Antoinette Schoar: Thank you very much for the kind introduction. I'm pleased to be with all of you. Indeed, what I want to talk about is what can we learn about crypto as an asset class and the promises of defi decentralized finance from the example of the Terra Luna run. So this is where we're going to focus on today. And I have to say, you know, kind of in contrast to the very exciting and positive pitches you just saw in a second ago, this is going to be maybe a little bit more of a cautionary tale of some of the fault lines in in crypto and defi that we should be aware of when either investing in this asset class or just thinking in generally. Right. What the structural promises are. So just, you know, kind of I'm sure you have all been engaged in reading about cryptocurrencies and you understand that one of the basic premises and promises of crypto and decentralized finance is to use the decentralized sorry, the distributed ledger technology to reinvent if you want, the architecture of finance, meaning this idea that we will use free access to the financial markets and basically a move away from large centralized financial intermediaries to reinvent finance. And of course, you know, many crypto enthusiasts have postulated that this should allow us to eliminate many of the unnecessary intermediaries and cut down on rents in the financial system. Right. I mean, that is the the the positive message or that's the hope, of course, for crypto.
Now, let me show you, in the case of Terra Luna, where some of the fault lines are. So in some sense, right, What I want to do today going through a deep dive and if you want a forensic analysis of what happened in the case of Carolina to dispel some of the common myths that we have seen in in crypto, and the three big ones really are, number one, that just the fact that there's open access to the financial system through the distributed ledger technology or the blockchain, if you want that, that by itself eliminate rents and eliminate many of of the risks. That's not the case. And you know, anyone who understands finance obviously understands that in a world of financial markets where there are economies of scale and scope, right, where there are network externalities, just opening up markets by itself is not going to bring down rents. And I'll show you this in the in the context there. Right. The other two thing I want to I will touch on is that this idea that because all transactions are on a public ledger, meaning are on a blockchain by itself, that that makes the system less prone to to systemic risk. As you will see in the case of Terra Luna, that's not the case. And the reason the simple reason for this is that information being publicly available is not the same right as people having the right models to use the information with.
If you want to update on the actual risks that are building up. And what I'll show you is that there was in the case of Terra Luna, but this is actually really a message for for most of crypto is because there's so much heterogeneity in the sophistication of the investors that are participating in the system, You get very different responses across, you know, these very heterogeneous investors. And that in itself creates systemic risk, as we saw in the case of of Terra and Luna. And then finally, as you know, kind of I mentioned upfront, is that open access is not the same as financial inclusion. So ultimately, right, financial inclusion is not just a matter of a technological solution. We know from a lot of research in fintech, in development, finance, in developing countries that technological, technological solutions also need people to trust in them. And trust doesn't just come from better and more convenient technological access, but it comes from people being convinced that the financial infrastructure itself, right is safe, is working for them and not against them. And that they are not at an either informational or systemic disadvantage relative to the insiders. And as I will also show you, in the case of of Terra Luna, that's really a big message that we need to take from this, that we see that there are there were large, large differences in the outcomes from the investments for the sophisticated versus the unsophisticated investors.
All right. Having said this. Before we go into a deep type of of terror, I wanted to get get a sense of where your guys head is all at when you think about Terra Luna. And so just simply I wanted to ask you to do a little word cloud, so to speak, to give us a sense of what terms what words do you associate with Terra Luna? Right. So good, right? So as I was, I was hoping that some of you would say basically tulipmania, you know, let's see where we go coming back to us. Well, degenerate. Wow. Fraud, Ponzi, right? These are exactly some of the words in retrospect, people associate with Terra Luna. Also obviously fraud. You see, Ponzi gets bigger and bigger. Now, what I will show you is actually that ex-post we might think, you know, and we might have understood that this was, you know, at heart a form of a Ponzi scheme. But what I want to you to understand is actually this whole system was much more complicated than a typical Ponzi scheme, because remember, in a typical Ponzi scheme, right, like a madoff met Bernie Madoff type of situation is we take new investors money to pay old investors. Right.
