The essay said chips were underpriced. The fund built $7.4bn of puts against them in a single quarter.

In April 2024 OpenAI dismissed a 22-year-old researcher from its safety team over an alleged leak. Two months later Leopold Aschenbrenner published Situational Awareness: The Decade Ahead, an essay arguing that artificial intelligence was about to consume more electricity and capital than almost anyone had planned for. It was read across the industry. Then he did the thing essayists almost never do: he raised a fund, named it after the essay, and bought the argument.

Bought it very literally. In the portfolio it reported for 31 March 2026, the fund’s six largest holdings were Bloom Energy, SanDisk, CoreWeave, IREN, Core Scientific and Applied Digital — a fuel-cell maker, a memory maker, and four companies whose business is putting computers in buildings and buying the power to run them. Those six were 84.8% of everything it owned.

The one sentence in the essay about buying anything

Fifty thousand words, and one of them is investment advice: a footnote in the chapter titled “Racing to the Trillion-Dollar Cluster.”

“What all of this means for NVDA/TSM/etc. I leave as an exercise for the reader. Hint: Those with situational awareness bought much lower than you, but it’s still not even close to fully priced in.”

Chips are cheap, and the people who understand this already own them. That is the essay’s entire view on what to own.

What the fund held

American investment managers above a size threshold file a form called a 13F once a quarter, listing what they held on its last day. It reports shares and options held long. It never reports ordinary short sales and never reports borrowing. For an option line it reports the value of the underlying shares the option references — not what the option cost, not what it is worth, and not the amount anyone stands to lose.

Here is Nvidia, across every filing the fund has made:

No Nvidia long line appears in any of the six except one: 2,855 shares, about half a million dollars, in the same filing as the $1.57bn put.

Nor was it only Nvidia. Counting AMD, ASML, Broadcom, Intel, Micron, Nvidia, TSMC and the VanEck semiconductor ETF, the underlying value of the fund’s semiconductor puts ran:

Read the last two rows twice. At the end of December the fund held no semiconductor puts at all. Ninety days later it held puts referencing $7.36bn of semiconductor stock, and the largest single line in the whole portfolio was a $2.04bn put on a semiconductor ETF.

The ninety days

That is the finding, and it is bigger than the reversal, because two other things happened in the same window.

First: the long column stops moving, and that is not what it looks like. $3,914m to $3,856m is a market value, not a measure of buying — it can sit still while everything underneath it changes. Here it did. Eight positions in the December filing are absent from the March one. Ten appear that were not in December. Of the sixteen names in both, nine carry more shares and three fewer — Bloom Energy, the largest holding, down 35.6% of its shares, while CleanSpark rises 648%, Bitfarms 188%, Bitdeer 92% and Riot 87%. Two quarter-end photographs cannot say when any of that happened, or whether a name left and came back. What they do say is that twenty-six of the book’s lines changed between them, and share counts are the one field on this form that is a count rather than a valuation.

Second: the protection inverted. At the end of December a $3.9bn book of memory, power and cloud equities had one put behind it — $8.9m of Infosys, an Indian IT services company in none of the businesses the rest of the book was in. Ninety days later the put column was 219% of the long column.

That ratio divides one quantity by a different one — shares referenced by puts, over the market value of shares owned — so it is not hedge coverage, not net exposure and not capital at risk, and nothing here should be read as any of the three. It is worth reporting for one narrow reason: whatever the puts were for, at $9m they were too small to matter to a $3.9bn book, and at $8.46bn they were too large to have been sized against it.

So one quarter holds all three: the wholesale turnover behind a flat market value, the swing from almost no protection to twice the book, and $7.36bn of puts appearing against the sector the essay had called underpriced. Not a drift across five quarters. One filing to the next.

What happened next

From 31 March, the quarter-end date of that filing, to 30 July:

Every one rose. One name in the put book fell: Oracle, referencing $1.07bn, down 13% — and Oracle appears nowhere in the essay.

In July the fund lost 67% of its value, by its own account to partners, and traded a block of its portfolio in the small hours of 30 July to remove all its leverage. It reports the year still up about 80%. Those are the manager’s own unaudited figures; the price moves above are not, and neither set explains the other.

Where this reading is weakest

A put may be an outright bet or a hedge against something the form does not show, and the filing cannot tell you which. The fund may have been short in ways no 13F reports. A quarter-end photograph filed up to 45 days late is a poor instrument for reading anyone’s conviction, and this one was filed on 18 May, seven weeks after the date it describes.

There is also a reading in which none of this is a contradiction. The essay argues about what will be built. It does not argue about who captures the margin in building it, and those are different questions: compute demand can be enormous while the companies selling the compute are badly priced. The essay only answers the first. On that reading the put book is not a reversal but the second question finally being asked, by someone who had never publicly asked it.

What the record establishes is narrower and is not in dispute. The fund’s one public statement about what to own said chip stocks were underpriced. In a single quarter, the fund named after that statement took put positions referencing $7.36bn of chip stock. The document and the book say different things.

And a caution about a number in circulation. The gross value of the March filing is $13.68bn, which is not the size of the fund: $8.46bn of it is put notional — shares the options reference, not money anyone put up. The only line that is unambiguously the market value of shares owned is the $3.86bn of long equity. Adding the two produces a figure that means nothing, and several published analyses have added them.

What would settle it

The 13F for the quarter ended 30 June is due by 14 August, and has not been filed. It will show whether the semiconductor puts were still there at the end of June or had gone the way the December ones went, whether the turnover continued, and where the concentration sits.

It cannot settle what the puts were for; no 13F can. But if it shows them gone by 30 June and the long book rebuilt in the same names, the reading here is the weaker one: what looks like a reversal of view was a single quarter’s position, and it closed. That is the version this piece loses to, and it is checkable by anyone, on a date, from a document.

Every filing behind this piece is laid out, priced forward against our own sessions, on the Situational Awareness fund page.

Sources

Position figures are summed from the position lines of six 13F-HR filings by CIK 0002045724, with the raw SEC documents archived alongside; every period reconciles to the filing’s own stated total. Share-count comparisons are matched on CUSIP between the 2025-12-31 and 2026-03-31 filings.

The March 2026 filing is accession 0002045724-26-000008, period 2026-03-31, filed 2026-05-18. As of publication no 13F-HR has been filed for the quarter ended 2026-06-30.

“Semiconductor” above means AMD, ASML, Broadcom, Intel, Micron, Nvidia, TSMC and the VanEck semiconductor ETF; CoreWeave, Oracle, Infosys and Corning are excluded. Prices are as-traded closes from our point-in-time archive, 2026-03-31 to 2026-07-30, unadjusted for splits.

The essay is quoted from situational-awareness.ai, chapter IIIa, “Racing to the Trillion-Dollar Cluster,” retrieved 2026-08-01. Performance figures are unaudited and as stated in the manager’s 30 July 2026 letter to partners.

Long stubs of a few thousand shares sit beside eight of the eleven put lines in the March filing, between 0.03% and 5.6% of the put notional; they are a habit of the book rather than a position.

A 13F reports long positions only, is filed up to 45 days after the quarter it describes, reports the value of the underlying rather than a premium or strike for option lines, and never reports leverage — so gross 13F value is not assets under management and is not treated as such here. Not investment advice.