The Role of DeFi Risk Curators | Adrian & Tarun | S9 E5

Transcript

And the difficulty with DeFi is that you know the the effective duration of DeFi has always been 12 seconds or the length of a block. And this is also why it's this is part of the reason why when you have arbitrary lengths of duration in fixed rate protocols, the liquidity completely dries up in that period, especially when there's an absence of other lenders or other users of that fixed rate at that particular maturity point in time. So yeah, leaning on the curator to essentially make that market a bit more liquid is a valuable service. Quick shout out to today's sponsor, Katana. We'll hear more about Katana and their pre-eposit campaign later today, but for now, let's get into the episode.

Hey everyone, quick disclaimer before we get into today's episode. Nothing said on Bell Curve is a recommendation to buy or sell securities or tokens. This podcast is forformational purposes only, and the views expressed by anyone on the show are solely our opinions, not financial advice. Our guest and I may hold positions in the companies, funds, or projects discussed. All right, everyone. Welcome back to another episode of Bell Curve. Uh, today I'm joined by my trusty co-host, Monae. Um, and today we're speaking to Terrun of Gauntlet and Adrian of Steakhouse. Guys, welcome. Hi.

Hey, glad to be here. Hey. Um, okay guys. So, this episode's going to be really fun. Um, we spoke to Paul Frambo of Morpho uh on their V2. Um, and we spoke mostly about borrow lend, uh, stablecoin trends, uh, vertical integration, that type of stuff. Um, and we touched on risk curation as a genre, but we thought it might be fun to go, um, one layer deeper even than we did on the Morpho V2 episode and speak to uh, the two largest risk curators on the market and um, Monet also is involved with BA Lab. So, I'm the only non-risk curator here, so I'll mostly be staying out of everyone's way, but um, we've got a lot of ground to cover. And maybe just to start off, um, Adrian, if I could kick it over to you first, can you just give us an understanding of, you know,

just define the basic role of a risk curator here and how this market has, you know, kind of started from basically nothing a couple of years ago to being a relatively large market today. Yeah, definitely. It's I think Morpho is definitely the instigator of this change. I mean there's always been an element of userdriven behavior in protocols like unis swap where anybody could sort of create their own pools for any token but I think morpho was really the first to have this clearly defined and nomenclature and role of a curator that I feel is kind of two things one of them is modularization so this fits very well with sort of morph's philosophy of having the smallest possible financial

primitive that you can And then when you stack those on top of each other, you can create more complexity and allow uh people to sort of build on top of it. And the second is outsourcing the role of the risk and you know putting that in a silo and allowing people to make decisions on risk without necessarily impacting for example other markets or or other curator activities. It it comes with trade-offs obviously, but I would say this new cohort of Boralin protocols is probably making the bet that you can scale much faster under this curator model than by running a c a sort of monolithic approach.

And Tun, I I'd love to get your take on this as well. And maybe if you could give a little bit of an overview about how Gauntlet got into this because I think most people myself included until relatively recently have thought of Gauntlet mostly as kind of an advisory or a consulting sort of group. Um and recently from my understanding it seems like risk curation is a larger and larger part of the business. So could you just kind of give us you know the evolution from the business standpoint and then anything to build on what Adrian said in terms of the development of the market? Yeah, for sure. I mean um so we we started in 2018 um at the time I think the word defi didn't totally exist like Zharma and compound v1 which doesn't really exist rip uh were kind of the first I mean maker was obviously

really the first but then those fast follow D5 protocols or sort of the compound at that time and you know there wasn't much dex liquidity there's no unis swap um and so you know we started actually in terms of like trying to optimize proof ofstake protocols like how should they think about issuance you know actually ironically a lot of the stuff that people talk about now um with sort of some of our earliest research papers and um writings and actually consulting with uh different POS protocols at that time was actually around like how do you choose issuance how do you adjust issuance in order to maximize throughput how do you do it in a way that doesn't just like centralize

a validator set right like subject subject to those constraints. And in those early days of what I'd say were protodei, it became kind of clear that proof ofstake networks will, you know, despite uh the 2017 promises of like they'll have governance and fast updates and whatever, that never really transpired for a bunch of systemic reasons, right? like a lot of the big asset holders don't want to be actively participating dot dot dot and DeFi was s was sort of the first place where you saw a more active role of members of the community and like designing the system and so when we started we we you know that's sort of how we kind of got into

this you know is realistically there were you know I'd say the entire DAO consulting advisory type of stuff started from like probably like 10 people in Discord who Monae and I are in group chats from that era uh in uh kind of just like talking about how you should model different things, what the risks are, etc. You know, obviously that evolved. I think you know, compound um kind of opened the spigot. You know, I think maker obviously had a particular structure, but then once compound made it easy to copy governance, you saw this kind of Cambrian explosion. So we've been doing that for a while and I think we you know written a lot of research papers on trying to formalize DeFi

models formalize the risk and then also you know have the numerical evaluations and and kind of stress tests and I think in late 2021 we started to think a little bit about some of the um adverse selection moral hazard type of stuff and we you know we at the time we were working with a compound and and We kind of tried to point out a certain like moral hazard issues with how their insurance fund worked. And it was at that time it was very clear that you know the DAO model has this very weird incentive where any kind of action you have to vote on that maybe has a shortterm negative price impact but in

the long term makes the system safer was like very hard to convince this group of people of. It was a little bit like proof of stake where it's like really hard to convince proof ofstake asset holders to like change issuance. You know, we saw this with Salana, we see with Ethereum, like Ethereum still hasn't even really fully made that change or attempted. And um I think in the Dow model you have a very similar thing where you kind of have this like weird incentive model where like the people who own the token might be willing to take might be willing to force either more insecure parameters on the users or like kind of not want their token to be the actual backs stop insurance if like something goes wrong. And I think you know there

since there have been many cases like that but I think our experience with that in general amongst other things which are all kind of you know documented so if you care you can read all the forums from 2021 to 2023 kind of made us uh excited to try some new systems and new models um in particular because I think we had written a lot about different types of models of like how you know most DeFi systems in what I'd call D5 1.0 know of you know like maker compound unis swap generally we're like tripartite agreements between three to four types of participants there's like a liquidity

provider or a lender there's a borrower or a swapper there's an arbitrageer or a liquidator and then you know maybe there's like an oracle provider right like those are kind of the four principal agents you really need to think of in all these systems um and one natural question was like you if you had a fifth agent who is able to coordinate them more efficiently, would you do a lot better? Would you be able to make it more efficient? Would you be able to make the utilization in the protocol better, improve capital efficiency? Right? Like you can kind of think as Unis swap v3 as sort of doing that because you split liquidity providers into active and passive and you were able to improve capital efficiency. And so it started to become clear that if

you want capital efficient DeFi, you kind of had to go to this model where there was another type of agent who is, you know, being kind of very active in allocating. And that's sort of that's sort of the philosophical reasons we got into it sort of uh and then the path into that. Um but then you know I think once we started doing it it made a lot of sense. So yeah, sorry I know that was a little long answer but no no it's always good to get deep into it. This episode is sponsored by Katana. If your assets are sitting idly, they're dead. Katana is built from the ground up to bring your bags back to life. Be early to Katana by joining the pre-eposit campaign before mainet opens at the end of this month. Follow the link in today's show notes to get started.

Today's episode is brought to you by Ledger, the global leader in digital asset security. Celebrating 10 years of innovation, Ledger is trusted by 7 million users, secures 20% of the world's digital assets, and remains the most secure hardware wallet for your crypto and login. Follow the link in this episode's show notes to grab your ledger today. As always, investments in blockchain technology involve risk terms and conditions apply. Do your own research. I'm kind of curious. So, you mentioned that there was like a bit of like uh I'm sure you guys have both seen this cuz you you both have done, you know, sort of curation but also some more like advisory engagements as a risk provider. Um you mentioned that there's like some conflict sometimes between trying to reduce risk but then because

it has you know price impacts maybe maybe the Dow is resistant. Um but I've also seen interesting um sort of incentive gradients as a curator as well like if you take more risks and you have uh higher yielding vault maybe you can capture more fees. So like your your sort of revenue position as a curator is better. Um, so yeah, I'd be curious uh for either of you if you have like a view of how how does the incentive structure as a risk provider change based on this, you know, kind of this business model aspect.

