I get a sense that when they make a movie about this incident the sex doctor will steal the show. I picture him as a combination of Tobias Fünke and Dr. Jacoby from Twin Peaks.
This was the best explanation of the whole catastrophe I have read so far, and I especially appreciate the righteous indignation throughout and the sensible advice at the end. This is why I subscribe. Thank you.
To me the biggest question watching the past 6 years of crypto-mania has been:
Why are all these ostensibly sophisticated Venture Capital and other traditional finance firms "falling" for it? Did they truly drink the kool-aid and believe in the magic beans? FOMO? Or did they just believe they could find a greater fool?
1. How is it that no mainstream or finance journalist noticed this gaping sign of fraud?
2. Maybe crypto is too marginal/weird to attract the notice of normal journalists, the same way that I'm a biologist and I do not study Pokemon development.
3. If Steven Levitt is writing about fixed sumo wrestling matches, someone in finance ought to have been reporting on bizarre crypto scams.
4. We're going to learn about a lot of similar scams soon, aren't we?
So part of me is inclined to think it’s weird that journalists didn’t notice. But if the venture funds that invested in FTX didn’t notice, when they had a lot of money at stake, then is it that odd people with no skin in the game didn’t notice? Like, a journalist did figure out Theranos, but it was one journalist — John Carreyrou. The rest of the press basically fell for it. If he hadn’t been on the case, then I guess we would have found out Theranos had no product once many many more people got false test results through Walgreens. The investors had no idea. Rupert Murdoch had $125 million at stake and even he didn’t do his research.
It is indeed horrifying that the VC people didn't notice. I suppose that makes the journalists look better by comparison but makes the whole situation stranger to me.
I read Carreyrou's book and it makes a little more sense to me that Theranos didn't get caught for so long. They lied about proprietary technology that almost no one had access to and they were insanely legally aggressive toward former employees. Because of internal information controls at the company, surprisingly few people knew how badly their technology worked and most (though not all) were terrified into silence. The strange part was that Walgreens never did due diligence. The book interviewed some Walgreens guy who proposed obvious science (do 50 tests with a Theranos instrument, 50 with a regular instrument, compare) and no one listened to him and he stopped getting invited to meetings. The only way I can put it is Holmes' crusading sincerity was so compelling people didn't even want to question her. I suppose SBF had some of the same thing going on, but I still feel like this was a more obvious case. It was non-obvious that Theranos' proprietary technology didn't work, whereas the 8% return on savings account should have been a huge red flag.
I mean, I've seen people call crypto itself a Ponzi scheme for a long time, but I was aware of SBF and didn't know he was paying 8% interest on deposits at his exchange while claiming that he wasn't even investing deposits. Given his mainstream-media prominence, the obvious scamminess of his operation should have drawn more attention don't you think?
I always feel like "too big to fail" really just means "Is it too big for the government to allow it to fail".
Considering crypto in the US is difficult to regulate effectively and competes for market share with the USD (and whichever CBDC the US eventually unveils) the federal government and IMF likely have no interest in protecting crypto currencies from failing.
No, that's exactly what it means. It means that the downdraft created by a failure would take down too many innocents, and so the government has to step into to prevent it from doing so. Theoretically, one of the goals of financial market regulation should be to never allow that to happen, but then we have a thing called "regulatory capture" and so it does. But maybe worse than regulatory capture is just regulatory neglect resulting in the same thing, which could be the case here. If all crypto going to zero would create a depression, I don't know that the government would in fact allow that to happen.
I might just be cynical, but I tend to see most government regulation as regulatory capture that occasionally has the side effect of accidentally leading to a good market outcome. From that view, if the government hasn't picked their winners for the industry then it doesn't hold any value to bail people out of said industry. A collapse like this is likely to increase calls for regulation and they can pick said winners, then I think it will become too big to fail.
It depends on how you see bailouts. One view is it's just more regulatory capture, a transfer of money from the government to those who control the government. The other view is that it's a panicked, short-term focused response to a real threat of economic dislocation. Like, if the economy is going to crash, then the President will lose their job. So the President will hand out cash to whoever friend or foe to prevent that from happening. I suspect the truth is a mix of the two, but that would mean that even if crypto has no "friends in high places" they might still get bailed out. But I really don't know how big it actually is, maybe it could crash and not take anything else with it. I don't know that answer to that at all.
