Morpho & The Modular Application Thesis | Dan Elitzer & Paul Frambot

There is a high correlation between how distracted funders can get within the space where you can get a lot of those, you know, advisory deals or investments, etc. And so together as co-founders of Morpho, we signed an agreement uh which, you know, essentially prevents us from advising and investing in any other project. At this stage where you were currently like married to Morpho. So we're not, you know, allowed to work. Have you ever heard of

anyone doing that before. That is

the level of focus we're at with was Morpho, just lending in Morphos. Wow.

Welcome to Bankless, where we explore the frontier of money markets and modular DeFi applications on Ethereum. Today on the show, I have Dan Elitzer, GP at Nascent VC, stepping in as a co-host for this episode, as we both host Paul from Bro from Morpho on the podcast today. Morpho is a newer borrowing lending application on Ethereum, competing with a gargantuan incumbents that you're probably familiar with, Ave and Compound. Dan and his VC firm Nascent are particularly focused on both new innovative DeFi primitives and security. How do we make more capitally efficient DeFi apps? And how do we make sure that they don't get hacked? Morpho is at the intersection of both of these things. And a recent tweet thread from Dan sparked my interest in Morpho, who is going after one of the hardest to disrupt sectors in DeFi money markets. As a disclaimer, Nascent is an investor in Morpho, which is why Dan knows so much about Morpho and the design space, and why he's particularly useful as my co host on today's episode.

To me, Morpho is interesting not just because it's going after such a large market, but it's also its architecture design that matches a particular pattern that we're seeing emerge across the industry. Uniswap, for example, went from a monolithic to modular design architecture when it went from Uniswap V3 to Uniswap V4.

MakerDAO is doing the same with its sub DAO architecture and Sparks as its own lending model and Sparks as its own lending module for DAI.

The whole entire Frax ecosystem is chuckling since they did all of this modularity years ago and have already evolved into the Frax chain. And the concept of a Morpho native app chain was also discussed in this episode. In addition to specifically focusing on Morpho this episode, there's also an overarching theme of the modular application layer evolving inside of DeFi. And we use Morpho as a case study to explore this rabbit hole of app design. So let's go ahead and get right into understanding Morpho with Dan Elitzer from Nascent and Paul Frembeau, founder of Morpho. But first, a moment to talk about some of these fantastic sponsors that make the show possible. Bankless Nation, I'm excited to introduce you to my technical co-host for this podcast episode, Dan Elitzer, the GP of Nascent. We've had Dan on a number of times in the past. Dan has made some pretty remarkable calls, uh, not many calls throughout his life, but when he does make a call, a prediction about the future, it tends to be pretty damn accurate. I believe I remember Dan writing about the future of staked ETH tokens in 2019 before Lido even existed, just conceptually uh saw that coming. Uh in 2020, wrote about yield farming uh right before DeFi summer. Uh so whenever Dan makes a prediction, I always uh perk up. Uh and so we are going to hear about one of Dan's uh incoming predictions for a theme about DeFi here on this episode. Dan, it's good to have you back.

Thanks for having me, David.

Uh and then also with Dan, we have Paul Frembeau, the co-founder and CEO of Morpho, which is going to be one of the main subjects of this episode today. Paul, welcome to the show for the first time.

Hey David. Uh thanks for having me today.

Okay, so Dan, uh, you're gonna uh step in as my technical co-host. I'm gonna do a little bit of the hand holding to make sure that we kind of set the foundations, the 101s, 201s for the episode, but then slowly I'm gonna cede power and reigns over to you. First, I want to just ask you some like high level questions. Um, what should listeners be paying attention to as we go throughout this episode? What are your goals for the episode? And overall, like what inspires you about this particular content matter uh specifically?

Sure. I I've always been a big fan of DeFi. And what we're going to be doing in this episode is really talking about the evolution of DeFi, where it started, what some of the early attempts may have been like, and then what it was that really allowed DeFi to start taking off for the first time with DeFi 1.0. And then how that's evolved and how we've added complexity and added features. And what are the new things that are happening in DeFi today that are really going to allow DeFi, I think, to scale and achieve the vision of allowing anybody anywhere to access or create any imaginable financial product or service.