We definitely took a like I I echo the the adverse selection issue like our our history at stakehouse is is in maker. It's in like the DeFi 1.0 so to speak. Uh and there the sort of old core unit model is one where maybe the fact that maker operated as the DAO worked to its benefit from a risk perspective because it meant that basically decisions were much harder to make and enforce. And in particular, when you have a group of core units that are advising a DAO, it the more difficult things are to change, the safer, like the less reactive the Dow is to market conditions, but also on

the long run, the safer the the system just out of a virtue of lack of change. And I think that there's an interesting element in the morpho curation system. We've we've kind of taken part of that philosophy into how we're designing the morph bolts because there's a there's definitely a trade-off like there but on the other hand those trade-offs are more easily segmented. So with maker you kind of buy wholesale the risk of maker when you use the stable coin or when you open a CDP position and you're exposed to the full spectrum of of path dependent decisions that people have made as as a community. Whereas with a a more fragmented curator ecosystem, the user is able to also, you know, define their

risk appetite and select the vaults that that match those. Uh, nonetheless, I think there's really interesting things that you can build on top of it. So, in our vaults, for example, the LPS are essentially a mini DAO through the Aragon DAO construct. And this closes the loop quite neatly on on that sort of principal agent tension and makes our vaults harder to change. That's our person that's our sort of company philosophy on on making things more immutable and harder to move. There's definitely an incentive.

There's an element of it's funny because in crypto there's an element of risk premium that that consistently gets underpriced. I feel like people participants in DeFi nowadays generally underpric the level of risk that they take with different collateral assets and so there's there's maybe not enough maturity around sensitivity to the different types of collaterals or different types of vaults but there's definitely an opportunity for creators for example to differentiate along those vectors and you see that with you know sort of blue chip vaults like the gauntlet prime or the steakhouse USDC versus other more edge vaults or ecosystem and bolts that are catered to a specific use case and to a specific risk profile.

Yeah, I mean I I think I more or less agree with that that like specialization will come from how you trunch risk. Um, I think that sort of inherently is somewhat sometimes the problem in the DAO model in that it's really hard to get DAO members to accept the tanches because some DAO members are benefiting potentially from the highest risk assets because it's like, hey, maybe they're the largest holder and they're getting the highest per share of fees or they're a participant in that market and the, you know, the Dow's effectively subsidizing them by giving kind of like a cheaper loan for that type asset than like the true value should be. And I think fundamentally having having

basically people be able to compete to try to offer different types of verticals and specializations over time should converge to something that um does a lot better. And I like there's there's a sort of you know empirical sense in which you can kind of see this if you look at sort of morpho yields across um time versus you know a compound yields like the utilizations will generally be much closer to the the kink level or to the target level um than they are in kind of poolled protocols. Uh and I think you've seen the poolled protocols add isolated modes to emulate kind of some of the successes. I think you've seen in in in

the the vault model. Um I think one other important question you're you ask is about sort of like hey is there kind of yeah like this like systemwide principal agent problem. I think the nice thing about the vault model at least for morpho and then we can talk about other vault models like in perpetuals or in uh restaking where they kind of have a different I would say you have to think of them uh quite differently. Um, in the morpho case, the really interesting thing in my opinion versus the, you know, a or compound original pulled model is you you segment the borrower and supplier sides where

the supplier sides are sort of managed by the curators and the borrowers are just managing their own positions, but they're pulled. And this kind of matching problem is sort of almost like a one to many matching problem in the sense that like a single pool of borrowers is matched to many suppliers. And that type of matching can be done relatively efficiently. And I think you know you talk about morpho v2 that like kind of breaks that and goes you go back to this one one kind of matching mixed in. Um in general historically D5 protocols that have worked have always had some form of passive liquidity provision. um you know any protocol you could think of that works ha h that has worked in the sense of like survived for a long time hasn't been exploited um has real users has generally not been pure

RFQ so like unis swap x you know yes there's an RFQ element of people doing the intent solving but there's still quite a bit of quite a few trades that basically get go to the main chain and don't get fulfilled by the market makers effectively um I I think like pure RFQ intent type of things that require lots of interactions just don't generally work on chain. And I think what you're doing with curator competition at different levels of risk is effectively building in that kind of like auction mechanism for each transaction which you don't really have in kind of the original DeFi protocols in the Dow.

Well, Tan Holland, can I ask a question about that? So the criticism of that RFQ mechanism like having the RFQ layer and then the onchain more passive layer is that that should be bad from a markout or lever perspective for the onchain portion, right? Because the market makers, the professional market makers who participate in the RFQ just get to pick off all the trades with the juiciest spreads and then they kick the scraps on chain. So I don't know if that respond. So the thing is RFQ models have this one fat pure RFQ models have this one fatal flaw which is they they lose livveness. Like if there's if there's a really big market shock like all the market makers will stop quoting. Um and the average user who exactly at those times is like wow I want to get out of

all these positions. they will never come back to your protocol if like they tried to hit swap and it basically never was fulfilled at all. So you need the you do need that backs stop and usually at those times that's when you do actually see LPS make um non-trivial percentage of their profit. But yeah, there's definitely this this trade-off of having the hybrid models. But I would argue that the vaults right now are sort of a a kind in lending it's a little bit different because you know the you know when you talk about a markout in a AMM you're sort of talking about many trades in some time horizon and you don't know the number of trades. Whereas

when you're talking about the loss inured in a liquidation, it's sort of a fixed size and duration and you can actually exactly measure your your loss, right? So for lending, it's actually a lot cleaner to have this separation. And I do think there's less adverse selection provided there's some some amount of competition because basically you could think of the curators as every time a new loan enters like like one abstraction of it is every time a new loan enters they are bidding with their liquidity of how much they're allocating for that loan, right? For the the rights to the fees of that loan. But they don't realize the fees immediately, right?

They realize the fees over the direct. So they're taking risk to do that in a way that the market maker on a DEX is not really taking that much risk, right? They're they're taking the the duration of both legs of the arbitrage. And that extra duration risk is something real that you have to be compensated for and it's but it's also providing a lot of direct utility to the user because they just got a cheaper interest rate from the people bidding with their liquidity. Well, Tun, I'm glad and Adrian, I would love to get your response here, but I'm glad you brought this comparison up actually in between, you know, AM at like, you know, market making on or providing liquidity rather on Uniswap versus Curator on Morpho cuz they seem

actually somewhat related to me um in a sense, right? it it there's a the connection here being it looks something like on a form of onchain market making obviously providing passive liquidity on like a unis swap amm is basically exactly that um and then in moro it's it's a little bit uh you have to squint at it a little bit more but I'm sure there's quite a bit of control that the curators have over um you know this is these are the assets that we'll allow within our vault uh there have sometimes have to be uh gaps bridged so you have to inject liquidity maybe it's not quite the same thing But you kind of squinted and it seems like a similar activity. Is that roughly correct or talk to me about that comparison agree? Right. But there's there's

definitely the I mean yeah like they don't said they are they are different to a degree and the creator is yes sort of bidding for like the curator has to balance between bidding for fees or interest rates and being able to secure well the the number of variables that's open to a curator is I would say more there's more dimensions across which curators can optimize for their for their lenders relative to an AMM and fees is just one compon component of it. But for example, you if you consider uh hitting the utilization rate and not incurring the risk of a liquidity trap, for example, as being a factor or a variable, there are some trade-offs between generating enough lending return

for the vault users versus vault users being able to withdraw. And so it becomes a little bit more complicated. I think in time it's more yeah the the atom the at an atomic level the borrow market looks a lot like an AMM if you had curators allocating to individual markets let's say but the moment that you start pooling uh I think a lot more optionality opens up uh and and here curators will have different objectives you'll have curators that try to max yield and they rip the face off the borrowers by reallocating aggressively and then you have other creators that are a bit more prudent and try to equalize the rates in a in a more balanced way.

I I by the way the the auction analogy I make is is not uh a random or casual one. I like a few months ago I tried to actually like write out so so there's kind of this these kind of classic results in auction theory that's like how many biders do you need for an auction to be mostly competitive like not you get 90% of the maximum revenue or 90% of the welfare. Um, and it's like the number you get for a lot of cases is like under six. Like you don't need like a ton of biders in order for it to be super competitive.

