Who Are the Hot Investors of 2026? What Nine Real Q2 13F Filings Actually Show

📊 Who Are the Hot Investors of 2026? What Nine Real Q2 13F Filings Actually Show

Welcome back, everyone! 📊 I’m your Quant Analyst, filtering out market noise using data, statistical modeling, and systematic insights. 👩‍💻✨

Every few months a fresh crop of “investors you must follow” lists appears, and they are all assembled from the same nine or ten names. So instead of writing another one from headlines, I downloaded the actual Q2 2026 Form 13F filings from the SEC and parsed them myself. To cut straight to the chase: the filings say something the lists never do — the “smart money” is not a herd with a signal, it is a crowd that openly disagrees — and by construction you read every one of its trades between 45 and 135 days late.

▲ Everything the public knows about these portfolios arrives as paperwork, on a delay, and with a lot left out of the form

---

📌 First, the actual numbers

Every figure in this post comes from the raw XML of Form 13F-HR filings, not from a tracker site. For the quarter ended June 30, 2026 (filing deadline August 14, 2026), Berkshire Hathaway reported 89 table entries totalling $299,253,556,246. Pershing Square reported $19.47 billion across 14 positions. Michael Burry’s Scion Asset Management reported nothing at all — its last 13F covered September 30, 2025.
Source: SEC EDGAR, Form 13F-HR filings, filed 2026-08-14.

Hold on to that August 14 deadline. It is 45 days after the quarter closed, and it is the single most important number in this entire post.

1. The list, with real numbers attached

Here are nine managers whose names dominate the “hot investor” conversation, with what they actually disclosed. The dollar column is the reported value of US-listed long equity positions only — I will come back to why that caveat matters more than it sounds.

Manager	Principal	Q1 2026	Q2 2026	QoQ	Issuers
Berkshire Hathaway	Warren Buffett / Greg Abel	$263.10B	$299.25B	+13.7%	26
Coatue Management	Philippe Laffont	$29.06B	$48.63B	+67.4%	63
Tiger Global	Chase Coleman	$22.85B	$23.98B	+5.0%	46
Pershing Square	Bill Ackman	$13.71B	$19.47B	+41.9%	14
ARK Investment Mgmt	Cathie Wood	$12.86B	$15.40B	+19.8%	184
Altimeter Capital	Brad Gerstner	$5.70B	$9.83B	+72.5%	20
Appaloosa	David Tepper	$5.93B	$7.47B	+25.9%	25
Third Point	Daniel Loeb	$2.08B	$4.65B	+123.3%	38
Duquesne Family Office	Stanley Druckenmiller	$2.94B	$4.35B	+48.2%	85

Notice that every single one grew. That is not nine correct calls; it is mostly one rising market lifting nine books at once. Any list that presents a rising 13F value as evidence of skill is measuring the market, not the manager.

2. What they actually agreed on — and why you were last to know

Comparing new positions across all nine filings, two names dominate. SpaceX (SPCX) was initiated by six of the nine — Coatue, ARK, Altimeter, Tiger Global, Appaloosa and Third Point — for roughly $4.37 billion combined. Cerebras (CBRS) was initiated by four — Altimeter, Coatue, Tiger Global and ARK — for roughly $3.98 billion. Nothing else was bought fresh by more than three.

That is a real, measurable consensus, and it is worth knowing. Here is the problem with acting on it: both of those companies IPO’d inside the quarter being reported. Cerebras listed on May 14, 2026 and rose about 68% on its first day. SpaceX listed on June 12, 2026 and closed its debut up roughly 19%. The filings that revealed this crowd were published on August 14.

The 13F Disclosure Lag, Q2 2026

*Calendar arithmetic, not a forecast. A Q2 trade could be executed any time between April 1 and June 30, and all of them were disclosed on the same day, August 14. The reader cannot tell which end of that range any given position came from.

The mechanism is worth stating plainly, because it is the whole reason 13F copying underperforms the funds it copies. A manager can buy on April 1 and disclose on August 14 — a 135-day lag. Or buy on June 30 and disclose 45 days later. The form does not say which. So the expected age of any position on the day you read it is about 90 days, and today, at the end of August, closer to 107 days. Worse, the information is censored: a position opened in April and closed in July never appears at all, because 13F is a snapshot of one date, not a record of activity.

3. What they violently disagreed on: Alphabet

Eight of the nine managers held Alphabet at quarter end — the single most widely held name in the group, about $43.6 billion across the sample. If crowding were a signal, this would be the loudest one available. Then you look at the direction each of them was moving, and it falls apart.

One Stock, One Quarter, Opposite Directions

*Change in Alphabet Class A + Class C common share count between the Q1 and Q2 2026 filings. Duquesne and Altimeter are excluded because both went from zero, which has no percentage.

Let me argue against my own chart before anyone else does. Percentages flatter small positions and hide scale, and the two most extreme bars are the least meaningful ones. Third Point’s +486% is an increase of 850,000 shares — roughly $300 million. Berkshire’s +83% is 48.1 million shares, something on the order of $17 billion. Those are not the same act, and the chart draws Third Point’s bar six times taller. Pershing Square’s −100% is a complete exit, but of a position that was only about $120 million to begin with, which for a 14-name portfolio is closer to a rounding error than a statement.