Beautiful. So let's go ahead and kind of get right into the core subject matter because we are talking about money markets, lending. The DeFi money market sector is one of the most mature in DeFi. Like currently, there's $11 billion in Ave. There has been for a long time. Ave is the dominant money market in DeFi. Before Ave, like compound was the dominant one. And these two monoliths are both like behemoths in their own respect. Ave specifically has like built this entire economy and ecosystem around it. And it seems like sufficiently well entrenched that its own network effects more or less kind of protect it. And so the money market uh sector in DeFi is like not for the faint-hearted to try and tackle, let alone disrupt. Uh, money markets are subject to exploits, to oracle attacks, to governance attacks, market manipulation. The risks around building money markets scare away like even the most daring of founders. Yet, out of the corner of DeFi comes this project called Morpho, which recently, congrats, Paul, passed $2 billion in assets compared to Alave Ave's 11. So really starting to show up on the scene here. Uh, and specifically just after passing $1 billion last November. So something.

Is perhaps stirring in the world of money markets, and I want to know what's up. So, Paul, Morpho's still a long way off from dethroning Ave as the number one lending application in DeFi, but nonetheless, the growth of Morpho is impressive and on a pretty strong trajectory. I think we're gonna back up after I asked this question and start at the beginning and beginning of DeFi to end up uh once again back at Morpho. But I really want to give you the opportunity just to summarize the reason behind Morpho's success. Like what credit uh would you give to the rise of Morpho's TVL? Like, why is Morpho working?

The initial product of Morpho essentially was uh built on top of Aave and Compound itself. And it was thought as a product that was strictly improving the interest rates of Aave and Compound, whilst preserving the same liquidity and the same risk parameters. And so it was um very natural for us to be uh introduced to lenders and borrowers because we're providing this uh rate optimization algorithms uh that essentially improve the user experience of Aave users and and compound users. And this is how we went to market uh initially uh with Morpho, with this protocol called Morpho Optimizer that now has over $2 billion in deposits. And from there, we we sort of learned a lot of building on top of lending and borrowing protocols. We're probably like the largest lending and borrowing uh protocol users ourselves, uh, with Morpho Optimizer this first product. And this is how we learned, you know, and that that guided the second product that was released a few weeks ago, which is MorphoBlue, uh, which is thought more as a landing infrastructure rather than a landing uh product, right? And uh and yeah, but uh we'll have the chance to to discuss this uh later today.

Okay, so the first product to Morpho Optimizer, maybe to articulate how it works, is there is a gap, there's a spread between Ave and compound rates. And Morpho Optimizer just like takes advantage of that margin. There's a margin there. The optimizer takes advantage of that. It produces better efficiency for its depositors, and it like arbs the gap between the interest rates between these two ecosystems. And so like it's kind of started off as a I'm gonna invoke layer two, a layer two around the application. Uh, so that we're not talking about like Ethereum blockchain layer twos. We're talking about like an application that sits on top of Ave and Compound and it optimizes between these two things. And so that's kind of where it started off. Dan, you want to say something?

Yeah, it's actually to be to be uh very specific though, it doesn't optimize like if the rate is higher on Ave and lower on compound, uh Morpho optimizer isn't about arbitraging between those rates. It's about the spread inherent in these pooled lending models, where there's more people who've deposited looking to lend out their asset than have borrowed from that. And so the interest being paid by a borrower, if a borrower is borrowing one dollar and there's three dollars uh that was deposited into the pool, uh, then you know they're they're paying that across all three dollars. The interest from being paid on one dollar is paid to three dollars of lenders. And so that inherently creates a spread in the borrow late, borrow rate and the lend rate. And so what Morpho Optimizer does is finds kind of the midpoint between those rates and matches borrowers and lenders at that rate.

Beautiful. Okay. But I really want to put a pin on the fact that this is a layer two application on top of layer one applications being the monoliths that are compound and Ave. And Dan, this is where I kind of want to back up to the beginning of DeFi and really kind of just set the foundations for progressing through this conversation. Like compound and Ave were quote call it like DeFi 1.0, DeFi 1.0 breakthroughs. We had tried money like lending in Ethereum before compound and Ave, but no one really remembers those names because none of them worked until Compound and Ave worked. Maybe you can talk about the lessons that we learned back in this like history of DeFi 1.0. Like, why did why did compound and Ave work where many others had failed before them? Like what made them, what about that made them work?

Yeah, Compound and Ave worked for some of the same reasons that Uniswap worked. It was this evolution from trying to recreate order books on chain to saying rather than trying to match up people in a peer-to-peer way when there are so few users, it made a lot more sense to aggregate that liquidity and do this kind of peer-to-contract model or peer-to-pool model. And so the same way we've gotten used to Uniswap, you know, uh X X times Y equals K, we're instead looking at um, you know, what can we do on the lending side? And so the idea was rather than say, I have ETH, I want to borrow uh a stable coin, uh, and I'm gonna put up ETH and uh say, hey, this is the ratio I want, this is the duration I want, um, you know, somebody come lend to me.