Um, and yeah, I wrote this paper on like, hey, for these curator models, how how many curators do you need to get to the point where you are, you know, very close to the optimal utilization or target utilization and it it's it's not that many. So in some sense there is a version of the world where um you know that yeah there's a lot more dimensions to take into account but also by just having a few a little bit of competition you're already going to be like significantly better than the the static poolled model. Um and I think that you know if there is there is this question of like hey how much of a trade-off you know in AMM that's not so

obvious and in fact in a lot of cases that's not true where just because you have more biders doesn't actually necessarily mean that uh you get that much better prices and like you can you've you know there've been a bunch of recent academic papers comparing cow swap and new swap x kind of showing this type of thing where like the improvement for the user isn't that much relative to the losses to the LPS. PS um but in the lending market you can show it's actually much cleaner in some ways. So well you see that in the difference between like the bluechip lending markets against scrap Bitcoin or Rub stick versus some of the more tail assets where fewer creators allocate and there's more active rebalancing and more of this auction dynamic which is interesting. Shout out to today's

sponsor Katana. Katana is a DeFi first chain built for deep liquidity and high yield. No empty emissions, just real yield and sequencer fees routed back to DeFi users. Incubated by Polygon Labs and GSR. Katana is launching with notable ecosystem of primitives, assets, and partners like Morpho, Chainlink, and more. Be early to Katana by joining the pre-eposit campaign before mainet opens at the end of this month. Follow the link in today's show notes to get started. Ledger is the global leader in digital asset security, trusted by 7 million users and securing 20% of the world's digital assets. In recent years, Ledger has built a comprehensive ecosystem of products and services, making digital ownership more secure and accessible for everyone. Ledger's newest

devices, Ledger Stacks and Ledger Flex, introduced the world's first secure touchcreens, making it easier and safer to manage your assets. Alongside Ledger Stacks and Ledger Flex, they have also launched the Ledger Security Key app, offering a safer alternative to traditional password management. Follow the link in today's show notes to get started with Ledger today. As always, investments in blockchain technology involve risk. Terms and conditions apply. Do your own research. To kind of follow up on that, I'd be curious. I I imagine you guys are both uh deep in the weeds of how you'll be setting up your strategies for Morphobe 2, which was uh just recently like publicly announced.

It's um dissimilar in a lot of respects where it it basically like um asbalt curators that are um sort of selecting the sort of risks that they're they're addressing on behalf of uh depositors but it's based on fixed rates. So, um, how how do you kind of see like the design space as a a curator changing? And, um, do you think that there might be like more maybe like more similarities to market making or active like sort of PVP aspect to it?

would I mean we we we we have like this trifi lens of looking at things and the morphob or the fixed rate model component of it at least looks a lot like the there's a there's a useful model in sort of on the run and off the run treasuries where you have this sort of open market per treasuries that's going on all the time but it's very ill liquid unless it's at specific points in time when all of the bids concentrate trade and that's usually at the sort of primary issuance. It'll be interesting to see how this market develops. I expect that the blue chip assets will settle very quickly into a

cadence. It'll be interesting to see what that cadence is where curators almost coordinate in an automated way to make sure that the auction markets for specific collateral assets at specific points in time are the most liquid possible. uh it'll get probably a lot more chaotic around longer tail assets. Uh so it'll be interesting to see how how the market responds to this dynamic and and where the yield curve settles. But I think one of the one of the points that made earlier which I think Morphob V2 makes really interesting is this ability to sort of switch from the lender perspective between the fixed rate vaults and the existing V1 vaults. So there's a very clean transition from the

lender perspective from having this potentially more efficient or more competitive auction market which is a bit more discreet in in terms of its structure and defaulting down to a more continuous rolling uh model for the V1 markets which will hopefully provide a better lender experience as well. Yeah, I mean um I I guess I also take a step back and think of all of the failed fixed rate protocols uh in the last four years. No, more than that. Sorry, it's 20 years seven years cuz like actually even when compound v2 came out I remember there were all these people who were like trying to make uh fixed rate

D5 protocols like probably that was around the time a was probably switching from ETH lend to what they are now. Uh and there was there was already like kind of this huge demand of like hey we want to have fixed rate lending protocols and what people realized was hey the only way to kind of replicate fixed rate stuff at that time uh was to kind of have an AMM uh style pool where you could trade the variable for fixed rate and somehow like do a duration transform. And I mean I I think a lot of the people working on this wouldn't have called it that because I think they most people were developers and not traders at that time.

Um but there then you went to you had this whole line of protocols who were trying to do fix and I I can tell you as an investor in some of them I I understand I I lost my shirt in many of these so I understand the the uh the the dynamics that occurred but basically what happened was the AMM pools were sort of very inefficient. So compared to like the the lever arbitrage loss type of thing you see in a spot AMM the arbitrage loss in these fixed rate AMMs were really large. Um but an interesting thing that came out of the fixed rate AMM market so maybe there was like a hund00 million invested in it in 2019.

Why was it so large? Sorry Tun what was and lever by the way for Yeah we've also I realize we've just been using that word. Could you just define Yes. what that loss is for the LPS and then why is it bigger in a fixed rate versus like a unique yeah so lever and you know there's many arbitrage losses it's not the only one but the simplest way to think about is like you're an LP you put two assets in um into a pool and someone makes a trade against you generally speaking they're making the trade if on average the reason they're making the trade against the pool um in the direction that is opposite to the mo more valuable asset like are arbiting a more liquid pool, then you as the LP are taking an arbitrage loss. Like you're

you're owning the asset that's worth less, more of the asset that's worth less and less the asset that's worth more. Now, you can fractionalize that into components that depend on, you know, how much of the order flow is bad in that like toxic like that versus not. And and you can think of lever as quantifying the toxic versus not toxic. Um, and the name loss versus rebalancing is because the arbitrageer is rebalancing this portfolio you've given to AMM, but they're charging you a fee for it, and their fee is larger than the benefit you're realizing. Um and so the reason the fixed interest protocols uh arbitrage loss was so high was that you know a lot of times you had you have

to hold the duration risk over the fixed rate loan over you know whatever the duration is. Um and I have to look at the current onchain interest rates the variable rate and say like okay how much am I willing to pay in basis points for holding this fixed rate loan. But if the the variable onchain rate, you know, say a whale took a huge position out, like say there's a new yield farm and someone, you know, spiked utilization a lot, you would suddenly be taking this gigantic loss. And so if you look at the interest rate markets, they're much more spiky the demand. It's like less, you know, smooth like uh spot trading. And those spikes cause you like all your loss basically. So that that's I'm trying to give like the ELI 10 or five

or whatever version of it. Um and the interesting thing is the most successful thing to come out of all the fixed rate interest rate pro protocols is actually Pendle. So Pendle which probably is five billion TVL something like that right now four or five uh they they started as a fixed rate protocol and just kept pivoting and iterating until they realized that hey we're going to offer this fixed rate type of exposure but we're not going to do variable to fixed rate trade-offs. We're just going to take a yielding asset split it into the yielding part and not yielding part and and offer this kind of uh trade. And so the interesting thing is that whole era of the early fixed

rate protocols I would say the the crowning achievement was Pendle worked um and it worked a lot right like it and and they grinded for years so like I think everyone thinks of the success of the last two years for them but they've kind of it took them a long time to to get to that so the reason I bring up this history is this is also I think why people stopped being ambitious about fixed rate protocols because they were like everyone tried and like tons of money went into it and it didn't work out. Um, but I think the interesting thing about uh Morpho V2 and a v4 and and likely changes oiler, you know, I think everyone is going to kind of do

this thing is by segmenting the market into the kind of poolled variable part and then kind of allowing the same liquidity to be used uh in both fixed and variable. you're avoiding this kind of inefficiency of needing the AMM. You're making the curator figure out sort of the ratio, the price versus like having a fixed AMM that like has tons of me that is is static and everyone can see how to trade against it. Um, and I think that will be, you know, hopefully this will prove out that like you can do fixed rate, but you kind of need to have this more active management to get to work.

So, okay, let me see if I can just repeat that back to you. So basically the the the loss and the reason why this hasn't worked in the past is maybe another way of this saying this as well is that yeah if you are a lender you're taking variable rate risk in which in a very cyclical very spiky sort of environment means you know you can LP at you know whatever 40% APR and then it can drop to zero that feels like a very different risk than if you uh lend at a fixed rate and then for whatever reason the market falls out you could actually experience loss of capital which is like a very different type of risk and there was a lot of uh lever which bleeding in the early models here. Now if you take the idea of a passive pool which needs

arbitrageors to rebalance it and they get to extract a lot of fees, you have more active management like a UNV3 or a or you know hooks or something like that. Doesn't that also just mean that is that who's going to take that margin? Is that the curators that get that margin instead? Like is this great for you guys? Is this a bull market for curators here or what happens to that? But it's a service, right? similar similar to how curators today like allocate between different markets like what what you're what we're describing at the moment is the difference in liquidity between off ther run treasuries and on the run treasuries and the concentration of liquidity at different points and the difficulty with DeFi is that you know the the effective duration of DeFi has always been 12 seconds or the length of a block and

this is also why it's this is part of the reason why when you have arbitrary lengths of duration in fixed rate protocols the liquidity completely dries up in that period, especially when there's an absence of other lenders or other users of that fixed rate at that particular maturity point in time. So yeah, leaning on the curator to essentially make that market a bit more liquid is a valuable service and it's comparable to the role of a curator in a V1 model, even if the there's there's an extension of moving beyond sort of borrow lend on site to borrow lend on a on a on an arbitrary duration basis, let's say.

