Curated Credit: How Maple and Morpho Approach DeFi Lending | Sid Powell & Merlin Egalite

What's up, guys? Just wanted to let you know about the Block Works Digital Asset Summit, which is back in New York March 24th through 26. We'll have top speakers from leading asset managers, financial institutions, DeFi protocols, crypto companies, and policy makers all under one roof. Think of Black Rock, Coinbase, Robin Hood, and more. Follow the link in today's show notes to buy tickets, and make sure to use code 0x200 for $200 off at checkout. See you there.

Nothing said on Zerox research is a recommendation to buy or sell securities or tokens. This podcast is for informational purposes only and any views expressed by anyone on the show are solely our opinions, not financial advice. Pokio Ryan and our guests may hold positions in the company's funds or projects discussed.

What's up guys? Welcome back to another episode of Zerx Research. Today I'm joined by Anastasia from Blockworks Advisory and Sid from Maple and Merlin from Morpho. Thanks for joining guys. I hope you're doing well.

Before I let Anastasia give a short introduction about what we're going to talk about, I wanted to ask Sid and Berlin. I wanted to say thank you guys for coming on the show and give you a chance to potentially introduce. I know a lot of people are very familiar obviously both Morpho and Maple, but in case you guys wanted to give a short introduction of yourselves, I think that would be helpful. Or I guess the team and the protocol would be better. Whoever wants to go first, shall I pick?

I can go first. Maple is an onchain asset manager with a focus on lending, yield, and credit. We've been around since 2020. We run around 4.5 billion AUM. We have about two billion of loans and deposits on the platform.

And our primary business is institutional lending where we take Bitcoin, ETH, Soul, XRP, large cap crypto as collateral. And we're a little bit unique among onchain lenders in that it's a bit of a hybrid model. So the loans all occur on chain. They're funded in stable coins from our smart contracts.

However, we can take native BTC or any other asset that we choose either directly or we have triparty arrangements with custodians. So, the collateral it's a little bit different in that the loans are managed through the smart contracts. The collateral is more of a more of an offchain or kind of CFI arrangement, but that's a little bit about us.

Great. Cool. Thanks thanks for the invite. So, I'm Berlin, co-founder of Morpho. Morpho we're building the the credit layer for the word I would say. So we started maybe a bit after in 2021. We released the the first version of the protocol now there's 10 billions of total deposits and and the goal is to allow anyone to build yield and borrow product on top of a neutral permissionless financial infrastructure.

And so you can create the the product that you want. You can own everything from top to bottom. Select the exposure for your users and customers. It's pretty flexible and neutral. So you have full ownership on what you're building on top of Morpho.

So we're building that we are going to release a new version in the next month following month. That would be our intent-based lending protocol fixed rate based so that people can create offers or like make offers or take offers and then those focus offers can be very flexible in the sense that you can set require any type of KYC or you can like say that you're willing to lend against any type of codals or basket of collateral.

So you can be be very expressive around that and I we think that it will be it will unlock a lot of new capabilities in the Linux space. So that's that's it for for Morpho.

Perfect. Anastasia do you want to give a short rundown of the work that the work that you guys put out at block works advisory and then we can start our questioning?

Yeah sure. So first of all thank you very much guys for your comments. They were super helpful. We very much appreciate it. And about the paper. So what we did we showed how decentralized credit is changing and what we argue is that the defy lending has become two layer system where base protocols provide like rails or infrastructure but the underwriting decisions are made by WS and third party creators who decide on collateral caps launch urations etc.

So we in this paper we looked at the data and it's a sample that is covering one year from October 2024 to November 2025. Yes. And we measure three things. The first capital utilization in the field. A second how concentrating laden liquidity is across chains. And the third one, how interconnected creators are through overlapping positions.

So generally speaking, we're looking at the surface of the D5 credit at at this version, but we will follow up in with more details on contation and network between creators.

What we find? We find that the risk migrated upwards that the small sets of curators intermediate a larger share of walt's TVL we also find a co- movement in the downside risk so the tail risk and we also find that the fee capture vise a lot even though collateral is similar generally allocation assets and collateral set is very narrow

What we conclude? We conclude first of all that defy credit is the future and because it's the future we find that it's important to have a transparency framework and therefore we offer to disclose things like portfolio concentration liquidity coverage metrics rehypotheation this word maps so the connection maps and attestation for offchain exposure is if if you have an offchain exposure and we think that this transparency framework will help users and also institutions to compare WS and risk or or curators risk on a more consistent basis.

