00:00:00
00:00:01
Ep 34: Stablecoin Runs and the Centralization of Arbitrage
86 Plays2 months ago
Yiming Ma from Columbia Business School and Anthony Zhang from UChicago explore the intricacies of stablecoins and their impact on financial stability. Learn about the critical role of arbitrage in maintaining price stability and the potential risks associated with these digital assets.
Check out their paper here.
Transcript
00:00:06
Speaker
Hello and welcome to this Owl Explains Hootenanny, our podcast series where you can wise up on blockchain and web3 as we talk to the people seeking to build a better internet. Owl Explains is powered by Avalabs, a blockchain software company and participant in the avalanche ecosystem. My name is Silvia Sanchez, project manager of Owl Explains and with that I'll hand it over to today's amazing speakers.
00:00:34
Speaker
Mr. Howell explains Hootenanny. We ah do these podcasts with lots of different guests from all over the world. and As some of you may know, we did a special series with CBER Forum, which is the Crypto and Blockchain Economic Research Forum, a group of academics from around the world who do economics, finance, and other related research and writing on a wide variety of topics related to blockchain and crypto. ah We've got two of those academics here today, and they've been doing some really interesting work, one paper in particular on stablecoins that we're going to focus on. So Yiming and Anthony, why don't you each introduce yourself? Yiming, why don't you go first?
00:01:22
Speaker
Sure. Hi, everyone. ah My name is Yimeng Ma. I'm an associate professor at Columbia Business School and my work is generally on financial institutions. Hi everyone, I'm Anthony Zhang. I'm at UChicago. I work on a couple of topics, including crypto, and then have a paper which we're going to talk about with evening unstable planes, other than that I do some work on NFTs, automated market makers, competition between crypto exchanges, um and I also teach a blockchain class at UChicago. Excellent. Well, two universities that I was not good enough to get into when I was applying to colleges many, many years ago, but good to see that
00:02:04
Speaker
you all are Those universities are keeping up the high standards and keeping me out. um But at least they get to talk with professors from those colleges now. um So Yiming, why don't you tell folks ah just a little bit about some of the other work you've done ah here. And then Anthony, I'll ask you to do the same thing before we dive into the stablecoin paper, just so that people can hear the background that you come from. And it doesn't have to be just blockchain related work. It can be work in anything on ah you know financial services, financial institutions, et cetera. Absolutely. Yes. so As mentioned, my work is on financial institutions, as well as the interaction between financial institutions and the pass-through of monetary policy and financial stability. and In particular, over the past years, we've been looking at institutions
00:02:53
Speaker
that do what we call liquidity transformation, in the sense that they're liquid their assets are not as liquid as their liabilities. So traditionally, we think of a bank that has loans and then that has deposits that are very liquid. But more and more, ah you know that kind of activity has been moving outside of the banking sector. So we've been looking at bond mutual funds and bond ETFs that engage in liquidity transformation that at the same time, however, brings some risks to the financial stability. And so ah the latest one, the kind of institution we've been looking at are stablecoins, because as we will talk about, actually, there's a lot of similarities um in the sense that they also have assets that are not 100% liquid.
00:03:41
Speaker
Great. And how about you, Anthony? Give us an idea. I know you sort of mentioned NFTs and automated market makers in your intro, but tell us a little bit more about the particular focus you've had. So I'm going to talk about sort of some of the things I work on and how I got into working on crypto. um Before crypto, I was working on some more traditional topics, what we call market microstructure within finance. And so market microstructure is the analysis of basically how you would set up the mechanics of a market and how that influences market outcomes. For example, like let's take a limited order book. right What determines whether a limited order book works well or not? What determines what the bid-ask spreads are? um Who makes money? Who loses money? Can we redesign it to work better?
00:04:22
Speaker
And then so I had a paper too in that area of market microstructure. And in the meantime, I started hanging out with some um some people on crypto Twitter. And so that was very exciting. It took me a while to decide that this is actually something I wanted to research on. But I think the promise was basically there were all these people like inventing all these like really interesting new constructs that we really hadn't really seen before and tried by. And then so just as an academic, this is like there's so much low hanging fruit here to try and understand. Does this make any sense through the classic lens of the theories who we wrote before? And sort of are these mechanisms better or worse potentially? um And in what ways than the traditional things that we've analyzed for decades and decades in that academic literature? And then so that's how I kind of got into crypto literature. And then basically it was a very natural fit um for Yiming and I to work together because of the fact that sort of um her expertise is very much in this area of liquidity transformation, which is a very important component of stablecoins. And then I have some expertise in my microstructure, which is also another aspect which kind of shows up in our research on stablecoins.
