Tweet
1/ Very rough and speculative sketch of what I increasingly think happened at FTX as more info comes out…

The central question is where did the money go? Yes malfeasance and fraud is necessary, but at one point in the cycle cash actually has to go out the door
2/ At 3AC we knew it went to losses on leveraged long positions. At Lehman it went to bad mortgages. At Enron it went to boondoggle mega projects.

Some FTX money obviously went to seed rounds in bad or illiquid projects. But AFAICT nowhere near enough to explain the hole.
3/ Let’s rewind to 2017/18. Alameda the prop firm is a big fish in a little pond. They’re mediocre traders (there’s a video of SBF bragging about how their quoter latency is down to something like two seconds). But crypto is still a weird asset class that most won’t touch
4/ But inefficient markets rarely stay that way. As more actual pros enter the space their alphas are starting to decay to zero.

It happens. It’s a self-contained process. You size down and eventually shut off. You don’t lose (much) money, you just stop making money and move on
5/ I think at this point Alameda hatches a much more audacious exit plan. (Linear utility is a hell of a drug.)

The prop trading operation as an independent entity won’t survive. But it’s liquidity can be used to bootstrap a lucrative consumer facing crypto exchange
6/ The conviction is probably reinforced by rising tech valuations of and the falling multiples on trading firms around 2019.

Alameda will continue as a market maker, but instead move liquidity to its own exchange. Why make another CEX owner rich on the back of your trading?
7/ Sam has now turned a -EV trading strategy into an exchange valued at tens of billions based off growth metrics being basically generated by Alameda itself. Riding off nosebleed valuations and VC demand for crypto tech exposure.
8/ At this point, nothing here is terrible. Plenty of businesses run a loss leader type strategy to generate growth.

Where I believe this goes off the rails is that Alameda’s strats become a lot more -EV harder or faster than they anticipated.
9/ Instead of losing a little money trading every day, they’re losing a lot. But unlike an independent prop firm that would just turn off the quoter, they can’t get off the bus.

FTX valuations are dependent on Alameda liquidity. And Alameda liquidity is burning money.
10/ Supporting. Price discovery was almost entirely dominated by Binance, implying FTX liquidity was getting picked off left and right.

Two, trading firms made a killing in 2021, implying someone was on the other side.
11/ Three, trading firms seem to be disproportionately exposed to FTX losses. Implying that a lot of those halcyon trading firm profits were occurring at that venue.
12/ Alameda made the mistake of assuming it was operating in a random environment when it in an adversarial ones

But the market is filled with sharks The more you expose money losing behavior in size, the more the market will adapt to exploit you, despite what your sims say
13/ I think most of the Solana sphere was basically Alameda’s DeFi version of the same strategy. And why they were funding every Solana project that came along like drunken sailors.

Flood -EV liquidity and trading across the chain to pump your massive SOL alloc
14/ At one point or another Alameda lost more than it could internally fund. At that point it had to dip into FTX reserves.

This probably seemed innocuous at first. After all SBF, on paper was worth near $100bn between FTX and Solana at peak market cap. What’s a few million USD
15/ But crossing this rubicon was the point of no return. If Alameda ever stopped running the quoter, FTX liquidity would collapse, people would pull reserves (if not just because the easy alpha would disappear).

No matter how much Alameda loses, FTX has to keep supplying it
16/ Imagine a casino where all the games were +EV, except the house doesn’t actually have money to cash out the chips. Nobody would notice because instead of cashing out, they’d just keep pouring bee funds into the money machine.

FTX was a ponzi targeting trading firms
17/ Crypto trading firms had the illusion they were skilled. Instead all they were doing was winning against the Washington Generals.

All of it stops working if Alameda ever turns off the quoter and donating money to the market everyday
18/ I believe that’s where “the money went”. Mid-tier trading firms who struggled everywhere else but found FTX highly profitable

They made huge profits, but ironically many have lost even more because it seemed like a money machine, so why not pour even more capital in
19/ This coincides with a personal observation that a lot of crypto trading firms that were successful in 2021 seemed like larps that barely knew what they were doing.

This made sense in 2017 when it was crypto natives only. Big fish little pond.
20/ But how can all these crypto native firms make money now that we know the most sophisticated and skilled TradFi HFT firms are in the game?

The answer seems to be that Alameda was a giant money faucet.
21/ A ponzi works best you trick the marks into thinking the profits are because of how smart they are.

Very few people will ever stop and think “wait, there’s no way I’m smart enough to be making this much money”
22 (Addendum). Many smart people have pointed out that it’s not just the liquidity but probably pay also the liquidations which Alameda had to handle to keep FTX alive.
23/ Normally liqs are easily profitable layups. But if you have mediocre trading systems, a big liquidation with a lot of toxic one way flow in a fast market is bad

Seems likely that Luna pushed them over the edge, not because of prior exposure but just the FTX liquidations

Recommended by
Recommendations from around the web and our community.

Good thread. Rings true.

Excellent discussion on what may have happened at FTX, how it found success, how it lost money, and where the money went.