Epicenter - Learn about Crypto, Blockchain, Ethereum, Bitcoin and Distributed Technologies - Paul Frambot: Morpho Labs – Peer-to-Peer DeFi Lending Protocol

Episode Date: January 25, 2023

DeFi lending protocols operate accordingly to their smart contracts and are perfect examples of ‘Code is law’, even if some recent exploits have not quite abided to this harsh truth. From well established protocols to degenerate DeFi farms with astronomic APYs, they all mainly use liquidity pools. Morpho Labs proposes a peer-to-peer approach that operates on top of another protocol’s liquidity pool (i.e. Aave, Compound), offering better rates for lenders as well as borrowers.We were joined by Paul Frambot, co-founder and CEO of Morpho Labs, to discuss about the benefits and challenges that arise from a peer-to-peer lender-borrower matching system and what failsafes are in place.Topics covered in this episode:Paul’s background and diving into DeFiFounding and funding MorphoThe concept behind MorphoImproving lending & borrowing APYsHow lenders and borrowers are matchedManaging gas fees in a P2P settingSecuring collateral in peer matchesSharing liquidity with Aave and CompoundMorpho token economyManaging liquidationsUser experience & integrationsMorpho’s business modelFuture roadmapEpisode links: Paul Frambot on TwitterMorpho Labs on TwitterMorpho LabsSponsors: Omni: Access all of Web3 in one easy-to-use wallet! Earn and manage assets at once with Omni's built-in staking, yield vaults, bridges, swaps and NFT support.https://omni.app/ -This episode is hosted by Sebastien Couture & Friederike Ernst. Show notes and listening options: epicenter.tv/480

Welcome to Epicenter, the show which talks about the technologies, projects, and people driving decentralization and the blockchain revolution. I'm Sebastian Quicchio and I'm here today with Fredike Ernst. Today we're speaking with Paul Frambeau, who is CEO of Morpho Labs. Morphal Labs is building the Morphor Protocol, which is a DFI protocol that allows you to improve your borough and lend APY on lending protocols. So before we talk to Paul, I'd like to tell you about our sponsors this week. Omni is your new favorite multi-chain wallet. Omni supports more than 25 protocols so you can access all of your assets in one place. But what's really special about Omni is what you can do inside the wallet.

If you want to get yield, Omni allows you to get the best APIIs with zero fees in three taps. If you need to swap, Omni aggregates all major bridges and dexes, so you can bridge and swap across all supportive networks in one transactions directly in your wallet. If you love NFTs, Omni offers the... the best NFT support any wallet has, so you can collect and manage your favorite NFTs across all chains in one place. Omni truly is the easiest way to use Web3 and it's fully custodial, meaning you never

have to trust anyone of your assets other than yourself. And they support Ledger. Give Omni a try at Omni.com. So Paul, thanks for joining us today. It's been a long time coming. And so yeah, happy to have you here. Yeah, definitely. Thanks for having me.

I'm super excited to join today. As I mentioned before, I've been listening for Epicenter for as long as I was into crypto. So very excited to be here today. Super nice. So how long exactly is that, Paul? Tell us about your background and what you were up to be formal for. Yeah, definitely. I think my journey in crypto began when I was in high school.

Like basically, you know, it was pure technical interest into blockchain at first. So it was more into consensus. algorithms and Bitcoin in 2015-16. And really, I went into my studies, studied computer science and math, did a lot of work into like blockchain technology and how those consensus algorithm could scale and ended up being interested specifically in Ethereum and more the application layer, smart contracts, and then DFI. And DFI was really, to me, it's such a,

an interesting use case of blockchain technology that I could not do anything else but work into that space. So I just thought I put some research efforts there. And yeah, this is how, you know, step by step I came to play around with protocols. And I mean, play around theoretically. I mean, like playing with the concepts of the protocol. I never really used defy in my previous, you know, before going into awful. And then I started designing protocols. like Morpho and came to do Morphol apps to develop the Morphal Protocol. So that's how I got into this, basically.

How come you never used Defi beforehand, despite the fact that you were quite knowledgeable about it? So basically, what was missing for you to actually go, you know, head first into it? I think it was as simple as gas costs. You know, at the time I was like, I never invested in crypto specifically because I was too young to do so at the time. Like I did not pass my 18 years old. And basically when I went into crypto, I could see, you know, all those girls exploding,

but they could not participate, which was quite frustrating. But then when time come to play around with Defi, I was like gas price was completely outfitting my studio portfolio, basically. So that's probably the reason why. But, you know, in practice, I was more interested in reading white peppers and, you know, trying to make sense of what, you know, Yudiswap was doing and what, why it's different from traditional finance, what it brings to traditional finance, what could it, you know, what value it could bring to the world in general.

I think it's not a question we, we ask sufficiently often, especially in DFI. Like, what is the specific added value that we bring to the end user? Like, if DFI scales to millions of users. But yeah, it was more intellectual interest, I would say, than, you know, playing around. So I remember when you guys were just getting started and you gave a talk at the kiln office, I think it was probably one of the first talks you gave and, you know, explaining the concept of Morpho and, you know, one of the things that I thought was just, like, wild about your story and the story of the team is that when Morpho launched, you guys are all still in school.

