Become a Creator today!Start creating today - Share your story with the world!
Start for free
00:00:00
00:00:01
Ava Labs x CBER Ep 4: Economics of Lending Platforms
6 Plays11 months ago
This episode discusses lending platforms on blockchains. Thomas Rivera (McGill University) and Quentin Vandeweyer (University of Chicago) explain that lending platforms generate sub-optimal welfare due to under-utilization of funds that are lent to the platform. To provide more context, lending platforms are specified in such a way that the level of borrowing (relative to lending) necessarily fluctuates with market conditions, resulting in instances where borrowing is significantly below lending. Importantly, when borrowing is significantly below lending, the total interest accrued from borrowers (relative to lending volume) is necessarily low, yielding low interest rates for lenders and thereby discouraging lending. Potential improvements for lending platforms are discussed.
Recommended
Transcript
Introduction to Series
00:00:06
Speaker
Hello and welcome to Crafting the Crypto Economy. I am Silvia Sanchez, Project Manager at OWL Explains by Avallabs, and today we bring you a transformative podcast series in partnership with the Crypto and Blockchain Economic Research Forum. This series features leading faculty from renowned global universities exploring various elements in the blockchain ecosystem. These episodes are a bit longer than our usual hootenannies, since we will be getting very deep.
00:00:33
Speaker
and also each episode will have its accompanying paper posted on our website for further reading. And with that, I will hand it over to our moderators Fahad Saleh and Andreas Park.
Focus on DeFi Lending Markets
00:00:43
Speaker
Hey everyone and welcome back to another edition of the Crafting the Crypto Economy podcast series organized by the
00:00:51
Speaker
cryptoeconomics and blockchain research forums, CBER and OWL. Today, we're going to talk about DeFi lending and DeFi lending markets. Just as a reminder, our first podcast covered the trading of crypto tokens. If you think about a wealth of finance, a natural thing is that people need to be able to trade things. And the second iteration, of course, is that people in the financial world want to borrow on
Meet the Experts
00:01:15
Speaker
lend. So we're very happy to have the authors of a new paper
00:01:20
Speaker
titled Equilibrium in a Defile Lending Market. We have Thomas Rivera, Quentin Vanderveer, and Fahad Saleh, who will have two hats on today. One will be as the host, and the other one will be as one of the authors of the paper. So welcome, everybody. Thank you for having us. Yeah, it's great to be here. Thank you.
00:01:42
Speaker
So maybe we can start very briefly, starting maybe with you, Thomas, to introduce yourself so that people have some background on where you're from and what your journey is to crypto.
00:01:52
Speaker
Sure, yeah, I'm a assistant professor of finance at McGill University. I have a background in economics, a PhD in economics, particularly focusing on the microeconomics. And yeah, recently I have a deep interest in the economics of blockchain, both in terms of consensus protocols, as well as decentralized finance and these are very interesting applications. Very nice. Thanks. Thanks so much. So Quentin, how about yourself?
00:02:22
Speaker
I'm also an assistant professor of finance. I'm at the University of Chicago in the business school. So my journey to cryptocurrency is really coming through stablecoins because my background is more in asset pricing and monetary policy in banking. So stablecoin was kind of like the way I got interested in the DeFi environment and then from there on expanding, I guess.
00:02:46
Speaker
Very nice. All right. So then maybe let's just jump right into the topic area.
Understanding DeFi Lending Platforms
00:02:52
Speaker
So I think what would be very useful is for people, for listeners to just generally understand what we actually mean if we talk about defile lending. Now,
00:03:02
Speaker
There's actually various different forms out there. If you think about some stablecoins, you can think about those actually also as their products. But I think we have something very particular here in mind. We have in mind lending platforms where people deposit tokens and other people can take up loans. But before I blabber too much about this, Thomas, going back to you, how would you describe DeFi lending in a very succinct way?
00:03:27
Speaker
Yeah, so I think you sort of nailed the very basics of it. People can deposit crypto assets into these smart contracts. Once they're deposited, others can come and borrow those assets and there's a sort of way that interest rates are determined almost mechanically through the design of the smart contract.
00:03:48
Speaker
You did touch on this fact that trading is a very natural part. It's a natural feature that you'd want to see in the blockchain ecosystem. I would sort of bear that when you think about DeFi lending with a type of securities lending. So if you really begin to understand what are these
00:04:06
Speaker
DeFi lending platforms used for, it's not quite the home equity type story where you put your poster houses collateral and get some cash to spend. Instead, it's more along the lines of when you have information about cryptocurrencies and you want to trade it in a certain direction. Let's say you want to short a cryptocurrency
00:04:25
Speaker
That's precisely the service that these DeFi lending platforms are providing right now, and this is a service that you can't find elsewhere on the blockchain, so we do see it as sort of a vital service.
00:04:39
Speaker
Okay, that's very interesting. Now, taking a step back though from the usage of them ultimately, because as you can imagine, you can use crypto assets in a variety of different forms. Let's try to get more a little bit at the core of the function analogy.
00:04:56
Speaker
Let's say, would it normally work? Say, I am a user. I have, let's say, a stable coin. I want to put it to good use. How would I go about this if I want to become, can I become a lender by myself? How would that work?
00:05:12
Speaker
All right. So yeah, I think that's exactly right. So that's, that's what, uh, that's the kind of service that, uh, defile any platforms going to provide. So you have a smart contract that is out there that allows everyone that has some form of a crypto assets, basically deposit this asset into the smart contract.
00:05:29
Speaker
lock it so that somebody else can borrow that particular asset. The role of this smart controller, this platform, is going to pull all these assets together and ensure some form of safety and some form of remuneration for you to land something called an interest rate, normally in also in this case.
00:05:48
Speaker
So in some sense, you know, I pool my resources and it's actually not very much unlike what happens in a bank, right? So people put their deposits in the bank and then the bank lends the funds out to third parties for people who want to take up the loan, right? Exactly. Okay. And so then how do you, where does the interest rate come from? I mean, how does that work?
00:06:09
Speaker
Yeah. So this is sort of a unique feature of these DeFi lending platforms is that the interest rate is set sort of internally based on a parameter, which, which is effectively the fraction of the loanable funds that are being borrowed. So when this, when let's say you have 50% of the funds supplied to the contract are being borrowed, that's going to equate to an interest rate, this sort of interest rate at 50%.
00:06:37
Speaker
When the fraction of funds being borrowed moves, then so will that interest rate. And you can very easily see this functional relationship, for example, by going on to the website of Aave or Compound. They'll show you exactly as the utilization of funds changes how the interest rates are going to change.
00:06:57
Speaker
Another important feature, sometimes unlike a bank, which is the fact that the interest of the borrowers, most if not all of this interest is going to be passed through to the lenders. And so this is also something that you can very sort of concretely see as a lender if 50% or 80% of the funds are being borrowed, exactly how much interest will the borrowers pay and exactly how much interest will the lenders receive.
00:07:26
Speaker
Maybe we could provide a little bit more context in terms of the actual sort of why it's designed this way. So can either of you talk a little bit about sort of the frictions that are at play here that cause this to be the process or that led people to specify a process like this?
00:07:46
Speaker
Yeah, I think I can maybe take that one. So Thomas and I probably did a little, I think the way we think about that is like the mostly two main frictions that matters here.
00:07:57
Speaker
For me, there's one that's going to be something called liquidity. The fact that you would like at some point in time, when you're a borrower and you need cash, it's convenient for you to have a platform that actually has some cash on hands that you can withdraw from. That's going to be a big advantage from you as compared to just a platform where you can submit a bid and then wait for a day so that they collect bids on the other side, then eventually they match you together and then they give you the loan tomorrow.
00:08:25
Speaker
So I think that's one of the reasons why they find it convenient to set up. Everything is a function of something called the utilization rate. So the pool of cash that's available to the lenders so that the cash is sitting there and it's available whenever it's ready.