We had we had Maker initially doing that with Dye as a stable coin. Then we had, and that was that was fine, but it was a stable coin. It was new, it was different. People were still figuring it out. Compound came along and created these money markets where you could post any, I think it was originally around like five different assets. Um, and then you could post over collateralized loan in any of these assets, borrow any of the other assets. And that was a real breakthrough because it was a place to both earn interest, um, earn yield on your idle assets, and also be able to um easily access borrowing. And it really worked because there was also that constant liquidity. And so that that trade-off that we talked about before, where the borrowers are paying more than the lenders are receiving, that's not just because there is uh there was some kind of fee or spread in there. It was just inherent in the model as it was built. But the benefit of that was ease of use and not guarantees, but much, much higher assurances of liquidity in a short time frame should lenders want to recall uh the capital that they had had supplied into the protocol. Uh, and so this was not maybe the most efficient in every ways, right? Because you are making sure to maintain that buffer. But the flexibility, the ease of use was what really allowed these protocols to take off. Um, and after Compound's initial um success in there, Aave, which was originally Ethelend, also adopted a very similar model. Um, and and they've you know grown by leaps and bounds, uh, continuing to be, I think, more aggressive in their in their asset listing uh in terms of inclusion in these pools and adding additional features. And that's why they've grown to be the size that they are today.

Uh correct me if I'm wrong, but I think ETHLend had this prior model that Dharma also had. That we're really getting into some like DeFi history here, which is like it had it matched in like an order book style fashion, it matched borrowers and lenders on a one-off uh basis. And so like this, it was basically a matchmaker protocol. And then you and your match, whether you're the borrower and you get matched with the lender or vice versa, then you guys would go off and make your own private agreement. Uh and then money could change hands that way. And that uh never took off. And what you're saying is like, well, the reason why Ave Compound did work off is because these put really the autonomous into these systems where there was a central pool of liquidity and there was a global market rate. Uh, and that made uh matchmaking much more efficient. It was uh at demand of the users of both uh both sides of the market could instantly get what they wanted out of the protocol rather than having to wait for a counterparty. Uh and so really there was a central operator, call it a smart contract, that matched uh both the the both sides, both the demand side and the supply side. And this was far more efficient than the previous model, which was like two people getting matched and going off separately. Um and the arc of DeFi has always been one towards efficiency. And this was a massive step in improved efficiency. Uh, but this was DeFi 1.0. And so now some of these efficiencies are still now being discovered to be costs. So maybe we can talk about like the inefficiencies and the risks that uh came out in this like DeFi 1.0 lending model, because DeFi 1.0 is now old. Uh, and so like what was efficiency breakthroughs are now like efficiency frictions. So, Dan, maybe you can also kind of go through some of the constraints and the risks specifically um that have been discovered to be had in this like kind of DeFi 1.0 lending model.

Yeah, I mean, the the big thing has been around asset listing, right? And we've we've seen this, you know, as an investor in the space. We'll often have uh, you know, portfolio companies that we're working with that are saying, okay, our tokens out there, we want to get it, um, whether it's some kind of, you know, stable coin or stable coin-like product, or it's some kind of governance token. People want to get it uh available as collateral on compound and Ave. And then there's this big governance process to go through to say, like, is this appropriate to add to the pool? Is the is the risk worth taking on? Uh, how much demand is there gonna be? Uh, do we think there's gonna be to make it worth these risks? What should the uh max loan to value uh allowed on this uh asset be? Um, which Oracle uh should we use to integrate to add this new one? Like there's there's just so much that goes into it. And then you're constantly reassessing, subject to governance, where should the right loan to value ratio be? Uh, how do we look at other lending protocols and what are they offering on this asset? How do we need to be competitive with them? And that causes you often to move further and further out the risk curve in the good times, especially. And then in the in the bad times, it's everyone's like trying to contract and and clamp down as quickly as possible. And it's all being done uh in this, you know, honestly, pretty, pretty messy way uh that subjects everybody to the same levels of risk. And I think that the idea that somebody with kind of institutional capital and deploying at scale is going to have the same requirements and trade-offs in terms of risk and reward versus a retail user coming in wanting to deposit a few hundred or a few thousand dollars, just isn't the way forward. Um, we we really need people to be able to pick and choose where they want to sit on the risk spectrum by something other than do I use this protocol or not, or how much leverage do I take on within this protocol or not? Uh, there needs to be a lot more, I think, discrete controls and options available, uh, especially in ways that don't cause uh massive friction or decay in the UX associated with using this infrastructure in these protocols.