I guess one other thing I would add is is it it's sort of like it's like being a primary dealer in the on the run off their own analogy, right? Like you do end up having to have these people participate in the auctions for treasuries to make sure that you can deal with this kind of demand shock and I think it's the kind of similar role as as you would see in traditional finance. Okay, last question here and then Monet, I'd love to transition into maybe you can lead the institutional part of this conversation, but what should the introduction of uh V2 or basically fixed rate borrow lambda cross maybe it's V2 and Morpho but let's say Oiler and A V4 adopt this as well how does that impact the market of risk curators and you know

in I think in a lot of these intentbased um protocol designs right that could be everything from V2 to like an across or you know unis swap RFQ style uh construction. There's always this idea that hey all of these builder solver curators are going to pop up and uh pick up the the free dollars that are lying on the ground, but they never fully seem to. So Tun this is kind of coming full circle to your question about how many different, you know, curators in this instance do you need for a market to be competitive? It's consistently surprised me how few actors there are in each of these markets. So I'm just curious if all of you, you know, Monae, you included, have thoughts on why this market has remained relatively small.

And I'm also curious, you know, how it seems like the fixed rate borrow lens adds a dimension here of complexity. And I'm wondering how that will transform the market. Will it make it larger, more concentrated? I'm just curious to get everyone's thoughts. Where do you want to go first? because you got it's one verse three in that the answer to this question. I know. Yeah. Sorry. I mean at least in the V1 space it's it's very competitive. The there there are quite a lot of curators.

There's like 20 or 30 or something like this. But um when you get to a space that starts to become more competitive and you see fe compression, you know, curation still require the barrier to entry is a lot lower thanks to tools like Morpho, but it still requires some degree of technical expertise and domain knowledge to succeed. So I would be surprised if the market doesn't end up concentrating around the top five curators say of uh at least for a specific vertical like around the V1 bolts but it's true well that said this is very focused on sort of morphos specifically when you have if you had for example a

or what have you introduced their own version of a fixed rate um lending protocol I I don't expect them to integrate or be interoperable at least in the near future. Similar to how with stable coins, you don't see singleness of money. So, you're not able to take a USDT to circle and redeem it for a dollar or for USDC and and vice versa. So, this kind of forces there to exist a multiplicity of stable coins across different protocols. It's likely that there's going to be much more opportunity for creators to build smaller verticals. But even then you'd have to abstract to the sort of lending market space and see which one of those is going to acquire the most volume and

and and how that evolves over time which will determine essentially the number of curators that are available. I I think like in general a lot of markets have this structure where they don't have that many participants or they have like kind of the curator market where it's like a small number and then a lot of long tail. um you know for instance like pay the payment for orderflow market now that the auctions are public you don't see that many market makers applying even though there's a lot of firms that could there's actually very few people who bid in like the auctions for like Robin Hood order flow for instance um there's a lot of reasons for that uh I think one for dexes at least um I think maybe two years ago I wrote this paper

on on like if there's sort of a a congestion cost from having more people bidding at the same time. So like in the UNISOP case X case, say there's a user who's like, I want to buy one ETH for, I'm making up a price, $1,000, and each, you know, the bids that are being sent, it's a Dutch auction, is like I'll I'll sell you for $9.99. I'll sell you for9.98. I'll sell you for $9.99.5. If as the people are bidding closer to the true value, they're also increasing their costs externally like, oh, all of them went to Binance to go buy the asset to fulfill with at the same time. They actually worsen the overall price that's

received because basically they're there's this congestion cost of all of the solvers having to go source inventory at the same time in order to deliver it. Uh, in lending, that's like not as clear where that would happen. So it might actually be that you could have more than I think the dex market really concentrated to like three very quickly like almost instantly and sort of this has been one one reason people have suspected um I think in the lending market it's going to be a little more like what and especially in the fixed rate world what's the differentiated borrower borrow demand you can generate right because I think at some level yes there is curation competing for the the loan rates But there's also just a sense

in which you are likely to have very different types of borrowers than you have for V1 when you can have fixed rate large size borrows but like you as the curator have to be like okay this large-siz borrower comes in and uh wants a duration one month 10% loan um but they take up 50% of my liquidity so that means I get you know I have to kind of look at oh well what would in hindsight how much would I have regretted giving them that loan for a month versus keeping it in the V1 pool and taking the the up and down interest. And so I think there's going to be a lot more of that type of stuff which will generally mean that the type of borrow demand and quality of borrow demand will decide I

think a lot more than in the V1 markets how how many curators can survive. But in general if you look at lending businesses in traditional finance versus trading businesses trading concentrates more than lending. there's a lot more special purpose lenders, tail lenders, stuff like that. And I I I do think like as as the the types of lending products grows, you might actually see the market grow. It just might not grow linearly. Um it probably grows sublinearly. The margins, right? is the the I see the curator role more as the wholesale lender in a traditional context where at least for a v1 construct potentially for a v2 there's a degree of origination

cost that's borne by individual players that have like the tail and large scale you know banks or balance sheets do the sort of topline senior lending and they kind of allocate across everything though it's true that with the v2 model there is an element of cutting down especially niche assets or or tail assets. I think this is definitely going to be an interesting dynamic. Yeah. A and also I think it'll move you know to some of the other things. I think I think there's more than just lending also that people are basically doing curation for now. So that widens the market quite a bit. Like in the perpetuals market it's a totally different set of participants. I think we might be the only ones who are doing

both right now to be honest. Yeah, putting my curator hat on for a moment. Um, yeah, I would say for the the Morpho V1 markets, it's um, in a way it's actually been just the way the the Wii won markets work like you all all the curators sort of need to consolidate around a few different pools with similar parameters to make the markets liquid enough u to even be usable for borrowers. So it's tougher to be differentiated but I think that'll be an interesting uh development for V2 is and also with some of these other you know curation verticals is just that uh the design space is wider so you'll be able to kind of uh compete on product a

little bit more rather than competing on on brand or or you know distribution or what have you. This episode is sponsored by Catria. Catria is the first zero knowledge rollup to enhance the capabilities of Bitcoin block space, enabling Bitcoin applications and unlocking Bitcoin's full potential as a programmable financial platform. With Citria, developers can now tap into Bitcoin's liquidity to build scalable payment systems, privacy solutions, and more. Citria is currently in its test net phase and is approaching its mainet launch. To learn more about Citria, follow the link in today's show notes.

As always, investments in blockchain technology involve risk. Terms and conditions apply. Do your own research. But yeah, I would say maybe this is a good point to kind of segue into um talking about like sort of the institutional side of the curation business. Um I know there's been really huge demand from institutional borrowers to access fixed rates. Um, and you know, Tin, you mentioned like maybe that'll be like sort of a differentiator is you're actually sort of taking on more of like an originator role rather than just allocating capital to opportunities that are already out there existing in the world. Um, so yeah, how do how do you think um, you know, the the sort of BD

side and the the sales side of curation is going to develop in an institutional context? I guess like in general I think the interesting thing between in these curator protocols is like the BD is sort of half shared by the uh core protocol as well as the curator I would say you know in the in these early ecosystems I think there's obviously more collaboration and then you know over time things change although I will say you know running perpetuals vaults Um the perpetuals market is like the people making the vaults don't like talk to the curators at all. They're just like do whatever

you want and and that market because it's so large com in terms of like the amount of yield produced the people just show up like like I think there is way more longtail curation in in perpetuals than there is in lending. Um mainly because I think you're taking way more risk on behalf of the users. So you can differentiate your strategies quite a bit like they're not all delta neutral. I think there's actually that's actually kind of an interesting thing is like I don't think the data on the perpetual side is like particularly transparent yet. Um but you know as as that improves I think we'll see more of that. But I think the lending market is kind of interesting because yeah, there is this wholesale lender aspect of like you do

want to be an aggregation point for borrowers and lenders and you do want to try to find differentiated borrow opportunities. I think you know you know we we I guess we in this podcast we've talked a lot about how Morpho maybe was sort of the first curation protocol. I would actually slightly edit that to say Morpho was the first successful curation protocol. I think there are a lot that failed before or that were like kind of doing the same thing. You know, Maple V1, um, you know, I think like some of the older kind of people trying to do like RWA lending RFQ style where you had like a curator who was like an underwriter originator and there was not really much

that the smart contracts were doing. I think in in you know those failed obviously for in the case of maple v1 is adverse selection but I think you're you are kind of you know what v1 and v2 are doing v2 is really doing is like blending in that model with the the existing kind of pure onchain model and the question is what is the relative sizes of those and I think that will change the sales like if the v2 model can actually bring more borrow demand than the v1 model can but the v1 model brings in all the excess liquidity, then you might actually see a very different sales cycle where people are selling to the institutional borrowers more directly. But if the the capacity is