And in fact, I think the idea for Morpho came from one of your professors. Can you maybe just retell that story and, yeah, how did the idea come to be? Yeah, definitely. I think, so back in, I would say, 2020, I had really two passions. Like, the first one was blockchain technology and the second one was entrepreneurship. And so, you know, in entrepreneurship, we often tell you, like, you have to get some sort of unfair advantage such that you have something that you make you, you know, different from others and make you stand out. And this is how you become successful in entrepreneurship. And with that in mind, I really love blockchain. So it made sense to do blockchain stuff. And I decided to, so my goal at the time was pretty ambitious. It was like, I want to be the most knowledgeable entrepreneur. about blockchain, about the technical specifics about blockchain in France. And so I decided to take every single course on like theoretical course on blockchain and

DFI in Paris that existed. So I had like, you know, maybe 10 or 11 courses on top of my existing courses at university that I would take. And I came to to meet some of the best researchers in that space in Paris. And together we formed the think tank. And in that think tank, we basically met every, you know, twice every week and basically, you know, just chat and discuss about mechanisms, what we can do better about them. And I would say like, Morphor is a construction, a collective construction of so many different ideas that all came together. And that the premise of it was during this thing tank. And like, we had so many people saying, hey, like, okay, compound is doing the things this way. Why not how they're doing this this way?

and what are the technical limitation to be able to have, you know, better 100% capital efficiency while still being liquid, this kind of questions. And we ended up with the premise of Mofo and then I said, okay, like let's get serious about this. Co-funded Mofel Labs and really regrouped a team of four co-founders that, as you mentioned, were students at the time, which was quite funny. But I mean, somehow, you know, spending 10 years, 20 years in investment banks can sort of, you know, probably constrain your branch too much such that you would not think of, you know, doing things in Defi the same way a student with no experience in traditional finance. I mean, no professional experience would do. But yes, this is how we started. And actually, until six months ago, I was still. going to school and I just finished my my hand of study internship few months ago.

And I'm yet to get my diploma, by the way, but that's another story. Yeah, that's incredible. And you guys, you know, I think what's also been really interesting about your journey is that, that like you guys have built this protocol and just a few months ago closed a very impressive of funding around with some pretty top tier VCs. How did that come to be? And what was the kind of journey to getting A6CZ and all these investors on board?

Yeah, I think it's a good question. So maybe first, just like to mention that it was during market conditions that were quite favorable. Of course, it does not do everything, but that's probably not something that would be as easily doable. I don't mean it was easy.

hard to do, but it was maybe even harder in those market conditions, obviously. But basically, at the time, we just had the concepts, but it was such as zero to one in terms of mechanism design and in terms of like how the team internally at Morphalabs is conceptualizing and thinking about DFI is like maybe we'll have the chance to talk about it, but it's radically different from what other protocols are doing in terms of, you know, lending, but also in terms of Dexos, et cetera. So we really wanted to be very, very innovative. And at the same time, so we had very, very innovative mechanisms. And at the same time, we had a product that made so much sense, right? Again, we'll probably talk about it, but it was a pure improvement of something that

already had a huge market fit and probably the biggest market fit in Holy Defi at the time, which was basically collateralized lending on Av and compound. And most of the most of the Mofo is literally like the pure improvements of that. And so this is where like, hey, like this thing is working. It has a huge market. This is Avent compound. But Mofo is here. They have like fresh new concepts.

They think about things differently. But on top of that, they provide a product that is same risk parameters, same liquidity, and has better rates. So that could not fail. And I mean, this is what they said to them when investing at least. But I think this is like pretty. much the story. It was the round, the last round of funding that we did is mainly composed of American VCs, which I think is worth mentioning because we struggled in Europe before being successful

in the US. The story somehow had more fit with the US VCs. Like European VCs was like were probably, you know, maybe not as as well as funded as the American ones. But I just thought it was interesting to see how of a gap that was between their reaction. in the US versus in Europe. So yeah, probably something, I don't know, some sort of fits. I don't really explain it to be honest. Cool. So we've talked a bit about the setting.

Let's dig into the protocol. Maybe before we kind of get into the weeds in a nutshell, you already said that basically it's a protocol that centers around lending and borrowing and it strictly improves on compound and Ava. or at least that's what you claim. So can you talk about what the protocol does in a nutshell? Yeah, absolutely. So Mofo is a landing protocol.

So basically you lend and borrow crypto assets on Mofo, the same way you would do on the AVE compound order, whatever lending protocol, other collateralized lending protocol that exists in D.EFI. So basically, when you're lending, there is a borrower that comes, that put some collateral to work and borrow your money. And it's a very low risk, low reward profile because every loan that is taken is secured by some collateral.

So in case, the collateral cannot be backing fully the loan, then the liquidation happened. And so in every case, the lender should, if liquidations are working well, be able to have some sort of risk-free yields. So in terms of product, this is what it is. And now what's different from avian compound is actually that morpho is built on top of avian compound. So if you have compound, you have an instance of morpho called morpho compounds that is working on top of compound. Same thing for avi.

There is avi protocol and there is morpho avi protocol that is built on top of morph of avi. And basically, why is that so is because morph is going to optimize avi. is going to optimize compound. So the question is, what do we have to optimize? And if you go to Avey website or to Campan website, if you look at the rates, you quickly realize that you have a huge spread between the lending and the borrowing rates,

like 1% to lend or 3% to borrow. Okay. Basically, the reason why that spread is is because those protocols, the way they work, deeply needs to have a ton of idol, liquidity in them that is not put to work and that is going to induce this spread. So it's, let's say capital inefficiency. We can dig into that maybe if necessary.

But without getting into the details, they have, you know, this is materialized by a spreads, which more for is going to optimize by just taking the mid-rate, for example, and say, hey, instead of lending at one, using more-for-a-vis, you'll be lending at two. And instead of borrowing at three, you'll be borrowing at two on Morphiwayway. So that's the overall concept. You get better rates on AVV-on-Morph-A-V-A-than-A-V-A-V-A, and you preserve all the risk parameters that you had on AVE, so you get the same collateral factors and these sort of things.

And you get access to the same liquidity, which is like billions of dollars on Avey. How do you manage to actually get the API down from the borrow API from compounding up from the lend API from compound. Yes, that's a good question. So in order to understand that, we have to explain how the mechanism works in Avey. So basically on Avey or compound, basically, let's set a pool because they're working very,

very similarly. You have basically many lenders all supplying to a same pool of capital. And you have very few borrowers. And those borrowers are paying three, and those lenders are earning one. And the reason why that is, is because the numerous lenders are going to share the earnings generated by the few borrowers. So the three is going to be divided into multiple ones given across all the pool.