Perfect. I honestly unless you have a very like direct question, I actually kind of wanted to ask something to sit specifically about the because like Maple's been around for quite some time. How has this shift been for you guys? Like because I feel like you guys have always been kind of at the curator side as opposed to offering like broad infrastructure, right?

Yeah. A little bit of both. We we kind of transition. So if you look at the first iteration of Maple which went live in 2021, we were more trying to position ourselves as infrastructure. So we had the model of pool delegates which is what you would now call a curator and they would run a lending pool. Select borrowers approve them and the loans at the time were more working capital loans that were under collateralized.

So you had overall I would say more ability for one of the curators or delegates as we called them at the time to you know to kind of influence the the protocol. Then post 2022 in 2023 when we started doing over collateralized lending we subsumed that role of delegate inhouse and now I would say we're kind of like a vertically integrated we're almost like a vertically integrated curator.

We have the protocol infrastructure layer which permits us to bring in capital issue LP tokens i.e. the the you know the tokenized positions in the lending in the lending pools and then we make the underwriting decisions the commercial decisions on pricing and and set the the collateral limits and and what is eligible collateral. So now it's much more of a kind of vertically like vertically integrated product where probably the most important role we do at the moment on on Maple is that kind of curator set of responsibilities as opposed to to being infrastructure. So we've we've kind of you know traversed from one side to the other.

I guess I want to do the same thing for you Merlin. You guys also kind of started off I guess always at the curator level kind of an amorphos shifted. How did you guys notice that you had to play on the on this other land?

I actually I I think we didn't really started as a creator but more like build on top of existing landing pools that were doing the the creation side of things. So we were not creating ourself but more enherating the decisions of another creator that was a DAO and everything was decided through governance process.

So on our side it it was mostly that we realized that building overlanding protocols on top of the kind of like lending pool design was a big challenge. We had a lot of issues. It was great for bootstrapping liquidity because we could tap into the underlying protocol. But it proved to be very challenging because at any time like the the underlying protocol could upgrade pose etc and change the logic over time and that proved quite challenging.

So we decided that hey let's create that infra infrastructure that will allow not only one like creator to operate but open that to a broader market. Because our core thinking about this is if you upon open like you delegate the risk management and you open that to the competition then you'll be able to offer better terms to end users because there will be like healthy competition on top of the protocol and and and so yeah as I said better better terms for users but to c to create that space for creator to compete.

You need to create really a very natural flexible infrastructure on top of it which it's easy to select the collaterals the the the create the markets that you want. And and that without having a third party that can like change the rules of the game. So that's that's why we changed it we changed the the course of trajectory and we built our own protocol that was was kind of solving the issue that we faced when when building the very first version of the protocol.

Perfect. Sia, do you want to?

So I I started I will start with the questions that I prepared which are related to the structure of the paper and the first one would be if I am an investor or a depositor, what risk did I used to take by picking a protocol that I now take by picking a curator?

I I can start this one. So previously I think depositors would many like like took the the protocol governance risk that was like trusting the governance and the DAO to manage the risk do the upgrades and change the parameters over time etc. But that process was kind of implicit in the sense that it was not really open. It was not really clear for the user that all those things were happening in the background.

But if you have to go to a protocol where now you need to select a vault and and the risk exposure then all those parameters and those changes are now becomes explicit and I think this is the power of blockchain is to have the capacity to fetch on chain what's happening and and display that to all whole stakeholders and and participants on the on on the given market and is and to surface all those attributes.

So you need to have to look at the collateral, the utilization, the allocation of a specific vault etc. and and with the the V framework and creator framework, you can select a creator and the V that really match your risk profile. And everything is transparent and even one creator can can offer different Yel product I would say. So from like very low risk to high risk and the hand customers can look at what's the the market exposure of that specific vault and determine like oh okay this one is is a low risk and and high risk.

Of course that demands to be quite tax he savvy and kind of like financial quite financial person to understand that kind of risk and this is where well surfacing the data educating people but also maybe some some from time to time like doing if you if you're a distributor and you create a yield product you can be opinionated and say hey we want only a low lowrisk product and then you abstract all the complexity and and also it could be like a risk rating. So all these can things all those things can be helpful to and really will really depend on that who is your end customers.