limited by the V1 model in some sense, the V2 capacity, then you may see kind of more deals on the supply side. So, the reason I'm bringing this up is like I'm saying it's actually quite uncertain as to whether you're marketing the demand side or supply side until we see how a little bit more about how the V2 market develops. I know that's like a sad nothing burger answer because I said, you know, we have to see how it goes. But I'm I'm just saying I don't see a strategy where you're trying to do both at the same time. I think it's you either end up going all demand or all supply. And we we actually take the that's interesting and I agree that the morpho v2 option space opens up dramatically in

particular in with respect to the opportunities but kind of taking the Mike will be familiar with this but at stakehouse we wear balance sheet tinted glasses so everything is a balance sheet to us and all balance sheets start with the liabilities with the users so in the morpho context that's the lender essentially where the If you have to if to match to effectively match a two-sided market between the cash portfolio manager, let's say, and the risk portfolio manager, it kind the problem kind of starts at the lender and the the I think I mean when I originally asked something about like institutional like for us the institutional user or the DeFi user have our our thesis is

that it kind you kind of cater to everybody if you build something that's set up for institutional use later on. This meant that we had a bit of a slow start at the beginning because the way that we configured the vaults, for example, was quite conservative and boring. Um, you know, longer time logs and all this, but I think these things really lean into things that will matter in an institutional context. And then again sort of deriving from the stable coin experience when you think of the when you think of uh a regular bank for example where you have a separation between the operational asp the the operational execution of a bank and of running a a deposit balance sheet and of

allocating those assets and the legal enforcement of some of the claims on those assets. They're separate. I think the big uh difference in a DeFi context is that those two are fundamentally combined. And I think this is ultimately the value proposition for fundamentally moving uh more traditional finance uh transactions on chain where you can have the guarantee of enforcability and rule selection up front. And the more that you lean into these features and these characteristics, the more attractive you make it for, yeah, quote unquote institutions. So, in other words, I think and I think this is where, you know, in in the different trade-offs that the Morpho team aimed at or selected for when they were building the

the the early vault and and lending market, they they lean very heavily on immutability and lack of governance for certain market parameters. Uh I think to a degree part of this was as a way of concentrating liquidity. So for example, the LTV values are hard-coded and that kind of prevents you know infinite fragmentation across LTVs for curators to allocate to. But it's also kind of part of the value proposition also because you can very transparently disclose what the rule set is. And the immutability around governance is in our view a type of embedded supervision where the the more trustlessness you have in a vault, it essentially equates to, you know, less counterparty

exposure. And like these are the things that large scale capital allocators are looking at and they translate immediately into a benefit that an individual user will also have. So like something that works something that's immutable and governanceless or trustless that that makes it work easier and it's easier to explain to a trady user for example. But that benefit translates immediately to the person with $1 who who allocates the morph because they get the same benefit. And there's an element of because of this specialization of the vaults, you know, again relative to more monolithic platforms, the cost to being a minority user is lower in our view. So

if you are a a small staked ETH holder for example in Lido the the cost of a large LDO holder is larger until you have the mechanism like some with something like dual governance to mitigate that cost. And with Morpho because the vault the vaults will specialize and users will pick different vaults even a small user is proportionately a larger share of the of the LP base. and can therefore have an outsized impact on the curation decisions or collateral onboarding decisions. And so I think the the future of this business really is institutional. I I hope to see it in a in a place where we don't make that

distinction anymore. I think there's a lot of features even in V2 where just from a loan servicing perspective, it would be really interesting to see, you know, traditional lenders start to use that infrastructure to to originate stable coin denominated loans to to businesses or to individuals. I I definitely get the sense that that's where the morpher team wants this to to go. In the near term, it'll probably be more around uh still sort of very DeFiheavy cryptofocused, you know, speculation on on yields, the emergence of a yield curve around, for example, Bitcoin or Ethereum. Like these things are definitely interesting and they're valuable in creating a capital markets on chain. Erh but anyway sorry to to wind up a long-winded answer like

from our perspective the all balance sheets start from the users and that's on the liability side. Um and then everything else is kind of optimization around the parameters that you allocate or expose those users to and the more immutable you make it and the more trustless you make it the stronger a value proposition it becomes for all cohorts of users not just uh institutions. Afterwards, there may be requirements around KYC and things like this which you can build in more easily into a modular protocol versus a monolithic one and that presents certain advantages. Um, but I think fundamentally this is this is something that resonates uh with a with a traditional audience like people people get it very quickly.

I've got a question for you guys on maybe a potential future state here. um sparked a little bit, Adrian, what you were just describing, but also to reunion on your observation that there's a longer tale of more specialized lenders. Um I'm flashing back actually to a conversation that I had in like 2018 or 2019. Shout out um Ali Hamemed uh if you're listening to this but uh you know he was describing some of the opportunities around consumer lending and leveraging unique data sets like something like like let's say Tik Tok wanted to create a platform or a a platform where it lends to its creators right they have very specialized data they have a good sense of they can underwrite um hey I think I I once you said that I think I actually know this

person who you did the shout out for yeah he's great yeah sorry I can't remember the name of this a lot of this like YouTube creator lending for a long time. Yeah. But the reason I'm super interested in it is because Yeah. I there's a little bit of a connection in between. I've thought about this too from a Blockworks standpoint. Like I've seen a lot of backends of like new shows. I have a very good sense, right? Like just kind of looking at it, not even being particularly sophisticated, like is this person going to be able to make money based on what the show does in the first six months or something like that. Uh it's pretty easy to do. Um, and I would imagine you can get very very sophisticated if you had a data set as wide uh and deep as something like

Tik Tok. So today, you know, I think we're all looking at this from the standpoint of, okay, we're, you know, kind of slicing and dicing different loans so we can fund uh, you know, leveraged um, looping trades or something like that. Or maybe then the next thing that we're dreaming bigger is we're going to bring community banks. But I'm wondering if there's an even bigger opportunity here for especially fintexs or something like that or like a social platforms or something like that that have very unique data sets that they could leverage something like um a vault or plat like a you know infrastructure as neutral as a vault and I'm also curious um to Adrian and Monet how much of a value proposition and a cell from the institutional standpoint and I mean institution in the sense of

someone like a black rockck or potentially someone like a Tik Tok how influential is it that they have a say in controlling the infrastructure themselves as well, right? Like this is kind of the DeFi mullet question. So, that was a bit of verbal spewing there, but I'm curious if you ever see a future where there are some more differentiated vault curators and um I'm sure uh you know, BA Labs uh Steakhouse and Gauntlet will be right up there with the the Tik Toks of the world, but I'm curious if you see those if you play this out long enough that you could eventually be competing with some of these types.

Tune, you little very far out into the future here. Adrian, sorry. Yeah, whoever. I don't know that we're gonna start lending to Tik Tok creators immediately. But in the future, again, like projecting way out, there's no reason why like to make it a bit more grounded. For example, think of something like a theme lending for example for something like trade receivables. like this is a a loan the loan servicing and loan origination happens at the last mile and there's a lot of specialization required and and shoe leather that you need to wear out to make SM loans basically and underwrite them correctly.

It's possible that one of these platforms may eventually become a curator and try to source capital on chain, let's say, or use onchain infrastructure to try and do do whatever it is that they're doing. But like the the data aspect, like all all of what you're describing is kind of what we would consider sort of underwriting or last mile underwriting. Yeah. And you know and I don't I think I think the curation business segments very quickly into senior and junior and you start to see a lot more tunching around risk where you might have a curator that does a specialized vault or Tik Tok lending or theme receivables financing or whatever.

Uh but then they themselves again thinking back to the sort of balance sheet analogy they need to raise like they have a that balance sheet has a funding cost and that that funding cost needs to be met by people lending to it and I think curators will still have a role in as sort of wholesale lenders to these types of markets. So I I see it more likely of a system where it coexists rather than one that's strictly competitive because I don't think it's likely that CLA will create a morpho v2 market for you know buy now pay later loans and expect why not though there's no reason why they should and expect a lot of people so and expect a lot of people to lend to it directly by its own

right where they may be more success uccessful is in raising wholesale capital on chain from different sources stable coins or uh you know more for curators and then do that allocation themselves while paying the funding cost. I also uh I feel like maybe there's a bit of a challenge with like DeFi, you know, it's been so uh sort of locked in on like variable rates that you know, Morpho V2 is planning to have fixed rate loans, but the liability side of it is still basically you're funding everything with demand deposits. And I think in some respects uh that might cause some challenges once you get into

let's say like riskier assets like you know a Tik Tok creator uh like term loan secured by their show or you know by their profile like um there's probably going to be you know some amount of defaults for those sort of products and if you have demand lot like you know demand deposits funding everything then it's always going to be like a race for the exits as soon as somebody defaults. Um So I almost think that uh I don't know maybe there's an aspect where like uh there will be some future curation protocol that is set up in a way that's that's more suitable for that.