So yields generated by borrowers is diluted across all the pool. And so the reason why Mofo is more capital efficient is basically saying, hey, you're a lending pool user. You are borrowing from, from Avey, for example, or your lending. from AVA. Well, basically, you could be lending and borrowing it too. Like, I mean, the concept is simple. Like, I'm lending at 2%, someone theoretically could take my my deposit for for 2% of a borrowing employee. And this is what Mofford does. Moffo is building a peer-to-peer matching engine that is built on top of the landing pool that enables landing pool users to be matched pure-to-peer whenever there is an opportunity

to increase their rates if they're a lender or to decrease their rates if they're a borrower. So it's a small improvement. It represents basically what it means is instead of having a diluted share of their capital that is put to work, they get all the capital that is put to work, which results in an increase of like plus 0.5 plus 1% EPI, which is reasonable in those market conditions, to be honest. How have people dealt with this spread before Morpho? Were there any other solutions or like other attempts to try to reduce this spread? I mean, you guys are doing it by matching borrowers and lenders directly in a sort of peer-to-peer way.

Are there other approaches that one can implement here to have a better efficiency of capital when doing lending? Yeah, absolutely. There's many different ways that could work. We went for a very specific implementation. Before Mofo, one of the reason why nobody went for something, I would say, Mofus, similar. is because actually this spread was hidden by liquidity mining. So as of context, in 2007, 20, 21, 22, lending protocols spent a lot of money

distributing tokens to their users. And you had a 1% USDC API, you're 3% borrowing API in USDC. But on each side of the market, you would get distributed 2%, in which case you'd get a much better API. this spread, I would say, virtually or artificially is disappearing. And probably the reason why no one came, you know, for a solution like Mofo before is because they were doubtful about the fact that rewards was ever going to stop a day or not. And we really started the innovation of Mofo by betting that, you know, rewards were not going

to stand like forever, right? Basically, protocols are losing money, emitting rewards. so they could not do that forever, which they stopped pretty much the day we launched morphos. So that would be one of the reasons why, second reason why it would probably be the complexity of it. It's not so easy to do, obviously, but does not mean like no one would have ever done it, especially if we managed to do it to do it. But yeah, I would say those are the two biggest fraction to actually make it. Okay, so I kind of, I want to talk about how it works on an nitty-gritty-gris.

level before I go to like the economics of it. Does it work for you, Paul? Absolutely. Fantastic. So let's talk about the matching engine. So basically say I want to post a position to lend some capital. What do I do? Do I post a message or do I already, I mean, do I have to pay gas for my transaction?

Yes, absolutely. So the matching engine is fully on chain. So basically, Mofo is a smart contract. it's completely on chain and it's a smart contract plugged on top of AVE or compound where users, instead of interacting with

Rave directly and using the supply borrow, withdraw, withdraw, repay functions of the pool, they would be using the supply, borrow, withdrawal, repay functions of Mofo. And then Mofo is going to, let's say, as a lender, let's have the very first person to lend on Mofo,

I'm going to deposit 100 die, I'm going to be put in a list or any kind of data structure. And my liquidity is going to be put on the pool. So I'm yielding at 1%, which is the API of the pool. And there is no counterparty to match me. So basically, like, there was no other way just Mofo puts you into the pool, but you are in the list.

And now a borrower comes in. And when the borrower comes in, he's going to say, hey, I want to borrow one. So he deposits some ethos collateral and he's going to borrow 100s, and Moffo is going to look into this data structure. So let's say it's a queue to simplify. And he's say, okay, the first member of that queue, he's posting liquidity on AVE. Let's withdraw this liquidity from AVE in order to give it directly to the borrower. And this is how broadly the matching engine works. Okay, so when I post something to the queue, it's immediately matched against Avoa compound.

And if there's a counterparty within Morpho, I am matched against the counterparty and basically the over collateralization of the lending pool on Avoa compound, that kind of results in the in the yield being split several ways between different lenders, goes to the sole lender that is now matched. peer-to-peer. Is that kind of a good summary? Yes, I think it's a nice way to think about it, yes. How do you stop people from submitting lots of small orders that the borrower will then have to pay gas for matching? Yes, that's an next question. And that's actually the reason why you can do for a first-in-first out model, as I was describing. And actually, you have different ways. you can solve the problem.

One way would be, okay, let's just do a pool inside a pool. And this way you don't need any queue and it's just, you know, a pool with a higher utilization of the capital. And so, but that works. But at the same time, for different reasons, it's not optimal in terms of capital efficiency. So the queue does not work because basically someone can, as you mentioned, supply one cents, like 1,000.

and basically DDoS the matching engine. So you have to come up with different ideas. One of the most simple one, and this is the one that we started with, is basically to sort a part of the list. So in a blockchain environment, you have a limited complexity that you have available in order to do operations,

otherwise gas would basically explode. And so we came up with what we call semi-sorting lists, list, a semi-sarted list, where basically it's trivial. We just have a heap, so basically a data structure that is able to get us the biggest users across like the top 10. And then the rest is a fee-fo. So basically, when I loop into the matching engine, I'll get to, you know, if I'm borrowing, I'll be match first with the biggest lender, second with the second biggest lender, etc., etc. So I'm sure that I'm going to get liquidity out of it and I won't be DDoS.

And then I go into the FIFO. And so this is kind of the simplest implementation that solved the specification of the problem that we were given in the first place. And this is the one we decided to implement. But yes, there's other more complex and interesting way to solve the problem that we've been working on as well. Okay, so basically larger lenders at lower rates are prioritized, if I understand correctly. And then basically my question kind of that follows from that. If there later comes a better lender or borrower to Morpho, is everything else reshuffled? No, it is not.