Okay. Sit you also want to share your point of view?

Yeah I'll just build a little bit on what Marilyn said. But I I think for me the simplest and most the simplest and most direct change is has been the collateral well two two risks collateral risk which is a subset of credit risk and then liquidity risk. So previously all of these parameters were set at a protocol level. pricing was determined to allow for liquidity at a protocol level and then the collateral was what was eligible collateral was determined by like a Dow governance vote.

Now it's much more decentraliz like it's a further layer decentralized again because you have the collateral that's eligible at a protocol level but then the collateral that a vault curator can determine is eligible. So now you're relying on them to to on board the new collateral whether it's a new type of stable coin or some other DeFi token to to to originate loans against it. And then also they're going to kind of manage the liquidity as to you know as as to how much effectively how much liquidity do they try and have in reserve for people to make withdrawals versus trying to maximize the yield through high utilization.

So I think those are probably the two most salient things and of them definitely the collateral side is probably the most important thing. So that's why I think it's very important that you get a feel for what type of collateral underwriting process your vault curator uses.

Okay, makes sense. A lot of W growths seem to come from the distribution partners like wallets, custodial integrations and so on so forth. But the fee split is usually private so it's not disclosed. I mean it is possible to kind of approximate how much TVL is coming via partner and yeah looking at the deposits and so on so forth.

But I'm curious if you would support to to have like a label on on Walts to show that this is a direct deposit and this is the W let's say incentivized by the partner sourced even if the exact terms would wouldn't be disclosed to make it kind of comparable to to see the fee split and fee that creators are earning.

Are you talking more about the split of TVL that's coming in or more about the economic split of who gets what share fees?

I think it is kind of connected to to each other. So I I would say both.

They're definitely connected. I would say as you know as as a lending protocol, we wouldn't want to show the fee split for certain TVL partners because what happens is you end up in a race to the bottom. You get a you get a partner that might be a big Asian exchange and then they say, you know, they're getting X amount and then they look across at somebody else and say, well, they're getting that deal. We want that deal. So, it ends up being, you know, it's kind of it just ends up a race to the bottom on on fees.

And we've we've never found that. We haven't we haven't found that to be to be a good practice. We prefer to keep it private so that you can you know you can try and get the best deal for your users and ultimately we want to pay out the highest yield to to users. But I think it's interesting to have a TVL split so you can see where it's coming in from. Like is it coming in from this wallet or this exchange? I think it's good in terms of creating a bit of FOMO so that you can get other partnerships.

That being said, I can see why certain, you know, certain parties might not want that. They might not want to be disclosed if they're a wallet or an exchange because then people might try and infer risks about them. Let's say, you know, let let's say an exchange was you know had an earn program partnership with Morpho or with with Maple and they might not want the size of that disclosed because then if something's happening in DeFi, their users might might panic. So you kind of want to respect the you want to sort of respect the perspective of of the other party and be conscious there.

But I like it in terms of generating FOMO to help get more partnerships. I mean one of our strategic focuses this year is we want to be establishing partnerships with web two fintexs with more of the centralized exchange earn programs and having a giant you know dashboard that says that we've got a billion in from XYZ wallet or exchange is a fantastic way to do that.

Yeah, I think I agree with with Sid. And anyway, at the end of the day, if you're using for for for instance, if you deposit into MOVA, like the performance fee, the management fee, everything is on chain. So, you really know what they take and what what the yield that you you get. you don't necessarily like know how much percentage of the vault is coming from that specific distributor.

Some like you can know I don't know for instance we have like a a very large integration with coinbase they are using a smart wallet so you can see on chain that this is the coinbase smart wallet and you can attribute and we have like a for instance like a public dashboard where you can see oh there's two billions of BTC as collateral borrowing one billion of us this is coming coming from Coinbase for instance so you can do that attribution but a lot integration that happening on mofo are doing like through mofo v and and the creator is taking the fee and they have like a private deal with that specific distributor we don't even know what's the the fee speed the fee sharing in that and we and I mean it's private information even if we knew I it's not an information that we should share I think be just more like a respecting privacy of those creators.

I think that makes sense. I have I think a bit more on the product side. As like a very attractive thing that you guys can offer is this earn product which I think Morph was kind of doing through the base app. Is that how much of that focus is like Sid you mentioned you guys want to maybe get integrated with web two fintex. How much of that is like offering your own front end and trying to have like users exclusively use you versus like just trying to get integrated into maybe in Europe it's Revolute or I don't know what they have in the US to be honest with you but like those types of or like I guess the base app etc.

