Investors in Situational Awareness deserved to lose their shirts
economist.com
economist.com
Mind you I don't particularly care about Aschenbrenner or his fund but it feels to me like typical journalists reporting on hindsight without any sort of skin in the game. "Hhm, how could you be so dumb so as to trust your money to a 22 year old".
Peter Thiel did the same with Zuck and like him or hate him he did pretty well.
If you want to show how smart you are hash your predictions, post them publicly with a commitment to reveal by a certain date and then show the world how this was so obvious in hindsight.
>> What, then, to make of Situational Awareness’s plunge? ...Mr Aschenbrenner was hardly alone in such bets; over 80% of fund managers responding to Bank of America’s latest monthly survey named “long global semiconductors” as the most crowded trade. The big worry is that this trade’s whiplash-inducing reversal will have set other investment firms teetering, too.
If Citadel hadn't stepped in, Situational Awareness may have had to fire sell tens of billions of dollars of assets. Assets others have leveraged positions in. That, in turn, could have triggered margin calls and a potential credit or even bank crisis.
I'm still trying to figure out what their gross and net positions were. But taking reported figures at face value, yes, $45bn is more than enough to start a credit crisis, particularly if everyone is crowded into a small set of leveraged positions that begin fire selling.
The piece is published in a weekly magazine. It may include some interesting things that nobody knew one week ago.
Its purpose is entertainment
It highlights the age of the Situational Awareness founder and what he was doing before he started the fund
"There is nothing breaking or remotely interesting which we didn't already know."
It's a weekly column not "breaking news"
The column author shares his perspective on current events and usually reflects on the past. It's unlikely that readers "already know" which current events the columnist will select, what he will write about them, including which comparative history he might reference. As such, it does provide something readers "don't already know", namely, the columnist's views on the current event he has selected. As for "remotely interesting", that's an opinion. Surely this column has a readership that finds the author's writing "remotely interesting"
The Economist does at times try, incorrectly, to predict the future
But that's not what is going on in this column. It's looking backward, not forward
If the parent comment is coming from a 20-something recipient of Silicon Valley VC funding and hoping to profit from an "AI" startup, then perhaps a bias disclosure is in order
If, hypothetically, no predictions were ever shared with the public over the internet, what would that do to Silicon Valley. What would happen to the volume of HN submimssions/comments
Of course there is no way to know unless and until it occurs
He's up, but he's wiped out because he was 4x leveraged, why was he 4x leveraged? Because he's a dumb kid who doesn't understand that markets have down days or how margin calls work. The same failure mode doesn't happen with facebook or rather, when giving money to a hedgefund in a bull market you expect at least some returns or that they will actually hedge. Not fuel a mid 20s gambling addiction. As of now most of the cash is in a private position with the worst of the AI companies.
If you want to show how smart you are don't waste your time predicting the future, since it's a fools errand. The belief that intelligence makes you better at prediction ironically makes you stupid.
> which posted a loss of 67% in July
So it's still up 44% this year? The article notes this, but seems unnecessarily adversarial against an investor who is still wildly successful.
It's also interesting that Citadel is mentioned everywhere as the buyer, as if it's strange that a huge market maker firm would be involved in a large forced sale.
Seems the fund also has Anthropic shares, so it's not like their entire portfolio got margin called.
(Spicier claim: When Citadel bailed out Melvin at the hype of the GME craze, that was also just business as usual, not a conspiracy as the redditors believed.)
Another way to parse it when they put "Citadel" in their headlines, is that they're trying to communicate to knowledgeable investors that there's no need to panic. The book is in experienced hands now, with enough capitalization to weather any further attacks, and the contagion probably won't spread.
Devil is in the details of what Citadel paid for its positions, whether there are any performance tails/clawbacks, how and when the GP charges fees, et cetera.
Aschenbrenner is almost certainly up. I'd be surprised if his median LP is breaking even.
Does anyone know if this refers to AUM or the gross size of his positions after leverage?
Oversee: to be in charge of how an amount of money is invested or spent.