00:08:43
Speaker
So in many ways, if I may just interject here. So if you go over, bring this back to a lay person. So in a certain sense, it looks very much like what normal financial institutions are doing. You put your funds into a particular pool. The pool is used to send money out. And I presume, as you describe this, Thomas, there is some form of increasing costs for a loan depending
DeFi vs Traditional Financial Systems
00:09:07
Speaker
on what the liquidity set up in the market is. I mean, we see this also in other markets.
00:09:13
Speaker
If people require a lot of money, then eventually the price for those loans would increase. Although we can argue in the real world there is, because the central banks had a benchmark, they set basically a floor for how much you can get elsewhere. There's a bit of a, the market is, there's not full market utilization if you want. So in this case here, what is really the critical part, if I can just take the different items that you and Quentin
00:09:41
Speaker
Thomas and Quentin have said is, it's number one is the immediate availability. So there is no question that I have to look at a loan officer, it's just entirely programmatic and automatic. And the second part is that, instead of what we, you know, in our world as economists, we like to think of people come together on the market, and they put their bids in, and then miraculously, you know, the invisible hand creates the equilibrates demand and supply. Here, this is really just hard-coded.
00:10:09
Speaker
Right? And so the question that arises then is how good is this hard coding? And is this any fair reflection of what a market would come out? Is that roughly what you're trying to accomplish? Just to clarify. So what we're talking about here is that effectively you can have on demand loans, right? And they're trying to ensure that I want to borrow, I don't have to wait four days to borrow. I can get my loan priced up basically immediately and get my capital for the loan. Is that right?
00:10:56
Speaker
Yeah, so there's a lot of sort of features that go into making lending work on a blockchain, right? One of the big concerns is this is there's anonymity. So when I come and borrow, you don't know who I am. And if somehow I could get away with not paying the loan, obviously I would take advantage of that.
00:11:15
Speaker
So I think it's worth mentioning that there's also all of these loans are over collateralized. And so this sort of levels the playing field where provided that you have enough collateral for your loan, the lenders are not worried about you sort of running away with and never repaying them.
00:11:33
Speaker
Now, once you have that in place, the lenders, you know, let's say a well-functioning collateral mechanism in place, the lenders are basically going to be thinking, okay, all of these loans are going to be repaid, and sort of the only challenges left are how do we set interest rates.
00:11:53
Speaker
So the design for the utilization, so it starts with a very simple idea, right? As you have fewer funds, as Quentin was alluding to earlier, as you have fewer funds available to be lent, you would like to see interest rates go up to make sure that it's the most promising loans, the loans that generate the most value that are the ones that are borrowing those last few precious dollars of loanable funds.
00:12:17
Speaker
So I think it's very natural to say, okay, let's take utilization of funds. Let's create interest rates, which basically say when utilization goes up, more and more of those funds are being borrowed. So do the interest rates similar to a traditional sort of market forces that would have interest rates go up when there's a high demand. So
Risks and Liquidity Management in DeFi
00:12:36
Speaker
now I have a little bit of a sidetrack question, and I'm wondering if you actually can answer this. This all sounds really
00:12:46
Speaker
well thought out when you're, say, far from 100% usage of all the funds. But there is a concern that you might have, which is that if somebody, so I imagine the following happens. As you take up a loan, you accumulate interest balances. So depending on what the interest rate is, over time, you just owe more and more and more to the pool.
00:13:10
Speaker
And you kind of have to worry about whether or not the pool is actually well capitalized. Is there actually enough money there to satisfy, say, withdrawal requests? So if I want to take my money out, can I actually get that out? In particular, can I get it out with interest? Can you speak maybe for just a minute or so, maybe Quentin, I've seen you nodding about how that is actually handled on a practical level?
00:13:36
Speaker
Right, so before getting there from mine, so I'm going to get there, not sorry, but I think the idea is that you're touching to a fundamental trade-off. I want to introduce that one first. A fundamental trade-off is that, in a way, having a lower authorization rate than 100%, one might see that as being wasteful, because now you have some people that are lending some money that that money is just hid out.
00:13:57
Speaker
not lend back to somebody else. So that actually has some very practical implication, means that now the interest rates that you're going to take from the borrowers and then move on to the lenders is going to have to be reduced.
00:14:13
Speaker
because you are going to have more lenders than you have borrowers. That's some kind of an efficiency. But then on the other side, as you mentioned, when you are lending to the platform, you also want to mitigate some kind of liquidity risk or even like some sovereignty risk for the platform. You accumulate these interest rates, but you don't know if you're actually going to be able to withdraw.
00:14:35
Speaker
So the way the platform is handling this kind of situation is actually by doing some form of redemption gates, or they are preventing you from
00:14:48
Speaker
taking your money out whenever the utilization rate is too high. And so that's actually a source of liquidity risk that whenever you're lending to this platform, you have to think about. And then beyond that, if you imagine like some very bad event would happen on the collateral asset or whatever, and you cannot withdraw, then you might also pay some actual risk where your money is actually locked there and then there's not enough money for everyone.
00:15:10
Speaker
Again, that kind of like draws to the analogy with banking where you might have like some forms like a banking alone, banking run, or some kind of issue of that nature. Just to ask, is there also some form of equity buffer in any form?
00:15:30
Speaker
Yeah, so that's correct. So it depends on the exact platform. Maybe I know Farhad has better knowledge of the institutional details here than I do. My understanding is that there's two part of the buffer. The equity buffer is decomposing to two parts. One part is just the digitization rate. You can think about the digitization rate or like the other side of the digitization rate, like the 20% of non-itization rate that you might get.
00:15:56
Speaker
That's going to serve as implicit equity on the platform. That's going to protect against like no-claim price and the collateral assets. But also, on top of that, you might have some additional equity trenches in some of these platforms. I believe there's like some reserves, but I don't remember.
00:16:16
Speaker
I can tell you, for example, Aave, they take about 15% of all the interest paid and they put this into a reserve fund and it's used precisely for buffers to worry about insolvencies. Again, I don't want to get into too many of the details about the collateral mechanism. The idea is that it works and so basically loans
00:16:38
Speaker
whose collateral is dropping in value, they will be sort of closed out mechanically before the collateral value drops below the loan value. Now, this is only going to hold in, let's say, normal times. There is slight risk that when there's high blockchain congestion, you might miss some of these loans, and this might lead to insolvencies. And so that's also a sort of use for these reserves, which is to say,
00:17:05
Speaker
if the collateral mechanism fails because of things out of their control, let's say, high congestion on the blockchain, these reserves are here to cover losses, to provide, to have some extra liquidity if people want to pull their funds off of the platform. And so you have in tandem the reserves as well as the fact that all of these platforms are targeting utilization rates strictly less than one.
00:17:30
Speaker
Well, that means if your target is 80%, that means you're always going to have 20% of funds available. Those are funds that people can come and borrow when they really, really want to. But these are also funds that lenders can use to sort of pull out their loanable funds when they also want to stop lending to the platform.
00:17:52
Speaker
So we're talking about now sort of a couple of different things, I think, as it pertains to these lending platforms. But I wanted to kind of take a step back and think sort of about some of the more fundamental aspects here again for a moment and kind of relating to some of the discussions that just happened now. So Thomas just sort of referred to this point that, you know, part of the interest payments from borrowers, they don't necessarily go to lenders.
00:18:18
Speaker
Right and that seems to be kind of an important thing, right? But then, in fact, 1 way to think about so we talked about we were discussing what how interest rates are set. And so, we said, well, they're set as a function of this utilization rate. And so that's where the bar interest rate is set. What do we do with the borrower payments?
00:18:36
Speaker
Well, I guess the ideal thing is that these borrower payments are going to be given to the lenders, but apparently not entirely because there are other things to be concerned about. Like, for example, if there's actually some capital still left over. But then, so I guess maybe asking kind of a big picture question here to either Thomas or Quinn.