Yeah, our position on on our front end is more like a facade or some sort of like explorer and and mostly toward pro proumers versus something that fintex or even partners would would integrate directly. It's more like oh you can see like all the vaults what you can do the markets you have like metrics and you you know what's what's like the liquidity network that is being built on mofo but then we encourage partners to build their own interface because then because like mofo is immutable etc you can own the product like the vault the the the market but also the interface and you can really this way provide the best experience for your users.

So that's that's mostly the goal of our interface. Yeah, I think on the interface question it's the different you know it's the difference between partner like channel partnerships let's say and and just going direct to to consumer or customer. So our front end is how we go direct to customer. So anyone can go to you know to maple.inance and go and click on you know click on the earn page and start using it.

However, the the value for partnering with whether it's a Revolute or a Lemon or Robin Hood or one of these other or Stripe, one of these other fintech earn programs or centralized exchange earn programs is that you get embedded in their front end. That's where you want the distribution. They have way more daily users than we do and we want to be visible. You know, it's like getting access to a a prime piece of real estate. We want to be on there in their app where their users can can easily use our product through their interface. So I think you know if you get one of those partnerships you don't care about directing people to your site. You want you actually want your product distributed in you know in their real estate.

Makes sense. I want to go a bit away from interface and ask something related to the APY and the question is is creative competition a race to a better risk adjusted underwriting or a race to a most markable APY? What do you think?

I would say short term is APY that attracts attention and deposits. Over the long term, I think it's risk adjusted. So, we have like had some events that like I don't know collateral that blows up and then that blows like some vaults that could be more risky than others. But because the risk is isolated, it only affect the high-risisk fault that were exposed to that specific market while the lowrisk one like when are not exposed and and are like completely fine.

And as these kind of rules that are higher risk blows up then you have like a flight of capital towards lower risk strategies and and and so over time I would think that risk risk adjusted vault will kind of win. Obviously there will always be customers that will search for higher yield and risk is kind of a spectrum. There's like a it's continuous and you need like a solution for anyone and not like one a one size fits all model where you have only one risk profile. We don't think it's the scalable way especially as the there's like more assets and and type of asset that enters on chain like I I don't know like the the Hwood asset tokenized equities etc that comes on chain.

Then then we need like a more scalable way to to handle all those kind of assets. Yeah, I I tend to agree with Marilyn that I think in the short term it's probably more APY, but if you were to look at it over a cycle, it's probably going to be risk adjusted. But the the thing is you only know what risk adjusted was in hindsight, right? It's it's an expost it's yeah, it's an expost assessment of like how much risk did I actually take when I looked at how many times things blew up over a multi-year period.

Howard Marx has this great concept where you know you have your CAPM type model for anyone who did corporate finance you have risk and and return it goes up but what he says is on this on this dotted line where you are getting more return as you take on more risk there's actually a bunch of bell curves and your risk or or your true return might be anywhere on a bell curve but the line is going up on average but here you might take a lot of risk or or rather get a lot of return without blowing up But sim simultaneously you might also actually realize zero or a negative return if you take a lot of risk and blow up. But it's the uncertainty that is the the defining point of his graph.

And I think it's also largely influenced by the macro environment. So when yields were very low in 2021 and early 2022, people were pushing out the risk curve without knowing it to try and get higher APYs. and it encouraged you know it encouraged a lot of like looser credit practices and since you know since rates were tightened credit practices became tighter and more conservative but I think there's always a push to try and get the higher APY and you won't know that it was not risk adjusted until you see a blow up whether that's with you know like a a stable coin that doesn't have the right backing or whether it's with you know you took some collateral that actually turned out to not be that liquid or you know or fraud or something else.

Yeah, makes sense. I have maybe two follow-up questions with respect to that. In relation to corporate finance where everyone is climbing efficient frontier to get the u high return versus higher risk. Would you then say that we should be sharing risk adjusted APYs when it is possible or like approximated?

And then the next question, the second question would be where does APY most hide the tail risk? Is it then correlated collateral or is it then liquidation liquidity? What what do you think?

