# What's Next for Vaults? | Steakhouse Co-Founder, Adrian Cachinero

*Author: David Hoffman*
*Published: Jul 30, 2026*
*Source: https://www.bankless.com/es/podcast/whats-next-for-vaults-steakhouse-co-founder-adrian-cachinero*

---

## TRANSCRIPT
David Hoffman:
[0:04] Bankless Nation, I'm here with Adrian Cacinero. He is the co-founder of StakeHouse. StakeHouse is a curator in the vault space, and we're going to learn a lot about vaults today. Adrian, welcome to the show.

Adrian Cachinero:
[0:14] Hey, nice to be here.

David Hoffman:
[0:15] Adrian, how bullish are you on vaults as a sector?

Adrian Cachinero:
[0:21] Obviously very bullish, but I think it would be cool to try and unpack a little bit what vaults means. It's a very it's a very jargon loaded term it can mean a lot of things um we're obviously very bullish on the types of vaults that we're building and there's lots of and we're bullish on types of that nobody is building yet as a idea the idea that you can intermediate a value exchange without an intermediary or without relinquishing custody or without or doing it in a way that's purely ministerial or without discretion. I think this is quite new for finance and to the degree that a vote enables this is bullish, I think, because it represents a more efficient market. And we believe one that offers better investor protections as well.

David Hoffman:
[1:16] Yeah, maybe I think it's worth going into just the word vault, the vault word. Maybe the best parallel to talk about what a vault is, is to like talk about the token. It's like, what's a token? Well, a token can be anything. Like it's really just a symbol, a token of whatever we want to ascribe to it. And we can ascribe to it meaning, value, mechanisms via code. And so is a token a security? Well, that depends. You know, is a token a utility? Well, that depends. And like vault, a vault is very similarly as broad and undefined as the word token. And maybe that as like a starting place helps illustrate just the immense breadth of category of what a vault can do. Maybe you can take that and run with it. And also maybe give us a little bit of a history lesson about like the vault sector, how vaults came to be and how it's grown into what it is today.

Adrian Cachinero:
[2:13] Yeah, I think that's definitely the right way to think about it. Very much like along the lines of the definition of like what's a token? Oh, it can be so many things. similarly, what is the vault? The vault can be so many things.

Adrian Cachinero:
[2:28] I think DeFi summer with the food farms, the food farm era and the yams and all this, that was really where you saw the first vaults with the Y-Earn. They had

Adrian Cachinero:
[2:40] The first smart contract architecture that would allow people using vaults to delegate basically the selection of, let's say, a strategy to a smart contract. But they only really took off I think they really took off with Morpho Morpho invented the, well, maybe didn't invent, but Morpho pioneered the concept of an isolated risk market for borrow and lend. This had obviously been tried before in different contexts, but Morpho really, let's say, formalized it, hit it with an appropriate level of go-to-market, you know, found a good fit. The difficulty with an isolated borrow-lend market is that you lose a lot of the network effects from a pooled lending model. So something like Aave, where all the liquidity is pooled, has the benefit of rehypothecation of collateral. So... You can deposit the collateral and borrow against it at the same time. With Morpho, you lose that. And so they use the idea of a vault as a way of aggregating liquidity around these isolated markets to recapture it and rebundle some of that network effect again.

David Hoffman:
[3:48] Would you say that Morpho kind of produced the basic atomic units of markets that when composed together via vaults, which is what a vault does, turns it into something bigger? But it was important for the foundation of the vault industry to have these isolated lending markets to become very specific and opinionated and kind of like fit the demands of the market. And with these atomic units going down just to the bare basement of like the borrowing and lending markets, we can start to build up more stable structures because we have all these atomic units of these isolated risk markets. Would you say that's a fair description?

Adrian Cachinero:
[4:30] Yeah, for sure. Sure. It's a philosophy that we very much agree with. I think it speaks to like the experiments with DeFi that work the best tend to take this direction. Smart contracts are very sensitive infrastructure to build anything on. They can go like when they go wrong, they go very wrong. So you should aim to reduce the risk surface as much as possible. You have governance heavy, heavily controlled like DAO governed whatever models from the early stages of DeFi and the Morpho approach is to try and boil it down back to what's the simplest possible it's like a smart contract as a spreadsheet like I just want the spreadsheet formula to calculate how much interest is accrued on one side and owed to the other and that's it that makes the surface much smaller and you know It loses all of the network effects, but it also makes it recomposable and you can re-bundle it later on.

