Disclaimer: This is an evolving situation. It might take months to get an accurate picture of what exactly happened. This post is my attempt to simplify what is currently the general theory of what happened at FTX as of November 16, 2022. The goal of this post is to explain to those not well versed in investment finance how it all (potentially) went wrong. If you have a sophisticated understanding of investment finance, this post is probably too elementary for you. 

This is a story about lending. 

But before I begin the story, I would like to explain how certain financial entities work today.

A bank, in simple terms, is an entity where a customer can deposit their money (and get paid interest), or can borrow money (and pay interest to the bank). The way a bank makes money is that it charges more interest to the borrower than what it pays to the depositor. A bank is allowed to lend your money out without your explicit permission. Because of this, governments heavily regulate banks. They are limited as to who they can lend to and how much. They are audited frequently and have robust external oversight. 

A brokerage, in simple terms, is an entity where a customer can deposit their money and buy investment assets, in hopes to grow their money over time. Typically, a customer can open two types of accounts in a brokerage: 

  1. Cash account: This account can exclusively buy and sell assets
  2. Margin account: This account can not only buy and sell assets, it can borrow against those assets. For example, you can deposit $10,000 USD and buy Apple stock. You can then use that Apple stock as collateral and borrow against it. Typically you could borrow up to a certain amount. For a stock like Apple, you could borrow up to 70% of the value of the stock, so in our example, you could borrow up to $7,000 USD, which you could use to buy whatever you want. However, there is a caveat: if Apple stock falls 50%, you would be faced with a margin call: Your account balance is $5,000 USD in Apple stock, but you borrowed $7,000. When this happens the brokerage will call you and ask you to deposit $5,000 USD in collateral so your account is covered. If you don’t have the money, the brokerage has the right to sell your remaining Apple stock to cover your bad debt. However in our example, there is a shortfall of $2,000 USD ($5,000 from the sale of Apple stock - $7,000 loan you took). In that case the brokerage will try to get that $2,000 from you through legal means, but that would cost money and time and they might not be able to recover it. What would happen is that the brokerage would then realize a loss of $2,000 against their capital.

How does a brokerage make money? Typically, there are four ways:

  1. Charging commissions on trading. You buy $10,000 worth of Apple stock? That trade costs you $10. 
  2. Charging a fee on your account. You have $10,000 in your account. The brokerage charges an 1% annual fee for holding your assets, but your trades are free. So if your account balance stays at $10,000 for a year, you would pay $100 for having your assets in there. 
  3. They charge for making a market in an asset. In simple terms, they sell the asset to you at a price slightly higher than the purchase price in the open market, or buy the asset from you at a lower price than the sale price in the open market. The best example of this model for making money are currency exchange houses. 
  4. Lending. They will lend cash and other assets to margin accounts, and charge an interest fee to the customer that borrows. Like I explained in our Apple example above, this adds considerable risk to the business model, so brokerages that engage in margin lending must have robust risk management processes to ensure that there are no shortfalls in the customers accounts. No brokerage wants to be constantly calling their customers asking for more money. Furthermore, a brokerage must have capital on hand to be able to lend to their customers and to be able to absorb shortfalls. A Brokerage (in most countries) is not allowed to lend customer funds out, unlike a bank. It must hold all customer deposits 1:1. This means that if you deposited $10,000 USD, it must always hold that $10,000 USD. If you deposited 10 Bitcoin it must hold that 10 Bitcoin at all times. 

Enter FTX

FTX was a crypto exchange and brokerage, with a slick trading platform. FTX’s founder and CEO was Sam Bankman-Fried (SBF). What separated FTX from the rest of the firms that offered similar services was how fast and efficient their platform was, how liquid their market making business was (This means, in simple terms, that FTX was one of the best places to buy and sell cryptocurrencies. It was fast and the prices they offered where some of the best), and their margin lending business. At FTX, you were allowed to deposit cash or cryptocurrencies, then borrow against them. Their systems were so advanced when it came to risk management that the sale of collateral when margin was breached was automatic, no need to call the customer asking for cash to cover. 

Enter Alameda Research

Alameda was an investment firm founded by SBF prior to his founding of FTX. Alameda invested in cryptocurrencies and its strategies were considered wildly successful in the crypto space. As a fund it only managed investments of its employees and a few early backers. SBF was the CEO of Alameda before becoming the CEO of FTX. Alameda was a major client of FTX. 