00:18:58
Speaker
What do you see as sort of the high level pitch for people in the crypto world as to why having this sort of mechanism in the first place is useful? When we compare pricing, let's say, in this setting to any sort of more traditional setting,
00:19:19
Speaker
I think we've sort of been talking around this point that there's an effect like more discretionary abilities in the traditional settings. And a lot of people feel like, well, that's great because the world is really dynamic and it's good to incorporate all of that in a discretionary way. What's the trade off to that though? So what's the crypto enthusiast answer to somebody saying,
00:19:46
Speaker
Hey, the world's really complicated. I don't like this idea that you're going to sort of make this thing mechanical, even if it's a function of utilization. I mean, lots of things can happen in the world. What is the flip side to that?
Challenges in Setting Interest Rates
00:20:01
Speaker
Yeah, so I would say when you look at the way interest rates are set in practice, in many markets, especially small markets, markets with very few banks, very few competitors, these interest rates may not always fully reflect the underlying states of the credit market.
00:20:21
Speaker
So you may get a high interest rate, even though in other parts of the economy, other lenders can get much lower rates in a more competitive environment. So the first thing would be, if you take discretion out of large intermediaries hands, especially those with large market power, there's a world where you can actually make the end user better off because they're getting competitive rates.
00:20:46
Speaker
Now, does the DeFi lending platform actually offer you this? Well, it's not clear because effectively, this boils down to a thing called the Oracle problem, which is that the smart contract, ideally, you'd like it to be like a bank where the interest rate it sets is some variable rate, LIBOR, the LIBOR rate plus some percentage, right?
00:21:10
Speaker
The problem with blockchain is that feeding that information to the blockchain is very, let's say, difficult, especially if you want to really rely on the information that's being fed. So while smart contracts can read information, it's much more difficult for them to read information that wasn't created on the blockchain.
00:21:31
Speaker
Now utilization rates, which is basically what is determining the interest rates in DeFi lending, that's mechanically information that's produced by the DeFi lending smart contract, whereas other rates like the Fed's fund rates or LIBOR, this is something that's produced off of the blockchain and therefore becomes very difficult to incorporate onto the blockchain. So if I may, I would like to take another step back, right? So just to summarize what I've heard so far, right?
00:21:59
Speaker
Again, we create a lending pool where people submit their assets so that they can be borrowed out. There is a function that determines the interest rate. This interest rate depends on the mutualization rate, which is really just the ratio of the amounts borrowed and the amount made available.
00:22:22
Speaker
There is an interest rate spread being charged and the spread amount if you want to go to a reserve pool or if you want to the equity of the platform. So all of that actually sounds pretty much like a bank in many ways. Now, as Thomas, you pointed out, one of the issues is we don't really know everything about the market. And we would assume, for instance, if we say think about we have the various different platforms,
00:22:46
Speaker
And there would be different, you know, there would be, you know, essentially you could say that in many ways to emulate banks, except that there is a little bit that is missing, right? And as you saying, essentially, this is some information that they're missing, that they can't fully incorporate. Now I have a question here. So there is an interest rate, which depends on the utilization rate. Now that's a function. Now, very naively, I would imagine this is maybe some form of a upward sloping function. So who sets the slope?
00:23:16
Speaker
Oh, that's a very good question. So the answer would be like getting back to the banking energy, this would be the equity owner or like the manager of the bank or the platform. So the way it works, I guess very often these defile environments, there's also another token that's going to be issued that will serve the roles as an equity token.
00:23:39
Speaker
And we'll serve the world as an equity token in two different fashion on the two different roles of what we think about an equity share typically. So we'll give some entitlement to the underlying profits that the firm is going to make. We're also going to give you some role in the management of the company or in this case of the platform.
00:24:06
Speaker
And so here, the difference is that it's going to be a little bit more decentralized in the sense that there's going to be an ongoing voting mechanism that happens online and are quite efficient. But at the end of the day, I think this is actually quite close. So you have, in the first place, you have this function that says interest rates. That's automatized. But you still need to decide on the slope. And the decisions on the slope are going to be made by these equity orders.
00:24:34
Speaker
I see. I think this is very interesting because fundamentally, this brings me to another difference that's there. In a bank, you can argue that the decisions of the lending rate that is available to borrowers for whatever it is that they want to borrow is really determined if you want by a committee.
00:24:52
Speaker
It could be by the board, it could be by a lower level manager, but there is a certain outsourcing of information of the decision from the shareholders to a manager. What you're saying here is this is really very direct, it's like a direct democracy. The owners of the token, I imagine you think this is the DAO token, so for a decentralized autonomous organization.
00:25:14
Speaker
Those guys actually get to vote, I presume on a regular basis, like maybe what is it, biweekly or so, to determine what their respective interest rates are. So it's not on a, it will be on a schedule I presume in a way, right?
00:25:29
Speaker
So, just to jump in here, I think what we're alluding to here is the governance tokens, right? And in some sense, this reflects the difficulty of the whole process, right? Because you initially specify this function, and in principle, you could leave the function as is, but in practice, what happens is frequently they decide for whatever reason, it's not exactly ideal.
00:25:52
Speaker
And then there's a proposal that's put forth about changing the function, maybe for a particular market in isolation or something to that effect. But there's a bit of a sort of a trial and error that goes around it. So, you know, we've been talking about slopes. And I don't know whether the more pedantic of our viewers might be sitting there thinking, well,
00:26:12
Speaker
Are you assuming linearity or something like that? But the funny thing, of course, is in many cases, they had actually started with the linear function. And then they decided to get really intricate going to a kink piecewise linear.
00:26:31
Speaker
Now, there are, of course, platforms that do, you know, more involved things. But if we're talking about, for example, Abe, right now they're using a kink function and it is a bit kind of ad hoc, which I guess gets back to this question about how easy is it
00:26:48
Speaker
to specify this interest rate setting protocol in a way that it actually does what you want it to do. And in some sense, even like what exactly do you want it to do? I mean, how do you know, what, stepping back?
00:27:05
Speaker
What is it that these lending platforms are trying to achieve through that infrastrate setting mechanism? And yeah, how hard is it to actually achieve those things? Yeah, this is a great question. So yeah, I mean, basically one goal of these platforms is to always have loans available to be lent
Balancing Interest and Utilization Rates
00:27:31
Speaker
out.
00:27:31
Speaker
Again, borrowers may come and they may have different sort of proclivity to borrow for different purposes. And so, again, you want to design the interest rate function. So this determines how interest rates move with the utilization of the underlying funds.
00:27:48
Speaker
You somebody mentioned it earlier you want it to be upward sloping because when more funds are taken out this is sort of a market signaling that there's higher demand and so naturally to make sure that there's not too much demand relative to supply you always as economists we use the price as the equalizer right so
00:28:07
Speaker
demand increases, now more of the funds are being taken out, you're worried that you might no longer have available funds, but why don't we just increase the price? Some people will decide at that higher price they no longer want to borrow, and others will say, no, at this price I'm still happy to borrow.
00:28:24
Speaker
So this sort of mechanism here is is designed again with this underlying goal of having the funds on hand ready to be borrowed now an interesting fact about this is that it's very hard to think about the stability of these interest rates right so when when when let's say demand in the underlying market increases.
00:28:46
Speaker
utilization is starting to go up, there's a question of when will it stop, right, if ever. And I think this is a fundamental question for the design of these underlying platforms. So this is one thing that we show in our paper, which is that, again, provided that you have this sort of upward sloping,
00:29:03
Speaker
interest rate function, you can actually ensure that whenever there's a change in the underlying supply or demand for loanable funds, you'll always, the utilization will adjust, but eventually it will stabilize until there's another change in the underlying supply and demand, rather than just bouncing around all the time. I know when we first sat down to study this, we were unsure whether that would be the case or not. Yeah, if I may continue on that, so I think that's great.