So I think on the on the risk adjusted returns what I would like to see more of is we kind of need ratings or you know I know S&P and Moody's have both a little bit of crypto in terms of stable coins or or cryptoback loans. you have you know risk assessors whether they be chaos gauntlet or others but if you could have at least an accepted benchmark of you know investment grade high yield and and and uh junk or or deeply subinvestment grade then you can at least look at you know for a given level of yield where am I you know where are third parties pegging the product in the ratings because otherwise the problem is you have somebody getting six and somebody getting 10 and they're viewing them as equivalent levels of risk because there there's no proper benchmark to use.

So I think if we if we get that then it'll be easier to to determine what a risk adjusted yield is. I strongly agree on this point absolutely we will come to more maybe in detail a little bit in in the following.

The second question was where the APY hides the tail risk is it correlated collateral or liquidation liquidity?

Ooh, I think it varies. Like I would say probably middle middle of last year like Q3 I would have said it it was probably more like correlated collateral. Now I would probably say it's like liquidation risk. I think as you've seen more looping there's you know you have to consider the liquidity of potential liquidations and so that shifts it more to the the liquidity on liquidations.

But before you know when there were a lot of different new stable coins coming out that proposed very high yields they often had correlations because one might have taken deposits and put it in another one. And I think that was often not, you know, you couldn't tell what the backing was or whether there was, you know, crosscontamination there.

I I would add two things also is collateral reaation. Like for a lot of actors it's not super clear that they're actually like the collateral is being lended out and that has drastic consequences on like how much you can like or performant can be the liquidations. And that should be taken into account into like lending pools where all the all the the funds and assets are comingled and and and so you deposit then you're exposed to all the the assets and that's not super clear for a lot of actors that if you go to that kind of design then you get exposure to all the assets.

So if one blows up then you can be exposed even though you think that you're just lending USDC to a safe safe safe markets and the problem is that the yield that you get is actually not representative of the risk that you're taking because that it's all like like everything is pulled together. So you might have boys that are like boring and it's like at a very like it's super risky I guess a very risky collateral but and some other boys that will borrow against like I would say blue cheap asset and those rate will you'll have the same rate and I think this is kind of like very detrimental to for people to understand like the the risk they're taking.

Yes, I I agree with you. Based on this, if if a user sees two WS with the same collateral family but different net APY, how should the depositor decomposed into yield, risk and fees? And what should be standardized for this decomposition or is it is it possible to standardize it?

Well, I I don't know if we can standardize, but I I think we can like put that into different buckets. So, of course, there's the native yield that you're earning through like the the interest that being paid by borrowers and that goes to the lender. So, that would be the native yield. then on top of that you can have incentives and and from that native yield you can subtract the performance and management fee or whatever is used and and those fees I think should be made clearer to the to the end customers the end depositors so that when you deposit you know that you're earning this part on incentive this part for net APY but like subtracted from from the performance fee

I and and if you disagree as a user about like oh this performance fee is too high there's there should be like healthy competition and and maybe you can go to another creator that charge lower fees but then you need also to understand like those creators are are char are charging a fee because they need to like pay the risk managers like the infrastructure to run vaults to reallocate liquidity. So there's no free lunch. I would say but to standardize that I I I I don't think this is we should do it. It's more about like a like making sure that it's fully transparent for the end user.

Yeah, I largely agree with Marilyn's points. The only other things I would add, you know, I think if you you know, if you see that there is an incentive payment on top, it's kind of like a free kick. So, you know, there there's arguably not a lot of difference in the underlying risk if assuming it's the same collateral underneath it. So, I'd say look, you know, the the difference being attributable to an incentive payment is kind of a positive thing if you're a user. However, if you look at it and one of those pools is a billion dollars and the other is 20 million bucks, then I would be mindful. Are you actually being compensated because of liquidity risk? You're not going to be able to get your deposit out when you want because it's like you and one other whale. And if the whale goes, you're you're left you're left with a position that can't be redeemed. So I think consider consider the liquidity, consider incentives, but if it's otherwise, if it's the same collateral, then that's kind of like for like.

I I have a followup about collateral. Specifically white labeled stable coin providers. Does this add complexity on your end? if the back end is the same thing specifically. I'm thinking of M0 and maybe Aphina because they're pushing for it. I'm wondering if you guys I think Morpho on your end like Merlin on your end is probably more like the curator that's taking it but I'm curious how you guys think of that type of process because the back collateral is the same thing but you you don't really know how the SPVS are set up and yeah no it would be more the curator that perform the due diligence.