No. "Oversee" can refer to the gross or net position. If I raise $10bn and use that to buy $100bn in assets, I absolutely control and thus oversee the latter.
The leverage was 4x or 5x of that and highly concentrated. Citadel bought $16 billion public portfolio with 10% discount.
Do you have a credible source specifically claiming this? I want to know whose data they're looking at. (Specifically, I want to unpack how they're marking their private positions.)
I think this piece is pointing out two things:
That this bubble still has a long way to deflate if it is indeed deflating. The numbers involved truly are spectacular and unwinding could be catastrophic.
That there is a lot of blind faith in the market in unsubstantiated predictions - like AGI this decade (or indeed AGI at all based on LLMs), and a lot of follower behaviour among investors chasing the latest trade.
The Economist should change its name to The Populist with this type of screed.
Here's the reality:
1. Lots of money is invested willy-nilly for all sorts of reasons (or no reason at all) in all sorts of dubious and poorly-managed vehicles.
2. Except for the profile of the fund manager and that this was AI, this failure wasn't all that special. The seed capital for the fund is reported to have been around $225 million and it was highly leveraged on top of significant gains so there's a whole class of LPs who probably aren't even going to lose anything on this. It's the people who put capital in more recently who are going to lose.
3. The people who will lose money investing in this fund are the type of people who have a full wardrobe. It's going to be a very long time before they "lose their shirts".
>> people who will lose money investing in this fund are the type of people who have a full wardrobe
If Citadel hadn't bailed him out, Situational Awareness very likely could have triggered a cascade of fire sales. That would not only hurt ordinary investors' positions, it could have also triggered a credit or even banking crisis (depending on how the margin loans are held).
A big, leveraged, concentrated fund blowing up isn't novel. But it's the traditional way to start a recession.
Citadel didn't bail him out/rescue him. This was a distressed purchase, made at auction. Citadel believes it will eventually make money on the book, which was only around $16 billion.
If Citadel and Millennium (the other bidder) hadn't stepped in, the prime brokers would have liquidated the collateral themselves. The collateral here was mostly liquid mega-cap semi stocks. Probably the best and easiest collateral to deal with. And the exposure was split across three of the most highly capitalized banks (BofA, Goldman and JPMorgan) and they were already managing the wind-down when Citadel stepped in opportunistically.
If AI is going to take down the market, it isn't this. It's all the private credit exposure that isn't getting marked to market daily.
...you're describing a bail-out. The LTCM lenders and Fed in the financial crisis made money on their books. They were still extending liquidity.
> the prime brokers would have liquidated the collateral themselves. The collateral here was mostly liquid mega-cap semi stocks
Yes. By fire selling. Which triggers, in turn, further margin calls and potential failures.
I'm not saying it's a certainty. But claiming there is limited financial risk from any $10+ billion leveraged meltdown is absurd.
> If AI is going to take down the market, it isn't this. It's all the private credit exposure that isn't getting marked to market daily
Private markets move slowly. That gives time to mop up messes. Crises metastasize when they hit public instruments precisely because that's where the most volume, leverage and potential for panic live.
Sorry, but this is just not accurate.
LTCM was a coordinated recapitalization. The New York Fed brought together 14 institutions who put over $3.5 billion of their own capital into the fund in exchange for about 90% of the fund's equity, and then they wound the book down slowly over the following year.
Here, Citadel, without any regulator involvement, made an arm's-length purchase of Situational Awareness' assets at a discount in a competitive auction. It put no money into the fund.
Also, you need to look at what LTCM was versus what Situational Awareness is: LTCM was invested in fixed-income securities with highly-illiquid derivatives exceeding $1 trillion. Its leverage was 25x.
Situational Awareness was a long/short equity fund with 4x leverage in daily-marked, over-collateralized margin accounts, held against liquid large-cap semis, with three well-capitalized prime brokers who saw the failure coming and were prepared to manage the collateral.
> I'm not saying it's a certainty. But claiming there is limited financial risk from any $10+ billion leveraged meltdown is absurd.