I've read a lot of the accounts of what happened, but there's been a lack of specificity (understandable talk about "magic beans" and such). This is my guess (only a guess): FTX created its own crypto token called FTT. Alameda was SBF's incestuous "investment" company. FTX loaned Alameda $10. Alameda used the $10 to buy FTT (basically worthless "magic beans") from FTX. This created a market for FTT and increased its value by creating demand for it. FTX ends up with its $10 back, which it then loaned to Alameda again. Alameda turned around and bought more FTT... and so on. FTT benefits from all of this trading by seeming to have a value based on its trading on FTX's exchange, so that Alameda's FTT assets seem extremely valuable (billions of dollars). FTX ultimately loaned billions to Alameda (but it was always just that $10 loaned repeatedly), which seemed to give it billions in "assets" (loans collateralized by apparently valuable FTT).
Did you all see than Bankman communicated with employees using “applications that were set to auto-delete after a short period of time, and encouraged others to do the same.”?!!!
I think most writers in the world of money are frightened to come off looking like they"don't understand" crypto if they suggest this particular outfit looked like a "something for nothing" scheme.
Part of my thing is that I can't get too mad at this because everyone who was defrauded in this was trying to defraud others (b/c that's the only reason someone would invest in crypto -- to sell it to someone else before it goes down in value).
To the extent that I had ever heard of FTX or Binance before this story, it would have been just another crypto name passing in the wind. There's no reason to pay attention to any of these things except when they blow up.
Also, upon further, thought, I would for fairness modify your first rule of thumb (emphasis added):
1. Any crypto-related business *that is spending a significant effort to recruit "normies"* is a scam.
I think the modification is necessary to reflect that I do think there are a lot of companies out there not trying to scam people, but are genuinely searching for a fundamental economic problem that crypto can solve. These companies generally lose money during this search, which is often the case for venture-backed or boot-strapped startup companies. Most of these companies have and will continue to die out, which again is normal and healthy for the technology ecosystem, and doesn't indicate a scam, just an unsuccessful venture.
I think the crypto-space is particularly misguided, because almost all of the true zealots are have a classic "law of the instrument" bias (https://en.wikipedia.org/wiki/Law_of_the_instrument). They are enamored with a specific technology because of some elegant/nerdy properties it has, and are out searching for solutions to whack with their magical hammer. This is *almost* always a bad way to start a company, which should really start with a problem to solve, and then an evaluation of the entire solution-space.
Oh the absurd ying yang of novelty and greed that is crypto.
Another appalling dumb thing about crypto: NFTs! An NFT is a password protected token that crypto people want everyone to believe means you own something. Like how your car title is a thing that represents you own a car but instead of being paper it exists on the *waves hands* B L O C K C H A I N... Now there's nothing wrong with trying to convince people an NFT represents a certificate of ownership (although the first court case around that should be interesting), it has the right properties to behave like a car title with extra steps if that's what you're into. What's bonkers is: if no one wanted to own Jack Dorsey's first tweet before NFTs were a thing... why would you want to pay millions just to make the blockchain say you own it??? People were not running to notaries public with slips of paper saying they owned a jpeg before but now...?
What next? An NFT of Bart Simpson's soul???
All that said, I would like to play devil's advocate here. I love what Josh has on this, this FTX mess reveals how so much of the pandemic crypto "economy" is an appalling farce BUT I see comments asking some variation of "how could anyone like / invest in / support the grifter ridden cesspool that is crypto?" and as a denizen of that cesspool I have my take.
The anonymous creator of bitcoin solved a nontrival computer science problem that basically made it so you can have a digital thing exist that you can't copy, kinda surprising that that's even possible. From this you basically get the blockchains of today: always online immutable databases of who owns what and that can even run arbitrary scripts aka contracts.
So far all of this has amounted to NOTHING. Or probably less than nothing given all the scams and crime that have piggy backed on crypto.