An immediate easy pass through if you want from the from the new ones funding you know the old ones and there you know that the system is going to run out of money. I mean it has to right because at some point there will be reckoning here, as I will show you, is actually the the financial architecture was much more complicated. And in fact, and this is the stunning thing, if investors had been a little bit more rational about the system, they might have avoided a Ponzi scheme. But let me show you why why I'm saying this. And we can come back at the end. Right, to to take stock of how much was this a Ponzi scheme versus more like a tulip mania. So people being, you know, over overly optimistic, basically a bubble. Right. Or was it some some combination of all of this with this? Can we go back to the to the slides, please? So great. So let's very quickly I want to recap. What is the what was the Terra Luna ecosystem? And why this is useful is actually that it was very similar to a typical smart contract blockchain, say, as Ethereum, Solana, Cardano, etcetera, is and at its heart, right, what a smart contract blockchain actually tries to accomplish is to to solve a dual sided platform problem, meaning that what the blockchain wants needs to do in order to generate rents and returns for its its participants and for its coin holders is to attract businesses and customers to the blockchain, right? So you want to attract businesses that provide different types of services, for example, right? Say crowdfunding services like, you know, they had pylon or decentralized exchanges, which, you know, on Terra Luna were Astro Pod and Terra swap on Ethereum, right? You have the curve protocol Uniswap.
Et cetera. Right. And why do you do this? Is because the more attractive the services you offer on the on your blockchain, the more customers will come to your blockchain and engage in transactions on these blockchains. What happens when there are transactions on the on the blockchain, right. The transaction demand feeds demand for the different applications, but it also means that with each transaction that has to be encoded on the blockchain, right there is there are fees from from basically the staking right in a proof of stake protocol which then go to the Terra, to the Luna token holders, right. And then ultimately they can also be speculative demand, people basically betting on the future price of Luna. Right. And so Luna in this case was the the native token. If you want the currency in which all these transactions had to be embedded on the blockchain. Yeah. And so any if you want additional activity that this blockchain could attract to itself should actually increase the price of Luna.
Because it means that there are activities on the blockchain. And the hope, of course, here was the more those customers or those users who come to to to Terra Luna are sticky on the blockchain meaning. Right? Once they started using this blockchain, they will keep on using it because it is convenient, right? That might feed into the future rent to this blockchain. The same is true, right? The same logic is true If you think about why the price of Ethereum or why the price of Bitcoin is higher than the price of smallish blockchains that haven't attracted a lot of users. Right? This is exactly the at core, the idea of a network externality. Yeah. Now by itself there is nothing wrong. This is not per se a Ponzi scheme, right? If you think about it, that's what Uber did, what Lyft did, what most basically startups nowadays try to do, right? Any type of platform application often tries the same thing, right? Let's get a lot of users onto Uber so that the merchants, meaning the drivers, have an interest of being on that platform and the customers too. Right. And if you remember at the beginning, the pricing of Uber and Lyft was clearly below the market of a real cab drive because they were trying to attract demand away from cabs and were subsidizing the price of it. Right. Do we call that a Ponzi scheme? No.
Right. We might call it a failed business model because there were some businesses that tried this and went out of business. Right. And basically went bankrupt and others failed. So, so far, this is not if you aren't yet a Ponzi scheme. Right. It was an investment that was trying to basically provide potentially useful services to their customers. Now it becomes a bit more interesting because if you look at the very top here, this is USD, right? This is the famous or infamous infamous Stablecoin that Terra or TFL, which is the company behind Terra Luna was trying to introduce or did introduce. And so what is the appeal of USD? The appeal was that different from other stablecoins that I'm sure you know, like tether or USD circle, right? It was not backed by an outside collateral, but it was backed by an algorithm that was pegging USD to a dollar, but backed by Luna. Right. So the idea here was that we have an algorithmic stablecoin that lives entirely within the crypto space. It doesn't need the traditional financial system or what people often call tradfi right to to have backing. But it was hoping to be basically backed entirely by, you know, being crypto native. Now, how did this work? The way it worked was that on the Terra Blockchain, right, you had what is called a native swap. Think of it as a piece of code that allows any holder of Luna to swap, say, a dollar worth of Luna, a dollar verse of USD into Luna, right.