In the current implementation of Morphoid is not. So basically, if you get matched, no one can unmatch you, even though they propose a better rate, for example. Actually, in Mofo, you can't really propose a rate. There is one peer-to-peer API for everybody. But you can, you know, this is a limit of the design, right? It would be better, of course, if everybody was able to, you know, be more competitive about their rates.

And every time I repropose something that is better, everything gets reshuffled such that the market is more efficient. But the limit here and the reason why we decided not to go for this kind of solution is simply gas costs. It's like doing it on layer one, maybe it's doable on other chains or on layer two, like there's this chain or,

you know, ZK Sync or this kind of things. But on layer one, it's very hard to come up with a, I would say, constant time agarism that all those kind of matching in peer to peer. So that was probably the reason why we decided not to have

this reshuffling mechanism. Even though it's a good improvement to have, obviously, if you don't take gas. ask us into consideration. When you lend in a tradition, like Ave or compound, your liquidity is being posted up to, and it's basically collateralizing all of the loans that are being taken out by borrowers. As you said earlier, there is so much more liquidity that is covering, so it's capital and efficient.

But that means that as a borrower, I can come in. and out of that market and I can sort of always be borrowing assets across this pool of collateral. So as a lender, if I'm matched with a borrower and now that borrower wants to take their collateral out, how is my loan now being collateralized knowing that it's this like peer-to-peer thing rather than borrowing against this massive pool of collateral? Yeah, that's an excellent question. And that's probably the first cornerstone to the Mofo protocol and how all this mechanism, you know, unwrapped. So I'm going to attempt this as simply as possible.

Please, you know, feel free to ask me into many questions if you need. But basically, let's get the intuition for an example first. So if we come back to this example where Alice was providing 100 die and Bob provided one ETH in order to borrow 100,000 morpho. Basically, what's happening here is that Alice, her 100 die, is going to get into the pockets of Bob. And Bob can basically, you know, since the old collateral is much pure to pure, Bob can just, you know, turn off his machine, go offline and do whatever he wants with the 100 die. And at this point in time, let's say Alice is alone, there is no die. Like, morpho has access to no die.

So how could she be, you know, able to withdraw? And this is what Mofa solves. It's like, how do you have 100% capital efficiency without, you know, while staying liquid? And so what happens here when Alice is going to click the withdrawal button is the following. So Alice triggers the withdrawal function. And Mofo is going first to unmatch Alice and Bob. say, okay, Alice and Bob are not even, are not matched anymore because Alice is about to withdraw.

So Mofo needs to reconnect Bob with someone and Bob is going to be reconnected with the pool itself. So what's happening here is that Mofo is going to take a loan of 100 die on the pool with the heath of Bob. And give that 100 die to Alice such that Alice can withdraw at any time. So basically, yes, we have 100. percent capital efficiency. There is no dye, but since the borrower that is taking all the dye has put some collateral, then this collateral can be used to be actually borrowing on compounds and unmatch the position such that Alice can withdraw at any time. I hope that was clear. I can go into different explanations of that if necessary. So basically then what's happening is that compound or

AVE is acting as a backup mechanism in order, and so you fall back to, you fall back to compound or AVE rates if there is no lender, or sorry, if there's no, yeah, there's no lender to back your position. Exactly. Yeah, I think it's interesting. And maybe I did not mention sufficiently clearly in the beginning. It's important to mention that Mofo not always offers the period of pureture API, which is in the middle. it offers the period of purepioi if a match is found from the borough and the lander side.

And if a counterparty leaves or if there is no counterparty available, then there is a fallback to the underlying pool. So basically in terms of products, it's a lending protocol that provide the same risk parameters, the same liquidity, and at least the same rates as AVEA, right? Worst case scenario, you get the Avae yields from a borrower or from a lender perspective. and in average, you'll be much pure to clear and enjoy better rates. So if you look at Avent compound, the reason why the borrow and lend APIs are different from one another is to ensure there is sufficient exit liquidity, right? Because basically the user experience, if you want to withdraw your funds from a pool, and you can't because it's still borrowed and you can't force repayment unless the loan is underwater,

you can't withdraw your funds and basically what will happen then is that the rates are adjusted until it is economically unviable to kind of keep paying the interest on those loans and so on. But it's a bad situation for the lender because you're under the impression that in principle you can exit at any time. And that's why the rates are so different because basically they have to make the pool bigger so that basically in like 99.5% of cases, everyone who wants to exit can exit the pool. As morpho, you're kind of piggybacking off of that mechanism.

So basically you're using the Aver and compound exit liquidity and kind of take the most lucrative deals, the lend and borrow deals of large lenders and borrowers, because those are prioritized of large lenders and kind of skim that off the top. Yes. So I assume you guys are in contact with Arben compound. How do they feel about this? Yes, I think it's a next-time question.

And this is the question I get asked all the time is, I mean, this is true that morpho, the way it's design, is taking volume from Avey and from compound on the barring and the lending side. And Mofo is, you know, it's a bit of a weird relationship because it's a bit competitive in the sense that we're taking market shares. But on the other end, Mofo is, you know, bringing new rates. I mean, rates that are not artificial, like actual efficiency that is enabled into defy, rates that were not possible before. And so this is interesting because basically we're unlocking new use cases, new usage, where basically lenders would not come at 1%, but they would come at 2. And thanks to Mofo, they're now coming.

And Avey is getting a little bit of a share of that through the fullback borrowing volume or the fullback lending volume. And if you look at compound, for example, I don't know if that's still the case, but at least for many months this year, Mofo was the biggest bar of, of compound. And that's very interesting because actually compound does not have so many organic borrowers. When you look at the data of compound, most borrowers on compound, which is the most important thing for compound because they bring the money in, is actually artificial borrowers farming, comp rewards most of the time, not always, but most of the time. And so morpho actually

appear to be the biggest borrower, to the biggest money bringer to the to the compound protocols. So, you know, I don't want to hide the fact that Morpho is competitive. It is to some extent. But it's also growing the pie for everybody. And I think, like, Roberts from compounds and Stanley from Aviv got this right. And, you know, I had the chance to talk to Robert, like, to DM him about this a few times. And I think, you know, he and Stanley expressed sometimes themselves about it very briefly. But I think this is the right way to think about it.