Right. So with the efficiency breakthrough, that was the centralized liquidity model of Ave and Compound, we also have like centralized risk scaling. And so if we ever add, so like Aave started off with like ETH, wrapped Bitcoin, maybe a few other tokens, I can't remember, but like really just blue chip, super secure, super liquid tokens. But what you're saying is like in order to grow the protocol, in order to grow in profitability, grow in fees, grow in users, we need to add more tokens. And maybe at the very beginning, that was like pretty fine because there was still a good selection of blue chip assets to use as collateral. I think maybe AMKR came pretty soon, one of the largest Lindy tokens in DeFi. But really, you uh approach some frontier of risk that is acceptable for something like Ave or Compound. And because when you integrate a new token who can borrow ETH, uh which can be collateral to borrow ETH or stables from Ave, you have you are exposing the tokens risks to all the entire pool. And so if there is like an inflation bug mint in a particular token, that particular token can drain the entire Ave contract because it's one single centralized source of liquidity. And this is just the risks of that one particular token contract. There's also Oracle risks. Basically, any sort of risk that is applied can is a risk to the entire TVL that's supplied inside of this monolith style of lending applications. This is another way to articulate this.

Yeah, I think that's a great way to articulate it. And what I would add is that, you know, there are things that these protocols do in terms of kind of capping borrow amounts and and and setting different things there. So there are ways to address this to some degree, but it is still just a much bigger consideration when you're exposing a new asset or new new parameter shifts to everybody in the protocol unilaterally. And I don't know that kind of governance on behalf of all current users and potential future users is the right way to make a lot of those decisions.

Right. Yeah. You said there's a a cap, uh a ways to like constrain these risks. Well that's the that's the frontier that I was talking about, the efficiency frontier. Yeah. Whereas like the protocol could in theory be more profitable if it didn't have to constrain the risks, but it does have to constrain the risks because there's a trade-off here.

And we've we've seen this happen, you know, time and again with exploits, both big and small, um, on these these lending protocols, right? There was the uh infamous Mango Markets uh attack, right? Which which involves kind of manipulation of the Oracle price around a pretty illiquid um longer tail asset. And, you know, the the funny thing is, like, you know, yes, Mango Markets was the one exploited there. The same vulnerability um was present in both compound and Ave at the time. Um and actually I know uh a couple folks who had reported to those DAOs and tried to tried to claim a bounty for reporting this type of exploit being possible with logger tail assets. Uh, and they were, they were basically told, ah, we don't really think that this is realistic. We don't consider this a vulnerability. And then mangle markets happened, and they both very quickly moved uh to adjust their risk parameters because they realized, yeah, this is realistic. Um and you know, from their perspective, you know, it may have been totally reasonable at the time to say, yeah, we don't think this is a meaningful risk. But there was no subset of their users who could kind of opt themselves out of exposure to the longer tail assets and just access liquidity on the more core assets like ETH and WBTC. Um, and so I think the the weaknesses have been shown in the tens and hundreds of millions of dollars in terms of what can be exploited from these protocols based on either bugs in the code or misparameterization, um, misassessment on some of the risk um pieces. And so that's why I think we need to look at you know, well, what's next? What can be done other than just doing a monolithic product that opts everybody in to the same terms?

Yeah. Okay. I really like that you uh invoked that word monolithic. Uh, because there are some patterns here that I want to put in listeners' brains. Uh before Paul brings us into the world of modular lending, I want to really want to define uh monolithic lending, which is what we're talking about with this one size fits all. Put every single feature into the monolith that we need, try to optimize the monolith, but it's still a monolith at the end of the day. And so, like Ave compound, these are kind of like um Swiss Army knives for lending markets, where like if we discover a new tool uh that's pretty cool, let's put it into the Swiss Army knife. But at some point, like your Swiss Army knife just becomes so big and hairy and just like a large tool that it like the whole thing just looks crazy, looks like a mess. Um, and there's like I like we said, there's a constraint on to like really how much TVL this thing can really attract because a Swiss Army knife, it can be optimized for many, many things. But really, there's a uh like if you want a really, really good knife, you don't use a Swiss Army knife, you use something else. And this is like this uh these patterns and metaphors have been described to talk about layer ones. Now we're using them to uh describe them to talk about like Ethereum applications, DeFi applications. Uh and so maybe Paul, now it's time to turn to you and and talk about like how Morpho is different. Um, we talked about uh the way that Morpho uh innovated into the space with um the Morvo optimizer. We call I called it like a layer two application on top of the layer ones monoliths that are uh compound and Ave. Uh so like once again, just speed run us through that, the optimum the optimizer, and then we'll get into Morpho Blue.