Or Luna into USD. Now, why why is that useful? Because actually, if you think about what this is basically hoping to do this this type of swap. Right. It is it is a mechanism to bring the value of USD always back to par because what would happen, right is that imagine if the price of USD was above the price was above $1. What you can what you could do then, right, is you could actually take USD sorry, buy USD for $1 worth of Luna, but sell USD itself for a price above $1. Right now. What does this do in this swapping mechanism? Actually, because you are now swapping USD into Luna, you are sorry, swapping Luna into USD, you are increasing the supply of USD, which brings down the price of USD. You are reducing the price of Luna, the supply of Luna. And so you're bringing basically, you know, the two prices back to to par, right? And so therefore this type of swap mechanism was meant to bring. The the price of USD. Always back to $1. Now, if you think about like a kind of a finance person, that makes sense as long as there is enough value if you want in Luna to support the entire outstanding USD. Right. Or another way of thinking about this is like, you know, if you, if you think about, you know, kind of this really in financial term ultimately USD is like a convertible debt that is issued right on the value of Luna.
We do this all the time in start up financing, right? We we all the time use actually say convertible debt issued against the equity of a startup and we don't again, we don't call that a Ponzi scheme. Right. But here the one important thing that I want you to have in mind is that this type of system be can become extremely fragile and run prone. If the any situation where users are trying to swap USD into Luna leads to a drop in the price of Luna. Because if that's the case, right, if a if an investor who holds Luna thinks that as soon as there is a swapping of USD into Luna will have massive dilution, Right? A price impact on the price of Luna, then you're better off selling Luna in advance of the swap, which then drops the price even more. Which means that then when people are swapping USD into Luna, they need to get even more Luna, which is even more dilutive. Right? Then you get basically this negative death spiral, right? Or what people have termed a death spiral. Now, again, if you think with your finance head right on it through this, right is when would you expect should the the the swapping of USD into Luna have a big price impact on the price of Luna actually.
Right. It is most likely to be the case if investors who are holding Luna did not understand what it means that there is an that there is USD outstanding. Another way of saying this right in all the models we have on the pricing of convertible debt, we know that if investors are rational, they should you should see a price impact on the underlying equity. So in this case, Luna Not when there is a conversion of USD into Luna, but when there is an issuance of USD, right? Because forward looking investors should understand you are issuing a ton of USD, meaning you're issuing a lot of senior debt that has a claim on Luna. If I don't think that Luna itself is valuable enough to to actually sustain all that debt, right, the price of Luna should go down at the point of issuance, not at the point of conversion. In fact, at the point of conversion, then very little should happen because if investors had rationally priced the convertible debt before. Yeah. And so you see, right, that why am I being so tedious in going through this is because. Right. Depending on whether investors in that system are rational or not, even the insiders to the system have either more leeway in issuing more and more in US USD or not.
Another way of saying this is that in a world where where most investors, if most investors had been rational, the, you know, the founders behind Terra Luna would not have been able to issue a lot of this algorithmic stablecoin because the price of Luna would have immediately gone down because investors would have understood, look, there's not enough value in that system to sustain all that USD overhang. But what happened and that's what I'm going to show you now in more detail, is that that's not what happened in the system. So what happened in the system is actually that what you can see is that so if you first look at the right graph here is that over the the 2021 and early 2022 period, actually there was massive amount of USD being issued which most of it actually ended. The DUP on a protocol on Terra Luna that was called anchor. And so this is the last piece I need to explain to you now is what is anchor? Because that was actually part of, you know, where maybe now we can come a little bit closer to was this a Ponzi scheme or not? But so what is anchor? Anchor was a borrowing and lending protocol on Terra Luna similar to even a money market fund, if you want. Where holders of Luna could sorry holders of UST could deposit their UST and remarkably earn a 20% interest on the UST deposit.