Yeah, I I I kind of agree. If I look at like the the fund structures around these like lending vehicles for things like the YouTube loans like I guess like things Ali set up um it's like they're they kind of have a very weird structure. They're a little more like private credit, right? Where they like allow some amount of redemptions on some periodic timeout, but generally speaking, they're they're you like not Yeah. demand withdrawals at all. Um, on the other hand, a kind of interesting question is like how do you do syndication and you can argue that morpho is a very opinionated decisionmaking process on how you should

syndicate loans where the curator is acting as the like there's a bunch of you know pools of loans of SPVS of loans and I'm which are you know all sort of relatively homogeneous at least to some extent and we have this capital we want to allocate it And you by depositing the vault are buying the syndicate, right? You're buying the the those partitions of loans. And there is a question of how to make a syndication protocol that allows for mostly demand withdrawals but also allows the syndic also has some like immutable parameters that are like, you know, no syndicate can ever go over 10% into like non non-withdrawable, right?

Like stuff like that. So yeah, to Monet's point, I like you could imagine doing layering this on top in in a modular stack. You could imagine layering this on top as a set of rules on top of existing vaults. And I do think we are starting to see a lot more of these like bundle vault bundles. um you know like just like in normal finance like yes technically there are hundreds of millions of of rated assets but like realistically only like the top you know point 10 basis points are traded regularly you know in corporate debt and also in in other loans. So if you think about the fact that like yes there's going to be a ton of vaults but like

only a small number are going to be have this like active demand and withdrawal you may actually find that there is demand for these kind of like longtail um things. I think the first things to get right though are like a figuring out how to do these fixed rate types of things to like service that market because that market doesn't want to pay variable interest rates historically and then b figuring out the oracle issues for this type of stuff. I think that Tik Tok and YouTube influencer one is actually a much better alone than many RWAs from an oracle stand risk standpoint because you could imagine in a world where ZKTLS is sufficiently cheap. Um, all of these people who are doing ZKTLS lending like Ernify and stuff like that, they're they're kind of

examples of this where ZKTLS is like I go to the website and I generate a ZK proof that like, hey, this many views of the app were found or like, hey, this is how much my earnings were from Google and it I make a ZK proof that shows that like I logged in using my credentials and like this is the actual number. You can actually imagine that type of oracle working for a a web two company type of thing. So I don't think it's impossible for for but there are a bunch of moving parts like a figuring out syndication b figuring out the oracle stuff c figuring out how to make it so that like you can have depositors can like look like a homogeneous Bob in a vault but some of them maybe are not. So so so I think

like there's a lot of problems to solve but it's not impossible. So it makes sense. I hope I hope Run will listen to this episode because to the syndication point, I think his answer would be subdows would be the stars in the sky ecosystem because they're regulated exactly as you described like no in the Atlas rules, you know, the encyclopedia of things that a that a star can and can't do. And that's like one approach which is really interesting to see how this will evolve as more stars launch in the near future. And then on the rest of it, I it's funny because the there there are two points that that that that you uh Monet I think raised. One is the demand the so all of the deposits are

essentially on site and the other is that some of these loans yes like the in particular in like things like private credit in in in offchain SPVS the redemption windows are are gated to specific days or specific times which makes the duration impossible to match. There's an interesting parallel with uh V1 borrow markets that are stuck in a liquidity trap for for what it's worth. So like as long as the utilization is at 100, the vault depositors can't withdraw. And that that's similar element although resolves much quicker typically than than than let's say 90-day window. Um to me that sounds basically like a fractional reserve stable coin where what will be really interesting to to see is how the user

behavior evolves over time for vaults. So you may have for example a blue chip vault where the effective where even if the site the the deposits are on site the effective duration is much longer than maker has this exact problem or sorry sky sky has this exact problem because in theory the entire balance sheet of USDS can withdraw within 12 seconds. So the asset liability matching is an extremely complicated problem to solve if you want to rule yourself by the purely onchain constraints. But if you take a more behavioral view of how users interact with the vault and the and you measure the gross and net outflows, you can start to create a track record and a tape of let's say let's call it effective duration of the

deposit side which can then allow you to take more educated guesses as to what exposure you can take on the asset side. And I think in in our view that's how you will ultimately reconcile the ill liquidity premium of assets that offer you know additional risk or additional premiums due to liquidity or or discoverability or what have you. Um because also the the the even the Tik Tok loans will have some some amount of duration in terms of just in terms of revenue collection like the time between a view and an advertising payout even if it's the space of a few days still creates some some duration from the perspective of a DeFi user. I think understanding the asset liability management and looking at things more like a stable coin is is kind of the

model that we're we're we're going to be taking for that we take already for the sort of V1 bolts and that will will migrate into the V2 world. Um but yeah, personally will be fascinating to see the the option space of loan opportunities that will become available and how different creators are going to start underwriting them and who those are who is going to be ultimately the originator of these YouTube loans, SM loans, what have you in the future like it would be nice if YouTube say paid in stable coins so that you wouldn't so that you minimize the frictions and create and start to actually originate loans on chain, right? so that anme that finances a receivable receives stable coins and

uses those stable coins to run their working capital. Like this is kind of the ideal end state that we'd like to land in. Yeah. I mean, I don't want to derail us too much and talk about YouTube and Tik Tok loans too far and I know we got to wind down here in a second, but yeah, one other and Monet maybe we can talk about this bit in the wrap-up, but one it's actually an old business model um is of subsidizing the product with a financing division. Um, so like I'm thinking about cars actually or something like that. Like uh most of the time those are sold almost at cost and the way that you actually make money is by financing the car and that financing division is owned by a dealer or something like that. And you know if you take what morpho and risk curation is to

its extreme right you can see I mean you don't have to squint that hard at it to actually see something relatively similar there. Like and that that actually brings me to at least where I would love to end here is like where does okay I kind of see there are originators right there's originators there is our curators um and there are there's like the protocols themselves like the morphos of the world um and maybe Monae I can actually start with you um with this question but how do you guys see the evolution of the risk curation model is it at a comm this is the question everyone loves to ask in crypto is it at the commoditized part of the stack You know, I think that the uh the flagship vaults um at least uh for steakhouse and gauntlet on Morpho, I

don't think there are any fees being charged on the USDC vault. So, you know, I'm curious, do you view it ultimately as a commoditized space? Does the value lie with the originator or the protocol? Is there a way that we can monetize on the risk curator side of things that isn't super obvious yet? How do we see that value cool component? Well, I'll keep my my bit just really short and sweet here, but I think you see a lot of curators uh not monetizing much now, you know, running very low or sometimes even zero fees because they think that there's long-term value in acquiring deposits, sort of gaining brand value and then also just acquiring users. Um, so yeah, I think that's in a way like whoever can

get closest to the user and own like some sort of user relationship, whether that's, you know, with your depositors or maybe with your your borrowers. Um, that's sort of like where the intangible value of this curation business uh sits. So maybe maybe protocols will have that. maybe curators. I think there's actually a you know a good argument that curators are um gonna sit closest to the user and and end up capturing a lot of the value from this. I would tend to agree like the relationship with the user is ultimately where most of the value and brand is accured.

Um, I don't think it necessarily means that the curator won't ever get commoditized or like the or that the protocol will eventually get commoditized. There's an element of there's always going to be an element of differentiation possible, right? There's never going to be one single oligopy of like just like today, you don't have one bank doing everything. There's always going to be an element of brand. There's always going to be an element of differentiation. There's always going to be the ability to sequence risk and capture new segments of users. And you'll have curators that cater to those. And arguably in traditional finance, you have that distinction as

well where if you if you abstract away the regulatory aspect of it, you know, the investor in a fund or in a debt fund for example is one particular type of user segment and they may have a bank account as well. they may concentrate most of their wealth in fund management. Whereas a a retail bank deposit user even if they ultimately end up doing very similar things on the asset side rely on the bank for different reasons and they may not even necessarily be related to the size of the capital. Um so I don't think it's necessary the one advantage in in traditional finance is that a lot of the interoperability is abstracted. So when someone makes or issues a loan to a degree, they don't

when somebody takes out a loan or mortgages a house or whatever, they don't necessarily care who the originator is or what rails it's on or what, you know, bank deposit stable coin type they receive. Ultimately, they just care about getting best rate and and financing their home. in in DeFi you'll you know one interesting thing is going to be to see how fixed rate loans interact between each other because you know like I said earlier in the same way that we don't have singleness of money amongst stable coins even if we had if we if you had a a bitcoin fixed rate loan on morphon a bitcoin fixed rate loan on something else they're not immediately transferable and so there is immediately a a secondary market need for to to sort of arbitrage the price

and I think that'll ultimately to keep it always exciting and curious to hear what Don thinks. Turn close out here. Yeah, I actually think, you know, and maybe this is the the bias from my background, but like I I kind of think of the institutions that I view as being the kind of correct view of curator in my opinion is like more kind of like multistrategy hedge funds than like a bank in some ways where you know if I look at the history of multi strategy hedge funds over the last 30 years a lot of them have had lifetimes much longer than banks. They haven't had to do M&A.