00:05:24
Speaker
Terrific. So before I ask you to describe the recent work, this recent paper that that we mentioned already on stablecoins, I'm just going to remind our audience that ah This is a very timely moment to be talking about stablecoins in part because there's been proposed stablecoin legislation in the US, although the FIT21 Act was really more of a market structure as opposed to a stablecoin specific legislation. But more importantly, Europe is about to embark on the stablecoin portion of the marketing crypto assets regulation.
00:06:04
Speaker
ah They don't actually use the term stablecoins in MICA. They call them ART's asset reference tokens and EMT's eMoney tokens, where asset reference tokens are a basket of different types of assets that ah are used to create the token, whereas eMoney tokens are really what we would normally think of ah about stablecoins. That is to say, they are fiat currency and single fiat currency specific stablecoins. So the questions around what do stablecoins do and how do they function are very much alive with regulators in Europe right now. They're very much alive with the stablecoin issuers because of the regulatory regime that's coming into effect in Europe. And then, of course, in the United States, people have been thinking about these issues from ah a legislative standpoint.
00:07:02
Speaker
And then just one last thing, ah our friends at the New York Department of Financial Services have released material on stablecoins in the last year. The Japanese FSA has released ah material on stablecoins within the last year. And so this has been an area of particular focus for policymakers and regulators around the world. So with that little bit of ah policy and regulatory introduction, I'll leave it. Which one of you wants to sort of describe the basics of the paper? What was the hypothesis that you all came came to the paper with and how did that play out?
00:07:41
Speaker
I'm happy to start. Maybe let me start by just quickly explaining what a stablecoin is, because I think that was where we started off, which is I think it's important that everyone is on the same page as in what a stablecoin is, what it is not, and what it is doing ah before we can understand you know what are the implications and then eventually also so you know how to optimally regulate them. Because I think often in the news, stablecoins are mentioned in in various different different ways. So I think it's important to understand that if you are holding a stablecoin, meaning the audience, ah then 99% of the time, you are selling and buying those stablecoins in a secondary market. So think of this as like an exchange market like the New York Stock Exchange, like you know any other trading platform. You are just buying a stablecoin from someone else who happens to want to sell, or if you're selling, you're selling it to someone else. You actually have nothing to do with the underlying assets of the stablecoin.
00:08:39
Speaker
okay So when we talk about stablecoins you know as having some assets, that is the issuer holding a portfolio of securities. Think of this as bank deposits. Think of this as US Treasuries. And they are then issuing some shares. um based on these securities. Now, again, if you're just a normal trader, you're not going to be able to go to the issuer and ask for anything from the issuer. You're just trading with another seller or buy on the secondary market. And the linkage between those pools of assets or the issuer and the secondary market um is coming from a process called arbitrage.
00:09:21
Speaker
right So, there exist some specific people or institutions who are able to both trade on the secondary market as well as ah redeem shares with the issuer, meaning that if they see, for example, if the secondary market price goes very low, say below $1, what they can do is to buy up shares on the secondary market like everyone else does, but then they can go to the issuer and say, hey, here's a share. Can you give me $1 in cash in US dollars for the share I delivered to you? And so the issuer then would go to sell some assets and liquidate $1 in cash and give that to this arbitrage.
00:10:05
Speaker
and So that is generally the setup and it's actually really important to understand it for it for any kind of implication. Because the first thing we were able to look at is how does this arbitrage market look like, right? And because stablecoins are on the chain, in terms of the arbitrage market, we can actually figure out exactly how many arbitrages there are, how much arbitrage is being done. And ah the first surprise that we had was that there's very, very few arbitrages. So the biggest stablecoin fiat-backed, Tether, actually only has six arbitrages that are actively redeeming shares on average in a given month. um And that was that was surprising to us.
00:10:51
Speaker
So the fact that you have such limited arbitrage is surprising because you can think of basically the arbitrage setup maybe as the following. like right Under one hypothesis, you basically have this kind of two-layered market structure with secondary markets where most people trade stablecoins and primary markets where a couple people can just do that transit transaction with the issuer. You think of that as basically a logistical hack to basically make it easier for people to access liquidity. right It might be difficult for me to interface with an issuer and sort of deal with the redemption and creation process myself. So we delegate that to a class of people who are responsible for the logistics. But if that's the case, we should have as many of these guys as possible. So basically, um the arbitrage process is as efficient as possible. Right. And then in that world, what you would see is that you would go on Coinbase by an answer somewhere. You would sell Tether or something like that.
