SBF's Latest Statement: The Real Reasons and Process Behind Alameda's Hundreds of Billions of Dollars in Losses

Original Title: 《FTX Pre-Mortem Overview》

Author: Sam Bankman-Fried

Compiled by: Linqi, ChainCatcher

Abstract

In mid-November 2022, FTX International was effectively bankrupt. The legendary story of FTX is, at its core, a tale that lies between Voyager and Celsius.

These three events combined ultimately led to the implosion.

a) During 2021, Alameda's balance sheet grew to approximately $100 billion in net asset value, with $8 billion in net borrowings (leverage) and $7 billion in liquid funds on hand.

b) Alameda failed to adequately hedge its market risks. Throughout 2022, a series of large market crashes, affecting both stocks and cryptocurrencies, led to a decline in its asset value of about 80%.

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

Then Alameda's situation spread to FTX and beyond, similarly affecting Voyager, Genesis, Celsius, BlockFi, and Gemini, among others.

Despite this, there remains a potential for a more objective recovery of funds. FTX US is still fully solvent and should be able to return all customer funds. FTX International has billions of dollars in assets, and I have nearly dedicated all my personal assets to the customers.

Key Points

This article discusses whether FTX International is solvent.

This is unrelated to FTX US, as FTX US is fully solvent and has always been. When I handed over FTX US to Ray and the bankruptcy protection team, it had approximately $350 million in net cash on hand beyond customer balances. Its funds and customers are isolated from FTX International.

But absurdly, users of FTX US have not yet been compensated and have not received their funds back.

Here is the balance sheet record when I handed over FTX US:

FTX International is a non-U.S. exchange that operates outside the U.S., is regulated outside the U.S., is incorporated outside the U.S., and accepts non-U.S. customers.
(In fact, its headquarters is in the Bahamas, registered and operating locally as FTX Digital Markets LTD).
U.S. customers are all on the (still solvent) FTX US exchange.

Senators have expressed concerns about potential conflicts of interest with Sullivan & Cromwell (S&C). Contrary to S&C's statement that they "have limited and primarily transactional relationships with FTX," S&C was one of the two main law firms for FTX International before bankruptcy and the primary law firm for FTX US. The CEO of FTX US came from S&C, and they collaborated with FTX US on its most important regulatory applications, as well as on some of the most critical regulatory issues with FTX International, and they also worked with FTX US on its most significant transactions. When I visited New York, I sometimes worked in S&C's office.

S&C and GC are the main parties threatening to appoint their own candidate as CEO of FTX, including the solvent FTX US, which subsequently filed for Chapter 11 bankruptcy and chose S&C as the advisor for the debtor entity.

Despite being in bankruptcy and processing about $5 billion in withdrawals during its last few days of operation, FTX International still retained a significant amount of assets, approximately $8 billion in various liquid assets as of when Ray took over.
In addition, there were many potential financing proposals, including letters of intent signed after filing for bankruptcy protection, totaling over $4 billion. I believe that if given a few weeks, FTX International could likely raise enough funds using its illiquid assets and equity to substantially compensate customers.

However, since S&C pressured FTX to file for bankruptcy protection, I worry that these avenues may have been abandoned. Even now, I believe that if FTX International were to restart, there is a real possibility for customers to receive substantial compensation.

While FTX's liquidity primarily relied on Alameda starting in 2019, by 2022, FTX's liquidity had diversified, with Alameda's trading volume on FTX dropping to around 2%.

I did not steal funds, nor did I hide billions of dollars. Past and present, nearly all my assets are available to support FTX customers. For example, if the bankruptcy protection team is willing to fulfill my D&O legal expense reimbursement guarantee, I proposed to contribute nearly all or 100% of my personal shares in Robinhood to the customers.

FTX International and Alameda were both legitimate and independently profitable businesses in 2021, each earning billions of dollars.

However, Alameda's asset value declined by about 80% during 2022 due to a series of market crashes, similar to what happened to Three Arrows Capital (3AC) and other cryptocurrency companies last year. After that, its assets further declined due to a targeted attack. FTX was affected by Alameda's decline, just as Voyager and other companies were earlier affected by 3AC and others.

Please note that in many places in the text, I am still forced to make approximate judgments. Many of my personal passwords are still held by the Chapter 11 bankruptcy team, not to mention the data. If they wish to include the data in the discussion, I would welcome it.