They haven't had to get kind of their banking regulator slapping them on the risk. A lot of them just shed some pods, gain some pods, and they're they're able to to to kind of grow. But the way they think about risk management is very different in that they generally think about like what is the overall portfolio sharp? How concentrated am I in any strategy? How do I think about allocation across all the strategies? And how do I kind of ensure that you know hey when I get withdrawal requests that you know like that's my demand deposit type of thing where my LPs start trying to to redeem uh that I can exit and like some of these strategies are sufficiently liquid some of these strategies maybe are not liquid so I have to like borrow against them and I have financing costs like all

of this type of stuff I feel like to me is more like curators and that's why I actually view the curation model as more about aggregation than about like a single model you know we spent a lot of time talking about lending today but I think as the set of products on chain grows whether it's uh whether it's the perpetuals products whether it's like hey maybe one day RWA like stock type things actually get traded a lot you know I'm not going to I'm not going to say that I believe that or don't I'm just like hey if that market exists they're obviously going to need lending u and a lot of the staking market I think you kind of want to take the bet that in aggregate you will be able to find relatively good riskadjusted yield

at like a a sufficiently high sharp that that is that is your model versus trying to like um I think bet on a single thing and I think lending curation is the bedrock of everything else um in the same way that like you know if you're if you're a multi strategy hedge fund your structural advantage often time versus a competitor tor is what ways you have to borrow and ways you have to get financing to open new strategies or open new trades, right? Um and and that cost of capital and and in general I think like you have to be you know in the case of onchain lending like you have to be close to that to even to

get that right. It's like a little different than I went to play golf with a banker or whatever like however you you or like oh we just happen to have a big enough balance sheet that like this bank is desperate to lend to us. It's a very different thing. You actually really do have to be much more in the weeds and understand that. And so yeah, I just view the aggregation effect of curation as kind of the important piece. And I think it's it's a little bit like wholesalers in marketplaces who aggregate demand and supply and help match them. And often times they're not necessarily bigger than any single marketplace, but they're they can oftenimes if they aggregate across marketplaces be, you know, a large sustainable business. So I I guess I

take the view of like the institutions of sort, you know, Millennium has lasted longer than like every bank I see in the US other than like Chase, right? All the other ones have had crazy M&A. they've had to divest stuff. They've had huge fines. Like I think to me those are like better institutions at at kind of these kind of credit capital management. But that that's the way we view it is like lending is always the core. But as this ecosystem grows, as stable coins grow, like inevitably there's going to be tons of different types of opportunities that will be created. And not to not to belver the the the balance sheet point, but it it ultimately all does come down to what the funding cost is to raise the balance sheet and how effectively you a

curator how effectively a curator can manage those assets on chain to generate the necessary return on assets to be profitable and to pay that funding cost and that kind of the the world of onchain asset management has very strong elements of what Dr. was describing. I agree of like aggregation and flexibility also in terms of yeah at the moment it's easy to say flexibility just because from the infrastructure perspective it's easy to write smart contracts and everything is kind of interoperable today everything's still today at least in DeFi everything is still kind of circular to a degree but the really interesting components will emerge when you know these curation businesses really do start looking more like onchain asset management with

elements of immutability and and all of the things that make crypto an actual value proposition. All right, guys. Uh this was a really fun one. Thank you both uh so much for coming on here. Um Adrien Tun, if folks want to find out more about what you're doing at Steakhouse Gauntlet, just get more information. What's the best way to do that? Twitter, the website, and of course the vaults. Uh yeah, I think Twitter websites on chain. Beautiful. All right, guys. This was a ton of fun. Thank you so much for coming on. See you. Thanks.

All right. Good app. That was really fun to get Adrian and Tun uh competitors uh in a sense. I guess all three of you are somewhat competitive, I guess. Yeah. Yeah. It's it's like a cooperition, I guess. Like we're I think we're all uh you know, good friends and have mutual respect, but they're in a lot of respects probably our two biggest uh two of our biggest competitors. a few others as well. But yeah, but yeah, great to have him on. It was a lot of fun. Maybe we could start at the end there around where you see like how competitive you see curation getting and uh sort of do you see it as a commoditized business and I think your opinion is that it is relatively

commoditized if I had to maybe put words in your mouth or reiterate what I heard you say at the end of that episode there. you know within that context then I would love to understand how you think about brand within the because for any product that's commoditized the one lever that you have for the most part outside of some sort of specialized distribution is brand and um yeah we didn't really talk about that so much but it would have been interesting to get Adrian's take on this because I know smoke or steakhouse has different names for some of their vaults there's the steakhouse USDC and then they have smokehouse for some of the higher yielding stuff so yeah talk to Tell me about how you kind of think about the brand lever here in the uh land of curation. Yeah. No, I think I think you

clocked uh correctly sort of my views like uh the actual curation part of being a vault curator uh where you're identifying strategies or pools that you want to lend to or like sort of identifying like the riskreward of stuff. Um, I don't think there's much of like a long-term like moat or defensibility there because um, you're executing all these strategies on train on chain. So, there's nothing to prevent someone from just doing like a trend following strategy coming in after you and undercutting you on on lower fees. And I actually think you know so far on on morpho v1 we've uh seen a lot of that where someone will

develop a interesting like a new pool like lend against pendle or something like that and immediately you know everyone else is kind of just following them in and compressing the margins there. So, um, the actual curation part of being a curator, I don't think is is particularly defensible, but um, when you look at, you know, let's take Gauntlet for example. Um, they actually, I would say they're pretty well known. Um, that's a it's a very strong brand. I think Steakhouse as well. Um, and people, you know, they I see people go to Norpho to invest their capital, but I almost think in a way like um they're trusting in Gauntlet or Steakhouse or,

you know, BA, I think we're a bit less wellknown, but we're getting there. Um, you know, they're kind of trusting in the curator more than the protocol. they know the protocol is like this base layer and actually the differentiation of like you know who's taking what risk and who you're operationally like uh depending on to do a good job as curator. So um I think the brand value is a pretty big factor there. Um, and I als I mean I think there's also a huge there is a huge like BD and distribution aspect to it like um you know you see these you know tens of millions of dollars deposited into someone's vault.

Um generally it's not just some whale who like showed up on the Morpho front end and said oh that looks good like there's a whole sales cycle to uh close for these really big users. So, it's there's a lot of interesting stuff, you know, happening kind of below the surface with this. Yeah. You know, okay. So, I'm going to maybe push an analogy here and stretch it. Uh, but I want to get your take on whether or not you think it makes sense, but okay. If you were to look at, you know, because you mentioned at the end of the um at the end there, maybe it was Adrian, that yeah, it doesn't really matter, you know, who gives you the your mortgage.

is just the best rate. But you still do need, you wouldn't go to some random uh, you know, shopkeeper on your street and get a mortgage. You go to like Wells Fargo or Bank of America or someone like that. So brand and trust do matter a lot within finance. And I think there are strong I don't know if they're network effects, but they're definitely a moat. Um and when I look at okay let's say that the architecture of protocols like Morpho drastically take the decrease either funding cost by making depositors like anyone can just deposit globally or it's shared backend so you know you take some costs out of a bank or something like that but you still do need that where do I go to get this right like

where do I go to get a loan or whatever it is. So weirdly, I could see the curators if they continue to scale, right? Like this is a very weird far out future, but it could be lenders with a big brand, you know, even a big bank, even a bank that is kind of built on top of Morpho. I'm curious maybe reframing it from from that standpoint if you saw like a curator you know offer different types of financial products built on top of morpho like rails maybe brand actually does become a mo and you can create a really big business here and then you can become an originator right um yourself but yeah I think there's you know uh like there's not too many different curation opportunities right now but uh there's also like a kind of

interesting opportunity for like cross-selling I think when you have you know a developed brand as a curator and then you're also offering products across different opportunities. So like you know steakhouse they have uh like some ortho vaults um but they also work on restaking products you know you can go stake with symbiotic or something like this and uh it's curated by steakhouse so maybe if I've had a good experience with working with them on morpho then I'll be happy to you know as I'm looking for restaking I see oh steakhouse I recognize them I'll put my money in with them. Um, so I I yeah I think you know uh maybe if it's just

Morpho it's like uh it's a little bit tougher but as you as the whole space gets like kind of more fragmented with just more opportunities um it's possible that curators will be kind of like the focal point for people having brand recognition for these different products. Okay, I want to end maybe with talking about I feel like we talked a little bit about V2 and what fixed rate borrow lend does on the creator side of things, but I would also love to get a sense of how BD is now being split in between, you know, the protocol itself like Morpho who I know has a you know an active BD team, but also the curators do some amount of that uh as well is my understanding. So yeah, I kind of want

to uh you know, maybe putting back your BA Labs uh cap on here is uh is a lot of that outbound going to I don't know the fintexs like the Coinbases of the world to be like, hey, you know, deposit, you know, use our vault, right, on your back end? Um or is that still mostly Morpho? And how do you see the balance evolving over time? Yeah, I'd say it's a it's a bit of a mix. Like I think Morpho is uh in a lot of these curation protocols like uh the protocols they understand that they need their curators to like have some wins to like bootstrap a ecosystem of curators and make the whole protocol system work.