00:11:45
Speaker
And immediately, a very large number of people would compete to basically whenever there's a small dpeg, take that coin, buy it, redeem it, and then sort of um convert that to dollars, right? And then so basically that would be functionally identical to me just going to Circle or Tether or another issuer and redeeming my stablecoin for dollars, right? You would have an arbitrage sector that serves a purely logistical purpose that's extremely efficient. And so basically the law of one price, these coins are never worth less than a dollar. But the fact that we see so few arbitrages
00:12:16
Speaker
suggests maybe that this serves more than a logistical purpose, right? It suggests that there's not enough arbitrage, as we'll show later with more evidence, um to make this arbitrage process very efficient. We see, as we'll show later, that the prices of tether coins often de-peg upwards and downwards from a dollar. A lot of trading has to happen, a lot of buying and selling has to happen before these guys jump in and try and buy and sort of make these prices peg to the dollar. And so the fact that it seems intuitively obvious that arbitrage is more efficient, the more competitive and the more arbitrages there are, and nonetheless, these issuers who can decide how many arbitrages to include do not choose to make arbitrage perfectly efficient suggests that there might be some actually economic motivation. Our jumping off point is what if the choice of having so few people who are allowed to trade these for dollars is a purposeful choice? Is there any story why it could be?
00:13:08
Speaker
that the issuers are purposefully limiting arbitrage in this market. Could there be some economic function that this serves? So that's kind of the jumping off point of where we want to start in this paper. So let's leave that cliffhanger for a moment, um because I suspect you to have some answers to that to that question that you just posed, Anthony. um I want to just back up and and just make sure that we're ah talking terms correctly here. So, Yiming, at one point you said that that the stablecoin issuers are issuing shares. I think you were using shares in the colloquial sense, not in the sense of issuing stock. um right this This is not the equivalent of a money market mutual fund, for example.
00:13:52
Speaker
um that the stablecoin is is not a um a quote unquote share in some legal entity, but rather a claim that allows somebody to redeem. And that somebody, I guess in this case, Anthony, is these arbitrageurs, right? Because the redemptions You and I can't hold USDT and then take my 100 USDT to tether and get it redeemed for $100. It's these arbitrageurs, ah that that second market, that second level of the market that that is handling that those kinds of trades. Is that basically right?
00:14:38
Speaker
Right. I think, yes, maybe we can just, from now on, instead of shares, I think maybe we can just use stablecoins. so I think the issuer issues stablecoins um that only the arbitrageurs can redeem this issuer. Everyone else can only trade stablecoins on the secondary market. And when also we use the word price, and when the news refers to word price, what everyone is referring to is the price of this stablecoin on the secondary market.
00:15:11
Speaker
And so maybe the analogy is exactly with money market. Mutual funds arbitrageurs can go to the issuer to redeem stablecoins like everyone can do with a money market fund. But with the stablecoin, everyone can only trade stablecoins like people are trading ETF shares on an exchange. Right. And so so that brings me to the second point I wanted to make sure I had clear, which is Not only can I not take my 100 US DT or USDC or pick your favorite stable coin and go to the issuer to redeem it, I also can't take that to the arbitrageur and get US dollars for it. right All I can do is trade it.
00:16:00
Speaker
Right. I think the arbitrageurs also trade on the secondary market, so it's an exchange market. But yes, it is not a bilateral trade with the arbitrageur, it is a trade on an exchange. Oh, yeah. So just to jump in, I guess just highlight what I think is kind of interesting about this. I think there are two concepts which sometimes um are confounded a little bit in colloquial discourse on stablecoins. The concept that a stablecoin is backed by a dollar of assets and the idea that, oh, if there's not a dollar of backing, clearly this can't be stable versus the concept that a stablecoin is redeemable for a dollar of assets.
00:16:36
Speaker
These sound very close to each other but are actually very distinct, right? Sort of one concept is basically does the issuer have enough assets such that the value of the assets is big enough to support the amount of outstanding stablecoins? Naturally you might think a stablecoin should not be worth a dollar unless there's at least sort of enough backing assets, right? However, there's a separate concept that basically most people in the market cannot actually make good on that promise, even for a backed stablecoin. Right. And then so I think like one of the starting points of our work is to highlight basically that sort of how backed a stablecoin is and how many people can actually directly take these coins and redeem them for a dollar can be quite distinct concepts. And so maybe that is motivation. We'll kind of dive into a sort of framework which talks about the difference between these two concepts. And that's kind of the goal of what we want to do in the paper.