So Morpha Optimizer was uh imagined as a peer-to-peer layer on top of monolithic lending pools. So you we were talking about Dharma and East Land. Essentially, it's taking the peer-to-peer efficiency, but combining it with the ease of use and the liquidity of lending and borrowing platforms like Ave. So the idea is that we would match uh Morph Optimizer is matching lenders and borrowers of compound with a counterparty. But whenever there is no counterparty, well, the fallback is Ave. So the worst case scenario for a Morph Optimizer user is actually Ave itself, right?

And essentially building for two years on top of Av in compound, we kind of, you know, live through all the limits that that you guys just described. The most important of those being the risk, right? Like, you know, we've gone through all those war rooms of like, you know, either the economic attacks that were mentioned, but also a smart contract, right? Because we because we want to have all those features enshrined in one monolith, well, that implies a lot of lines of code, right? And well, at some point, uh, you know, you have some critical bugs that are reported either on imunify or or bilateral, and and as protocols built on top of those monoliths, well, you suffer from this because you have to uh maybe yourself going to upgrade because Ave has to perform an upgrade and and

We thought of uh Morpho Optimizer as you know a great protocol. It has grown a lot over the last two years, but uh definitely not scalable, at least in the way we wanted it to be. And so this is why we decided to to make the move uh from from Ave to Morpho Blue. And I guess the biggest learning from this experience is that

We think more of those monolithic platforms as funds rather than actual protocols in the sense that those are large pools of capital that are managed, but they're not predictable in the same way Uniswap is. For example, building on top of Uniswap, you can build maybe a protocol that's immutable itself. Building on top of Ave, you have to be reactive. You have to read the forums. You have to, you know, check the playload in the upgrade pattern. You have to make sure the protocol is not posed, etc. So it's more of building on top of an API rather than building on top of like an immutable core protocol. And this gave us ideas into how should we build true learning and borrowing infrastructure on top of which reliable layer twos can be built themselves and can grow actual businesses.

Okay, so Morpho Optimizer being the layer two, Morpho Blue is is kind of like, well, I I would like an alternative layer one to settle on. Uh and so Morpho Blue is like this new lending primitive. Uh that Morpho Optimizer is also a layer two for, uh, but now this primitive is is looking different. Can you uh kind of just illustrate us like what does this new primitive look like?

Yes, and I would even call it a layer zero in the sense that we're MorphoBlue is essentially focused at lending and borrowing and just this. And it unbundles completely the risk management part and the UX part, right? So on MorphoBlues, there is not even the concept of management or risk management. It's just 600 lines of code of landing and borrowing, one collateral and one loan assets, right? So it's much more like Uniswap than it is to Ave, but applied to lending and borrowing. So one MorphoBlue pool is, you know, maybe east collateral USDC loan assets with one specific risk and Oracle. That's just it, right? The upside of this approach is that we get rid of all the complexity and all the risk, and we can focus at doing one thing and do it in the most efficient way possible. The downside is that it's not a product, right? It's not a uh you have as a lender, you have to go to MorphoBlue and choose your pool, right? On Aave, it's all managed, you deposit, there is one option for all the USDC lenders in the world. On MorphoBlue, you have n different options. Should I be lending to the pool backed by BTC or should I be lending to the pool backed by ETH? And this is where you know MorphoBlue is not a great product in itself and is going to require layers of abstraction, layers of risk management being built on top, essentially rebuilding the on-chain fund approach that Ave has is providing all this passive user experience and risk abstracted experience. Well, it's going to be rebuilt on top of MorphoBlue, but it's not in the core of the protocol.

And Paul, can you, you know, explain to us exactly what parameters go into a morpho blue pool? What are what are the there's a small number of them, but what are they?

Right. So when creating a morpho blue pool, you have to select the collateral assets, the loan assets, the liquidation LTV, and the Oracle. There is also an interest rate model, but that's you know waitlisted by governance. So that those are the main parameters that you have to go through. Similar to when you you know you list a pool on Uniswap, you have to select two tokens and a fee uh tier, right, on Uniswap B3. And here you would select an L LT uh liquidation LTV tier, essentially. So that's the only thing you need. And when it's deployed, you can't change anything. That's completely immutable, set in stone forever.

