|
Matthias Knab, Opalesque for New Managers: Aswath Damodaran, the NYU Stern valuation professor whose annual data updates have become a fixture of the finance calendar, published a post-mortem on the Situational Awareness collapse on August 10. Readers expecting a takedown will not find one. Readers expecting a lesson about youth and hubris will not find that either, and Damodaran says so explicitly.
Instead he uses the episode to interrogate a word the asset management industry treats as unambiguously virtuous: conviction. His position is that conviction is not obviously a net positive, and that the mechanism by which it destroys capital is well understood and entirely predictable.
The four weeks
The facts are by now familiar. Leopold Aschenbrenner founded Situational Awareness in July 2024, building the fund around the thesis set out in his 167-page essay of the same name: long the companies supplying AI infrastructure, short the software and services businesses AI would disrupt. Early backing came from technology figures, with Jane Street later among the more prominent names involved. A private stake in Anthropic sat alongside the public book.
The returns were extraordinary. Damodaran, citing Portfolios Lab data, puts the fund up roughly 367 percent between August 7, 2025 and its high-water mark on June 19, 2026, with since-inception figures higher still. Assets peaked at around $45 billion at the start of July.
What followed was compressed into four weeks. Semiconductor and AI infrastructure names reversed sharply. Both legs of the book moved against the fund simultaneously. Prime brokers called for additional collateral, forcing sales into a falling market, which generated further losses and further calls. Reporting has put the fund's leverage at as much as 400 percent. By the end of July the public equity portfolio had lost close to two thirds of its value, principal was down more than 43 percent, and Aschenbrenner had agreed to sell most of the fund's listed holdings to Ken Griffin's Citadel in order to settle with lenders. Assets stood at roughly $10 billion. The Anthropic position, being private and unmarked by the daily tape, survived intact.
In a letter to investors reported at the time, Aschenbrenner accepted responsibility, telling them the fund had let them down in July and describing the losses as expensive scars. He has since made a roughly $400 million private investment, the first visible sign of a rebuild.
What Damodaran actually argues
Damodaran's contention is that the three caveats worth raising about Situational Awareness were all available on June 19, at the peak, and required no hindsight.
First, a two-year record is a shooting star rather than a beacon. In market time, separating luck from skill takes decades. Second, returns of that magnitude cannot be produced by being right alone; they require leverage, whether explicit through borrowing or implicit through options. Third, and this is where he is bluntest, the 2 and 20 fee structure not only imposes a long-term handicap on investors but positively encourages reckless risk-taking by the manager.
From there he builds a framework. Conviction, he argues, is a continuum rather than a binary. At one pole sits absolute certainty; at the other sits what he calls "investment mush," a view too weak to voice, let alone fund. Conviction requires belief in three things at once: that your assessment of fair value is better than the market's, that the market will correct, and that it will correct within your holding period.
He then decomposes each. Perceived mispricing can come from private information, superior information processing, better business understanding, or an identifiable pricing error. Whether correction actually arrives depends on whether the instrument has a maturity date, whether market frictions block the correction indefinitely, whether the market is liquid enough to generate catalysts, and how long the investor can wait. True arbitrage, he notes, lives almost entirely in fixed income and derivatives, where maturity guarantees the correction. Equity strategies are pseudo-arbitrage, and the risk never leaves the position.
The third set of factors is about the investor, and two of them will be uncomfortable reading for allocators.
On credentials: the more intelligent the manager and the more exceptional the pedigree, the greater the conviction they tend to bring - because it becomes easier to attribute an apparent market mistake to the stupidity of other participants than to examine whether it is a mistake at all. On track record: early success is among the worst possible teachers, particularly when it arrives accompanied by admiring profiles and inbound capital. Damodaran pointedly declines to add age to the list, saying he does not equate ageing with wisdom.
Truncation risk
Conviction matters, in his account, because it determines exactly two decisions: how large a position is, and how much debt sits behind it.
Leverage magnifies judgment in both directions, which is uncontroversial. The less obvious hazard he identifies is that debt can shorten the investor's effective time horizon by forcing liquidation before the market corrects. He calls this truncation risk, and it is the concept that does the analytical work in the piece. It eliminates not just the position but the possibility of the thesis ever being vindicated. A fund that survives is a fund that can still be proved right.
This is his central objection to Situational Awareness, and it is narrower than the objections most commentators have raised. He takes no issue with the AI thesis itself, conceding that Aschenbrenner knows more about AI than he ever will. He takes no issue with leverage as a tool. His objection is to the specific combination: a macro story facing substantial business, political and regulatory obstacles, funded with maximal debt. Run with less leverage, he argues, the fund would have had a bad July and lived to tell the tale.
He also overlays his corporate life cycle framework. Pricing is inherently less precise for young companies, where value sits almost entirely in the future and catalysts for correction are infrequent. Concentration and leverage are therefore more defensible in mature businesses than in early-stage ones. He applies this to himself: having valued SpaceX at around $100 per share, he says he would buy it as one holding among many but could never have sufficient conviction to make it his largest or only position.
Three lessons, and one that stings
The first is that investment actions inconsistent with the actual strength of your conviction risk ruin. The second is that momentum was the wild card: both sides of the book were aligned with what the market was already pricing, merely concentrated and levered, and much of the fund's rise and fall may be better explained by momentum than by anything specific to AI.
The third is that humble money beats smart money. Damodaran distinguishes between managers who attribute every basis point of outperformance to their own brilliance and those willing to concede that being in the right place at the right time did much of the work. Investors, he suggests, do considerably better with the latter.
Here he closes off the easiest reading of the episode. The temptation is to attribute the overuse of leverage to youth and inexperience. Damodaran points to Long-Term Capital Management, where John Meriwether brought decades of distinguished trading experience at Salomon Brothers and two Nobel laureates to the same outcome. Age was not the variable.
Why it matters for allocators
For anyone conducting manager due diligence, the practical value of the piece lies in its inversion of a standard question. Managers are routinely asked how much conviction they have in a position. Damodaran's framework suggests a better question: does the observable position sizing and leverage match the conviction the underlying analysis can actually support, given the asset class, the maturity of the businesses involved, and the plausibility of a catalyst arriving inside the stated horizon?
That is a testable question. It can be asked at the peak rather than after the fact - which is precisely his point about June 19.
Damodaran ends without malice, expressing the hope that Aschenbrenner returns to fund management, describing him as an original thinker willing to take a stand, and adding that both qualities are scarce among active managers. He hopes the next vehicle comes with a fee structure that gives investors a genuine chance of beating the market.
The full post, with charts and a reading list on the episode, is available at Musings on Markets.
|
RSS