00:17:22
Speaker
Yeah, and so I don't want to derail us, but i'm I'm going to make an observation, which is a lot of people think that banks have 100% assets backing the money that they have and That's kind of true, but it's also not like if, you know, that's how you have runs on banks. If everybody shows up at the bank at the same time and asks for their cash, the bank does not have enough cash at any given moment to to give that to to everybody. So there are at least some similarities with bank accounts too. Right. I think the similarity is that the bank is old they also holding
00:18:06
Speaker
probably even more illiquid assets, and that the claims that are issued, which are deposits for the most case, depositors can go to the bank and ask for the money back. And if everyone does it at the same time, the bank has to sell a lot of illiquid assets. and that is often not going to be enough to meet all the deposit claims. And I think that is common between banks, between money market funds, and with stablecoins in the sense that if everyone wants their liquidity at the same time, ah
00:18:40
Speaker
the stablecoin would have to sell some assets that are illiquid, and that is where the risk of liquidity transformation comes in. But as Anthony mentioned, it's important to notice that for the stablecoin, not everyone can go to redeem their stablecoin. And that is different because everyone can go to the bank and ask for the deposits back. But in this case, only the arbitrageurs can go to the issuer and ask for that one US dollar. Everyone can just hope that on the secondary market in which they're trading, the price will allow them to get the money back. But whether that is the case and what is the effect of that arbitrage sector um that is sort of in between the secondary market and the issuer,
00:19:26
Speaker
That's going to be very important, interesting in our paper. So that was the cliffhanger question that that Anthony, that i that i I caused everybody to have to pause on. So since you got us back there, Yiming, tell us what what what happened. What did you guys find? Okay. ah So, you know, again, the cliffhanger was the fact that there's so few arbitrages, right, which is a little surprising if you think arbitrage is good for making that secondary market price being close to one. After what all, that is what a stablecoin is promising, it's to have
00:20:04
Speaker
a coin whose value is relatively stable. So we can use it for payments, for doing transactions. So if you have a lot of arbitrage that's very efficient, that would mean that once the price in the secondary market is not one, right the arbitrage process should make it go back to one very, very quickly. Because the arbitrages can always ask for $1. for each stablecoin from from the issuer. so So that's why we would have expected that if the only thing the stablecoin issuer cares about is the stability of the secondary market price at $1, then they should allow for competitive and efficient arbitrage reflected in allowing a lot of arbitrages. But that is, again, not what we see. We see that some of the largest stablecoins only have very few arbitrages in a given month.
00:20:54
Speaker
and be at the same time see that their secondary market price is not always at one. So sometimes people use the word DPG, but again, that only means the secondary market price, which is a price on an exchange, goes above and below one. And so there seems, there has to be some other reason why the issuers do not want to have so much arbitrage, right? Because if price stability was the only goal, they should have allowed as many arbitrages as they could have. And so what we find then is that there is a cost of having efficient arbitrageurs, or having a lot of arbitrageurs. And that cost is, in a way, just the flip side of the coin of price stability. Because if you have more arbitrageurs, what it does is that if I'm selling on the secondary market as an investor, I'm thinking, oh, should I sell or not?
00:21:47
Speaker
I would want to think about if I sold right and what price can I sell. right And the higher the price at which I can sell, the more incentivized I am to sell because I'm going to get more for it for each sale. Now, what arbitrage does is to ensure that if I sell and I have a price impact right because I'm putting selling pressure into the market, the arbitrage can then very quickly absorb away that selling pressure. meaning that if I sell right with more efficient arbitrage, my incentive of selling is improved because the price impact of that sale is reduced.
00:22:23
Speaker
So think about it this way. is If I sell, I know there's a lot of buyers. right Then I would be more happy to sell because I know that I'm going to get a better price. and That is the world's efficient or more arbitrageous. In contrast, if I sell and I think there's no arbitrageous buying this up or there's no buyers in the market, then I probably would suffer a very low price and that makes me less incentivized to sell. And so having efficient arbitrage also means that people on the secondary market are more incentivized to sell. They're also more incentivized to sell when there's a lot of other people that are selling, because you always think that there's enough arbitragers in the background who are going to just buy up everything when you sell.
00:23:08
Speaker
And what we show in the theory is that in the end, this more incentivized selling actually leads into m amplification of there being a run on the stablecoin by the investors in the secondary market. Because what is a run? you know We know it's a bank run. It's when everyone runs to the bank and wants their money back. In other words, it's if everyone wants to run to the stablecoin and wants to cash out. Now, they can't do that directly. Their way of cashing out is just to sell. And the presence of more efficient arbitrage incentivizes selling.