And so that we're actually kind of invoking some part of the prior version of DeFi lending before Avang and Compound, which was the ETHLand and Dharma models, where we actually have specific one-off implementations rather than a centralized liquidity model. We have one-off implementations of pools. There are still pools, and so multiple users can deposit into pools. So it's not perfectly like one-to-one borrower-lender being matched. It's still there's still like centralized liquidity, but there's also fragmented liquidity in that there are multiple pools. And uh I think maybe just like the way that this works is like the market will determine what kind of pools are useful while also still creating different pools based off of different needs. So it's a little bit of a hybrid between the Ave compound centralized liquidity model and the earlier implementations of one off um one off agreements. It's like a little bit of a middle ground between these two things, correct?

Yeah, exactly. So like Morphoblue at its core is completely permissionless. So anyone is free to sort of spin off their own landing and borrowing pools, right? And then the free market will decide eventually which pool are too risky because they have a too high liquidation loan to value, for example, or you know, that just do not have sufficient utility, et cetera. And so that enables the protocol to capture, you know, every single new market that Dan was referring to, all those projects that real world assets, liquid staking and ristaking tokens, for example, but also, you know, very blue chip uh crypto assets like the short tail of the market. Everything can be deployed, and then the markets allocate liquidity where it makes sense to allocate liquidity, right?

So, Dan, question to you. Like this so we have like the one size fits all model, but from Aave, and Ave can change what that size is via governance, right? We can change the collateral types, we can change the risk parameters, we can change the oracles via governance. Um, and then in this model, in the Morpho Blue model, where there is no governance, if you want something different, you just spin up a new pool. This is like governance via the market. The market will actually determine the construction of the pools and also the TVL of the pools and the risk parameters of the pools. So we're actually uh taking away governance power and we're giving it towards um expressiveness in the morpho blue pools that are created. Is this kind of how you see it? And like what are the benefits of that?

Yeah, I think, you know, in some ways it it ends up being similar to the transition from Uniswap v2 to Uniswap v3, right? In that uh it's still at the core, you can do the same things. But for those who want it, there's additional complexity available and an additional discretion. And then you get to decide as a user, the same way in Univ3, do I want to go in directly and select a range that I want to supply liquidity into? So I want to just fall back to the full range uh supplying of liquidity, or do I want to choose a uh an LP manager that will automate moving that range for me? And I have to put some additional trust and stuff into them as well. Similarly, in Morpho with Morpho Blue, you can either go directly into the pool yourself if you have very strong opinions and the ability to manage it and select your own risk and do that. Um, or you can still choose to delegate to any number of other parties who will help make that decision for you. Um and uh and and Paul and his team have developed um uh kind of a companion uh uh protocol to go along with Morpho Blue that is is being used for that. So maybe Paul, this might be a good time for you to talk a little bit about Metamorpho and how that ties in with MorphoBlue.

Right. So MorphoBlue is the first step that essentially creates the core lending engine. But as we described now, lenders and borrowers are left with their risk management decision, right? So as a lender, I have to select which risk I'm opting into, which you know most lenders frankly don't want to go through. Right. And this is why compound and RV are such great models, is that they're super passive. And Ave have their risk managers that are consultants for the Ave DAO that are going to set all the different risk parameters that we mentioned. So, like the Oracles, the CAPS, the liquidation LTVs. And you know, they're they're paying Chaos Labs, for example, which is a company that is going to advise for changes to those parameters. And this is what provides the very passive user experience for lenders. In MorphoBlue, this is not set through governance. This is set through the market. And so we are seeing some players, essentially risk experts, that are building vaults on top of MorphoBlue, uh, which are super passive vaults, the exact same way you would have Ave, where you can deposit USDC, and they're going to do the risk management for you. They're going to select those oracles, those risk parameters by depositing the capital in the relevant market, into the market that makes sense, right? And so there is this new class of players that are running what we call metamorpho volts, essentially passive volts of liquidity that are very similar to Ave, but rebuilt on top of MorphoBlue. The general idea is that we're rebuilding the Ave risk profile, the compound risk profile, the Spark risk profile on top of one trustless and efficient core, which is which is Morpho Blue. And so, you know.

Those could be Block Analytica, like the Risk Team of Maker, or Gauntlet, formerly like the Risk Team of Ave, that now uh left Ave to to build on top of MorphoBlue, but also entire DAOs. Like very recently, Spark and Maker, uh actually two days ago announced that they were building on top of MorphoBlue, like the the next iterations of their Spark lending markets, where they're basically depositing DAI into their own vault. And the Spark DAO is going to manage uh the vaults, uh essentially.