And perhaps this is the beginning of the end for crypto. Maybe this sets off a chain effect of crypto institutional failure: binance blows up, coinbase goes insolvent, prices tumble and consensus forever settles on this being a misadventure of a weird hammer than never found its nail. I'd fault no one in delighting in this. There'd even be some justice in it. Totally plausible.
But consider the alternative: if crypto doesn't implode it'd need some purpose, maybe even something that'd be a boon for liberal democracy. I can offer two ideas on what that "killer app" might be:
1. The securing powerful technologies.
This one is a little sci-fi but hear me out.
Josh, I like your impressionist image of the FTX cuddle puddle from Dall-E. It legit looks nice. At least some tech exists that can do useful things we can appreciate and enjoy ;)
The creation of such an image wasn't possible a few year ago. Science advances. More likely than not that continues. Robotics improve. Maybe more mastery over biology is gained. Artificial intelligence is developed that rivals humans. Doesn't matter if it is 5, 10, 50 years, eventually a technology will be developed that, if it doesn't immediately go out of control and destroy us all, will give the entity that controls that tech an insurmountable advantage over any adversary.
When that happens, who will be the entity that owns the tech? Insert Elon/Zuck/Bezos du jour. What are the odds of a positive outcome here? If we'd don't all die then we'd get Earth's first (and maybe last) emperor.
Why not just have government halt the development of these technologies before they pose existential risks, regulate it? Then China would develop it. Plus, we'd want the benefits to humanity that this powerful tech could produce.
If only there was some kind of secure decentralized information platform where any interested party could audit the system and verify that certain transactions (like sending instructions to a super intelligent AI for example) could only occur when some quorum of keys acted in coordination...
It's sci-fi now but one day such a technology (super intelligence, nanobots, a super intelligence that can build nanobots, etc) will emerge. When it does we'll need to have a preexisting technology ready that secures it as a decentralized and ownerless public good. Either we coordinate to birth the tech into a secure container and only interact with it via a rigid protocol implemented atop some blockchain based contracts or we gamble with the fate of humanity.
2. A live fine grained view of everything happening in an economy.
There's no reason you couldn't have a public record of everything of economic and civic meaning within a city/nation/earth. With appropriate privacy features and what not. Obviously such software does not exist today. But it's merely an extremely dense engineering challenge, possible w smart contract enabled blockchains like Ethereum and Cardano.
"What would that get us? Sounds like how our economy operates today but with extra steps". I imagine similar things were said when someone first suggested we should write down laws.
It'd provide legible fine grained data about the economy and government that'd unlock new levels of economic growth. Sorta like the opposite of how China inflates their GPD numbers.
Imagine a public works project funded by tax payer money to build a bridge or something. How much funding is the project allocated? Who gets the contract and the subcontracts? How much is each contract awarded? Is the completion of the contract timely? Is the work good? None of this information is currently hidden. But it is also not in a single data source and of consistent form. Plus construction subcontracting for public works project hardly has open markets. Now take this concept and apply it too all other kinds of "projects" that happen in an economy.
The barrier to entry to becoming a firm in the economy would lower as you have a live view what firms are currently succeeding and how they are structured. Lenders would have more confidence when distributing capital as they'd have improved data about how successful firms similar to the one they are maybe investing in have performed and what kind of outcomes the individuals seeking their capital have had in the transactions they'd been involved in previously. In a way this addresses the problem with FTX, where investors YOLO'ed funds into a scam. In a world where a blockchain is able to serve live data about how firms are structured and interacting I'd imagine it'd be less common for scams to hand wave their way into capital.
Of course this would all be a huge engineering lift. It'd take time, take a while to dial in and get right. But the internet has existed for, like, 30 years? and how long have we been at the project of economic growth? Three millennia let's call it. So it is likely the current relationship among information technology, the economy and government is far from it's optimal state.
I get a sense that when they make a movie about this incident the sex doctor will steal the show. I picture him as a combination of Tobias Fünke and Dr. Jacoby from Twin Peaks.
This was the best explanation of the whole catastrophe I have read so far, and I especially appreciate the righteous indignation throughout and the sensible advice at the end. This is why I subscribe. Thank you.
Thank you!