00:29:32
Speaker
For me, it needs to be efficient and always be at the level that you want it to be, at the decision rate that you target, you would like to have a very high slope. That's for me what Thomas was saying. Now, in reality, there is actually some force that's going to push you against doing that, which is that as a lender,
00:29:53
Speaker
or as a borrower, you don't want to incur too high interest rates when you don't pay attention. Things that remember these interest rates are dynamic. You just put your money there and you don't know how much interest rates you're going to pay tomorrow because that depends on the evolution of the utilization rate. And so if you have now
00:30:13
Speaker
A slope that is too high, that means that there's just a little bit of an imbalance between demand and supply. You might end up with a huge increase in interest rates. And that's very good from an efficiency perspective, because at some point people will realize and they are going to pull out very fast.
00:30:29
Speaker
But from an executive perspective, you are going to have the lender are going to think twice before, sorry, the borrowers are going to think twice before borrowing in such a scheme because they will be worried that if they just leave it for a day, go for a walk and then they come back and then they don't have any money anymore because the interest rate would have eaten. They've moved so much higher that it would have eaten all of their wealth in the meantime.
00:30:54
Speaker
So I think it just became very, very detailed all of a sudden and very intricate. I think we have to break this down for the audience a little more. But I thought, actually, if I had you, you presented actually a pretty good segue to the particular paper, right? So that we're trying to discuss here beyond the general intricacies of the borrowing and lending market.
00:31:16
Speaker
One thing I will say, and I think this is also something which is a little bit misunderstood in particular when I talk about banks and the relationship to banks, because in banks, we usually think of deposits in particular as being quite sticky, so that the decision is really often a question of that the bank tries to pick an interest rate which makes it the most money, to the extent it's possible in a competitive environment. Whereas here, there's really lots of different parties together, so you have
00:31:44
Speaker
a change in the interest rate or whatever in the slope of the interest rate that could have effect on both borrowers and lenders. I think this is one of the problems that you're trying to get at in your paper, in how a market or how these smart contracts process that information. Now, going back to Fahad's segue to your paper, maybe you want to describe
00:32:09
Speaker
what problem you have identified that exists with smart contracts. And I'm going to throw this back at Quentin, because Quentin just gave a very intricate answer to something. And I would just like to see if we can break this down a little and make it a little easy for the audience. Yeah, sure. I can try that again. So to us, explain that it would be actually good to have no ports looking
00:32:36
Speaker
or like a function that decides on the interest rate, which means that whenever the utilization rate goes up by a little, you would like to increase the interest rates by a high amount. And the reason is that the more that's going to make the market more reactive, if suddenly the interest rate goes up by a high amount, that means that the borrowers are going to react to that by borrowing less, they are going to come back and reimburse their
00:33:05
Speaker
and repay their loan. And that's going to push back the utilization rate in line with what's your target. So that's a good thing in terms of efficiency if you have a target in mind.
00:33:17
Speaker
But that basically assumes that agents are perfectly rational, they have all the information, and they are very reactive. In practice, what's going to happen is that many of the borrowers, they are going to have other things to do in life, and are not going to track down at every second what's going on in this market. And what might happen is that if you have this very reactive function, it might be a situation where interest rate is going to shoot up to, I don't know, like 200%.
00:33:45
Speaker
within a day. And if that happens, then you're just going to lose a lot of money if you have a boring position in that market. And that might be a very bad news when you come back from home or from a walk. And then you just realize that the value of your collateral now has been really decreased because they adjusted for this very high interest rates. Now you need to pay back a much higher amount. So there is a really clear balance between the two. From an efficiency perspective, you'd like to have a very high slope.
00:34:15
Speaker
But from an example, they provide you with creation. You want to take into account that agents are facing some risk whenever you have a very high slope. That's the argument I was trying to make.
00:34:28
Speaker
If I may just interject very briefly, I think there's also an implicit assumption that you're making, which is that lending capital is also quite sticky in the sense of even if the industry goes up a lot, then it all of a sudden becomes very attractive to put lending capital in, but capital is hard to come by. It's not actually moving that fast. I think that would be another
00:34:48
Speaker
Partly. Exactly. If on the other side, if on the landing side, the landing side would be extremely attentive all the time, that might mitigate that concern because they will also like the utilization rate back in line. But if one of the two parties at the same time are somewhat sticky or inattentive, then you might get into some trouble.
00:35:10
Speaker
So you've now talked about some of the possible problems that could arise. You seem to be, is this sort of a practical nature or is this already something that you would explicitly take into account in your particular model?
00:35:24
Speaker
Yeah, so the way we have been working on this on this paper is we first build up the foundation, try to understand the nature of the problem. And then we came to this first extreme solution where you actually want to go as high slope as you can. And that's the way you're going to minimize efficiency losses.
00:35:45
Speaker
But obviously, now we are at the stage where we're trying to think beyond that and try to understand the kind of friction that might push you towards what economists call an interior solution. So basically, the idea that you don't want to go 100% positive or 0% positive. You want to go somewhere in the middle. And so it's basically the trailer between these two things. That's something we are currently working on, incorporating into the paper.
00:36:09
Speaker
So let me just very briefly for the audience. So you said this and I heard it within some of the subtext of what you said is you have a particular benchmark in mind of what is the optimal outcome. So what is the optimal outcome and explain why it is the optimal outcome. Maybe Thomas, you go for that.
00:36:26
Speaker
Yeah, sure. So I actually wanted to come back to this because from the platform's perspective, it's still not obvious what the objective is. You asked this question earlier, what is Aave or compound's objective here? It's not obvious. From the economist's perspective, when it comes to these markets, the objective is maximize efficiency.
00:36:46
Speaker
And what that means is to find the price that clears the market, right? The price at which nobody else would want to borrow anymore and nobody else would want to lend anymore. And you get this sort of thing that we allude to as the market clearing, right?
00:37:02
Speaker
And so this is what Quentin was alluding to when he said efficiency, right? So now from the platform's perspective, it's not clear that their goal is efficiency from this sort of economic perspective because their goal is also for the platform to grow. And so I think while efficiency helps,
00:37:23
Speaker
the users of the platform, make sure that the funds are being best allocated to the right borrowers. Coming to the practical side, adoption of these platforms should be easy. It shouldn't come with huge risks. And this speaks to this idea that if interest rates are moving around too much, even if you have, on average, efficient outcomes, this could be enough to dissuade users from wanting to adopt these types of platforms.
00:37:50
Speaker
So it's sort of like a dual purpose. You want it economically to be as efficient as possible, but you also want it to be user friendly. You don't want your interest rates to be very volatile, especially because before you borrow or lend from these platforms, you're going to have to make an implicit sort of projection or forecast of what you think the interest rates you're going to earn or pay are going to be in the coming days or months. Let me summarize if I understand this.
00:38:19
Speaker
In an ideal world, if we had a normal market, in a normal market, demand and supply balances. The amount of funds made available for lending or for borrowing is exactly what borrowers would want to take up. You're saying in terms of a platform, this is actually something that we can't really do because if there are shocks to the platform, the platform can't absorb it. As Quentin said, you go for a walk and then all of a sudden you're bankrupt afterwards. We can't have that.
00:38:47
Speaker
Now, what is the mechanism that prevents the full equilibration of demand and supply is simply, one, is it an information fiction? Is that the way to think about it? So there's something that you just don't know at the right time.
00:39:04
Speaker
the smart contract can't process information correctly. Is that what it is? Right. So this is, this is, so it starts first with the way that the interest rates are currently being set, right? And we're studying this mechanism. We're not studying other mechanisms, but if you take that mechanism at face value, what you're going to say is that the interest rate is always going to be determined ex ante depend on the changes that have occurred in the utilization of funds.
00:39:29
Speaker
This means that it's not like you're observing the changes to the credit market and then changing the interest rates. It's happening the other way around. And it turns out that when you have supply and demand that are moving at the same time, it's not going to be possible to just observe utilization and to understand what that means for the state of the underlying credit market.