I hope that was clear as an explanation. Yeah, that's clear. Do you have any actual data about how many of the users your vampire attacking away from compound and oven, how many you are actually organically bringing to the table? I never fetch the data, but I should definitely. I think that's interesting. And I think, you know, the market would decide in the end. And in any case, morphos inevitable.

So we'll have to deal with it anyway, whatever the result is. We've kind of glossed over the fact that in matching borrowers to lenders directly, you're also taking rewards from the pool. So basically not only are you taking share, but you're also making the rewards on the pool smaller than they otherwise would be, right? Yeah, so maybe you can add a little bit of reflection on that is users. So first, Mofo, if a user is put on the pool, if no counterparties match, then the user is put on the pool and is earning rewards. Mofo is not taking any cuts on that and is just giving it back to the user.

And so Mofo will just give the comparates to the lender or the borrower that did not find a match. If the user is matched, then the user is out of compound and experiencing a better rate, a better native rate, like in native, you know, in ETH, for example. And in which case, the lender or the borrower is actually removed from the pool. So there's more rewards remaining for the rest of the pool. I don't know if that's clear. Well, yes, but they're removed from the pool because 100% of the money their lending is borrowed. So basically, if that were to be, I mean, I mean, this is, I don't think this computes. What do you mean?

No, I think so there is a borer coming. So and extracting a lender or the opposite from the pool, in which case, those users that were previously earning rewards are not earning rewards anymore. And more further, they are not earning the company rewards if they're a match credit. I totally understand that, Paul. But so basically, to me, the question is the borrower that comes in a new, would that borrower otherwise have gone to Aval compound or would they have just refrained from taking a loan? So basically, if they had refrained from taking a loan, I told you.

get your line of argument. But basically, I'm skeptical that people who actually come to Morpho wouldn't otherwise have defaulted to going to Aveo compound. Oh, yeah. I think, like, you're right. Like, it depends on the use case. Like, there are some places where you only want to edge yourself or to borrow if the rates is lower.

And in some cases where you would have gone any way to Ave, because like, whatever the race you want to borrow. and in which case you go to Morpho, in which case, this is the vampire set of things, but there is a not-vampir set of things where you can only offer it to borrow at 1% or you just have a use case at 1%

in which case you come to Morpho, and there is the fullback of that, if there is no match that is found or only a part of the match that is found that is full-back into Ava for the borrowing site. Sure, absolutely. If you're priced out, you're priced out. I totally get that.

let's talk about the preferential treatment of larger lenders, right? So basically on compound and aber, everyone kind of gets the same rewards in proportion to how much money they put up. Whereas for morpho, whales are matched first because of, you know, gas efficiencies, because everything is on chain and it's not a pool mechanism. what happens to small lenders? Is it even feasible or attractive for them to actually come to morpho? Yes, that's a next question.

So maybe first something that I want to underline is that it's a default. Like it's a default of the mechanism that we had to sort users. It's because we were so constrained about a serum that we had to sort about of the user. But ideally, we would not want any preferential treatment on the size of the nodes. Now, this is solved by having small lenders regrouping together in a single smart contract, which happens usually, like,

usually, you know, end users are not interacting with more than, they are interacting with protocols that are themselves interacting with Mofo, in which case they can have, like, you know, bigger size and have more chances to be matched altogether. So that's one thing. And, yeah, I don't want to tease too much,

much about what we're doing, but like the problem is solvable. Okay. And what happens if someone from this larger, this pool of people who kind of just got together to kind of co-invest in Morpho? What happens if one of them wants out? How do you get them out? Oh, basically one lender. So to the eyes of Morpho, this vault is just a normal user. and so they can withdraw a share of, like on Mofo, you can withdraw a share of what you have. Okay, you can do a fractional withdrawal. Yes.

Oh, okay, good. That's super good. When you guys take a borough position on AVE or compound, like you mentioned earlier, that you were one of the largest borrowers on compound. You're earning compound like comp rewards for that. Can you talk a little bit about what those tokens are used for? And I think more broadly,

I'd like to understand the tokenomics of Morpho because as a morpho user, you're also earning more for tokens. And so what's the interplay here between that and broader tokenomics and how they work? Yeah, I think we're very minimalist in terms of token design and what we want to achieve with the token. We really want to, you know,

every proposal that we're going to make on the Dow to have an evolution of the token is going to be very minimal. minimalist, and we want that every step that we take makes sense is actually useful for the protocol and its users. So coming back to your question is, the comp tokens are all redistributed to compound, to the Mofo compound users. So if I'm lending and my position somehow, you know, abstractly is ended up on the pool, then I'm earning those rewards and Mof is going to account for those and is going to give them back to you. So there is no cut or whatsoever that

is taken by a morpho, like this is not the idea. So that's one thing. Then the second thing about the morph tokens, is that on top of that, whether you're matched, whether you're not matched, you'll always get moth tokens distributed to your line in borough positions. And those tokens are not transferable, for the moment, at least. And basically the idea here, and the motivation is not,

you know, to increase the API. I'm a firm non-believer on yield powered by governance tokens. However, it's here to decentralize the protocol. Like, in my opinion, it makes sense to have DAOs and protocols governed by investors, contributors, and users. Like, it's, in my opinion, a good equilibrium to have the three of those that can have a say in the governance in the long term. So that was important to us to, you know, kickstarts in some way, some sort of liquidity mining program as it is done on Mofo. And there is also this good effect of bootstrapping, right?

And in the very, you know, the first millions that we had, I think Mofo now is almost at half a billion. But the first millions that we had was like 100% of it, not 100%, but like maybe 99% of it was analyzed. Like we interpreted it as liquidity mining, right? which is not the case now. It's like because we had fair volumes, we were able to, you know, discuss more integrations.

