What happened? SBF published a comprehensive article to review the Alameda and FTX crash process

summary

In mid-November, FTX International was actually insolvent. In the final analysis, FTX's collapse was somewhere between Voyager and Celsius.

Three things combined to cause the implosion:

a) During 2021, Alameda's balance sheet grew to approximately $100 billion in net assets, $8 billion in net borrowings (leverage), and $7 billion in liquidity on hand.

b) Alameda failed to adequately hedge its market risk. During 2022, a series of large-scale broad market crashes occurred in the equity and cryptocurrency sectors, causing the market value of its assets to drop by approximately 80%.

c) In November 2022, an extreme, rapid, and targeted crash facilitated by the Binance CEO left Alameda insolvent.

Alameda’s collapse then spread to FTX and elsewhere, similar to how Three Arrows, etc. eventually affected Voyager, Genesis, Celsius, BlockFi, Gemini, and others.

Nonetheless, a very substantial recovery is still possible. FTX US remains fully solvent and should be able to return all customer funds. FTX International has billions of dollars in assets, and I have dedicated nearly all of my personal assets to my customers.

illustrate

This article is about the solvency of FTX International.

This has nothing to do with FTX US, as FTX US is fully solvent and always has been. When I turned FTX US over to Ray chapter11, it had net cash on hand in excess of customer balances by approximately $350 million. Its funds and customers are segregated from FTX International. Ridiculously, FTX US users have not yet fully recovered and gotten their funds back. This is a record of the FTX US balance sheet at the time of my filing:

FTX International is a non-US exchange. It operates outside the US, is regulated outside the US, is incorporated outside the US, and accepts non-US clients. (In fact, it has its primary headquarters in, operates from, and is incorporated in the Bahamas, under the name FTX Digital Markets LTD.) US clients are on the (still solvent) FTX US exchange.

The Senators expressed concerns about potential conflicts of interest at Sullivan & Crowell (S&C). Contrary to S&C’s statements that they have a “limited and primarily transactional relationship with FTX,” S&C was one of the two primary law firms for FTX International prior to bankruptcy, and is also the primary law firm for FTX US. FTX US’s GC (government communication) comes from S&C, they work with FTX US on its most important regulatory applications, they work with FTX International on some of its most important regulatory issues, and they work with FTX US on its most important transactions. When I visit New York City, I sometimes work out of S&C’s offices.

S&C and GC were the main parties, threatening me to appoint a candidate of their own choosing as CEO of FTX – including a solvent entity of FTX US – and then file Ray chapter11 to select S&C as the advisory debtor entity.

Despite being insolvent, and despite processing approximately $5 billion in withdrawals in the last few days of operations, FTX International still retains a significant amount of assets - approximately $8 billion in different liquid assets as of the time Ray chapter11 took over. On top of this, there are many potential funding proposals - including an LOI signed after Ray chapter11 submitted, totaling over $4 billion. I believe that if FTX International is given a few weeks, it could probably raise enough funds using its illiquid assets and equity to make its customers essentially whole.

However, as the S&C pressures FTX into a chapter 11 filing, I worry that these avenues may have been abandoned. Even now, I believe that if FTX International were to relaunch, it would be possible for customers to become essentially whole.

While FTX's liquidity began to rely primarily on Alameda in 2019, it had diversified significantly by 2022, with Alameda falling to around 2% of FTX's trading volume.

I didn’t steal funds, and I certainly didn’t hide billions. Almost all of my assets were and are available to support FTX customers. For example, I have offered to contribute nearly all of my personal shares in Robinhood to customers — or 100% — if the Chapter 11 team is willing to honor my D&O legal fee recovery.

Both FTX International and Alameda are legitimate and independently profitable businesses in 2021, each making billions of dollars.

Then, Alameda lost about 80% of its asset value during 2022 amid a series of market crashes — as did Three Arrows Capital (3AC) and other crypto firms last year — and has since seen its assets fall from targeted attacks. FTX was affected by Alameda’s decline, as Voyager and others were earlier affected by 3AC and others.