One theme that um I'm seeing emerge across the crypto space is this trade-off between sovereignty and um fragmentation or sovereignty and composability. Uh and so like at the very bottom of this morpho stack, we have morpho blue, which is fragmentation, so also known as sovereignty, uh, where each one of these pools gets to choose exactly what is best fit for them. And now the problem in the in if we're talking about the Ethereum uh role of centric roadmap design is well, well, we have the fragmentation, we have the sovereignty. Now can we get the composability? Can we get the centralization of liquidity? Uh, can it can we get the abstraction? Uh and so metamorpho is the abstraction layer on top of uh Morpho Blue that from the end user makes it perceived to be something like Ave or Compound, which is just this stupid simple lending application, deposit your ETH, borrow whatever. Uh, while there is underlying uh mechanics that allows for the sovereignty of Morpho Blue pools, uh, but the composability and uh recom like unfragmentation of these pools at this like metamorpho layer. Is this is this a way to like kind of uh uh illustrate the stack?

Yeah, exactly. And like Metamorpho is here to re-aggregate liquidity in the same way, you know, Ivan Compound would and provide this passive risk managed experience. Essentially, that's the product, right? And MorphoBlue is the core protocol, the core infrastructure on top of which specialized risk profiles, specialized on chain financial products can be built through Metamorpho Vault. And that's really how we unbundled and decoupled lending and borrowing.

Dan, can you go into this like risk management layer that Paul was talking about? Like I think everyone's pretty familiar who's paying attention to the DeFi lending space of this like tenuous relationship between uh uh Ave and Gauntlet. Uh can you talk about how that relationship changes? Uh and like what is what is this like um this middleware misc management layer that's uh between metamorphosis it. Uh sorry, what is it, Paul? Metamorphos.

Minimum.

Mm-hmm.

Metamorpho vauls, yeah. And then the metamorpho blue.

Yeah. So I mean, if you look at uh historically, um Gauntlet has been a consulting firm, a very, very good kind of like risk consulting firm and kind of economic analysis. And I think they first got on people's radar back when they started uh being one of the early governance participants for compound and helping to uh rec make recommendations uh there.

As the space started to mature and we started seeing more of this kind of decentralized governance happen, uh, they started going to Ave and to any number of other lending protocols and various types of protocols that needed parameterization via governance and saying, hey, you know, we are the smartest guys in the room, uh, you know, pay us a fee and we will make recommendations, you know, over the next year or whatever. And those are typically, you know, just flat fees, millions of dollars a year, but uh, you know, capped as a typical kind of consulting arrangement would be. And there are billions and billions of dollars um in these lending protocols today that were effectively, uh are effectively managed by compound and and similar kind of risk advisory uh agencies. And what Metamorpho does is say, well, okay, if these guys are actually doing the hard work, right? What the main thing is uh there's there's always a connection between risk and reward. In efficient markets, there should be a connection between the riskier, you're taking the rewards that you can get. Um and due to the uh, I would say, idiosyncrasies and and difficulties uh within DAOs, it's very hard to make that direct connection because you're always kind of quibbling over, well, how much of the TVL, how much of the interest earned is due to changes recommended by Gauntlet or one of these other parties, how much is due to the marketing and protocol upgrades and these other things. Um and uh now what Metamorpho vaults do is it allows these risk managers to directly manage risk. And if you trust uh Gauntlet or you trust B Protocol and Block Analytica or Re7 or any of these other guys, you can deposit as a lender directly into their Metamorpho vaults and know that you are being exposed to the risks that they say they are exposing you to in this vault. So it does give you that kind of easy mode to just deposit in and you know delegate those risk decisions to somebody else. But then uh it's just them. You're not also additionally subject to governance by a broader DAO and introduction of additional assets by other parties that that this risk manager may not have been able to uh you know properly advise on. It it all just gets concentrated at that layer.

And so this gives the risk managers like a way to express themselves on chain. Rather than being off-chain third parties that that propose inputs, this allows these risk managers to go straight to the chain themselves, put their put their opinions in on chain, have people ascribe to or not subscribe to though their opinions about risk, uh, and then actually um get paid some sort of service fee as a result as a protocol, not like a third party advisor, correct?

Yeah. And and Paul, can you describe how that how that works for them in terms of incentive alignment?