So I I M0 there's not so many asset that are using M0 on on on MOU. I think there's MUSD from from MetaMask. But I I have no idea. I mean there's some many things happening on MU I'm not even aware of now. So I'm I'm not sure I would be able to answer to be honest that that question correctly.

If I were looking, look, if I were looking at white labelled stable coins, I'd consider the underlying backing like let's say it's M0 or Athena and then we know that you know it's it's USB underneath it like it's effectively T bills. I think you can take a degree of comfort there. Then you just want to consider the liquidity. Can I lend out the white labeled stable coin? Because for us, we have to take it in and then lend it out. And so the yield we're going to be able to get determines on is determined by the universe of borrowers who will take it and do stuff with it. So that's where it comes in. You're probably going to end up with different pricing because if it's a white labeled stable coin that you know is only native to a single protocol, then it's harder for us to to find borrowers. So, you're going to get a lower price on it than a more widely used stable coin like a USDE or USDC or USDT or or USDS.

Yes. I I agree as well. I have another question related to the curators network and it's the following. So in in March 2020 money markets funds faced a dash for cash events where everyone wanted to exit to the safest liquidity at the same time. So assume we will replay this event but in DeFi. What do you think is the analog for dash for cash in DeFi and where do you think the cascade actually will spread first?

So you know you've seen a couple of instances over the last few years of of something like this. there was USDC in 2023 with the the community banking issues and it's either going to be a liquidity issue or an asset quality issue where it's a liquidity issue there is simply a mismatch between the time to redeem and and the the clearing price on the secondary market let's say u however an asset quality issue is when you think that there's going to be you know an impairment of the backing somehow so in in that case there was funds held at one of the community banks. I think in that case it was SVB and or silvergate and so there was a concern that there was going to be an impairment to you know to the asset quality there. So you see it from time to time. Other instances like October 10, we saw a bit of an issue on Binance. And so there was a a temporary disjoint there.

I think so we can we can accept that it happens and that it's likely to continue to happen in future because these are effectively you know token in a way they're kind of like tokenized money market or tokenized delta mutual positions. So it can happen. The second order impact is either going to be felt on centralized exchanges. So in October 10, you had per liquidations because it was like USD was used as as collateral for those positions. And then in DeFi, you had like lending markets where you had a lot of people went to borrow or to go short the stable coin because you know because they were worried about it deping. So they wanted to wanted to be short rather than long. So I think it'll naturally cascade to either centralized exchanges where there's a large exposure of it or it can come to lending protocols like us or or like Morpho you know because primarily we we we lend in stable coins I think so you want you want to have kind of a healthy level of diversification there but yeah but you I don't think you can limit the risk entirely. This is why, you know, I think this is why you really want to go with battle tested stable coins as you're building your lending infrastructure. You don't want to use a $50 million brand new stable coin. You want something that's that's been battle tested, has been through a few cycles. But I think it's just something that you can't you can't engineer out of the system.

Yeah. And like and this is why I think it's super important to to have protocol that are designed to contain that risk or at least like make sure that you're really exposed to the risk that that that you think you're taking. So if you're like taking risk on data set then it's explicit and you have selected the vault that is like allocating to those markets and you know that it's one of those collaterals or like the the asset you depositing blows up then you're affected by that. But if you want as sid said that if you only want to use stable coin like blue chip stable coins and like lowrisk then you're sure that maybe there will be like some crazy move in terms of liquidity but the there won't be default because of one asset that is blowing up and maybe endangering the whole Danny pool. and and this is like the requirement for I think like isolating risk at the the smart contract infrastructure layer which is like one thing that we're trying to to highlight and hammer people in the space and and because we think it's like deeply important on MOO and one of the key thing if you want to to create residient and and and and Yeah, resilient yet product for end users.

I'll just add one thing. I think we're not trying to solve or engineer out of the stable coin risk problem because customers who come to us, if they have USDC, they go into the syrup USDC pool. If they have USDT, they go into syrup USDT pool. But they are already exposed to this the inherent stable coin risk. We're giving them a yield built on top of that, but they have already self- selected into that kind of risk. So, it's not our job to take that away.

Totally. I have two follow-up questions on that. The first one with respect to diversification. So in general the assets under management set is pretty much similar what