David Hoffman:
[5:28] And don't the network effects kind of just come back around at a higher level in this stack later? So an individual Morpho market doesn't really have too much network effects, but the network effects emerge once the recomposition of these atomic units come together. And that's in the vault space.

Adrian Cachinero:
[5:46] Correct. Yeah, for sure. Sure. So like an individual borrow position in Morpho has very limited network effect. Network effect meaning the value of the network increases faster the more people use the same product. An individual borrow position on a Morpho blue isolated borrow land market has very little of that. When it's aggregated in a vault, it gains value faster the more people start to use the vault to aggregate that liquidity. Because more loans can be issued and the more loans are issued, the more efficient the market becomes and so forth.

David Hoffman:
[6:23] Okay, so Morpho produces the isolated markets. Vaults emerge because they start to become composed together. But where's the typical place that vaults are produced or accessed? Or, you know, I can go to Morpho.com and look at the markets, but like, where are the vaults?

Adrian Cachinero:
[6:43] Where are the vaults? Where are the vaults? Yeah. They're in lots of places. So the Morpho Vault is one flavor of Vault, right? Like, I think a simple way of thinking of a Vault is a token that aggregates liquidity for a goal. And that's where I think you can start to, you start to really suss out the differences between different types of Vault. We like the Morpho Vault a lot because it's very simple. It's very minimal and very governance light. Obviously, you can have Vaults that are more complex. Something like a VEDA vault is a more complex execution layer that has a vault as a liquidity aggregator, but that allows more actions or discretion at the operator level, let's say. Vaults.fyi is a good place to shop around. Beefy makes some very tasty vaults. It's a word like a token. It depends what you're looking for.

David Hoffman:
[7:39] So after Morpho produces a bunch of isolated lending markets and then also VEDA, there's a few other vault producers. What would you name the category of Bfee and VEDA and Morpho? What are these things?

Adrian Cachinero:
[7:54] Yeah, like vault infrastructure.

David Hoffman:
[7:56] Vault infrastructure. Okay, so first at the bottom layer is vault infrastructure, which produces many, many, many isolated lending markets or in the case of like VEDA, maybe a little bit more opinionated. But that's still nonetheless the bottom of this stack. Next comes the curators. Now, that's where a lot of opinions can be expressed. So maybe we can start, I just want to kind of get a lay of this land of this next highest level in the tech stack. Maybe we can start talking about kind of the very cypherpunky, totally non-custodial, very safe end of what can emerge at this vault curator stack. And let's talk about what kind of the products those are and where those end up and then we can go a little bit more into the more opinionated, more managerial end of that same spectrum. But kind of give us the lay of the land of this next higher order stack.

Adrian Cachinero:
[8:54] Yeah, I think the, obviously the simplest, if a user wanted to retain the maximum degree of custody and control, nothing is preventing them from borrowing and lending into a Morpho market themselves.

Adrian Cachinero:
[9:12] They lose all of... There are some trade-offs to this. So, for example... So an individual borrow land market within Morpho targets 90% of the liquidity to be borrowed at any given time. And that level of liquidity sort of oscillates depending on the supply and demand, but it's generally on equilibrium at around 90, which means that 10% is withdrawal at any time. So if your position happens to be bigger than, if your individual position happens to be bigger than that one slice of 10%, you may find that you are unable to withdraw the full amount because part of it is borrowed if you allocate to a single one. When you use a vault, you can access multiple markets and so you benefit from the liquidity across many different markets. So there is some shared liquidity benefit to aggregating liquidity higher up. If you wanted to do it yourself you could run your own vault where only you are the depositor it's a it offers a very simple abstraction it makes it very easy to interface with you regain some of the network effects around you know having multiple borrow land markets and some additional you know safeguards on the liquidity what you may find that this is maybe the most cypherpunk way maybe of doing it when you have like individual self-serve i'm just going to make my own vault, select my own collateral, reallocate my own markets.