You're failing to make a distinction between this fund melting down and the correlated risk. Situational Awareness made leveraged bets in highly crowded trades. While forced selling into an already-falling market can be problematic, it's a second order effect. A symptom, not the cause.
> Private markets move slowly. That gives time to mop up messes. Crises metastasize when they hit public instruments precisely because that's where the most volume, leverage and potential for panic live.
That's not how it works. Slow marks to market don't give you time to mop up messes. They are what allow institutions to keep lending against untested valuations.
And you have it backwards on volume. Volume isn't what makes crises metastasize. It's what lets them resolve. Bad public investments can be cleared in a block trade while private ones have no exit that doesn't set a mark for everyone still holding. This is why 2008 started in mortgage credit that was being carried at model prices. It didn't start in public equities.
If you look at the history of deleveragings, the ones that stay contained are the public ones.
> While at Columbia, he co-founded the university's effective altruism (EA) chapter.[5]
Just another smart, yet ethically challenged, individual. Nothing to see here.
As a brand it's tainted. You might argue unfairly, but that's never mattered in "public brand perception" before.
This is a joke of an article. It shows a few incidental deaths while ignoring the total number of saved lives due to EA. It literally points to ~2 incidental deaths while Give Well saves ~300k lives. Its remarkably ignorant.
Is there a word for this kind of fallacy? Cherry picking sounds too Reddit-coded.
In short, Effective Altruism says you don't engage in charity until you have sufficient means to actually make a difference. Basically, you leverage capitalism to accumulate enough wealth to actually be able to solve a big problem, then you go all in on solving that problem. Sounds great right?
In the real world these people never accumulate enough wealth to actually start solving problems. Even when they become billionaires. In practice it is a convenient excuse to act like the most depraved capitalist while still pretending to sit on the high horse and talk down to people who simply try to help with the means available to them.
https://www.youtube.com/watch?v=rQZWMmINqxo
https://www.youtube.com/watch?v=S54GrXDjokg
https://www.thephilosopher1923.org/post/a-mirror-for-tech-br...
https://www.persuasion.community/p/the-problem-with-effectiv...
In practice, it can be used as a fully-general counterargument to supporting most charities, since decisively and completely solving problems in this space is rare. It’s harder to split a problem into component parts than it is to throw up one’s hands and say, “well, this isn’t tractable.”
> You can only throw up your hands and say "this isn't tractable" if tractability is 0, i.e. the charity doesn't help at all, no matter how much money you spend on it.
Even I don’t believe this. You’ve gone too far in the other direction and made tractability a useless metric.
Because almost everyone prominent in the movement has had some combination of lack of judgment and morals. There various terms for rationally pursuing philanthropy; associating with effective altruism involves embracing its ideas, people and culture.
Going to need a good source for that crowd having done more good than the frauds of even just their most-notorious members.
Masquerading as an altruist to justify massive management fees for running a fund of concentrated, leveraged positions is one of the oldest lipsticks on financial pigs.
And it does not come cheap. Someone who says "I will not work for Facebook" is taking a stand at the expense of significant personal gain.
EA has many tenets, some of make sense per se, but collectively–particulalry when taking into account that EA isn't just an idea but a movement of people with a distinct culture and track record–probably do more harm than good.
Well, for one, the frequency of its members committing mass fraud.
Like, if a church preaches kindness and love but its preachers and members are constantly out there murdering people, it's a bit silly to point to the text alone when evaluating the organisation.
Like, if a church had 100,000 members, you would indeed expect them to commit a few murders each year. And GiveWell alone claims to have induced 150,000 people to donate... https://www.givewell.org/default/citations#Number_of_donors
Another branch is "I make lots of money at a job that's probably making the world a worse place. How do I reframe this to not feel like a bad person?" They do some math and conclude that it's net more good for them to earn a bunch of money and donate some of it than it is to volunteer or it is to work at a job that isn't making the world a worse place. This is an awfully convenient conclusion as it means that all they have to do is pay a tithe and make no real changes to their life.
The notion of making money specifically for the sake of funding good things isn't a completely terrible one, but a vocal portion of the people who espouse it are very obviously not primarily motivated by maximizing the good they do.