Enter FTT Token

FTT is a cryptocurrency created by FTX as a rewards system for its customers. You were rewarded FTT tokens for trading in FTX (Although you could also just buy the tokens in the open market). Having FTT in your FTX account allowed you to have discounts on trading fees, and FTX allowed you to borrow up to 95% of the value of your FTT tokens in your account. How did the FTT token maintain its value? FTX would use some of its trading profits to buy back FTT in the open market and destroy the tokens. This would theoretically mean that FTT would increase in value the more trading profits FTX generated. FTT token holders theoretically had a partial claim to the future profits of FTX, similar to a stock.

So, how does this all lead to the implosion of FTX?

It is assumed that, like many other crypto investment and trading firms, Alameda Research incurred significant losses in their portfolio during the crypto downturn of this year. Alameda traded and invested using loans, and at one point their losses became so big that their lenders asked for their money back, which Alameda did not have. For reasons still unknown to the crypto space, SBF made a decision to bail out Alameda. Now allegedly the hole was so big that SBF’s own money or even FTX’s own capital was not enough to cover these losses (It is assumed that the losses at Alameda could have been over $10 billion USD), so SBF decided to use FTX’s customer funds to bail out Alameda.

Now, how can SBF do this without raising any flags? Transferring billions of dollars in assets from FTX to Alameda would have raised all of the red flags, and it would have leaked to the crypto market fairly quickly. SBF knew that he had to keep this hidden from the majority of the FTX organization. Allegedly, SBF and two other executives of FTX did the following:

  1. They created a set of bookkeeping separate from FTX’s so that they could accurately keep track FTT tokens out in circulation.
  2. They created a back door system that allowed them to withdraw FTT tokens without the oversight of FTX’s internal controls.

With these two systems in place, SBF was able to take out millions of FTT tokens from FTX, lend them to Alameda (in secret, supposedly only a handful of people knew at FTX), then Alameda would deposit the tokens on FTX, use them as collateral to borrow assets and pay back its loans.

This mechanism in it of itself does not fail automatically. As long as FTT maintained a high enough price, the loans Alameda took were collateralized.

However, that all began to change on November 2, when CoinDesk, a crypto journalism website, revealed that the majority of Alameda’s balance sheet was FTT tokens. This raised a lot of eyebrows in the crypto space.

Enter Changpeng Zhao (CZ)

CZ is Binance’s founder and CEO, and one of the leading figures in the crypto space. Binance is one of the largest cryptocurrency exchanges in the world. A few years ago, CZ invested in FTX. Last year, it seems that the relationship between CZ and SBF deteriorated, and SBF decided to buy back CZ’s FTX investment. Part of that buy back was paid in FTT tokens (approximately $500 million USD worth).

On November 6, presumably after CZ learned about Alameda’s balance sheet, he announced on Twitter that he was starting to sell his FTT tokens. This led to a crash in FTT’s price.

FTT token price chart, 1 year, measured in USD. Source: CoinGecko

This spooked FTX’s clients, and many started to withdraw their balances. If FTX had kept the assets of their clients 1:1, this shouldn’t have been a problem. However, as explained above, FTX loaned customer funds to Alameda, collateralized by FTT tokens. Now those tokens lost their value significantly, and because there weren’t enough FTT tokens in open market circulation, the millions of tokens Alameda had could not be sold on time to cover their loan.

Eventually, FTX did not have enough money to cover client withdrawals. SBF and FTX tried everything to stop the withdrawals, but ultimately nothing worked. Finally, on November 11, FTX filed for bankruptcy.

What Happens Now?

It is yet to be seen how this will settle. It is going to take months or years to uncover exactly what happened, and what can be done to remediate all the stakeholders of this saga. I hope that somehow all the clients of FTX who have their funds frozen on the platform can get their assets back. I hope that a detailed investigation occurs, and if SBF and other FTX and Alameda employees engaged in fraud, they get sentenced for it. Based on the theory that I just described, it seems fairly clear to me that fraud did occur.

Many have compared this event to the fall of Lehman Brothers in 2008. It is similar to Lehman in that the risk management process was very poor when it came to properly assessing the value of certain assets (In the case of Lehman, Mortgage Backed Securities. In the case of FTX, small cryptocurrencies including their own token, FTT)

Others have compared it to the fall of Enron in 2001. It is similar in that both internal accounting controls and external auditing processes at FTX were so poor, that SBF and other FTX executives were able to not only lend FTT tokens off book to Alameda, but lend client funds.