Additionally, I have not operated Alameda in the past few years.

Thus, much of the information is pieced together in hindsight, based on models and approximations, generally based on the data I had before stepping down as CEO, as well as models and estimates based on that data.

Overview of Events

2021

During 2021, Alameda's net asset value skyrocketed, according to my models, to about $100 billion by the end of the year. Even if you ignore assets like SRM, which are significantly diluted, I believe it was still around $50 billion.

During 2021, Alameda's positions were also growing.

Specifically, I believe it had about $8 billion in net borrowings, which I think was spent on the following:
a) Paying about $1 billion in interest to lenders
b) Buying out Binance's holdings in FTX for about $3 billion from FTX's original investment list
c) About $4 billion in venture capital

(What I mean by 'net borrowings' essentially refers to borrowings minus liquid assets available to repay loans. This net borrowing in 2021 primarily came from third-party lending platforms like Genesis, Celsius, Voyager, etc., rather than from FTX's margin trading).

Thus, by early 2022, I believe Alameda's balance sheet looked roughly as follows.

a) Approximately $100 billion in net assets
b) About $12 billion in liquidity from third-party desks (Genesis, etc.).
c) ~ $10 billion in liquidity
d) ~ 1.06 times leverage

In this context, $8 billion in illiquid positions (with hundreds of billions of available credit/margin from third-party lenders) seemed reasonable and not overly risky. I believed that just Alameda's SOL was sufficient to cover the net borrowings. Moreover, it came from third-party lending desks, and I was informed that they all received Alameda's accurate balance sheet.

I believed that its holdings in FTX International at the time were reasonable, according to my models, at about $1.3 billion, backed by hundreds of billions of assets, and FTX successfully passed GAAP audits at that time.

So, by the end of 2021, to drag Alameda underwater, there would need to be a 94% market crash! And not just with SRM and similar assets; even without considering those, Alameda still had a significant amount of over-collateralization. I believed that just its SOL position was larger than its leverage.

But Alameda failed to adequately hedge against the risk of extreme market crashes: out of $100 billion in assets, only a few billion dollars were hedged. Its net leverage ([net position - hedge] / net asset value) was about 1.06 times; it was long on the market at that time.

Thus, theoretically, Alameda faced the risk of extreme market crashes, but it would require a crash of around 94% to render it bankrupt.

2022 Market Crash

So, the rough situation for Alameda entering 2022 was:

1. $100 billion in net assets

2. $8 billion in net borrowings

3. 1.06 times leverage

4. Hundreds of billions in liquid funds

Then, throughout the year, the market crashed time and again. Until mid-summer, Alameda repeatedly failed to adequately hedge its positions.

-BTC dropped by 30%.
-BTC dropped again by 30%.
-BTC dropped again by 30%.
-Rising interest rates suppressed global financial liquidity
-Luna went to zero
-3AC collapsed
-Alameda's co-CEO resigned
-Voyager collapsed
-BlockFi nearly went bankrupt
-Celsius went under
-Genesis began to shut down
-Alameda's borrowing/lending liquidity fell from about $20 billion at the end of 2021 to about $2 billion by the end of 2022

Thus, Alameda's assets were repeatedly hit. But this part was not specific to Alameda's assets. Bitcoin, Ethereum, Tesla, and Facebook all saw declines of over 60% this year; Coinbase and Robinhood's stock prices dropped about 85% from last year's peaks.

Remember, by the end of 2021, Alameda had about $8 billion in net borrowings:

a) Paying about $1 billion in interest to lenders
b) Buying out Binance's holdings in FTX for about $3 billion from FTX's original investment list
c) About $4 billion in venture capital

This $8 billion in net borrowings, minus a few billion dollars in hedges, led to about $6 billion in excess leverage/net positions, supported by about $100 billion in assets.

As the market crashed, these assets also collapsed. Alameda's assets, including a mix of altcoins, crypto companies, stocks, and venture capital, fell by about 80% over the year, gradually increasing its leverage.

Meanwhile, liquidity dried up in the lending market, public markets, credit, private equity, venture capital, and nearly every sector. Over the past year, nearly all sources of liquidity in the crypto space—including almost all lending platforms—collapsed.