The genuinely large move is Berkshire’s, and it is worth seeing as a sequence rather than a single quarter:

Quarter ended	Alphabet shares	Reported value	Move
Jun 30, 2025	0	—	no position
Sep 30, 2025	17,846,142	$4.34B	initiated
Dec 31, 2025	17,846,142	$5.59B	held flat
Mar 31, 2026	57,835,013	$16.63B	3.2x
Jun 30, 2026	105,979,600	$37.76B	1.8x

Alphabet went from nothing to 12.6% of Berkshire’s reported equity book in four quarters — its third-largest holding, behind only Apple and American Express and now ahead of Coca-Cola. One quarter is noise; a position tripled and then nearly doubled again, over three consecutive quarters of buying, is a decision. And it happened in the same 90 days that Bill Ackman sold his last share and Chase Coleman cut his by 45%. Two of the most closely followed investors alive looked at the same company at the same time and did the exact opposite thing. There is no version of “follow the smart money” that survives that.

4. Five things the form does not tell you

Parsing the raw filings rather than reading a summary site surfaced four problems I had genuinely underestimated, plus one I already knew about.

• It is long US equity only. No short positions, no bonds, no cash, no foreign listings, no private holdings. Ackman’s $19.47 billion is his 13F-reportable slice, not his fund. Every “% of portfolio” figure you see quoted — including the ones in this post — is a percentage of a partial book.
• Options are reported at notional, not at cost. Scion’s final filing showed $1.31 billion of puts and calls against a long book of just $68.1 million. The headline number everyone quoted — a ~$1.1 billion bearish bet — is $912.1 million of Palantir puts (5,000,000 underlying shares) plus $186.6 million of Nvidia puts (1,000,000 shares). That is the value of the underlying stock. The premium actually at risk is not disclosed anywhere on the form, and for out-of-the-money puts it can be a small fraction of notional.
• The units are not enforced. This one genuinely surprised me. Berkshire filed its Q2 2026 values in whole dollars ($299,253,556,246). Duquesne filed the same quarter, on the same form, in thousands ($5,210,860, meaning $5.21 billion). Baupost did the same. Sum the raw value field across filers without checking and you are off by a factor of 1,000 on some of them.
• The filer identity can move. For Q2 2026, “Pershing Square Capital Management, L.P.” — the entity most trackers have bookmarked — filed a 13F-NT, a notice stating it had nothing to report. The actual holdings appeared under a different registrant, “Pershing Square Inc.”, with five affiliated entities all filing notices pointing elsewhere. Watch the old identifier and Ackman’s quarter looks empty.
• The filer can simply leave. Michael Burry is still the most-quoted name in any “hot investor” conversation, and he has not filed a 13F in three quarters. Scion’s last one was submitted November 3, 2025; it deregistered with the SEC on November 10, 2025 and no longer has to file. Whatever anyone tells you Burry is holding today, it did not come from a filing.

5. So what is a 13F actually good for?

Not for copying trades — the lag arithmetic in section 2 settles that. But I read them every quarter anyway, for three narrower purposes:

1. Idea sourcing, not idea acceptance. Six funds initiating SpaceX is a reason to go read SpaceX’s own financials. It is not a reason to buy SpaceX. The filing is where the research starts, not where it ends.
2. Measuring crowding. This is the use most people have backwards. When eight of nine managers hold the same name, that is not confirmation — it is a warning that the marginal buyer may already have bought, and that a disappointment will find a lot of the same people reaching for the same exit.
3. Auditing your own reasoning. The valuable question is never “what did they buy?” It is “someone who had to commit real capital reached the opposite conclusion from mine — what do they see that I don’t?” The Alphabet split is a gift for exactly this. Somebody in that chart is wrong.

If you want the mechanics behind the disagreement rather than the tally, my Alphabet fair-value report walks through the forward-EPS and multiple inputs a buyer and a seller would have to disagree about, and this piece on the semiconductor cycle covers why the picks-and-shovels names in these filings — Lam Research, Applied Materials, Micron — behave differently from the chip designers.

Quick FAQ

Q. Can I just buy whatever Berkshire bought and hold for years?
You can, and the lag hurts a buy-and-hold copier far less than it hurts a trader — that is a fair point in the strategy’s favour. But you still inherit two problems: you don’t know Berkshire’s cost basis, and you don’t know their sizing relative to a balance sheet that includes an insurance float and a very large cash position. Owning 12.6% Alphabet inside Berkshire’s structure is a completely different risk than 12.6% inside yours.

Q. Which of these managers is actually the best?
A 13F cannot answer that, and I would distrust anyone who says it can. The form has no cost basis, no shorts, no cash and no fund-level returns, so you cannot compute performance from it — only position changes. Ranking managers from 13F data is one of the most common mistakes in this whole genre.

Q. Should I buy Cerebras or SpaceX because these funds did?
I can’t answer that for your portfolio, and no honest analyst can. What I can point out is that both were newly listed inside the reporting quarter, both moved sharply on debut, and you are seeing the disclosure 63 to 92 days after those listings. If you want to own them, the case has to come from the businesses, at today’s price — not from a filing describing what someone else paid at a price you can no longer get.

💡 Quant Strategy & Takeaways

The nine most-watched investors of 2026 agreed on two IPOs that had already re-rated before you could see the filing, and split violently on the one stock eight of them owned. Treat a 13F as a 90-day-old, partially censored research lead — never as a trade signal.

When market volatility spikes, remove emotion and focus strictly on the numbers! 🤖

Do you track 13F filings yourself — and have you ever found the disagreements more useful than the consensus? Let me know in the comments! 📈✨

Disclaimer: This article analyses public regulatory filings for educational purposes and is not financial or investment advice. Position data reflects holdings as of June 30, 2026 and may have changed substantially since. Always do your own research or consult a licensed financial advisor before investing.

Disclaimer: Educational content only — not financial advice. Read the full Disclaimer.