Adrian Cachinero:
[10:37] That's fine. The cool thing about Morpho is the barrier to entry is zero. Essentially, you just need to know how to interact with the smart contract and you can print your own vaults.

Adrian Cachinero:
[10:49] And where we operate is on the immediate next level beyond there. We wouldn't go as far as saying as we're intermediating a vault because we don't stand between the borrower and the lender. But we run backend infrastructure that keeps the vaults well balanced, let's say, or that makes the price discovery of the efficient borrow rates. Going. And that rebalancing activity has value, both to the borrowers and to the lenders, because they are able to meet each other better and quicker and at a lower spread. And then in StakeHouse's case, we mitigate the degree to which we can interfere with the bolt. The main vectors that you can interfere with a MorphoVolt would be on essentially, like the biggest one is onboarding new collateral markets. So if you, as a curator, propose adding a new collateral market, that's a vector for risk for the perspective of Vault user and is a governance, it's a governance lever that only the curator can operate. And yeah, so in Stakehouse, I don't want to make this a shilly.

David Hoffman:
[12:08] No, well, this is great because this is your domain. So this is perfect.

Adrian Cachinero:
[12:12] The Stakehouse view in Vault is that the reason you would use a Stakehouse Vault versus doing it yourself is because you don't want to bother with actually rebalancing price discovery, any of this stuff. It's not trivial. It is actually quite involved and complicated. But you still don't want to surrender, you know, intermediation, custody, risk. So we try to think of ways that will mitigate

Adrian Cachinero:
[12:36] Ourselves as a counterparty to the vault user. One of these ways is giving the user the ability to stop our decision-making through the veto mechanism. We build an Aragon DAO around the vault and we make the vault user participate, so to speak. So if they disagree with a collateral onboarding decision that we propose, they have the ability to stop it. Like these types of on-chain mechanisms would be ways for us to remove the discretion out of the equation. You can then... Take it further. There is certainly a space. At the end of the day, the vault is an abstraction for aggregated liquidity, not dissimilar to an ETF. An ETF is an abrogator of liquidity in traditional finance. You can imagine a world where some types of vaults are replacing or competing with ETFs or take custody, for example. Some vault infrastructure platforms run regulated funds. They could run it on something like Makina or Veda. And they just do it because it's simpler from an accounting point of view to account for on-chain positions. But the spectrum is obviously very broad.

Adrian Cachinero:
[13:48] Yeah. And so we it's worth being very specific about the type of vault that you are describing. In most cases, what all the vaults have in common is facilitated accounting in the ideal transparent exposure to the underlying positions and in the long run, hopefully better investor protections and market efficiency as a whole.

David Hoffman:
[14:15] When you're describing the role of steakhouse in the vault space, it seems very close, very overlapping, very congruous to both Aave and Yearn, because Aave has to do some sort of risk management with like onboarding collateral, like whether it's in an isolated market or, you know, a cross margin collateral. That that's a governance decision. And then also your role is to optimize for yield so that depositors maximize their yield, lenders pay the minimum or receive the most and borrowers pay the minimum. And so it seems like it's some sort of intersection between both Yearn and Aave. How accurate do you think that is? And is there any fix, any changes you would add to that comparison?

Adrian Cachinero:
[15:03] No, I think it's a fair comparison, but I would, not to speak for Aave, I would say in Aave's case and in our case, because we are operating borrow land markets or what we would call repo markets, it's not necessarily about optimizing for yield, it's more about optimizing for efficiency of the underlying borrow land markets, if that makes sense. But yes, both of them are governed you could say that Aave Labs is the curator of Aave if you want they have a governance process for onboarding new collateral they retain a certain amount of control for certain decisions and areas they have mitigants to themselves as counterparties, that's part of Aave's value proposition you can become an Aave token holder and vote on the decisions as well if you want You know, it's just as it's a slightly different approach to essentially the same thing done in very similar ways.

David Hoffman:
[16:06] Is it fair to call Aave just like one very big vault that's been verticalized and integrated?