And I think this is the way Tarkin should be used. I want to talk about the governance in a little bit of Movo Dow. I have one last question for the protocol. Sorry, to kind of dig deep here. So obviously a cool part of a lending protocols is liquidation. So if a position is underwater and a borrower is met, with a lender, typically you could, anyone can kind of liquidate that position and, and take a cut off of the cholesterol. So I mean, I assume I think it's exactly the same

and morpho, but obviously sometimes it goes wrong. So basically sometimes there's, I mean, there's still some risk involved. So sometimes prices crash too fast or no one actually liquidates because gas costs are too high or something. And on compound and Ava, Obviously, this is a communalized risk because basically this is just an unrepaid debt from the pool. So basically this loss gets redistributed to all lenders. What happens on Morpho if there's a liquidation failure? Does the person who's matched with the person with the borrower that has that mishap happen actually, you know, do they end? up empty-handed?

Yeah, that's an excellent question. So we often say pure-to-peer matching because it helps conceptualize the mechanism of Morpho, but in Morpho, actually, you never matched with a single counterparty. It's just the state. It's like, I'm in a pure-to-peer enhanced states or I'm not in a pure-to-pure and hand-states. And so it feels like, okay, there is one counterparty which I'm bearing the risk of, but this is not the case. The way it works is like when you have your money that is put in the pool,

then you bear the risk of bad debt under the pool. So if I've increased some bad debts, you're also incurred some bad debts. Your position is match peer to peer. You have shared loss. So first, something that I want to mention is that you have the same collateral factors, the same oracles, et cetera, as on AVE. So there is no specific parameter that would, you know, but yeah, I think that's important to mention.

But now if we put in the scenario where for some reason, liquidators are not able to liquidate, then the losses shares require all more for users. So it's not specific to one counterparty, but it's a smaller subset as the Aver protocol. Is that answer the question? Yeah, that answers the question perfectly. So kind of it's like a gated community within the within the, within the, uh, within the, uh,

within the other. Okay. Yeah, no, I, I think I understand that. I think this also solves a question that I had on the regulatory side, because basically the reason why protocols like Arvin compound are usually largely unregulated is because you always just deal with a smart contract. Whereas if you have a direct counterparty, this kind of, this kind of preferential treatment usually goes away in most jurisdictions.

So kind of actually matching people peer to peer also comes with legal downsides. Yes, absolutely. And I think this is why, you know, they really hate to talk about lending protocol. And they prefer talking about, you know, liquidity pools or, you know, you don't lend, you supply liquidity or you deposit liquidity. But yeah, here it's a pure to pure state. But at the same time, I want people to understand. And so I say put up to your matching.

But yeah, that's it's clearly a mutualized stage. I'd like to jump to using the product because I, you know, like yesterday, I, you know, looked at my Instadap and I was like, oh, hey, my, my Avey positions are not yet on Morpho. And, and, you know, maybe I should just move them over there and benefit from better rates. So what one question I had was when you're using Morpho, When you say that you use this phrasing of like enhanced, like an enhanced position where you're basically like, you know, Morpho, I feel it's sort of like this progressive enhancement of like a lending pool to appear to peer, but you'll always fall back. As a user, is it made obvious that your lending position is either matched on Morpho or you've fallen back to compound? Is that clear to the user and also the rate that you're getting?

Yes, it's a good question. I think it's hard to have a good user experience with such a complex, you know, mechanism, I would say. But basically on compound and avie, you have those C tokens that are basically not so useful except for user experience. There's no technical interest in having a C token or eight tokens. It's just a representation of something. Yes, when you have eight tokens, you see that.

the token yields like in your portfolio, you know how much it's worse. And it's, you know, a very user-friendly behavior. On more than all positions are treated equally. And because of that, you don't have like a, the RC20 for example. You could have in some world some sort of NFT that we did think that that was an even worse user experience to have an NFD and that people might throw it out, you know, not knowing what it was. So we're like, okay. maybe that's not a good way to go.

And so basically this is a work the front end has to do. So when you go to Instaab, for example, they do it extremely well. They tell you like what is your average. And if you go to morphos like interfaces with as well, the one developed by Morseolabs, the one developed by Mutative Team, like we have many, many different front ends. And usually what they're going to do is going to say, hey, You have 100 in peer-to-peer, 100 in pool. Okay, so it's like half-half.

So we're gonna make an average of your rates and such that the user does not feel the complexes. Okay. Does that make sense? So your position can be spread out. Yeah, so your position can be half, or there could be one proportion of your position that's backed by like a peer-to-peer loaned morpho

and the other would be backed by some, some position on compound. Exactly, yeah. And it's the work of the front end to average that and make it similar such that you have an average API. And why is it that there are like two separate, like morpho compound and morpho AVE, why wouldn't you just be utilizing the liquidity

in both of those pools and sort of finding ways to optimize? So basically, like finding the better rate, depending on the assets you're boring against for the user and sort of just, you know, adding the ability to add any sort of liquidity pool underlying like one morpho interface or contract. I think two reasons. The first one is complexity. Many, many protocols have tried, you know, aggregating learning protocols.

And they mostly all have failed to some extent. And probably the reason why is because they have not every protocol should. at the same risk and they're not the same, the sense that they take different collateral factors approaches. So it's easier for the landing side of things. You can say, hey, we'll select the pool with the more interesting API. But if you start also aggregating borrow positions

at the same time, then it becomes a bit hard because you have to decide on risk. It's like you have to decide which collateral factor I'm going to use. Like, am I going to use the collateral factor of Avey of compounds or the most conservative of the two. And as soon as you say, I'm going to use the most conservative of the two, you're not an improvement anymore. You're just a different product.