00:39:51
Speaker
So this is the informational problem which is that you can't just learn what's happening in credit markets and map that like a bank would a bank would say look at the fence fund rate or look at LIBOR use that as a benchmark right and when those things move we know we're going to move our interest rates. Here the platform can't do this and so this is sort of represents one of the biggest inefficiencies which would hold even for more general interest rate function designs
00:40:16
Speaker
but because of this sort of oracle problem or problem of feeding information to the blockchain from the real economy.
00:40:26
Speaker
So there's a real fundamental point I think that Thomas is touching on now that sort of comes out of some of this work, which is that, you know, we were saying earlier that ultimately there's a complicated world out there. Things are going to change. You need the interest rates to be able to respond to that. Loosely speaking, interest rates need to respond to supply and demand somehow. There are frictions that sort of restrict how they can, which is why these platforms use the utilization rate.
00:40:55
Speaker
But what that tells you then is that these platforms kind of need the utilization rate to be moving, right? So how else can the platform pick up? So in other words, if the platform is pricing based on utilization rates and utilization is supposed to serve as a proxy for supply and demand, then utilization necessarily must be
00:41:17
Speaker
moving around so that you can use it to figure out what's actually happening to supply and demand factors. And therefore, you can have interest rates reflect supply and demand. But the big problem with that, which is more general than thinking about the case that, let's say, you want supply equals demand, which means utilization would be one. Even if you say, as a lot of these platforms do, hey, we're going to target a utilization rate of 80% or 90%.
00:41:45
Speaker
Well, if you actually succeed in your goal of targeting that thing, doesn't that mean the utilization rate doesn't move? And if the utilization rate doesn't move, then how is it possible that your interest rates are going to respond to fluctuations in supply and demand when you encoded them to be a function of the utilization rate?
00:42:02
Speaker
Um, and so actually this, I think leads to sort of, you know, one of the questions that that's come to my mind beyond the paper, which is that, okay, there are these frictions about how you can't access absolutely everything in the real world. But is it the right idea to specify this thing as a function of.
00:42:19
Speaker
At least only the utilization rate, right? Like we've been talking about slopes and the function, but we've sort of been implicitly assuming that it must be a function of the utilization rate. I think one of the things that comes out of the analysis is, is that right? Should it be necessarily a function of the utilization rate?
00:42:40
Speaker
It might sound as that, I think it's kind of like the thing about it as being sufficient statistics, if you wish, for the right equilibrium of supply and demand. You don't really care if you're just platform. It's like movements coming from demand or if it's coming from supply, as long as you know that you can incentivize to remain at the target that you have for a given interest rate. I think that's basically enough.
00:43:04
Speaker
That's fine. In that case, how would you know? So in the last part, you were talking about the targeting again. But if you succeed in the targeting, then you wouldn't be able to actually know whether there's more demand or whether you've got a positive demand shock. How would you know if there is demand shock, et cetera? In a way, we'd say you don't even really care about that. I think to the extent that the market remains in equilibrium, there could be that
00:43:32
Speaker
All the agents would be anticipating each other. They would know like when demand is moving you have somebody already anticipate. The supply is moving exactly at the same moment as demand is moving so that everything just remains in equilibrium. You never see any change in utilization rate.
00:43:48
Speaker
And that would work perfectly. You always remain exactly where you're supposed to be. I don't think there'd be an issue in something like that. But are you assuming we do see a change in interest rates? So that's the usual, like if we think about the, let's call it the frictionless benchmark, we all started an intro to econ. Supply always equals demand, but the rates move, right? And so it seems to me the issue here is that if you embed a relationship,
00:44:16
Speaker
between the interest rate and the, let's call it, relative level of imbalance between supply and demand, which is essentially what the utilization rate is, then if you fix the imbalance, you fix the rate, and you've now lost the ability to adjust, right? Yeah, so it depends if you're using the Shoki Spellman Anto, it's temporary.
00:44:32
Speaker
So if you think the shock is a permanent shock, then we are getting back into the situation that Thomas described earlier, that you need something like Oracle or you need like the governance token holder to meet at some point and decide they want to shift the whole slope or the whole curve so that eventually you get back to having both the digitization rate that you want and the right correct market rate. Otherwise, it's not going to work in this case.
00:45:01
Speaker
If you have temporary shocks, you think it's going to come back, then you don't care. You have some arbitrageurs that basically anticipate that these are just noise trading. They might be able to bring the equilibrium back to where it's supposed to be, and that's actually good for you. Yeah, and I would just say, I think Fahad and Quentin are alluding to this fundamental problem, which is that you cannot always hitch your target utilization, let's say, of 80%.
00:45:29
Speaker
while still, sorry, you can't hit your target utilization of 80% if the underlying credit markets are changing. Because that means when, let's say, demand increases, the interest rate at 80% is going to be too low, and that's going to push up utilization, right? And now you're no longer at your target.
00:45:49
Speaker
Fundamentally, when the state of the world is changing, the state of the credit market is changing, you're not going to be able to set an interest rate function that always gets you to hit your target 100% of the time. So you have to accept movement in the utilization. That means you have to accept movement in the underlying interest rates. And it's sort of like how well can you track the real world interest rates is sort of a secondary question. Now, as a practical matter, how far off is it though? Is it like
00:46:19
Speaker
order of magnitude off, so if we're talking about, let's say, an average interest rate of 5%, are we at 15%, 20% off, or are we at the order of basis points off? This is a tough question because it's not clear what you mean in terms of what interest rate we should be using as a benchmark.
00:46:39
Speaker
Right. The blockchain is itself its own sort of closed economy. And so interest rates are going to be determined within that economy. And it's not clear that we should be taking the Fed funds rate, for example, as a benchmark or European interest rates as a benchmark. So.
00:46:57
Speaker
All that we can say is that interest rates, basically, under the current interest rate functions, the platforms have set targets. Those targets sometimes are grossly missed due to underutilization of funds. But yeah, again, I'm not sure there's a clear answer to this question of what's the right interest rate to begin with.
00:47:19
Speaker
I'm just wondering if you had some simulations of some kind, so that gives us a little bit of a handle of this. I mean, imagine there's a target interest rate, right? There's a target utilization rate. And then, you know.
00:47:31
Speaker
you basically have a statement saying that it could be better, right? So I will say in practice, there tend to be very large gaps between the target utilization rates and the realized ones. So as we're doing this, I'm actually looking at, for example, wrapped Bitcoin, which has a target utilization of 45 and an actual utilization right now of 10.6%.
00:47:59
Speaker
So, I think the ability to target these things is a practically binding concern.
00:48:08
Speaker
By the way, just for my, just so that I understand it, so the target rate, is that the same as the rate where you mentioned earlier, the kink of the function is, right? So that's kind of like the max that you're, so then that will be the max that you're aiming for, right? That is correct. I would say there is no fundamental reason for that, but it seems to be that's the heuristic that that one has adopted as a way to basically
00:48:33
Speaker
seeing that this is their target rate. They don't, they don't wish to go beyond that. So the king part, it's a big he decides to increase the slope once you pass this particular point that he thinks the right point. So just to clarify, they, so they actually, if you, if you look at their dashboards, like I'm on obvious website now in the background here, they actually put a vertical line right at the king point and right optimal right there.
00:48:58
Speaker
And in principle, you have to realize it's a continuous function, right? So it's not like you couldn't go above that rate. It's that the slope does get fairly steep. And in fact, one of the things I think that Thomas and Quentin and I have discussed in the past is that maybe they thought of it as a way to kind of collapse the rates onto that point by having it be kinked there. But in practice, what it really does seem to do is it tends to force the rates to be below the kink because it's just really steep once you pass the kink.
00:49:25
Speaker
But in theory, the rates could surpass the kink. And we've seen this in some very sort of extreme cases. Like, for example, when Tether was trading at this big premium and people started shorting it like crazy at the lending protocols. So then it did pass the kink.
00:49:47
Speaker
That was for, I would say, reasons that you shouldn't be thinking about when you were sort of thinking of the general design of these platforms. Now, again, from a practical perspective, I'm wondering one of the issues with these DeFi platforms is everything happens essentially in continuous time. And as Quentin pointed out, right, so your walk could be a risk essentially, right, if there's a major movement in the lending rates.