Adrian Cachinero:
[16:13] They're the biggest curator for sure. We like to say we are the biggest curator. In reality, Aave is the biggest curator.

David Hoffman:
[16:18] Aave is the biggest curator of the Aave vault. And what you've done is you have integrated governance with the depositors. And so the depositors into Steakhouse curated vaults also have that veto power that you were talking about. And so they have some sort of governance. Whereas if I deposit ETH or USCC into Aave, I don't have any governance. Only the Aave token holders do. And so that's a distinction. But beyond that, there is governance. It's done by some people. There is risk management. there is collateral onboarding or collateral discrimination and then there's like yield optimization as well. So it's all kind of the same activity but in a different form factor.

Adrian Cachinero:
[17:00] Yeah, and... This to me illustrates how flexible the vault construct is. You can build it in an almost infinite number of ways. So it's very difficult to say vaults and then refer to everything very accurately. Aave has a huge amount of complexity around things like the umbrella, the security module, the fact that they have the two tokens for the governance. It's a good point. We're building on the Morpho stack. We have a much simpler surface that has trade-offs as well, of course. Yeah, it's the one that we're familiar with and happy with, but users are able to decide and sort of allocate accordingly. But at the core, they are both functionally liquidity aggregating vaults and the underlying primitive is a borrow-lend market or activity.

David Hoffman:
[17:53] When Steakhouse does... curation, managerial decisions about how its products look and behave. Do you have like a philosophy as to like risk and collateral acceptance? And what are those parameters that need to be managed by Steakhouse? And like, how do you decide to draw a line between, you know, what's what is acceptable and unacceptable? Do you have like a philosophy to guide your guys' decision making?

Adrian Cachinero:
[18:25] Yeah, very much. So it won't be dissimilar to, you know, the collateral onboarding process that AVE undergoes and will be familiar to anybody like this. I think the simplest way to think about it is more in terms of a framework. So rather than it's not like there is an active ongoing managerial effort with the collateral once it's onboarded, the key inflection point is the collateral onboarding proposal. And then the market kind of settles around that. There are some collateral markets that we onboard that nobody uses. And it's just like, well, that was a flop. You know, Aave has a long tail of collateral markets that nobody winds up using. There are some that a lot of people wind up using. And the risk orients around, like the framework that we try to build around is functionally in a repo market, borrowers are seeking safety. Sorry, borrowers are seeking risk, lenders are seeking safety. We aim to minimize the risk to the principal as much as possible. There are many levers that will mitigate the risk to the lender.

Adrian Cachinero:
[19:28] The chief among them is the haircut. So the fact that a borrower can't take more than 100% loan-to-value against their position, and often they have to take a significant haircut to the collateral value, this is an enhancement that protects the lender because it gives buffer and time for the loan to be repaid. In our risk management, so we aim to do principal protection through collateral selection and then let the market kind of balance where it finds the most fit. And the selection of the collateral for our proposals orients largely around sort of credit risk in the underlying asset and then market and platform risk for the specific loan so you can have a

Adrian Cachinero:
[20:11] An asset with very little credit risk, but the loan position could be very risky. For example, wrapped Bitcoin, we would consider it to be very low credit risk. It's held in BitGo. It's very liquid, what have you. If you onboarded it with a 98% loan to value, this may be an insufficient haircut to prevent bad debt from accruing to the lender. And so is therefore too risky. So you would find, you would try to strike a balance in that's kind of the framework i don't think this would be alien to like the other risk teams you know they would evaluate collateral in a very similar way um what they might have is more of a community engagement with the token holder base around uh you know when do we onboard what do we onboard under what parameters and so forth with stickhouse we propose it's more binary we propose it and people are happy or they are not and if they're not they leave and in the extreme instance they veto.

David Hoffman:
[21:11] What parameters are other vault curators accepting that Steakhouse is not accepting in terms of like the risk profile? And that's just, and the reason why they're accepting it is just because they are further on down the risk spectrum. So like what's the, in the risk spectrum, what's the next most proximate thing that you guys have decided is not appropriate for Steakhouse vaults that the next most marginal risk taker says, we actually do want that in our balls.

