught my attention.

It does not necessarily mean Tiger thinks Nvidia is finished or Intel is suddenly the better company.

It could simply mean the risk/reward changed.

That distinction matters.

🔥 Signal #2: Semiconductors are not being abandoned

Here is where it gets even more interesting.

Despite some huge funds trimming individual AI winners, institutional positioning toward semiconductors was still positive overall.

Around 48% of reporting institutions were net buyers of semiconductor stocks, compared with roughly 34.5% that were net sellers.

So the message is not:

❌ Smart money is leaving chips.

It looks more like:

✅ Smart money is becoming much more selective about which chips, which valuations and which part of the AI supply chain it wants to own.

That is a very different market.

🐯 Tiger Global rotating from NVDA toward INTC is fascinating

This is probably the move I find most interesting.

Nvidia remains the dominant AI accelerator story, but the stock already carries enormous expectations.

Intel is almost the opposite.

Intel has spent years trying to rebuild credibility, yet the company is pouring capital into manufacturing and foundry expansion.

The company recently raised $20 billion through an upsized share offering, giving it more capital for that turnaround and its manufacturing ambitions.

That does not automatically make Intel cheap or safe.

But it shows why two intelligent investors can look at the same semiconductor boom and choose completely different exposures.

One might prefer the proven AI leader.

Another might prefer the turnaround where expectations are lower.

👀 And other big funds are moving differently again

Third Point reportedly exited Nvidia and Broadcom completely during Q2, while increasing exposure to Taiwan Semiconductor and Alphabet.

So even among sophisticated investors, there is no single “correct” AI basket.

Meanwhile, institutions were still net buyers across a group of AI-related companies including names such as Broadcom, Arista Networks and CoreWeave.

This is why I think simply copying a famous investor’s 13F can be dangerous.

You might copy their sale without understanding what they bought instead.

You might copy a position that was hedged elsewhere.

Or you might copy something they have already changed since June.

⏰ The biggest problem with 13Fs

13Fs are useful, but they are rear-view mirrors.

The latest filings show holdings as of June 30, not what those investors necessarily own today.

Institutional managers generally have up to 45 days after quarter-end to report their holdings. They also do not give us a complete picture of shorts and many other exposures.

That means I would never read a headline saying:

“Big fund bought XYZ”

…and immediately hit Buy.

I want to know why the position makes sense today.

💰 So what am I taking from the filings?

Not “sell AI.”

Not “buy Intel.”

Not “copy Tiger Global.”

My takeaway is simpler:

The easy part of the AI trade may be over. Stock selection matters more now.

And I think there are four questions worth asking before adding money:

🔹 Is revenue actually accelerating?

🔹 Are margins and cash flow improving with that growth?

🔹 Does the company have something competitors cannot easily replicate?

🔹 How much future success is already reflected in the share price?

A brilliant company can still be a poor investment at the wrong price.

And a struggling company can occasionally become an excellent investment if expectations become low enough and execution improves.

📊 How this affects what I am actually doing

This is also why I still have cash rather than feeling like I need to be fully invested immediately.

I currently hold small positions in $NVIDIA(NVDA)$, $Advanced Micro Devices(AMD)$, $Apple(AAPL)$ and $SPDR S&P 500 ETF Trust(SPY)$.

I like having exposure.

But I also like having the ability to add when evidence strengthens or when the market gives me a better price.

I do not need to perfectly catch the bottom.

I would rather miss the first few percent of a move than commit everything because everyone suddenly agrees a stock “has to go higher.”

🚨 The part I’ll be watching next

There is another warning hiding underneath these filings.

Hedge funds increased short exposure to AI-related stocks during July, after the June-quarter positions shown in the 13Fs were already locked in.

That tells me positioning can change quickly.

The next phase could be messy.

We might see great companies fall because expectations were too high.

We might see forgotten companies rally because expectations were too low.

And we might see AI money rotate from headline names into infrastructure, networking, memory, foundries and other parts of the ecosystem.

That is exactly why I do not think the biggest 13F story is what smart money bought.

🧠 My biggest takeaway

Smart money cannot agree on the winner.

And maybe that is the signal.

When almost everyone agrees on the same trade, the opportunity can already be crowded.

When serious investors start disagreeing about where the returns will come from, research becomes more valuable.

That is the market I would rather invest in.

Not blindly following billionaires.

Not blindly buying dips.

Not blindly selling AI.

Watching the evidence, keeping some cash available and making each stock earn more of my capital. 💰📈

🗳️ What would you rather own for the next stage of AI?

A) NVDA and the established winners 🟢

B) INTC and turnaround/value plays 🔵

C) Infrastructure and picks-and-shovels 🟡

D) Keep cash and wait for better prices 💵

Drop your pick below 👇🐯

Disclosure: I currently hold small positions in NVDA, AMD, AAPL and SPY. This reflects my personal investing approach and is not financial advice.