00:23:45
Speaker
and thereby incentivizes running. When you think other investors are running and selling, you want to run more as well. And so that is what we find as being the cost of more efficient arbitrage. And so therefore, you know it's like a two-edged sword. On one end, the an issuer wants to have more arbitrage so that it can maintain a better and more stable price on the secondary market. On the other hand, this stablecoin issuer knows that it cannot have everyone wanting to sell at the same time, right because that will eventually force the issuers through the arbitrage process to sell the assets. And as we discussed, as was the bank, some of these assets are illiquid. So then therefore, they also have this incentive to constrain arbitrage so that not all
00:24:32
Speaker
the investors on the secondary market want to sell, and not all the assets will get redeemed. um And in choosing between price stability and that run risk, they will determine the amount of arbitrageurs and the degree of arbitrage that they want to allow. I'll jump in and just quickly add an analogy to kind of illustrate that point. In the picture we're drawing of the stablecoin market, you can think of there as being maybe on the right here, like supply and demand pressure, like people who want to buy or sell the stablecoin. And you can think of here on the left being like the issuer. And the issuer responds to that ultimately by if people want more stablecoins, they buy more assets and create more backing assets and expand the stablecoin. And if people want less stablecoins in itself, then the issuer sells a bunch of assets.
00:25:21
Speaker
And the role of arbitragers is basically a pipe that connects these markets. So basically when people want to buy or sell, the arbitragers respond to that and then turn those buys and sells into creations and redemptions of the issuer's assets. And essentially, in our picture of the world, the issuer chooses how wide or narrow this pipe is and trades off the following idea. If you make the pipe very wide, then shocks here are very quickly translated to creation and redemption. That's good from the perspective that prices will never deviate from what? Because any arbitrarily small demand, shocks have very efficiently transmitted through this pipe into the issuer's creation and redemptions.
00:26:02
Speaker
However, you have the issue that if a lot of people are selling, then that immediately or very quickly forces the issuer to sell a lot of assets and creates basically the classic run on a bank issue. And so what we're saying is basically the trade off is if you make the pipe narrow, then you can have a lot of like buys or sells and the arbitrators don't actually transmit those very efficiently. Prices fluctuate a lot because of the fact that these are not transmitted efficiently and you don't have redemptions or creations very quickly. However, what that also means is a lot of selling pressure is going to lead prices to decrease, but the issuer doesn't actually have to sell that many assets. And so really the picture we're painting is that the arbitrage concentration is akin to basically choosing the width of the pipe that connects the secondary to the primary markets. And there is some economic motive not to always have that pipe unfold throughout all the time, basically because of the fact that you want to minimize sometimes or you want to lower the transmission of negative shocks because of the fact that a large, well-transmitted negative shock
00:26:59
Speaker
can lead to a bank run, and by making that shock transmitted less while you sacrifice price stability, but actually may increase financial stability and lower the risk of bank runs. So I have about 1,000 questions to ask you guys. I'm just going to start with one. ah and And I think, ah Anthony, that last point you you made ah perfectly anticipated this question that I was going to ask. So so the those more limited number of arbitrageurs is not a price stability mechanism, but a financial stability mechanism for the issuer. Do I have that right?
00:27:37
Speaker
and And when you're saying, and and so when I say not a price stability mechanism, in other words, it's not helping to stabilize the price of the of the coin. It's not helping stop a DPG. What it's helping stop is a run on the issuer and therefore the issuer having to sell a lot of assets. Now, wouldn't if the issuer has to sell a lot of assets, doesn't that have a an effect, again, on the price of the of the stablecoin itself?
00:28:14
Speaker
So I think this is where we think um our contribution is, that we're giving a result that in many senses is it unintuitive. Think about when various stablecoins depeg, what do you see in the news? What you see in the news is the stablecoin is trading below a dollar, therefore that must be a financial stability issue, right? Sort of depegs or prices trading below a dollar is a sign that the stablecoin is about to collapse, there is a run, right? And so the thing is we kind of like critically look at this and ask sort of in a model, does that make sense? What drives stablecoin prices to trade below a dollar? Right. And so in a sense, we're pushing back against that idea that sort of there are different forces that drive stablecoin prices to depeg and that drive runs to the issue. Right. So let's walk through the process. Right. Suppose on the one hand you have an issuer that is
00:29:07
Speaker
um very liquid, facing a run, facing the issue, that people are selling assets. And then if the issuer tries to meet all the redemptions, they run out of cash very quickly. And then there are other assets they can't liquidate quickly. Suppose, however, that um arbitrage is very efficient. What you're going to see is basically that as people sell, every time they sell, arbitrageurs can very quickly redeem these for a dollar until they can't. So on the path towards the issue of blowing up or refusing to meet redemptions, prices are very stable.