So I think Morpho and like some of these other protocols that are offering curation they do try and open doors. Um, it's challenging because they also want to be neutral between all of their, you know, they want to love all of their, uh, curators equally. Um, so, you know, I think, uh, you'll see cases where they'll they'll just try and like put you in touch with someone who's come in as like an inbound for them or that they've developed a relationship. Um, but then it's really just kind of, you know, it's a warm handoff and it's up to you of uh how far you can run with this discussion whether you're going to be the the curator that someone chooses to work with. Um, so yeah, I would say it's it's definitely a mix. Um, I expect that Morpho is going to probably do less and

less of this or you know, the protocols are going to do less of that over time. it'll be um more on the part of the curator because the curator kind of has more economic direct upside and like sort of incentive alignment um for for landing these deals. You know, they're the ones who are taking the the performance fees uh when they get more aum. um they're the ones who have to have the difficult conversation when they, you know, take a haircut on the vault and, you know, all these clients they brought in lost a little bit of money. Um, which hasn't really happened very much yet, but it, you know, it's bound to happen at some point. So, I I expect it'll be a lot more curator

driven. Um, and, you know, the infrastructure will be kind of neutral. And do you expect on what time I know we were talking about very futuristic stuff here like Tik Tok or YouTube or car you know car dealerships or whatever but okay here's an earlier iteration which is happening now is you have spark as a curator on morpho right that's kind of an in that's kind of a um you know instance of a corporate not necessarily an active like a curator but someone who's a protocol and the lender themselves actually leveraging one of these vaults so that you know, you can maybe look into the future and if you were to say, okay, this ends with, you know, banks or very differentiated types of originators, you know, leveraging

Morpho, maybe SparkCC is kind of the first iteration of that. That makes sense. Um, so yeah, I'm curious if I had to poke at you and make you guess like what is the evolution of this market look like? Is it more, you know, BA Labs or steakhouse type competitors or do you start to see u maybe protocols within crypto kind of leverage um you know the unique infrastructure here to create their own vaults or even you know corporates outside of crypto. Yeah. I mean so like Spark is maybe the biggest example of like a protocol or sort of quasi. It's like an onchain company you could look at it that way.

um who has run strategies on Morpho. They have the biggest vaults currently on both mainet and base. Um and BA we like in a way we're kind of like below the surface doing curation activities because we're um like the sort of risk partner for Spark. Um but there's been other examples as well. So u on base there's a few other protocols. There's uh Moonwell who we work with. There's Seamless who is another lending protocol that I think Gauntlet is kind of like doing this white label product for. Um there's Gauntlets had this compound blue partnership where you know Compound is kind of like white labeling a brand onto Morpho Vaults on Polygon. Um so it's you

know I think it's a it's a popular um it's it's a popular trend. you do see that more and more and I guess maybe like what I sort of drill that down to is like um kind of almost tying back into what Adrian was saying is like we're making balance sheets and the balance sheet starts with sourcing you know kind of the best possible cost base you can have on the liability side you know getting deposits um and I think protocols are very well suited to that because uh for some of these protocols they can run incentives.

So like uh you know compound uh Moonwell they're they're running like incentives to get in deposits which as an individual uh curator you know steakhouse alone going in they they can't offer that because they don't have a token. Um or with the case of Spark, you know, they kind of have this locked in extremely favorable cost base from Sky, which is kind of like this uh happy accident of history where there's just a lot of people who forgot their money in uh in die. um basically um so yeah I think it's you know in a way that's kind of the edge is like how can you get the best possible lowest cost of capital and commercial banks they have extremely

cheap they have like all these people earning nothing in checking accounts so I think that's a super favorable um source of funding people like PayPal who like when your money is sitting in your PayPal account I'm I'm pretty sure you're not earning anything or Venmo or you know these fintexs So, um I think yeah, if if you could partner with them, uh or if you know maybe these people just start coming and and operating vaults themselves, they kind of have a huge edge. You know, I I feel like, you know, sort of a side tangent here, uh is I feel like stable coins are leading me to a line of thinking that checking accounts are a scam actually in a sense.

Like there are there are products that already exist in crypto where you can just keep your USDC in a morpho vault earning yield up until the exact moment that you want to spend it which is you know the banking equivalent would be by default whenever you deposit into a bank it you just put your market in your money in a money market fund you earn yield and then whenever you want to spend it it gets converted. That's what it should be. It is. It's kind of crazy that banks just have the struct the structural advantage of this really low funding for deposits and it feels like crypto should fix this. Um, yeah. I mean, I think it's it's like a it's it it didn't start as a scam, but it's like

a a model that has outlived its usefulness where you there needed to be some way to kind of subsidize the cost of banks managing all these paper checks and having, you know, wheelbarrows of cash that they're sort of taking around the country and rebalancing all the place. Um, now like you know the actual sort of nitty-gritty of of processing a payment with crypto rails is like minuscule small um and it's small enough that the end user you know is basically able to just manage it with gas fees.

They don't need to have some sort of cross subsidy thing happening. So, um, yeah, it's it's like, uh, it's interesting to see that that's just like sort of an artifact of history where it it made sense maybe 20 plus years ago, but, um, it doesn't make a sense anymore. And, uh, we're just slowly slowly waking everyone up like, hey, you have money at Wells Fargo earning 0.01%. Wake up. Yeah, totally. I mean actually someone was pointing out to me I can't remember who that even in this is different in between the US banking system versus other banks in Europe I guess are much better at giving yields uh but also I guess there was a difference post 2008 the great financial crisis it was much more of the norm that

the yield in your checking account would track um fed funds but now that's just changed for whatever reason uh so um okay last question for you here I would love to get your thoughts a little bit if I ask you to wax poetic on you know what do you think that morpho v2 uh you know fixed rate borrow lend what does that do to the role of the curator we got into it a little bit with adrian and tun but yeah I would love to get your expanded thoughts on how you see that evolving different types of strategies that you might be thinking through give us some of that inside baseball special sauce here uh but yeah I'd love to get your thoughts yeah I think it's it's going to get a lot more challenging um to be a curator you

As it stands now, you have to identify kind of what pools you feel safely about and have some sort of like light infrastructure to be rebalancing between the pools that you whitelist um to, you know, kind of optimize the riskreward, not be overconentrated. Uh but it's it's in a in a sense it's fairly simple. Um when you look at V2 where you're you know basically bidding on these fixed rate loans you have to manage not only you know what sort of collaterals and LTVs that you're comfortable with as an example which is same as V1 but you also need to be thinking about how much duration you want to take. So like you know you're you have to like sort of

have an active view of how many of your users are going to want their money back in the next month, two months, three months. um you have to have an active view on where you think rates are going because if you take a fixed rate loan for three months now at let's say 6% which you know is more than you can get on a or variable rate lending right now but then suddenly next month we're in a bull market um you're giving up a lot of profit that if you had just kept it in variable you'd actually have done better. So uh it's you know it's you you just have to take a much more active view on markets and I think the execution also is uh more challenging because you know you're not just like adding or subtracting funds from a pool

where everyone is kind of getting paid at the same rate. So, it doesn't really matter very much if you're, you know, you're first or you're a little bit later to get into the pool or get out, except for rare cases where a pool is like, you know, potentially going to default. Um, but there's, you know, there's going to be all of this kind of PvP aspect of like if you can get your, you know, your offer into the order book, you know, one tick higher than everyone else, then you get all of the deals and the other person gets nothing. um you know it kind of it looks more like being a you know a bond dealer which is um you know it's a complicated and you know more challenging uh enterprise I'd say than than how

curation looks in V1. So um you know I think uh yeah just everyone's going to have to tool up. It's you know it's going to be a whole new ball game. It's the big leagues, baby. Um all right, Monet, this is a fun one, my man. Um, I will, uh, yeah, I'm sure I'll see you in person this week for Permissionless. And, uh, yeah. Yeah, man. It'll be a blast. And, uh, I I don't know. We, Jason and I said that depending on how many tickets Zerox Research sold, we might do the dunk tank. So, I think we might be getting dunked this week, actually.

We'll see. Well, I'll be there. Yeah. I'm not very good at pitching, though, so you're in luck. Pray for me. Pray for me. Um, all right, partner. I will uh I'll see you next week. Yep. See you.