Please note that I am still forced to make approximations in many places here. Many of my personal passwords are still held by the Chapter 11 team - not to mention the data. If the Chapter 11 team would like to add their data to the conversation, I would welcome it.

Also – I haven’t run Alameda for the past few years.

A lot of it was cobbled together after the fact, from models and approximations, often based on data from before I stepped down as CEO, and modeling and estimates based on that data.

Event Overview

2021

During 2021, Alameda's NAV has soared, reaching roughly $100 billion by the end of the year when it goes public, according to my model. Even if you ignore assets like SRM that have a much larger fully diluted share than circulating supply, I think it's still roughly $50 billion.

Alameda's position has also increased during 2021. In particular, I think it has about $8 billion of net borrowings, which I think are used to:

a) ~$1 billion in interest paid to lenders

b) ~$3 billion to acquire Binance from FTX’s cap table

c) ~$4 billion in venture capital

(By “net borrowing,” I mean basically borrowing less liquid assets on hand to repay loans. Net borrowing in 2021 came primarily from third-party lending platforms — Genesis, Celsius, Voyager, etc. — rather than from margin trading on FTX.)

Therefore, by early 2022, I think Alameda's balance sheet will look something like this:

a) ~$100 billion NAV

b) ~$12 billion in liquidity from 3rd party desks (Genesis, etc.) c) ~$10 billion more liquidity could potentially come from them

d) ~1.06x leverage

In this case, the ~$8 billion illiquid position (with tens of billions of dollars in available credit/margin from 3rd party lenders) seems reasonable and not too risky. I think Alameda's SOL alone is more than enough to cover the net borrowing. It comes from a 3rd party lending desk, and they all - I'm told - send accurate balance sheets from Alameda.

I thought its position on FTX International at the time was reasonable—around $1.3 billion, according to my model, backed by tens of billions of dollars in assets—and FTX successfully passed a GAAP audit.

Well, it would take a 94% market crash to drag Alameda underwater by the end of 2021! It's not just SRM and similar assets - if you ignore those, Alameda is still massively overcollateralized. I think its SOL position alone is greater than its leverage.

But Alameda has failed to adequately hedge against the risk of an extreme market crash: hundreds of billions of assets with only a few billion dollars of hedges. Its net leverage ratio — [net position - hedge] / NAV — is about 1.06 times; the market is long.

So Alameda theoretically faces an extreme market crash — but it would need something like a 94% crash to go bankrupt.

Market crash in 2022

So, here is Alameda's rough picture heading into 2022:

$100 billion is not

$8 billion in net borrowings

1.06x leverage

Tens of billions of dollars in liquidity

Then, over the course of the year, the market crashed — again and again. Until midsummer, Alameda repeatedly failed to adequately hedge its positions.

– BTC plunges 30%

– BTC plunges another 30% – BTC plunges another 30%

– Rising interest rates restrict global financial liquidity

–Luna drops to $0

– 3AC collapses – Alameda co-CEO resigns

– Voyager crashes – BlockFi almost crashes

– Celsius Eruption

– Genesis begins to shut down

– Alameda’s borrow/loan liquidity increased from approximately $20 billion at the end of 2021 to approximately $2 billion at the end of 2022

So Alameda’s assets have been hit again and again. But this part isn’t specific to Alameda’s assets. Bitcoin, Ethereum, Tesla, and Facebook are all down more than 60% year-over-year; Coinbase and Robinhood are down about 85% from their peaks last year.

Keep in mind that Alameda has approximately $8 billion in net borrowings at the end of 2021:

a) Pay approximately $1 billion in interest to lenders

b) ~$3 billion to acquire Binance from FTX’s cap table

c) About $4 billion in venture capital

This $8 billion of net borrowing, minus the several billion dollars of hedges it has, results in about $6 billion of excess leverage/net position, backed by about $10 billion of assets.

As the market collapsed, so did these assets. Alameda’s assets — a mix of altcoins, crypto companies, public stocks and venture investments — fell about 80% over the year, with its leverage increasing in small increments.

During the same period, liquidity dried up in lending markets, public markets, credit, private equity, venture capital, and just about every other sector. During the year, nearly every source of liquidity in crypto — including nearly every lending platform — dried up.