00:29:39
Speaker
right Suppose, on the other hand, you have an issuer that has no run problem whatsoever. That's quite liquid. However, arbitrage is very efficient. So what you can think of is basically these coins can be redeemed for a dollar by the arbitrator, but the arbitrator here is either lazy or trying to extract a lot of profits. So basically a bunch of people are buying. And in a perfectly efficient market, prices would be very stable, but the arbitrator isn't doing much. right You could have a situation where, due to pure laziness, basically, of the arbitrator, sort of laziness in a sort of colloquial sense, right you have a situation that a perfectly solvent
00:30:12
Speaker
Stablecoin, nonetheless, has big defects. And to give an example of arbitrage-induced defects, think about in the past in crypto markets, say, the kimchi premium. Think about sort of this premium, where basically, like, due to the fact that it is not that easy to arbitrage Bitcoin from Korean exchanges to, say, U.S. exchanges, you can have price deviations arise between these exchanges. Does that mean Bitcoin is insolvent? Not necessarily. All that it means is prices move. When somebody who is supposed to be doing the arbitrage trade, keeping prices in line is not doing their job. So that's essentially the view that we want to posit. That basically perfectly solvent liquid stablecoins can have prices depeg and that's most first and foremost a problem of there not being enough arbitrage. Whereas sort of stablecoins with liquid prices can actually be um
00:30:58
Speaker
stable wins with stable prices and efficient arbitrage still can have run risk because of the fact that runs are first and foremost driven by issuers liquidating a lot of illiquid assets, right? And so we're basically looking at these news articles saying basically, oh, this price is unstable. As a result, there's probably a run that is not first and foremost. The thing that drives price stability is not immediately a signal of run risk. It's most immediately a signal of not enough arbitrage. And in the light of our theory, not enough arbitrage has the interesting opposite effect of actually being a useful mechanism sometimes to limit run this. Yeah maybe let me just try to clarify I think Anthony give the full answer I would just stop like add in a little bit as a bridge I think Lee what you mentioned as an if the issuer sells assets what is directly changing is the let's say the value of the portfolio that it's remaining right So that is correct and that is you know directly happening. Now, what is important to notice is that no one is directly trading that portfolio. So and the price that everyone is talking about and that investors have access to is the price of these stablecoins and the secondary market.
00:32:13
Speaker
And so what we wanted to understand is the link between that price on the secondary market, the value of this portfolio, and importantly, the hype between these two things, as Anthony mentioned, which is that arbitrage sector. ah Notice that to this date, no matter how much has been sold by the fiat-backed stablecoins, um the claim is that the price at which the arbitrages were able to redeem from the issuer, they claim has always been at $1, regardless of what the secondary market price has been, and regardless of the
00:32:51
Speaker
in a real financial value of the remaining assets in the portfolio. So between the selling of the assets, the value of the portfolio, and how all of that actually goes through to the secondary market price, that is actually not clear at all. And so that is where ah we wanted to formally look at that link, both empirically and theoretically, and and and that link is what Anthony described. And and it sounds like what you were able to show was that the link that the arbitrageurs helped to cushion the the blow from that link and that if there were many, many arbitrageurs, in other words, if the arbitrage market was more efficient, then that link might actually be worse for everybody as opposed to better for everybody. And this gets to the the to the point of
00:33:46
Speaker
ah you know price stabilization versus financial stabilization, which are are the two different things. Very interesting. Yeah, exactly. So echoing that again, basically, um that's precisely our that's precisely what we want to argue, which is that there are two goals. And then the two goals are actually traded off in the choice of how, why do you want to make this pipe, basically. um You can get better price stability, you get worse financial stability, actually, um as you make the arbitrage market more competitive, exactly. And ah so now I'm going to ask you, were were you able to come up with an ideal number of arbitrageurs to to make the pipe the right size? And and if so, what what are the factors for figuring out what that ideal number of arbitrageurs is?
00:34:34
Speaker
Right. I think the key is in, I think we were able to come up with mapping between what factors, meaning what characteristics of the stablecoins and thereby what should you be choosing as your arbitrage composition. And I think that choice was also driven by the empirical observations we have made, which is we find that in the cross-section of fiat-based stablecoins, the more concentrated your arbitrage or the fewer arbitrages that you have
00:35:06
Speaker
In general, the more or the worst price ah stability you have, the more, let's say, quote, quote, DeepEx you have on the secondary market. So Tether, for example, you know, has larger price deviations than Circle. And Tether is the one that has on average six arbitrageurs. And Circle actually very differently. in an average month uses more than 600 arbitrages. So we see that there is a strong correlation between how stable your price is on the secondary market and what arbitrage sector you choose to allow. And certainly, if you know there's no concerns of financial stability, everyone wants to be like Circle. You want to have a very stable secondary market price by allowing the largest number and the more efficient arbitrage.