Adrian Cachinero:
[21:37] It's... Not to, like, there are a lot of collateral markets that are very successful and very,

Adrian Cachinero:
[21:48] And have sufficient credit enhancements and are probably fine, but we just don't have the time to get to them or we don't have the understanding to, like, go through it. So the fact that Steakhouse is not onboarded something doesn't mean that it's bad. It's just, it could just mean that we, whatever, haven't had time. Sure. In most instances, we are tending to prefer things that are regulated or supervised in some manner or have very strong on-chain guarantees. So we look for, we audit the governance structure. So of something like, like we went through this process with Etherfy recently, actually, where we had previously only, we had previously bucketed them as a high yield collateral. They engaged us very actively around, hey, we want to be considered prime. What is it that you would consider prime? What we would consider prime is your life needs to be made a lot more difficult. Your managerial discretion over the token needs to be significantly removed so that users have either sufficient time to exit or sufficient warning to understand what the changes that you're proposing are going to be made and that would be uh yeah and that would involve setting very long time looks putting you know governance guards around key roles like minter roles and things like this

Adrian Cachinero:
[23:13] And they went along with it and you know so that's actually one example of a of an upgrade rather than like a

Adrian Cachinero:
[23:22] A a denial let's say um

David Hoffman:
[23:27] I guess what I'm getting at, I did the podcast that kind of went around the vault space with Andrew from Heard. And we talked about this very basic incentive for vault businesses to add the next most marginal vault. And so like all vaults have some amount of risk to them. Some are very not risky. Some are further down the risk spectrum. And vault curators, vault businesses want to increase their fees that they take as a curator. And so they have like some share of the yield that's going through their vault. They take a cut of that. And so they're trying to increase a TVL in their vault and they're trying to increase the yield of their vault because that's how they get paid more. But then there's some sort of limit, of course, about the risks that they're willing to take. But the incentive is always there to try and find ways to increase TVL and increase yield and therefore implicitly increase risk. And so I'm always trying to figure out how do people buffer against this scope creep of risk inside their business? And so I'm sure you think about this a lot. I'm sure that's actually the whole entire alchemy of Steakhouse. But how do you guys buffer yourself against risk scope creep?

Adrian Cachinero:
[24:49] Yeah this is we we've thought about this a lot in the past i think the i'm hopeful that we have turned a corner so i understand the point we you know i definitely think that we've seen it play out in the past few years i also i also hold the view that defy has the ability to learn and self heal very quickly so whenever there are risk events like stream finance and elixir and back in 10.10, for me, that was one of the first catalysts of the last major clean-out of risk-taking vaults. When you see events like this, you... My hope is that the average DeFi user looks at it and then recognizes actually maybe the highest APY is not the best, because maybe I should be looking into what the underlying exposures are. And so, yeah, I'm hopeful that we've put this behind us to a large degree. The incentive is definitely very much in the direction that you pointed out, assuming that users only sort for yield. So in a world where users have an inelastic risk tolerance, and they are willing to take on any amount of risk as long as the API is higher, then there is a disincentive because the curator will just onboard more and more random shit. And then they will eventually explode because this is what shit tends to.

Adrian Cachinero:
[26:15] What I'm hoping is that that basic that base assumption that I mentioned that the user only you know the user doesn't care about risk and only looks at the API I'm hopeful that this is changing and that people are a bit more discerning and a bit more thoughtful you know people like Andrew are certainly doing a huge amount because they're empowering users to make their own risk assessments on the contents of these vaults so hopefully this era is behind us

David Hoffman:
[26:41] I agree that DeFi has this naturally self-correcting mechanism, but I am hesitant to put that responsibility on the role of the users because there's always going to be like the, more novice user, the younger user, the less experienced user because, well, what are we trying to do in crypto? We're trying to onboard the world. And, you know, only a very small percentage of the population has had some of this experience that you're talking about and why there will be some sort of self-correcting mechanism. And I don't know if giving the responsibility of risk management, of kicking the tires of vaults, belongs to the user. It should belong somewhere else in the tech stack for that risk to be appropriately scaled and managed. Because this is ultimately what we're trying to do. We're not trying to... Take away from one end of the vault ecosystem to give it to another because we think that's effective risk management. We're trying to optimize all parts of the vault's spectrum so that yield is maximized. Safety is also maximized. Everything is kind of maximized rather than trying to like constrain one part to benefit another. Yeah. But I don't know if we're going to do that. I don't know if the way to maximally scale the vault industry comes with the users coming into the industry for the first time and kicking the tires of the vaults that they are being advertised. So how do we solve this problem?