My husband and I both believe that when something is too difficult for the average person to understand, then be very skeptical.
Ding ding ding
😂👍
To me the biggest question watching the past 6 years of crypto-mania has been:
Why are all these ostensibly sophisticated Venture Capital and other traditional finance firms "falling" for it? Did they truly drink the kool-aid and believe in the magic beans? FOMO? Or did they just believe they could find a greater fool?
My thoughts in order:
1. How is it that no mainstream or finance journalist noticed this gaping sign of fraud?
2. Maybe crypto is too marginal/weird to attract the notice of normal journalists, the same way that I'm a biologist and I do not study Pokemon development.
3. If Steven Levitt is writing about fixed sumo wrestling matches, someone in finance ought to have been reporting on bizarre crypto scams.
4. We're going to learn about a lot of similar scams soon, aren't we?
So part of me is inclined to think it’s weird that journalists didn’t notice. But if the venture funds that invested in FTX didn’t notice, when they had a lot of money at stake, then is it that odd people with no skin in the game didn’t notice? Like, a journalist did figure out Theranos, but it was one journalist — John Carreyrou. The rest of the press basically fell for it. If he hadn’t been on the case, then I guess we would have found out Theranos had no product once many many more people got false test results through Walgreens. The investors had no idea. Rupert Murdoch had $125 million at stake and even he didn’t do his research.
It is indeed horrifying that the VC people didn't notice. I suppose that makes the journalists look better by comparison but makes the whole situation stranger to me.
I read Carreyrou's book and it makes a little more sense to me that Theranos didn't get caught for so long. They lied about proprietary technology that almost no one had access to and they were insanely legally aggressive toward former employees. Because of internal information controls at the company, surprisingly few people knew how badly their technology worked and most (though not all) were terrified into silence. The strange part was that Walgreens never did due diligence. The book interviewed some Walgreens guy who proposed obvious science (do 50 tests with a Theranos instrument, 50 with a regular instrument, compare) and no one listened to him and he stopped getting invited to meetings. The only way I can put it is Holmes' crusading sincerity was so compelling people didn't even want to question her. I suppose SBF had some of the same thing going on, but I still feel like this was a more obvious case. It was non-obvious that Theranos' proprietary technology didn't work, whereas the 8% return on savings account should have been a huge red flag.
So many people have been calling these crypto-related financial vehicles “Ponzi schemes” for years. Journalists have been reporting on those people.
I mean, I've seen people call crypto itself a Ponzi scheme for a long time, but I was aware of SBF and didn't know he was paying 8% interest on deposits at his exchange while claiming that he wasn't even investing deposits. Given his mainstream-media prominence, the obvious scamminess of his operation should have drawn more attention don't you think?
You're a hero Josh. More voices directly calling out the scam which is crypto are needed.
But is crypto "too big to fail"?
It is not, which is delightful.
I always feel like "too big to fail" really just means "Is it too big for the government to allow it to fail".
Considering crypto in the US is difficult to regulate effectively and competes for market share with the USD (and whichever CBDC the US eventually unveils) the federal government and IMF likely have no interest in protecting crypto currencies from failing.
No, that's exactly what it means. It means that the downdraft created by a failure would take down too many innocents, and so the government has to step into to prevent it from doing so. Theoretically, one of the goals of financial market regulation should be to never allow that to happen, but then we have a thing called "regulatory capture" and so it does. But maybe worse than regulatory capture is just regulatory neglect resulting in the same thing, which could be the case here. If all crypto going to zero would create a depression, I don't know that the government would in fact allow that to happen.
I might just be cynical, but I tend to see most government regulation as regulatory capture that occasionally has the side effect of accidentally leading to a good market outcome. From that view, if the government hasn't picked their winners for the industry then it doesn't hold any value to bail people out of said industry. A collapse like this is likely to increase calls for regulation and they can pick said winners, then I think it will become too big to fail.
It depends on how you see bailouts. One view is it's just more regulatory capture, a transfer of money from the government to those who control the government. The other view is that it's a panicked, short-term focused response to a real threat of economic dislocation. Like, if the economy is going to crash, then the President will lose their job. So the President will hand out cash to whoever friend or foe to prevent that from happening. I suspect the truth is a mix of the two, but that would mean that even if crypto has no "friends in high places" they might still get bailed out. But I really don't know how big it actually is, maybe it could crash and not take anything else with it. I don't know that answer to that at all.