And so in terms of product decision, it was basically a product decision that I took was, okay, I want to be an improvement of something. I want to be able to improve something that I know for sure has a market. And so that's why we decided to separate, you know, those have no tradeoff. However, like this is definitely room for, you know, more complex mechanisms where you could aggregate those different protocols in a more advanced way, I would say. Hmm. Okay. So there's like a lot of complexity and aggregating different liquidity pools that that doesn't exist, say, with like aggregating, you know, like there's so many defy a dex aggregators that, and that thing, that problem has been like mostly solved.

Yeah, yeah, definitely. I think it's a matter of, you know, risk reward. rather than just rewards. And so for that reason, it's, it's an additional dimension and you have to take into account in the aggregation process. And then compound and avi were not made to share the liquidity with the partner.

So when you can borrow on the one, you can borrow on the other. So it's very hard to have this sort of aggregated. It's visible, but it's complex. So we thought, like, as a first step, you know, our first protocol should be something fairly simple. This is why the matching engine has a lot of room for improvement. This is why we're not aggregating, but maybe in the future,

we'll be able to release more, more advanced versions. I totally understand where you're coming from. And basically as a defy user, you kind of, you want to be very specific about which protocols you're fine with. So basically, at least you'd actually have to disclose to the user, which protocols you kind of mix by default and have to have, have them, allow them to opt out of that and so on. So I actually do think that especially since everything is on chain,

this is probably too much overhead for not all that much reward. Yeah, definitely. I think there's room for, you know, actual defy mid-awarest like, you know, yearn finance, full finance, etc. to actually, you know, makes at least the lending side of things and makes different strategies or rebalance, stuff like that. But this is like a different user kind. We really wanted to be as low level as possible and just be, you know, as close to a primitive as we could be. And yeah, that's probably

And you also you also get natural rebalancing anyways, right? So basically it's not like no one's making the markets. I mean, there's tons of bots out there that kind of try to get the yields to commensurate numbers. Yeah, exactly. And actually, maybe one last thing to make about that topic because I think it's interesting is rates are not so well arbitraged. And the reason why they're not arbitraised that because it's not because butts are lazy or bots have not been coded. Butts have been coded. It's just that they don't have the same appreciation of risk on every single protocol

and that you have a premium in like the reputation of a protocol, et cetera. And this is like so complex to factor in a smart contract, obviously. Talk about the integration. already talked about Instadap that was announced last year. And what's cool about that is if you have like an AVE position, you essentially don't have to unwind that position. You can just have it be refinanced by Morpho. I haven't done that yet, but I think I'm going to hit the button on that pretty soon. But yeah, what other integrations do you have to like improve the usability and and sort of like people's ability to use morpho.

I think we have a bunch. Like we are talking to, you know, you have two kind of integrations like protocol integrations and, you know, more data analytics integration. So we're talking to a bunch of websites to get morpho in approximately everywhere. And you have a lot of funds as well, integrating with morpho and doing their complex quant stuff, like hedging and stuff like that. But we have as well protocols like Volt stablecoin or original US dollar stablecoin that are using their

collateral and they put their collateral to work on the Morpho protocol. They say, hey, like we want better yield for our stablecoin users or saving rate or whatever. We are going to use. We're going to use this. And so, yes, we have like, for example, defy a hedge that can help you, you know, hedge your positions automatically when you're managing unisobie free positions. We have spool that I was mentioning a bit before, which enables you to diversify your defy portfolio. We really have a bunch of those and many coming in the pipeline as well. But to be fair, and I think that's a good ending

notes as well, is that I feel like there is a ceiling to that course as well. It's like, okay, we're like I think we're top free lending protocol at the moment with like 500 million or close to 500 million in deposits on ether. But I do feel like, you know, those are just defy integrations for defy things that are often powered by defy tokens. And I feel like there's a lack of, you know, actual real world connected use case in all this. And I think that's going to be a challenge for us to reconnect somehow morpho to those real usage and, you know, take DeFi to the next level. Cool. So let's kind of shift gears a little bit and talk about the morpho business model. How do you guys make money, you know, other than raising money from American VCs?

That's a good way. But, no, of course, joking here. So Morphal Labs is. is a software and development company making money by doing contributions to DAO. And so ideally, like the ideal way of making money for us is basically proposing things to the more for Dow that are going to be approved, hopefully, and that we are paid by the Dow to produce this research and enhance the protocol, the same way Avi companies is doing it,

the same way PGD Labs is doing it for Ave, for example. That raises the question of how Mofo makes money. And so the answer is Mofo in its current form is able to take a cut on the improvement that is made from Avey or from compounds. So it's basically you're making 1% API on Avey, 1.5 on the Mofo have it, then Mofo is able to take, let's say, 20% of that improvement and taking 0.1% of the volume that is matched. So it's just a cut of the improvements.

There is nothing like that is actually lost. And this is theoretical. So it's implemented, but it's not turned on. So currently the protocol is not making money. So how, I mean, basically if it's just a fee that's turned on, how do you guard yourselves against me just, you know, know, taking your code and fulking it and kind of lowering that to 0.1%. And how do you make sure there's extra structural value in the protocol itself, in the user base?

And this doesn't actually end in a race to the bottom. Yeah, that's an excellent question that I think every defy contributor should reason about in somewhere or another. So you have different ways to make this happen. one obvious way is to have a BUSL license, which is the case of IVV-Free, Campan V-Free and Unisovv-V-3. This is not the case of Mofo, Mofo GPL, Copy-Left license. So you can fork it if you wish to do it.