So that right this is the blue line here I plotted for you, you know, during the same time period. Say the return on us on circle or tether was much, much, much, much lower. Right closer to 3 to 4%. Now, the why was this interest rate on terra on ust so high? Remarkably, it's because the founders of Terra Luna decided arbitrarily that's where they wanted to set the interest rate on anchor. Now, what happened with this very high interest rate is that all the USD that were issued that I showed you here, right. Were absorbed by anchor because it was such an attractive return on on on your USD especially remember this was 2021, early 22 where you know, kind of the interest rate environment was still extremely low. Right. So getting 20% anywhere should have raised anyone's red flags, right? Because it has to come from somewhere. Now, you might say, oh, maybe they had the magic wand. In fact, if you go on to Cryptospace or, you know, some of the chat rooms of Terra Luna, some people said they invented some magic wand of how to create 20% out of nothing. All right. Let me show you. Right. Is that really true? Would be nice. But of course, as you can imagine, the answer is no. But how did they sustain for more than a year this high interest rate? Because they did pay this high interest rate to the depositors.
The way they did actually was that and this is if you look at the the graph in the middle, what we show is, you know, this is I'm not bothering with all the painstaking, tedious work we did on the Terra Luna blockchain blockchain, but this is all you get this information from the blockchain because we can see how much subsidy was needed to subsidize the 20% deposit rate on Anchor. Yeah. And so what I'm showing you here, forget, you know, all the, the, the colorful lines, if you just look at basically the black line, you know, going down, it shows you on a daily basis the amount of subsidy needed to support anchor. And what you can see is, you know, by the end of by the beginning of May, which is, you know, a week before there was the crash, right? Tfl Or Terra was spending more than $6 dollars a day to subsidy to subsidize. Anchor Now, interestingly, it was even more stunning because they were not just subsidizing the deposit rate. In fact, they were even subsidizing the borrowing rate of anchor through airdrops and many other things. That's too tedious to talk about right now. But what it means is really smart investors could do something even more hypercharged than just depositing their USD on anchor. In fact, you could borrow from anchor.
This is my black and and yellow line here. So this is the fraction of people that would borrow from anchor deposit immediately back onto anchor to benefit on the borrowing and the lending side from the subsidy. Yeah. Now why am I you know, again, why am I showing you all this? Because given that all of this is on the blockchain, we can even show you how the subsidies were provided from Terra to the anchor system. And so my graph here on the graph on the right is basically looking at the community wallets on the Terra Luna blockchain. We can we can show you how much money there is basically reserves that had been deposited at the ICO of Terra Luna right into the community wallets and how this money was spent to subsidize anchor. And in fact, what you can see right in March 2022, they were already running out of this community subsidy. Then what happened is actually that TFL with some of their their main investors actually re upped the amount of that subsidy money in those community wallets. But you can see, look, within two months, by March 2022, they were already again, almost running out of that money. Yeah. So why is that, you know, kind of interesting in a way from a you know, from the stability of a financial system is that all of this was in principle observable on the blockchain, right? And any user not just, say, a fancy large financial institution that has access to the internal data of a of a financial institution had access to it.
No. Anyone. Right. Including, you know, kind of researchers like us. But unfortunately and that's the last part I want to show you, is that the way people use that information? Right. And in a way, all these massive warning signs about the. Ability of that system was very different. So as you might remember, in the 7th of May, 2022 is when the run started. And here is just I'm showing you on the black line the the pegging of sorry, on the right red line, the pegging of Terra Luna from the from the $1 USD price. And on the black line basically the withdrawals from anchor. So what you can see let me show you this in detail actually is that the first people to run even before there was any pegging, were a few very large wallets. So famously right now on on crypto Twitter, this has been termed wallet A and wallet B, but we know that A is probably Jane Stream and Jane Street and B is Celsius. Right? But what happened is basically that in the morning of May, the 7th of May 2022 is basically when Jane Street withdrew first $50 million, send it to the Covid protocol and sold. Now, interestingly Celsius in response. Right.