00:35:56
Speaker
Now, what is then making Tether and Circle's choice different, as it turns out, is how worried they should be about the run risk, about the financial stability. And as we know from the bank's case, the bank should be more worried if its assets are more illiquid. If everyone wants their money back and they sell, they have to suffer steep discounts for that selling. And we find that that same logic can also be seen across these stablecoins, because on average, Tether's assets are much more illiquid than Circle's assets, meaning that if there's a run, Tether should be much more concerned than Circle is. Therefore, because it's more vulnerable to runs, an investor wanting
00:36:44
Speaker
to take their money and selling the issuer selling all the assets, it is forced Tether is forced to choose a more concentrated arbitrage sector so that it can reduce the risk of rents, even though it does go at the expense of some price stability. Now, CIRCO is running a different business model. On average, its assets are more liquid. It's not 100% liquid, but it's relatively more liquid than that of Tethers, meaning that it's relatively less worried about having to sell these assets.
00:37:16
Speaker
Hence, it can allow itself to have more arbitrageurs because even if people sell their stablecoins, arbitrageurs come and redeem, and some asset sales happen, it's not as illiquid as that of Tether. That's why Circle can allow for a more competitive arbitrage sector ah to have more price stability because the assets it is choosing are a better buffer already against that run risk. In short, What arbitrage concentration should you choose? It depends on what assets you are holding. So to two follow ups there. One follow up is it sounds like ah you did not reach a qualitative conclusion, right? It depends on a number of factors and each stable coin will make a decision as to whether it wants more price stability versus more financial stability. Is that basically right?
00:38:16
Speaker
Yes, i think you you know I think we started from the starting point that there's many constraints on what assets a stablecoin issuer has access to. For example, Tether is offshore. It doesn't have access to as many US dollars nominated assets as Circle does. So to a large extent, there is a constraint on the starting point and what kinds of assets and how illiquid these assets are. So that's why we mapped the choice of the arbitrage concentration that is optimal based on the starting point of the assets. Now, of course, one can go you know one layer above. If one you know wants to be more theoretical and and less practical, one way to say is, OK, what would you do if you want to have minimum amount of run risk?
00:39:00
Speaker
right There, the ah answer is is not surprising. You want to have 100% in 100% liquid assets, and then you can choose fully efficient arbitrage, and you would still not have right risk, and you would have full price stability. right ah So that, I think, is the let's say the absolute answer, perhaps. But it's important to realize that in in real life, in practice, there is no such thing as a 100% liquid acid. right Even you know if before Silicon Valley Bank, you know what's been written across the board is that OVR 100% in FDIC insured banks.
00:39:41
Speaker
And you may think, okay, deposits, it's as liquid as you can get, it's as safe as you can get. And you know after Silicon Valley Bank, we realized, well, FDIC insured deposits may not be insured banks, may not be you know every dollar is FDIC insured. It depends on the amount in the account. And no further notice that there's also a difference between your checkings account and your time deposits. And if you need to take your money out of time deposit accounts, actually that also incurs a discount. The other commonly cited thing are treasuries. Treasuries, it depends a lot on what market conditions we are in. so you know I think broadly speaking, the assets that stablecoins have are not the most illiquid assets ever, but it's also the case that it is near impossible in the world we live in to have 100% of a liquid asset. Maybe that is the only thing you know I can think of is like central bank reserves, which currently stablecoins do not have access to.
00:40:35
Speaker
but So therefore, we can never be in that, I think, the probably the the the theoretical first best, which is to have 100% liquid acids. And therefore, you know that where you are in that acid choice becomes such an important starting point to where you should be in terms of choosing your arbitrage concentration. and And I was going to ask you about the Silicon Valley bank situation because obviously that had impacts on on USDC and some of the other stable points as well that had ah had large deposits there. ah So you anticipated my other question, um so that's great. OK, so I'm going to ask one last last question and it's going to be purely theoretical, which is based on this research.
00:41:21
Speaker
um Can you say anything about how a banking sector in a country functions like the arbitrageurs these between the you know population and the central bank? Is there an analogy to be drawn there, or did I just make something up that is completely ridiculous? Maybe I can jump in and try and talk about this a bit. um I think there is a sense in which this is true. And it's true in the sense that basically there's a broader area, which Yiming and I have a paper on and Yiming has more papers on, which basically asks the following question. Monterey policy in many jurisdictions is implemented through the central bank sets an interest rate. But that interest rate is only directly available to a couple dealer banks.
00:42:15
Speaker
And the thing is the idea is that this should propagate to the economy through the dealer banks and offering that rate to their clients. Right. And then so a larger research, which muing has contributed a lot to basically asks, is that transmission efficient in their role as, in a sense, arbitrators? of the central bank rate into the economy, um to what extent do the dealers actually one-to-one pass through and keep that price linked to the price the central bank says? So maybe even you can just take that from there. I think she's much more into this area than I am.