I also found it similar to the fall of LTCM in 1998, in that the executives of FTX and Alameda thought they were smarter than everyone else when it came to risk management, then the market taught them a very expensive lesson. Finally, SBF’s actions bailing out a failing Alameda reminded me of rogue traders (Nick Leeson at Barings Bank, Jerome Kerviel of Societe Generale, and Bruno Iksil of JP Morgan), who all doubled down on their losses in an attempt to make them back, leading to disastrous consequences.

What Happens to Crypto?

There are people out there who believe this is a devastating blow to the crypto space. In reality I think it is good that these bad actors get flushed out. While I do believe that less capital (both individuals and investment firms) will flow into the space in the short term, the engineers working on figuring out ways that the technology can have utility are still working on the space.

I hope that this saga leads to more due diligence in the space from institutional investors. It is embarrassing that prominent venture capital firms like Sequoia and pension funds like the Ontario Teachers Pension Plan did not due their due diligence when it came to internal accounting controls and good governance at FTX. It is likely that they won’t recover a single cent of their investment.

People will always say that this is a story about crypto. In reality, this is a story about lending. It is a story about very poor risk management, poor governance, poor oversight, and poor due diligence. I am interested to see how the crypto ecosystem changes due to FTX’s fall.

Disclosure: I own ethereum in private wallets and via an ETF held in my brokerage account. I have used both Binance and FTX in the past, but never left assets there in the long term. Nothing in this post constitutes investment advice.

New Comment
10 comments, sorted by Click to highlight new comments since: Today at 12:16 AM

People will always say that this is a story about crypto. In reality, this is a story about lending. It is a story about very poor risk management, poor governance, poor oversight, and poor due diligence.

 

Considering that crypto is an ecosystem with deliberatly poor risk management, poor governance, poor oversight, and poor due diligence... this is very much a story about crypto.

I strongly disagree with the upshot of this post. In my view, crypto is fundamentally a Ponzi scheme based on various forms of electronic tulip bulbs. FTX issued debt backed by e-tulips, and then people finally realized that FTX's e-tulips aren't actually worth anything. Meanwhile, FTX's non-tulip assets are long gone. (where, exactly?)

What does this mean for the e-tulip industry? Surely although these e-tulips have been shown to be worthless, other e-tulips will hold value....

Perhaps you are right: Creating a financial system for e-tulips was dumb in the first place. 

But that discussion goes beyond the scope of this post. The fact of the matter is the financial system for crypto was created, and a major player crashed, and people not well versed in finance and crypto are somewhat interested in how it happened. 

I think it's directly relevant to what you wrote in the "What Happens Now" and "What Happens to Crypto" section, which I take to be the point that the rest of the post is kinda building up to. These sections seem to be written with some assumptions which I think are unwarranted and, at least, untrue in my opinion.

Comparisons with Enron, Lehman, etc -> Assumption: There was some real value here that customers put in, that was lost/destroyed/stolen by fraud. I think this assumption is not true for other crypto assets, only for the USD that people put in.

"In reality I think it is good that these bad actors get flushed out."  -> Assumption: there are good actors, i.e. that after bad actors get "flushed out" there will be a significant number of non-flushed actors

"the engineers working on figuring out ways that the technology can have utility are still working on the space." -> Assumption: crypto can have significant utility and thus has value

"I hope that this saga leads to more due diligence . . . . when it [comes] to internal accounting controls and good governance [at crypto firms]. . . ." -> Assumption: It is possible for there to be crypto firms that have good internal accounting controls, good governance, and be well capitalized. IMO, doing due diligence on a crypto firm is like entering a cat in the dog show - sure you can list all the ways it does or does not conform to breed standards (pointy ears, upright tail, etc) but fundamentally the problem is that it's a cat.

"People will always say that this is a story about crypto. In reality, this is a story about lending." -> No, I think it's a story about crypto.

[-]lc1y84

It is possible for there to be crypto firms that have good internal accounting controls, good governance, and be well capitalized.

Clearly this is possible because Coinbase exists, and in fact does the exact same thing FTX did except they don't steal user funds. You may not like the fact that Coinbase exists, or that Bitcoin can currently be traded for 16,000$; those facts may confuse or frustrate you, but that does not make them untrue.

[-]tgb1y30

Why would CZ tweet out that he was starting to sell his FTT? Surely that would only decrease the amount he could recover on his sales?

One can speculate that he had already sold his FTT and started shorting when writing the tweet.

Maybe, but a big player selling without explanation can also cause a panic.

There are also reputation effects with either choice.

and their margin lending business.

Seems some word is missing, the whole sentence is hardly readable.

pension funds like the Ontario Teachers Pension Plan did not due

*do