00:50:14
Speaker
In the real world, we have a lot of, we have a checkpoint, right? If you think of, say, repos, for instance, they get negotiated every day, right? And so they're valid for a certain period of time. And so there's a certain discreteness to the market in which the market can actually work out, you know, what happens when.
00:50:32
Speaker
And for the DeFi protocols, I think the mechanism by which the utilization targets are changed and the mechanism by which the rates are, or if you want the slope of the rates are changed, is at a different time interval compared to
00:50:49
Speaker
what people make decisions at. So is that the fundamental problem? And is that a way to improve it? So if you had longer lending terms, or stable rates for a particular period of time? So I would just clarify that when interest rates move on a DeFi protocol, it doesn't mean that when you arrive, you're getting locked in at that rate.
00:51:13
Speaker
The idea here is you're going to pay whatever the platform's rate is, block by block. So even if you have short-term fluctuations in the rates, provided that they settle relatively quickly to, let's say, some stable rate, it's not going to be extremely expensive for you because even if it's 80% APY, if you turn that into what is the percentage you're paying for that particular block, it's going to be very low.
00:51:41
Speaker
So then you still have hopes, even when you walk away and interest rates blow up, that market forces will push those interest rates back down to normal levels, let's say within a couple of blocks, so that at the end of the day, it only costs you a few cents. What's more fundamental is if these rates and their fluctuation do not eventually stabilize, and that's where you're really concerned in terms of the walkway risk.
00:52:06
Speaker
But by the way, just to provide some background context here, so Aave actually does provide a loan product that is not floating block by block, the stable loans. It doesn't have a fixed duration, it has a reset point.
00:52:24
Speaker
Now, of course, I guess from the developer side, that might be a little bit more complicated, but I guess part of what you're getting at is, well, what's the economic benefit potentially of offering more than just a loan that varies every 12 seconds block by block?
00:52:47
Speaker
And so here I think our banking experts can jump in. So, Quentin, what do you think of platforms differentiating by offering these sorts of loans that are not just
00:53:04
Speaker
floating at the block by block level. It's interesting you mentioned repo, because I think that's exactly the right analogy. It's also where my expertise actually is. And so I just want to remind everyone here that those markets are not perfect either. On the repo market, we saw exactly that kind of problem in September 2019, where the repo is actually mostly overnight. So it also has this very short duration.
00:53:30
Speaker
And so because of many frictions in that market, you ended up with a very popular trait among hedge funds being to hold some long-term treasuries and then sell some futures contract against that and then basically roll over these traits in the repo markets every day.
00:53:52
Speaker
And then that rate completely blew up around that time because the repo rate suddenly was jumped up to almost 7% overnight. And they were like, these funds were all very highly leveraged. And so that meant that they made very large losses on this repo rate on a trade that was actually supposed to be perfectly safe, except that they didn't take into account that there might be some frictions, creating some inequality issues in the funding market.
00:54:23
Speaker
So these issues are not specific to defy lending at all, actually. They are also there in normal finance. So to answer your question, if I had it more directly, I would say there's again a trade-off here.
00:54:36
Speaker
The trailer is that if you want to offer a longer duration loan, and that's something you only do at the platform level. But on the other side, the borrowers will still get this overnight, not overnight, adjusting interest rates, floating interest rates. And that means that actually the platform is going to take some form of duration risk here.
00:54:59
Speaker
And the platform is taking some duration risk. That means that now the equity of the platform is bearing some more risk. And so that's also a risk that both sides of the platform has to take into account.
00:55:10
Speaker
So again, I think there is some tradeoff between stabilizing the market on one side, but then taking some more risks, more sources of risk in this particular case. But if we take a step back now and look at the work that you've done, is there anything that you can say in terms of how can we learn from your paper of how to improve the existing defile lending mechanisms? Is there something that we get from that, Thomas?
00:55:36
Speaker
Yeah, so this speaks to this sort of fundamental trade-off that Quentin was alluding to earlier. But basically what we're showing is that it's very easy to leverage the market forces to get your utilization within certain target rates.
00:55:53
Speaker
The problem with this approach is precisely that in order to leverage these forces to, let's say, ensure that your utilization is always between 80 and 81%, you need to basically impose a lot of potential volatility on the underlying interest rates.
00:56:10
Speaker
So there's sort of a trade-off there where if as you want to target utilization rates that are closer and closer to your targets, right, and not deviating from that, you need to set a steeper and steeper slope on the interest rates. And so now this compounds this sort of walk away risk problem,
00:56:29
Speaker
especially if there are some traders that are just not attuned to what's going on in the market and they're just kind of haphazardly borrowing or lending, that could send your interest rates way out of whack and this can create a lot of problems down the line. So if you didn't have this risk and the market perfectly snapped back into place every time there was some extra borrowing or lending on the platform, we show you can do very well setting these interest rates in terms of the sufficiency standard I mentioned earlier.
00:56:59
Speaker
In contrast, again, we want to weigh off the efficiency of the platform with adoption of the platform. And this sort of creates some more practical concerns about the designs that at least we propose in our paper. So one practical point that I think comes very clearly out of the paper is precisely one way to say it is utilization targeting has significant limits from the perspective of welfare.
00:57:27
Speaker
Now in terms of, and just to remind a little bit in terms of the intuition of it, it's just that the utilization is how the platform is actually, variation utilization is how the platform is sort of internalizing supply and demand shocks. And so when you target it, you sort of lose the ability for it to respond to those. But if you take that then, which I think is a clear insight that comes out of what we have,
00:57:55
Speaker
It actually leads, I think, to an important discussion which is maybe we should be thinking a little bit more broadly about the design space of these interest rate setting protocols. Does that mean that we need to rely on oracles for pricing? Does that mean we need to slow down the extent to which the pricing changes? Quentin alluded earlier to this idea that the governance process of setting the curve maybe should be a bit more active.
00:58:23
Speaker
There's certainly, so an immediate takeaway again is there's certainly a limitation to the current, let's say, parametric setup if you're thinking about it in a static context, meaning you set the curve and you leave it there forever. But there's a good discussion I think worth having about what exactly to do then and it's not totally obvious, I think.
00:58:44
Speaker
Well, there is also a bigger question and a related question, which
Comparing DeFi with Traditional Finance
00:58:48
Speaker
is as follows. I mean, in the papers, I understand that you're looking at an idealized situation, right? So an idealized frictionless market, where the efficiency, the efficient solution is very well defined and so on. And so the question is,
00:59:01
Speaker
i mean this for me now and then the thing that goes along maybe also a little bit with the discussion we had in the last podcast for heart is what is the what is the increment what's the improvement actually is there an improvement over the existing market right so what what do we see here what what advantages that we have and
00:59:19
Speaker
Is there actually a gain or is this just worse? Yeah, so if I could take that, I think there's a nice analogy. I mentioned at the beginning that you kind of want to think of DeFi lending as a securities lending market. And so the right analog to sort of hone in on would be securities lending, let's say equities lending in the real economy. And so there's a recent paper showing that Vanguard is a very active and
00:59:46
Speaker
a powerful player in this space. And according to their estimates, they're not passing through the interest that they get from lending these securities directly on to the providers of those securities. So to elaborate, basically every time you buy an ETF that's managed by a large fund, they're going to custody the underlying stocks in that ETF for you, and they're going to give you basically an ownership of that.
01:00:12
Speaker
Now, while they're holding on to those underlying stocks, they reserve the right to lend them out to other lenders. And they try to assure you that even though they're making money lending out your securities, that they're passing along those profits to you by charging you lower rates for managing the ETFs.