And basically, the reason why I think we wanted to go that route is we believe in a system grow more and more complex. We believe in network effects, being able to retain users, even if it's at the cost of a 0.1% fee. And the reason why that is, especially true for complex systems, is that security and, you know, the trust in the Dow and the contributors themselves can be valuable for users. Like, okay, there is a brand behind it. And when you look at lending rates, it's actually interesting to see that, for example, Avey is capturing a ton of of the lending volume, even though they're not offering, you know, as beneficial rewards as compound, for example. And some analyze it. I'm not saying this is what's happening, but I'm just

saying that some analyze it as basically the network effect of Ava, where people trust more Ava, the brand and their effect. And so you have, you know, quantitative questions here is that how much does that represent in practice and how much like this network effect is going to retain value and how big the fee could be, et cetera. But I think, you know, in the end, we have not built something that is yet to revolutionize the world and that is legitimate in really extracting a lot of value out of it. I think the current thinking is for us, yes, live thanks to fundraising, not turning out a fee because there is legal, also legal considerations on that and work as much as possible to build

something that has, you know, even greater traction and where it's actually worth taking a great attraction, there's a greater network effect. So I think we're not, we're more thinking about, you know, designing things that, you know, are able to get a sufficient traction such that we should legitimate to turn on a fee, or actually to propose to turn on the fee. But I think that's not not even the case at the moment. I hope it answers, I went a bit far, but I hope it answers the question. No, no, it answers the question. So actually, I also see the stickiness of, of DFI applications. I think it's a really difficult topic. So what actually makes something sticky? Clearly, it's not just the best price or the best security. It's kind of like, it's a user behavior question. And it's

not 100% clear to me what exactly actually entails the stickiness. Yeah. I think there's a lot of, you know, it all comes back to trust at some point, which is a bit weird, but that's how it is. And I do think the example of compound and av is a good example. Like I literally talking to users, I literally had many calls with people saying, hey, I'm never going to put a penny on compounds. And I'm like, okay, I mean, compound has been sure for a long time.

But, and, you know, it's, it's, the network effect here is, is, is really playing a role. I think it is, is major in terms of, you know, how to retain users. And also there are some products where by design, because you have more volume, your products is working better. It's a bit just the case of curve in that case. So tell us about the roadmap and what's coming for Morpho. I know that a lot of this is governed through. governance, but what are you most excited about?

Yes, I think like, you know, everything that will be decided, will be decided through governance and will be voted, et cetera. When it comes to more labs and the research that we're doing on lending in general, I would say like we've been thinking a lot about what makes sense, what does not make sense in decentralized lending. And I'm going to disappoint you a little bit, but as of now, we prefer, you know, keeping things close to the chest for the moment in the sense that we really like to release things that we are very sure that they are very neat and clean, at least scientifically.

So we'd rather not tease anything for six months and then release a big paper that we've been working on for more than a year rather than try to figure out roadmaps, etc. Because it's a constantly evolving space. I'm quite doubtful about roadmaps in... general. I think, you know, we are a team of 20, a bit more than 20. We're mostly researchers. So what we're doing is research. And what I can say, though, is that so far we have not something sufficiently cleans and that we're convinced of to to tease or really see anything in the public. And also, I have the team of researchers that are really, you know, firmly against

the fact that we should tease this kind of things. for them it's science and that we should not be teasing science and I respect that a lot. So I comply to the standards of the company and I just don't tease things. Okay, then maybe let me kind of zoom out a bit. What are your personal hopes for the ecosystem for 2023? So my personal hopes are rationalizing. So a lot more defy. I think 2022 and the events are grim and, you know, are grim.

And, you know, obviously, I don't wish this would have happened to a lot of people that suffered from it. But on the other hand, it has good aspects in the sense that we are rationalizing the mechanisms we're designing. I think a lot of time has been lost in token design while we should have focused on mechanism design, for example. That's a personal take, not the take of morph labs. But I think like we've seen mechanisms that are impressively relying on token incentives that inherently do not create value. And I'm hoping that 2003 is going to once and for all, you know, at least largely diminish the impact of those, those mechanisms in the space and how, you know, how talents are

sometimes, you know, working on those designs, what they could be working on more some, you know, know, more interesting primitives, for example. So I think this is my first hope. I also think that this is the perfect year for building. I'm very excited about what's going on. I talk to teams and they're all building something very exciting. Like, I'm really excited to see like the different announcements on Twitter that are going to be made in the next few months.

And I would say my last hope is that we regulations do not say, you know, two fast decisions on the topic of defy. I think one very important key thing for 2020 is that they take the stance of, you know, regulating apps rather than protocols. It's a very deep, you know, I think this is the way they should behave. And if they don't, again, personal take, but if they don't, I feel like this could endanger largely the system. And finally, I hope for more institutions.

on, you know, starting playing around Defi. This does not like a good start for 2023 because like with the FCX events and everything, people do not feel, you know, inclined to get into Defi again and he soon. But I'm hoping that with, you know, cleaner mechanisms, you know, protocols that have been sufficient trust, like I'm especially thinking of Gnosis safe, for example, that are safe now, sorry, is, I think once we have, you know, sufficiently battle tested prudives like this or uniswap, we can start thinking about onboarding like the next order of magnitude of users. And I'm really hoping one of those companies, you know, go big on some release to actually connect us to the rest of the world. And yeah, I have a lot of

hope in those industry leaders. And I hope that more will also, you know, be part of those very soon, which is, it isn't a good way to do so much. Cool. Thank you so much for coming on for our was very interesting. Yeah, definitely. I think probably one of the most interesting and in-depth postcast I've ever been, and so it was a pleasure

sharing that moment with you guys. Thanks, Paul. Thank you for joining us on this week's episode. We release new episodes every week. You can find and subscribe to the show on iTunes, Spotify, YouTube, SoundCloud, or wherever you listen to podcasts. And if you have a Google Home or Alexa device,

you can tell it to listen to the latest episode of the Epicenter podcast. Go to epicenter.tv slash subscribe for a full list of places where you can watch and listen. And while you're there, be sure to sign up for the newsletter, so you get new episodes in your inbox as they're released. If you want to interact with us, guests, or other podcast listeners, you can follow us on Twitter. And please leave us a review on iTunes. It helps people find the show, and we're always happy to read them. So thanks so much, and we look forward to being back next week.