I've read a lot of the accounts of what happened, but there's been a lack of specificity (understandable talk about "magic beans" and such). This is my guess (only a guess): FTX created its own crypto token called FTT. Alameda was SBF's incestuous "investment" company. FTX loaned Alameda $10. Alameda used the $10 to buy FTT (basically worthless "magic beans") from FTX. This created a market for FTT and increased its value by creating demand for it. FTX ends up with its $10 back, which it then loaned to Alameda again. Alameda turned around and bought more FTT... and so on. FTT benefits from all of this trading by seeming to have a value based on its trading on FTX's exchange, so that Alameda's FTT assets seem extremely valuable (billions of dollars). FTX ultimately loaned billions to Alameda (but it was always just that $10 loaned repeatedly), which seemed to give it billions in "assets" (loans collateralized by apparently valuable FTT).
Ultimately money created from thin air.
Did you all see than Bankman communicated with employees using “applications that were set to auto-delete after a short period of time, and encouraged others to do the same.”?!!!
I think most writers in the world of money are frightened to come off looking like they"don't understand" crypto if they suggest this particular outfit looked like a "something for nothing" scheme.
Part of my thing is that I can't get too mad at this because everyone who was defrauded in this was trying to defraud others (b/c that's the only reason someone would invest in crypto -- to sell it to someone else before it goes down in value).
Magic-beans is a much better metaphor than shovels!
To the extent that I had ever heard of FTX or Binance before this story, it would have been just another crypto name passing in the wind. There's no reason to pay attention to any of these things except when they blow up.
Also, upon further, thought, I would for fairness modify your first rule of thumb (emphasis added):
1. Any crypto-related business *that is spending a significant effort to recruit "normies"* is a scam.
I think the modification is necessary to reflect that I do think there are a lot of companies out there not trying to scam people, but are genuinely searching for a fundamental economic problem that crypto can solve. These companies generally lose money during this search, which is often the case for venture-backed or boot-strapped startup companies. Most of these companies have and will continue to die out, which again is normal and healthy for the technology ecosystem, and doesn't indicate a scam, just an unsuccessful venture.
I think the crypto-space is particularly misguided, because almost all of the true zealots are have a classic "law of the instrument" bias (https://en.wikipedia.org/wiki/Law_of_the_instrument). They are enamored with a specific technology because of some elegant/nerdy properties it has, and are out searching for solutions to whack with their magical hammer. This is *almost* always a bad way to start a company, which should really start with a problem to solve, and then an evaluation of the entire solution-space.
Oh the absurd ying yang of novelty and greed that is crypto.
Another appalling dumb thing about crypto: NFTs! An NFT is a password protected token that crypto people want everyone to believe means you own something. Like how your car title is a thing that represents you own a car but instead of being paper it exists on the *waves hands* B L O C K C H A I N... Now there's nothing wrong with trying to convince people an NFT represents a certificate of ownership (although the first court case around that should be interesting), it has the right properties to behave like a car title with extra steps if that's what you're into. What's bonkers is: if no one wanted to own Jack Dorsey's first tweet before NFTs were a thing... why would you want to pay millions just to make the blockchain say you own it??? People were not running to notaries public with slips of paper saying they owned a jpeg before but now...?
What next? An NFT of Bart Simpson's soul???
All that said, I would like to play devil's advocate here. I love what Josh has on this, this FTX mess reveals how so much of the pandemic crypto "economy" is an appalling farce BUT I see comments asking some variation of "how could anyone like / invest in / support the grifter ridden cesspool that is crypto?" and as a denizen of that cesspool I have my take.
The anonymous creator of bitcoin solved a nontrival computer science problem that basically made it so you can have a digital thing exist that you can't copy, kinda surprising that that's even possible. From this you basically get the blockchains of today: always online immutable databases of who owns what and that can even run arbitrary scripts aka contracts.