01:00:30
Speaker
Now, this paper is showing that basically the rates that they're passing through, once you account for everything, is not actually the sort of fair rate, and that these intermediaries are profiting a lot off of the fact that they simply are providing the service of costuming your assets. So here's a market where you could imagine this type of design might
01:00:53
Speaker
might be useful, right? We've been saying up until now, effectively, DeFi lending is inefficient if you compare it to this idealized frictionless world where markets are always clear. But in reality, in these markets, there's a lot of market power, monopoly power by intermediaries. And that means that, for example, the deposit rates that you get at your bank need not represent anything close to that idealized competitive rate.
01:01:20
Speaker
So this is a sort of follow-up question, which is, you know, yes, you cannot do as good as a perfectly competitive market, but maybe you can do better than some markets where there's a lot of rent extraction by these non-competitive. I'm going to disappoint a little bit. I think banking is a sector where there is extreme market concentration.
01:01:46
Speaker
And so one way of thinking about like the benefit of DeFi and all that things that's going to help deconcentrate that market because the technology is fundamentally quite different. Like there's a lot of barriers to entry in like getting into banking and doing these things, a lot of regulatory burden, a lot of fixed costs involved in like getting into this market. That's why CIGU is lending markets extremely segmented as a market.
01:02:15
Speaker
And so we are hoping that maybe by giving this much more accessible technology that's easily reproducible, that's much more transparent as well, that could provide some more de-concentration of the market and therefore that the markups that these players are making might get reduced down the line for that reason as well.
01:02:39
Speaker
Now it really boils down to why do we think that there is market position in the first place? And the rate is going to be a combination of different things. So it's going to be a combination of various countries, but also something like a natural network, economies, effects, meaning that you really want to interact with big players because you know big players have the pockets as well. So this can be, and that provides some form of natural monopoly, and then executives can revamp the value of this natural monopoly.
01:03:07
Speaker
So I guess like the benefits of the consumer trading in the market is going to depend on whether you're saying it's more going from the former or from the latter force. So, you know, I have an immediate follow-up question actually, because both of you brought up the issue of market power in particular, you know, you brought up the issue of delegated portfolio management and then, you know, essentially the portfolio manager keeping many of the gains for themselves.
01:03:32
Speaker
So that kind of points at two things here, right? So, and I'm going to make a much broader point here then just looking at the foreign lending market is we have a lot of regulations in place, often predicated on the agency problems that arise in the interaction of funds and so on and so forth. And there's usually a lack of transparency and the like.
01:03:53
Speaker
And in many ways, a DeFi protocol of this nature, say in the securities lending business, that would be where you would be able to see, for instance, what Vanguard puts into the protocol and earns the money off. This would be giving rise to a much more transparent environment where you can learn and actually see directly whether or not Vanguard actually, say, for instance, passes on the benefits to the people who they own the funds for. So that would be all very transparent, which
01:04:21
Speaker
regulators should be applauding, if I may get this right. I hope that nobody should have a contradiction to that one. The question here comes another one of market power. What does Vanguard actually have its incentive, say the world will be tokenized to actually use this platform because it's much better for them now to lend it out and not pass on the benefits to their customers?
Smart Contracts and Regulation
01:04:41
Speaker
What mechanisms do you think would bring them in the fold? This does boil down to a question of whether we expect the real-world version of this DeFi lending to be run on the blockchain or not. One nice thing about smart contracts is that any firm that produces a smart contract is effectively competing with future versions of itself.
01:05:07
Speaker
So, for example, you might worry, okay, Aave has this platform. Let's say Vanguard takes it over, and then they want to start increasing the take that the platform takes from this interest payments. Effectively, because the original Aave protocol already exists as a smart contract on the blockchain,
01:05:30
Speaker
When they want to make this adjustment, make this change, they're going to need to deploy a new smart contract that says, hey, here are the new rules of the game. And they're going to have to convince everybody that's using the old contract that they should also use the new contract. This is what I mean.
01:05:45
Speaker
I mean by competing with future versions of yourself. And I think this is kind of a nice thing because if you're going to do something malicious or something like extracting more rents, you have to provide equally value added with the new version of the smart contract or nobody's going to use it and it's not going to be useful.
01:06:07
Speaker
I think that's something that smart contracts in general is a feature that they have that is very interesting for the nature of competition in these types of markets. I see. So what you're saying is there's two things. So I think always that we should all be grateful for the existing DeFi world because people are willing to basically, I wouldn't say gamble, but experiment with their money.
01:06:31
Speaker
on trying to figure out ways how maybe programmatically and algorithmically we can solve some problems in markets. But you're also saying that not only that, but there's actually you create a history and you create benchmarks that any future version actually has to be measured upon. I think that's a very interesting thought here.
01:06:48
Speaker
Now, if you take a step back and you look at these tokens that are governance tokens here, and fundamentally, as we say, the reserve pools essentially create a buffer. They collect fees, create a buffer against possible fluctuations in, say, collateral
01:07:09
Speaker
any errors that can occur and so on and so forth. And the way to see this, I think, functionally is its equity. And so the owners of these tokens effectively have at least an implicit claim on this equity. Now, if we see it that way, then this particular type of token is a security. Now, what would that do, actually, to the usability of these protocols? Because all of a sudden, you have a completely different mechanism in place, and you have a complete different disclosure requirements,
01:07:39
Speaker
and ownership proofs and so on and so forth. What would they do? Would you be able to actually operate these protocols in this way? Or would you actually have to think of a way how you can separate the governance again from the reserve pool ownership claims? That's a big question. But I think it really goes beyond DeFi landing. It would go like any DeFi protocol you can think of. When you start increasing the regulatory burden, you're also going to increase
01:08:05
Speaker
the cost of operating this kind of business. And then your ability to go back to what I was describing earlier, there's like huge barrier to entry in anything that's financial industry related. And for, for that reason, you all have, you have a little market power. So there's definitely like one part for the market power bias frontier coming from. And so there is, there's a trade-off of regulatory scrutiny as well on regulatory, which is regulatory costs.
01:08:32
Speaker
And so that's definitely something that the regulators would need to think deeply about, whether if they start regulating governance token as equity and hence everything falls beyond that, that might be a big problem for anything we are trying to do here.
01:08:47
Speaker
Actually, I'm going to chime in on this one because I think, so one of the things that you hear in the regulatory space is always, you know, same role, same risk, same regulation. But I think one of the things that we learned from this experimentation in the DeFi world that we see is that maybe there's not the same risk, even though the roles kind of look similar. And we learn here that, you know, maybe
01:09:10
Speaker
different levels of disclosure and activity level and regulation would be required here. And I think this is actually probably something that comes into your face and that we see here. Yeah, this is all a very sort of tough question to answer in the regulatory environment that we're currently living in.
01:09:28
Speaker
And I think that part of the regulator's fear is this idea of, let's say nowadays we trust Uniswap, right? And so you could imagine a future where Uniswap is a registered company, their token pays dividends, everybody's happy with that world. The problem is that Uniswap's sort of intellectual property is this smart contract.
01:09:50
Speaker
and anybody can go and copy that smart contract and create their own version of uniswap let's say i take it and i make a small tweak to the future where i say we're uniswap is gonna stop collecting funds and we're gonna stop paying out dividends in that world there's nothing would keep people from using my new smart contract.
01:10:10
Speaker
And just because I deployed the contract doesn't mean that I have any ownership claim over that contract. Just because you use the contract doesn't mean you own it, doesn't mean you're liable for disclosures of that contract. And so it sort of creates this sort of weird situation where you can undercut these well-established firms that are documenting everything by simply copying their code and creating your own version that takes out all of the sort of
01:10:39
Speaker
finance part about paying dividends and things like that. Just you can run these protocols without collecting any fees in sort of a side pool. So this also speaks to the innovation aspect, which is how can Uniswap keep developing or Aave or Compound? How can they keep developing the next version of these contracts that improve upon previous versions when there's always a threat of somebody coming and copying and just deploying their own version of that same support?