Withdrew $150 Million from anchor but actually just didn't sell it yet. Just sat there waiting to see will they be more of a run if you want. And then when they saw that Jane Street actually kept withdrawing and selling, that's when Celsius also started selling. And then lots of other big guys started selling. And here what I'm showing you is now trade by trade over the the beginning of the run, meaning May 7th. Right. How large trades versus small trades looked like. So red are the large trades. And look what you see. Right? The large trades are all sells. The small guys keep buying. In fact, I can show you this now basically in terms of how people reduced the their wallets, broken down by the size of the deposit. What you can see is right. If you go from the lowest, which is the sorry lowest line, which is the the largest wallets, you can see. Right. The large wallets start running much faster, much earlier. And the other thing I can tell you is they are running much more decisively, meaning when they run, they take 100% out and they're done with it because they understand this is now a run. What you see for the small guys is not only they are running slower. In fact, we can see that some of them even come in back in in the middle of the run because they think it's buy the dip.
Right. They don't understand. This is a full run in full scale going on. Right. They think it's becoming cheap. And I can tell you, even some of the this medium lines say like in the 10,000 range. Right. People actually often withdrew half and let the rest sitting in there almost like thinking that they're diversifying. Right. There's nothing to diversify in a run. Once you understand this is a run, you want to get out ahead of people. But why am I saying this? Because it shows you, right, that even though everybody had access to the same type of data in principle, of course, in practice, right. It's not just that people couldn't process the data. They also didn't have the right model of what really is that system that they're getting involved in. Now, You know, we do a bunch of other analysis in this research to show you even that the it's not just a matter of size, meaning of wealth. It's really a matter of sophistication because we can remember, we can see all the traits that somebody did in their wallets prior to the run. And I can see what type of protocols they used, what kind of trading strategies, so I can classify people into how sophisticated or non sophisticated they are. And it's really sophistication that's driving, you know, how you act conditional on a run. So, you know, kind of to finish up right ultimately.
So we see this run on UST on anchor. We also see the run on Luna by people swapping out of Luna. But let me not get into those details. What ultimately then I'm sure you are familiar with, right, is that by the end, basically of that week of the run, so by the 13th of of May 2022, the price of UST crashed basically to, you know, to almost nothing. And we have several trillion dollars worth of Luna outstanding because of all the dilution that happened. Right. So the and then the platform gets stopped. Now, why did I want to show you kind of this as an example, you know, maybe as a warning example of the crypto sphere is the following. Um, you can say, right, maybe this was a Ponzi scheme, but hopefully you understand this is more complicated than a Ponzi scheme. Because, yes, the founders of Terra Luna were hyping Luna were maybe over claiming how stable this system is. But there was information out there. In fact, there were some hedge funds and so on that were warning people against the instability in the system. And still, because traders were hyping up the price of Luna, it allowed the insiders, the insiders to keep on issuing more and more USD without bringing down the price of Luna, which once the concerns started right, this actually led to like then a massive run.
And so I think it's really important for us to understand that by itself, if open access and transparency doesn't bring stability, if people don't have the right models to price those type of systems with. Right and in fact. Right. Ultimately, let me finish by saying this run had very differential effects on small versus large traders, as I showed you. Right. Poorer, less sophisticated people lost way more than some of the big traders. Now, one thing is that this run had right now. Right. Limited spillover effects on the real on the traditional financial system, because if you want the on and off ramps to the traditional financial system are still limited. Right. We do see that say Silver Lakes Circle and SVB did actually have some knock on effects. Right? And of course, Celsius and even FTX were ultimately affected by the run. But I think we should not how do I say think that the there it is inherently no spillover effect from crypto to the traditional financial system is right now the fact that this didn't have such a massive knock on effect is just because those often on rails do not exist in this world yet so much. But the more, say, ETFs like grayscale and so on are allowed to actually provide these off and on ramps. The more risk there will be in this system also for a traditional financial system. Thank you very much.
Speaker2: Answer is fascinating. Thank you. We do have a couple of minutes for questions, so if anyone has one, please raise your hand. Anyone. Is there one?