This meant that by the fall of 2022, Alameda's liquidity fell from hundreds of billions at the end of 2021 to single-digit hundreds of millions. Most other platforms in the sector had either collapsed or were in the process of collapsing, with FTX becoming the last survivor.

In the summer of 2022, Alameda finally executed significant hedges on some combinations of BTC, ETH, and QQQ (Nasdaq ETF).

But even after all the market crashes in 2022, shortly before November, Alameda still had about $10 billion in net assets; even excluding SRM and similar tokens, the net asset value was still positive, and it was ultimately hedged.

Margin Trading

During 2022, due to the surge in margin positions, many crypto platforms went bankrupt, including Voyager, Celsius, BlockFi, Genesis, Gemini, and ultimately FTX.

This situation is quite common on margin trading platforms; in addition, it also occurred in:

Traditional finance: LME (London Metal Exchange), MF Global (MF Global), LTCM (Long-Term Capital Management), Lehman (Lehman Brothers)

Crypto industry: OKEx, again OKEx (essentially every week throughout the year), CoinFlex, EMX, Voyager, Celsius, BlockFi, Genesis, Gemini, and so on.

The Collapse in November

After months of extremely effective public relations campaigns against FTX and its collapse, CZ released a decisive tweet.

Until the collapse of FTX in November, the volatility of QQQ was about half that of Alameda's portfolio, while the volatility of BTC/ETH was about 80%, indicating that Alameda's hedges (QQQ/BTC/ETH) were somewhat effective. Before the collapse of 3AC, the hedges were not large enough, but by October 2022, unfortunately, they finally became large enough.

However, the crash in November was a targeted attack on the assets held by Alameda, rather than a broad market fluctuation. In just a few days in November, Alameda's assets fell by about 50%; Bitcoin dropped about 15%, only 30% of Alameda's assets, while QQQ remained unchanged. As a result, the larger hedges Alameda had ultimately executed in the summer did not work. They helped with every crash that year, but not this one.

During the course of November 7 and 8, things went from tense but mostly under control to clearly insolvent.

By November 10, 2022, Alameda's balance sheet had only about $8 billion (half liquid) in assets left, while current liabilities were roughly the same, about $8 billion:

A bank run requires immediate liquidity, and Alameda had run out of liquidity.

This fall, Credit Suisse's stock fell nearly 50% due to the threat of a bank run. ++The Threat of a Bank Run++. At the end of the day, it did not succeed in a bank run. But FTX did not.

Thus, as Alameda became illiquid, so did FTX International, as Alameda had a margin position on FTX; and the bank run turned this lack of liquidity into an inability to repay.

This meant that FTX joined the ranks of Voyager, Celsius, BlockFi, Genesis, Gemini, and others that suffered collateral damage due to the liquidity crunch of borrowers.

All of this is to illustrate: no funds were stolen. Alameda incurred losses due to inadequate hedging of the market, just like 3AC and other companies this year. FTX was affected, just as Voyager and other companies were earlier affected.

Epilogue

Even so, I believe that if we work together to raise liquidity, FTX has the potential to compensate all customers.

When Ray took over, the company received billions of dollars in financing offers, and subsequently received over $4 billion in financing offers.

If given a few weeks to raise the necessary liquidity, I believe FTX could have provided substantial compensation to customers. I did not realize at the time that Sullivan & Cromwell, by pressuring Ray and filing for bankruptcy protection under Chapter 11 (including for solvent companies like FTX US), could potentially derail these efforts. I still believe that if FTX International were to restart today, it could provide substantial compensation to customers. Even if it does not, there are still plenty of assets available for customers.

Many of these are data from companies (Alameda) that I was not operating at the time. I regret that I was slow to respond to the public's misunderstandings and significant misstatements. It took me some time to piece together what I could, and I do not have and cannot access much relevant data, much of which is data from companies (Alameda) that I was not operating at the time.

I had planned to provide a substantive explanation of what happened when I testified before the House Financial Services Committee on December 13. Unfortunately, the Justice Department took action to arrest me the night before, seizing my testimony with a completely different news cycle. In any case, the draft of the testimony I was prepared to provide was leaked.

I have much more to say about why Alameda failed to hedge, what happened with FTX US, what led to the bankruptcy protection proceedings, S&C, and so on. But at least this is a start.