01:11:09
Speaker
You know, I don't want to open up a can of worms, but I'll just, I'll brief, this is of course, a very deep topic. And I'll just briefly say that it's not clear to me the discussion when we get into the legal stuff is so, let's say well-defined. And to clarify what I mean here, for example, when we talk about Uniswap's relationship to Uniswap's governance token, the thing is the governance token doesn't give you a right over Uniswap labs.
01:11:37
Speaker
They use the same name, but there's a bit of a detachment there. There's not a classical issuer in the traditional sense of most securities, where what you're talking about when you're talking about the security is a claim on the issuer. It seems to me that there's more fundamental legal questions that need to be addressed before the consequent discussion is well-defined.
01:12:06
Speaker
That's actually a very good point. There's also a discussion in the regulatory space, not everywhere, but in a significant portion of it, actually, whether or not a fully decentralized project in any form requires regulation and actually falls under it. For instance, if you have a governance, as you say, in Uniswap as an issued governance token, if it is simply decentralized, no more new tokens are going to come out
01:12:30
Speaker
I mean, ultimately, if there was be a compliance requirement, who's actually complying here? Who's supposed to be doing that? That's actually a very interesting question too, but I think we're actually already, the worms are spilling all over the place. Maybe I can rein in the discussion a little bit. I want to go back to something that Quentin said earlier. Before you do, I just want to throw one more worm out there.
01:12:58
Speaker
So putting our economist hats on for a moment, if you were to think about the incentives of the developer community supporting these protocols, pass the initial deployments into subsequent versions and all of that, and you were to somehow force them to partake in regulatory requirements, disclosure requirements, etc.
01:13:20
Speaker
you change their incentives maybe they decide you know what we'll just launch it and then we're not part of this thing there is no entity or we close the entity because we can't deal with this stuff and that has implications for you know for innovation etc so.
01:13:36
Speaker
Yeah, you can rein back in the worms now. No, no, but you're absolutely right. I think we're living in a world where there are differences. There's different approaches. The fact that every, as Thomas pointed out, everything is public information. All the IP is public information.
01:13:55
Speaker
it's, you know, this is a different world. And I think that it requires a very careful thinking of what is really useful in terms of regulation other than trying to, you know, force every round peg into a square hole, right? But I want to still rein it back in and go back to Quentin. And one of the things I think that I wanted to
01:14:17
Speaker
think about and explore just a little more as the question of what are the extreme risks that have occurred. We've talked many times over about the spiking rate during your walk. Now, we've seen probably a little bit of that. As Thomas said, well, yeah, you have a spiking rate, but you pay this literally for like 24 seconds and who cares? It's like 28 cents extra or so. What have we seen in terms of
01:14:44
Speaker
drama events and drama events is what the financial sector really excels at. And what could you imagine could be a drama event? So if you think about the worst case scenario, what would that be? What would that actually look like?
DeFi's Catastrophic Scenarios
01:14:58
Speaker
I think the worst case scenario would be something like there is the value of what most users collateral asset they say is just like suddenly dropping very, very fast. And then the consequence of that is that
01:15:14
Speaker
Many people will want to short ease. And because everyone wants to short ease, because they expect it to drop further, they will want to borrow a lot. And that's going to increase the utilization rate on the other side, on the cash market, on stablecoin borrowing market. So where ease is just a collateral asset. The utilization rate now just goes very high. Let's say at some point it reaches one. When it reaches one,
01:15:42
Speaker
That means that when you're on the other side, when you're actually landing in that market, you're still a coin, then you cannot withdraw anymore. They cannot withdraw anymore. And the value of the collateral asset that's on the other side is just keeps falling. The value keeps falling. And then there is just not enough value. You reach like some form like liquidation threshold for the whole platform. They start liquidating these ease. They put even more pressure, downward pressure on the price. Price really collapsed.
01:16:10
Speaker
a few cents on the dollar on disease, and then you cannot recover whatever is at your length. I think that's like the ultimate catastrophic scenario you can imagine. This has never happened before. There's like some safety without building to try to avoid this kind of situation that we discussed before, but that's like the bad case scenario.
01:16:33
Speaker
I would follow up with this. So this spike in ETH borrowing and ETH interest rates, it actually has occurred, but I view it more as sort of a testament to the value that these types of platforms create, because this happened right before the merge. So basically right before the merge, you saw this huge spike in ETH borrowing, right? On these platforms, when you see ETH borrowing, you imagine somebody is borrowing it, selling it, let's say for a stable coin, and now you have a short position in ETH.
01:17:03
Speaker
And so why would you see this happen before the merge? Well, one would be you think the merge is not going to be successful. That's going to lead to a subsequent drop in the price of ETH. And you sort of want to bet on that. But another way to think about it is if you are one of the stakers during the merge who had locked up a large amount of money staking, you would want to find a way to hedge that position.
01:17:24
Speaker
You would want to find a way to say, if this doesn't work out, I'd like to have some sort of compensation for the risk that I'm taking by locking up my ethanous protocol. And this is exactly how you could do it. You go borrow the eth, you build that short position and effectively hedges anything that you're staking in the protocol against the risk that the merge wasn't successful.
01:17:45
Speaker
I would actually just like to ask all of you for a little bit of you in the future.
Future of DeFi and Tokenization
01:17:51
Speaker
So imagine a world in which we find a way how we can tokenize existing assets. Because ultimately, we have crypto assets all very well, but ultimately, there's several trillion dollars worth of real assets existing that could be tokenized and could be run on a blockchain. Imagine that's possible.
01:18:09
Speaker
What are your views on the usability of the landing platforms, the default landing platforms as they are? Is there a future to it? Will we see much more of a sort of like case-by-case centralized solution where the blockchain is more used as a settlement infrastructure?
01:18:25
Speaker
What are your views on the future of this? So let's start with Thomas. Yeah, so I'll go back to my Vanguard example because it's the easiest one for me to equate, you know, like that's the next, I think, big iteration, right? So when I buy a Vanguard ETF, right?
01:18:40
Speaker
they're costuming my assets and they're able to lend out those assets. And they're basically trying to assure me that I'm getting enough of the pass-through through the very low rate of fee that they charge me to manage that ETF. Now you can imagine once we have a technology to tokenize equities, right now I could effectively, you could facilitate ETF buying and the lending of the underlying securities.
01:19:05
Speaker
through a smart contract where now you have some intermediary that basically deposits those tokens. I can buy an ownership share of the pool of tokens, but when you want to lend those tokens, we do it through a defined lending platform built within that protocol.
01:19:23
Speaker
Now, again, it would be very transparent exactly how much of those lending revenues are being passed through to me, the owner of the underlying assets. And I think that that would be, you know, let's say at the very least a good thing for the buyers of ETFs.
01:19:38
Speaker
I know for me, after reading this paper and finding out that maybe I'm not getting the pass-through that I thought I was getting for my ETFs, I'm still not choosing to go buy someone else's ETF because there's huge implications of selling and then having to report the taxes and things like that. So here I see a clean solution to this type of problem. Whether or not it gets implemented will be another open question.
01:20:08
Speaker
All right, so in order to talk about the future, I'd like to segue, starting from the present, I think the present of the financial infrastructure in the United States actually not great for many different reasons. So it sounds like we have like settlement at T plus two, T plus three, there's like huge concentration, very high markups. Most people, they just.
01:20:32
Speaker
don't get any interest on the deposit accounts for market power issues. Things are very untransparent. And I think going forward, there is actually generally a lot of things we can do. And the technology improvements such as blockchain, DLT can really help. And so I would envisage a world that's actually improved on all of those dimensions, thanks to this technology improvement. At least I hope so.
01:21:02
Speaker
That's my wish. Well, I think this closes this discussion and this podcast. I'm really grateful for your insight, Quentin and Thomas.
Conclusion and Additional Resources
01:21:11
Speaker
I hope that the audience finds this insightful and you'll hear from me soon. We hope you enjoyed this podcast. Thank you for listening. As a reminder, you can find additional materials on owlexplains.com and can stay updated by following us on social media. That's all for today.