So essentially on in the AVA model, you have hundreds and hundreds of risk parameters that need to be set. And this is a decision of the token orders. And obviously, token orders are not qualified, right, to determine if the LTV of you know STE on polygon should be 86 or 87%, right? So they contract with the risk managers like chaos and gauntlet. And this is where they come in with propositions in a daily or weekly basis. They come up with like updates of like 50 different risk parameters. But in this model, we can clearly see that incentives are not properly aligned. First, because none of the algorithms used to propose risk parameters are open source today. Like all the different risk parameters that are set in the AVI platform today are not open source. And for us as a builder, on top of AVID has been a problem in the past because we could not make, you know, for rezoning completely behind why certain parameters were decided. And the reason why it's not open source is because they have this consultancy fee model where if you know, for example, chaos or gauntlet were to open source their model, where you know they would not be able to charge a fee anymore, simply because the DAO could use the algorithms uh on their own. And so

This is how the model has has has been working so far. And

in the MorphoBlue model, this is an open marketplace. Like this is

A free market for risk managers. Anyone can spin off their metamorphos and decide to do risk management the way they like, whether it is completely closed source and super centralized or completely open source and completely decentralized as well. And so what we're witnessing right now out of all the different metamorphos that we have is that, for example, I think Block Analytica and RISDAO are going the fully automated and open source way, right? Which acts as a new value proposition for users. Like as an end user, I'm going to deposit into the Block Analytica vault because I know it's fully open source and fully automated in a way that I could not get elsewhere. I don't have the same guarantees in any other landing platforms. But you also have like Curve and Maker building their vaults with the same governance mechanisms that they used to have in their landing platforms. So I would say that the big advantage of Metamorpho uh on this front is really to opening the market for risk management. And from this will emerge the best practices and what the market truly wants, right? It won't lock users into like a one size fits all risk model with one specific business model approach.

Okay, so I want to kind of summarize the structure that I'm seeing here. And then Dan, I'm going to hand it over off to you to get into the really technical nitty-gritty questions that I don't know how to ask. And so starting at the very base, right, we have this like this older one-sized monolithic model that is Ave compound, where one governance decision impacts every single uh collateral and every single uh borrowing asset inside of Ave. Uh and then we're moving towards this um more siloed pools model, which is Morpho Blue, uh, where there are four different inputs into spinning up a pool: the uh lending asset, the borrowing asset, uh, the oracle, and then the fee structure. Uh, and those are the only inputs. And this goes from like having like one universal one size fits all to like uh sizes small through XXL, women's and men's sizes. Like everyone gets a size, right? And so like we are thinking that this increases the frontier of efficiency for Morpho because we have more high-fidelity uh pools that can um be fine-tuned to specific market demands. And the cool thing is that the market will actually determine what pools are created and what pools have outside success. And when the market is free to express its things, uh express its desires, we are seemingly finding a more efficient frontier for DeFi lending. But then we have the constraint of um not every all the liquidity is centralized into one single pool like Ave and Compound, which is where we saw their success come from in the first place. But this is what the next layer on top of Morpho is, which is Metamorpho. And this recentralizes all the liquidity. And then we have uh the risk managers really allow for that uh that centralization of liquidity to be uh constrained in a in a favorable way that is optimizing between the interests of the pools, the interests of uh the borrowers, and the interests of the actual risk managers themselves. And this is where we get back to these, the the scalability, where like we have all of these different sizes and now we can scale them out with Metamorpho. Uh, and then we also have just like a lack of dependency on the centralization of Chainlink. Uh, we have risk managers become like a part of these systems. And this is like the a new stack that metamorpho that Morpho is building. Um, Dan, is there any like part of this stack that I missed that you want to uh talk about? Or like where should we go next here?

I I think those are the parts of the stack as they exist today. And I think there's there's some more things that will come in the future. Um what I think might be interesting to do is actually let's break down what goes into a lending protocol. What are the kind of core design decisions that need to be made? We've talked about the kind of like modular versus monolithic, but there's there's a lot that goes into it. And so you know, the place I'd like to start is, you know, Paul, from your perspective, like what is like the number one mandate uh or or even one in top one in two mandates for a lending protocol? What must they get right uh to earn the right to to be used?

The first requirement could not be anything else than security, right? Like just looking at the number of exploits, but most importantly the number of war rooms and incidents behind the scenes that happened is really orders of magnitude behind what we should expect from you know the future of the financial infrastructure that we're sort of claiming DeFi will become.

I want to add that it does make sense that security comes first because in my mind security is where you get moneyness properties from. Uh Ethereum highly secure layer one. ETH is money. Uh and so it makes sense that a money market optimizes for security first and foremost because that that preserves the moneyness of the assets.

Yeah, no, I absolutely and and