So far all of this has amounted to NOTHING. Or probably less than nothing given all the scams and crime that have piggy backed on crypto.
And perhaps this is the beginning of the end for crypto. Maybe this sets off a chain effect of crypto institutional failure: binance blows up, coinbase goes insolvent, prices tumble and consensus forever settles on this being a misadventure of a weird hammer than never found its nail. I'd fault no one in delighting in this. There'd even be some justice in it. Totally plausible.
But consider the alternative: if crypto doesn't implode it'd need some purpose, maybe even something that'd be a boon for liberal democracy. I can offer two ideas on what that "killer app" might be:
1. The securing powerful technologies.
This one is a little sci-fi but hear me out.
Josh, I like your impressionist image of the FTX cuddle puddle from Dall-E. It legit looks nice. At least some tech exists that can do useful things we can appreciate and enjoy ;)
The creation of such an image wasn't possible a few year ago. Science advances. More likely than not that continues. Robotics improve. Maybe more mastery over biology is gained. Artificial intelligence is developed that rivals humans. Doesn't matter if it is 5, 10, 50 years, eventually a technology will be developed that, if it doesn't immediately go out of control and destroy us all, will give the entity that controls that tech an insurmountable advantage over any adversary.
When that happens, who will be the entity that owns the tech? Insert Elon/Zuck/Bezos du jour. What are the odds of a positive outcome here? If we'd don't all die then we'd get Earth's first (and maybe last) emperor.
Why not just have government halt the development of these technologies before they pose existential risks, regulate it? Then China would develop it. Plus, we'd want the benefits to humanity that this powerful tech could produce.
If only there was some kind of secure decentralized information platform where any interested party could audit the system and verify that certain transactions (like sending instructions to a super intelligent AI for example) could only occur when some quorum of keys acted in coordination...
It's sci-fi now but one day such a technology (super intelligence, nanobots, a super intelligence that can build nanobots, etc) will emerge. When it does we'll need to have a preexisting technology ready that secures it as a decentralized and ownerless public good. Either we coordinate to birth the tech into a secure container and only interact with it via a rigid protocol implemented atop some blockchain based contracts or we gamble with the fate of humanity.
2. A live fine grained view of everything happening in an economy.
There's no reason you couldn't have a public record of everything of economic and civic meaning within a city/nation/earth. With appropriate privacy features and what not. Obviously such software does not exist today. But it's merely an extremely dense engineering challenge, possible w smart contract enabled blockchains like Ethereum and Cardano.
"What would that get us? Sounds like how our economy operates today but with extra steps". I imagine similar things were said when someone first suggested we should write down laws.
It'd provide legible fine grained data about the economy and government that'd unlock new levels of economic growth. Sorta like the opposite of how China inflates their GPD numbers.
Imagine a public works project funded by tax payer money to build a bridge or something. How much funding is the project allocated? Who gets the contract and the subcontracts? How much is each contract awarded? Is the completion of the contract timely? Is the work good? None of this information is currently hidden. But it is also not in a single data source and of consistent form. Plus construction subcontracting for public works project hardly has open markets. Now take this concept and apply it too all other kinds of "projects" that happen in an economy.
The barrier to entry to becoming a firm in the economy would lower as you have a live view what firms are currently succeeding and how they are structured. Lenders would have more confidence when distributing capital as they'd have improved data about how successful firms similar to the one they are maybe investing in have performed and what kind of outcomes the individuals seeking their capital have had in the transactions they'd been involved in previously. In a way this addresses the problem with FTX, where investors YOLO'ed funds into a scam. In a world where a blockchain is able to serve live data about how firms are structured and interacting I'd imagine it'd be less common for scams to hand wave their way into capital.
Of course this would all be a huge engineering lift. It'd take time, take a while to dial in and get right. But the internet has existed for, like, 30 years? and how long have we been at the project of economic growth? Three millennia let's call it. So it is likely the current relationship among information technology, the economy and government is far from it's optimal state.
Well stated.
Doesn't seem at all to be an over-simplified explanation.
Just put in terms most folks can understand.
Wonder if there will be a bailout like in 2008.
My preference is the Madoff Model.
Take what they can recoup for you, and learn from the loss.