Antoinette Schoar: I scared them all.
Speaker2: Yeah. Oh, I thought you'd taken things away. We do have a couple on here. First one is, are there any other systems, blockchains at the moment that you're worried about? Are there any flashing signals right now that we should be concerned about, such as the the terror one?
Antoinette Schoar: So I would say, you know, kind of the what we have seen is in in case of collateralized stablecoins, like, you know, kind of usdc, like circle, right? The quality of the collateral really matters. And actually, initially, when we had the Terra Luna run, we saw a lot of people basically fleeing, quote unquote, into circle. So Circle was one of the big beneficiaries from the Terra run. Now, of course, I'm sure you're aware that when SVB Right. Silicon Valley Bank, you know, was basically going into receivership, three and one half billion dollars of Circle's money was in SVB. Right. And only after the the Fed and Treasury announced that they will be made whole. Right. That actually circles valuation or peg come back. But I think that you know kind of where. So why do I say this? Because it actually it shows you how sensitive the value of a collateralized stablecoin is to the quality of of the collateral. And where I am still a bit concerned is basically tether, right? Even though it's per se a collateralized stablecoin, it has been very opaque about really showing all the assets that are backing tether. And so my concern would be that, you know, if if something happens to the reputation of tether, that to me right now is the biggest red flag.
Speaker2: So this might not be the last one we see.
Antoinette Schoar: Worrisomely No.
Speaker2: Okay. So a question here. What do you think about the SEC's approval of a crypto ETF? Is that a good sign?
Antoinette Schoar: It depends on what you think a good sign is. So I would say to me, I believe that currently this system is still too opaque to really that you want to allow institutional money to flow into this. So I think, you know, kind of remember, I'm a Chicago trained economist. I do believe if people want to gamble with their own private money, you know, kind of that's up to them. But I don't think said say that in people's pension money, in money that gets passively invested. Right. We should allow massive inflow into into crypto meaning. So into these type of cryptocurrencies where we where we are not sure where the value is generated. And so I feel that you know, the ETF, the SEC's hands maybe are somewhat tight given that obviously the kind of the CME and Cboe are offering already, you know, kind of futures on, you know, kind of on crypto. So, you know, that's kind of an aspect of the US regulatory environment. But I personally I would be extremely concerned if we allow a massive inflow of institutional money into into a system that has so little transparency for the average person.
Speaker2: Anyone here? I think we've got a question at the back there if someone's got a mic. Ali, thank you.
Speaker3: Yeah. Thank you. I would like to know, how would you assess the current tug of war between governments issuing crypto versus private industry? And where do you think that that future is going to go?
Antoinette Schoar: When you say governance, meaning cbdc. Yeah, because governments are not I mean, most governments are not issuing crypto. They're thinking about issuing cbdc. Right. So, so number one, I think the Fed does have it right in the following sense that I would be extremely concerned if it is central banks issuing cbdc. Right. I mean, we don't want them to be the kind of the retail banker to to the world. So the Fed has already said in its cbdc white paper that if the Fed ever went into, you know, into issuing a digital dollar, it would be distributed via the private banking system. So that, I think, is is probably a good, good thing. I think you see the issue I see is we need to understand much better what a cbdc would do to the stability of the US banking system. Now this is obviously a very large discussion, maybe not for today, but the truth is that if there is a cbdc, there is a concern depending on how it gets implemented, that at the slightest worry of a run. People will take money out of the banking system and run into the Fed. But why? Why? That would not be a good thing, right? We already. Why would that not be a good thing? Because if they run into the Fed and the banks are then undercapitalized, it means the Fed has to be the recapitalizing of the private system system in in continuity. Think about how bad it is when we have one run like Silicon Valley Bank and then we have to think about was it a smart idea to make all the uninsured depositors hold whole? Et cetera. Right. So I feel that there will have to be a much more sophisticated discussion of how this government provided digital currencies should be implemented smartly to avoid these type of political pitfalls.
Speaker2: Antoinette another time. Thank you very much. That was fascinating. Thank you.