This means that Alameda’s liquidity — tens of billions of dollars at the end of 2021 — has fallen to single-digit billions of dollars by the fall of 2022. Most other platforms in the space have closed down or are in the process of closing down, leaving FTX as the last remaining one.

In the summer of 2022, Alameda finally hedged heavily with some combination of BTC, ETH, and QQQ (Nasdaq ETF).

But even after all the market crash in 2022, shortly before November, Alameda still had about $1 billion in NAV; even if you exclude SRM and similar tokens, it's positive, it's ultimately hedged.

Margin Trading

During 2022, many crypto platforms became insolvent due to margin positions blowing up, potentially including Voyager, Celsius, BlockFi, Genesis, Gemini, and eventually FTX.

This is quite common on margin platforms; among others, it occurs when:

Traditional Finance:

London Metal Exchange

MF Global

Long-Term Capital Management

Lehman

Cryptocurrency:

OKEx

OKEx again, I hold flexible coins almost every week

EMX

Voyager, Celsius, BlockFi, Genesis, Gemini, and more.

November Crash

After months of an extremely effective PR campaign against FTX and the crash, CZ posted the fateful tweet.

Prior to the last crash in November, QQQ was moving about half as much as Alameda’s portfolio, while BTC/ETH was moving about 80% as much as Alameda’s portfolio — meaning that Alameda’s hedges (QQQ/BTC/ETH), for as long as they existed, worked. Unfortunately, the hedges weren’t large enough until after the 3AC crash — but by October 2022, they were finally large enough.

But the November crash was a targeted attack on Alameda’s holdings, not a broad market move. Over a few days in November, Alameda’s assets fell by about 50%; BTC fell by about 15% — just 30% of Alameda’s assets — and QQQ didn’t move at all. As a result, the larger hedges that Alameda ultimately put in place that summer didn’t do anything. This would be true of every crash that year — but not this one.

Over the course of November 7-8, things went from stressed but mostly under control to clearly insolvent.

As of November 10, 2022, Alameda's balance sheet will only have about $8 billion in (only semi-liquid) assets, and roughly the same amount of current liabilities, about $8 billion:

A bank run requires immediate liquidity—liquidity that Alameda no longer has.

Credit Suisse’s stock price fell nearly 50% this fall as the bank faced a run threat. At the end of the day, its run on the bank failed. FTX did not.

So as Alameda becomes illiquid, FTX International also becomes illiquid, because Alameda has an open margin position on FTX; a bank run turns illiquidity into insolvency.

This means FTX joins Voyager, Celsius, BlockFi, Genesis, Gemini, and other companies that have suffered collateral damage from the liquidity crunch for borrowers.

All of this is to say: no funds were stolen. Alameda lost money from the market crash but didn’t adequately hedge it — as Three Arrows and others did this year. FTX was impacted as Voyager and others were earlier impacted.

Conclusion

Even so, I think FTX could benefit all of its customers if a concerted effort was made to improve liquidity.

There were offers for billions of dollars when Mr. Ray took over, and more than $4 billion later.

Had FTX been given a few weeks to raise the necessary liquidity, I believe it would have been able to leave customer funds substantially intact. I did not realize at the time that Sullivan & Cromwell—through pressure on Mr. Ray to come on board and file Chapter 11 , including solvent firms like FTX US—might have unwound those efforts. I still believe that if FTX International were to relaunch today, it would indeed be possible to leave customers substantially intact. Even without that, there are a ton of assets available to customers.

Regrettably, I was slow to respond to public misunderstandings and material misstatements. It took me some time to piece together what I could - I did not have access to a lot of relevant data, most of which was about a company I did not run at the time (Alameda).

I had been planning to give my first substantive account of what happened when I testified before the U.S. House Financial Services Committee on December 13. Unfortunately, the Justice Department moved to arrest me the night before, preempting my testimony with an entirely different news cycle. It’s worth noting that a draft of my planned testimony was leaked here.

I have a lot more to say – why Alameda failed to hedge, what happened at FTX US, what led to the Chapter 11 process, the S&C, etc. But at least it’s a start.