00:42:46
Speaker
Yeah, no, I think it's a great question. and I think I've never thought about the connection between it that way. I think, as Anthony mentioned, for the transmission of the central bank's policies, both conventional and unconventional, it is certainly you know ideal if that pipe through the financial intermediary sector is more efficient or it's broader or bigger, let's say, so that things pass through ah efficiently, equally to the population, let's say, or to the firms and the individuals in the economy. um I think that in most cases in that scenario, though, that transmission is as
00:43:32
Speaker
it's always ideal, or it's there's very few downsides of having an efficiently transmitted monetary policy. I think that is maybe about the difference here, which is that the transmission efficient transmission through a big pipe is like efficient arbitrage. It is non um-debatably good for price stability. But there is this downside that comes with it which is that of a higher run risk. And and the reason again comes in between there's these eliquid assets and there is that promised stable $1 fixed asset value to the arbitrators. That in the end, you know if you continue to sell and you sell at a discount, you are no longer able to meet that $1. Similarly to how the banks that they continue to sell, you can no longer honor that fixed deposit value contract.
00:44:28
Speaker
And so maybe the the the the downside is a bit of what's what's special in this. senate Yeah, so maybe to chime in briefly, I think the analogy, so I agree with your analogy that in a sense, these are both arbitrage questions, and the prices are more stable, the more arbitrage there is. Just building on what Yiming said, I think the nature of the debate on banks as arbitrators basically is that, as Yiming says, we would like perfect arbitrage. The banks are not perfectly or arbitraging for other reasons, which is that regulatory restrictions imposed on banks um prevent them from being perfect arbitrators. And so we limit their ability to arbitrate for reasons outside um and different from the stablecoin reason of wanting financial stability. And we have to deal with the consequences that regulations that have the benefit of making the arbitrators in their role as banks more stable and less risky and preventing perhaps things from o eight from happening have the incidental consequence
00:45:19
Speaker
um negative consequence of making monetary policy transmission more efficient. So I guess that's like the similarity and difference to this scenario, but so in a sense like the same like broad story is interpreted in kind of the light of different policy trade-offs and that's a pretty active line of work um that people are working on within finance academia. Maybe let me just add another, a bit of a narrower analogy, but we get asked this question a lot, which is, you know isn't this just an ETF? like First, I think maybe the first thing that people have meant, isn't this just a money market fund? And there the answer is no, because you cannot trade money market funds on the exchange. And so that's already a very, very big gap. And then I think the structure of an ETF is a bit closer to what the structure of a stablecoin is. Because the ETF is also you trade, if you're an everyday investor, you trade ETF shares on exchange, like you're trading trading stablecoins on an exchange. And there also exists a small group of people who can go directly to the issuer, say BlackRock, and go and redeem the ETF share, like the arbitrages can redeem the stablecoin.
00:46:23
Speaker
from the issuer. Now, the difference between the ETF and the arbitr but and and the stablecoin context, though, is if in the arbitrators for the ETF, they're called authorized participants, when they go to the issuer, what they get get is the actual asset. They get a basket or the like a portion of the actual underlying assets. And therefore, what they're getting, the value is going up and down with the actual value in the assets. right so Earlier, you mentioned le if you know the issue of selling some assets, that's going to be incorporated in what the arbitrages are getting for the ETF because you get what's remaining. in a way If it's no longer there, you don't you you don't get it.
00:47:02
Speaker
now The problem, perhaps, with stablecoin is that the arbitrageurs are not getting a fair share of the underlying value. You're getting a fixed $1 US per stablecoin value, regardless of what's left. in the underlying portfolio. And so that's the beginning of the friction that later then transmits through the arbitrage sector to the secondary market. But I think that is the reason why, for the stablecoin, there is this cost of having a very efficient type through the arbitrage sector.
00:47:42
Speaker
Well, listen, you two, this has been super interesting and super informative and ah even a non-economist lawyer like myself was able to understand your explanations. So you must have been pretty darn clear for us. I really want to thank you both for a great conversation. We will post the paper ah in addition to the podcast so that people can actually go to the source material and read the research and and conclusions and and everything. um This was really terrific. Greatly appreciate your time. Thank you very much.
00:48:20
Speaker
We hope you enjoyed our Hootenanny. Thank you for listening. For more Hootful and hype-free resources, visit www.owlexplanes.com. There, you will find articles, quizzes, practical explainers, suggested reading materials, and lots more. Also, follow us on Twitter and LinkedIn to continue wising up on blockchain and Web3. That's all for now on Owl Explains. Until next time!