Adrian Cachinero:
[28:07] No, look, that's a fair point. I'm not trying to disclaim responsibility and kick it to the user, but

Adrian Cachinero:
[28:14] When you describe that adverse incentive, that adverse incentive only exists if the users behave in a certain way.

Adrian Cachinero:
[28:21] The effect of going through a number of cycles is the curators that are more risk-taking hopefully die off and the ones that are left are the ones that have cumulatively cumulatively made better decisions we like to think that we will be in this you know that's kind of the the fire under our feet is to let's let's not fuck it up you know like let's keep making good decisions and keep going so in five years from now the space of curators that remains is the is like the one that has made the cumulative best choices for their users that's kind of how we think about it And yes, it is an interaction between the two. And the curator certainly has a very large responsibility in not making bad decisions. And those decisions largely orient around the degree to which you want to take collateral risk. For something like a repo market, we would argue you shouldn't be taking that much risk. Like it is actually quite simple. You don't need to be putting an exotic stream elixir stuff. Stuff. You just need to do a few things well. And yeah, again, I wouldn't characterize it as yield maximizing. It's market efficiency maximizing. Like the vault succeeds when it delivers a more efficient market for borrowers and lenders. And it does it through a vault construct that offers better investor protections. For us, these are the two key value propositions.

David Hoffman:
[29:50] Do you think that there is a principal agent problem, a gap between curators and depositors in terms of like the risk being taken? So like curators taking a performance fee of the yield. And so if the yield is six, seven, eight percent, curators take some chunk of that and that's their paycheck. But if that vault gets exploited, if it was improperly risk managed, the curators don't lose anything other than future fees, but they don't lose any principal, the depositors lose their principal. And so there's a different, there's a dislocation in exposure between depositors and curators. Is that a problem to be solved? Or how do we close the gap between the principal agent problem between those two parties?

Adrian Cachinero:
[30:34] Yeah, I would definitely agree there's some principal agent

Adrian Cachinero:
[30:38] Difference, like in the sense that the user is a principal and by using a vault, they are implicitly delegating to the curator, the agent, the responsibility of like running vault and selecting the collateral. And for less risky vaults, that difference is smaller. So, you know, a vault that only lends to BTC and ETH over collateralized lending is mostly about efficiency, is very, very little about a credit risk. Right. So there's no, yeah, there is some asymmetry, but it's very muted. For higher yield or, you know, higher risk, lower liquidity type repo markets, there's certainly more of an asymmetry. And yes, it's true. The curator and like, you know, Steakhouse as a business allocates entirely to our own vaults. And we've made a public commitment to withdraw lost. So if we did get hacked or if something did go wrong, we would want to take on the most, like the way these vaults work is unfortunately not perfect in the event of a default, let's say, because essentially the default will accrue to the, depending on how the vault is configured and how the bad depth is calculated, you can't have an instance where it accrues the most to the people who leave lost, which is suboptimal, right?

David Hoffman:
[32:04] Right, you run on the bank.

Adrian Cachinero:
[32:06] Yeah, you face the run on the bank, people exit whole, but the ones that are left behind face a bigger and bigger impairment the longer they stay in. And that's why we made this public commitment of like, look, our whole treasury is in our vaults. If something, if we fuck it up, then we will just take as much as we can.

David Hoffman:
[32:22] And it does close the gap, but because you guys are willingly doing that, it's not a structural fix. It is a fix, but it's not a structural fix. Correct.

Adrian Cachinero:
[32:31] Yes, I'm not satisfied by this because it's not cryptographically enforced, right? Like for us, the benefit of the vault is the cryptographic enforcement. It's a commitment. And look, it's not to say... You know, if a curator does incur losses on their depositors, it's not like they face no economic damage if they didn't hurt because their business prospect is significantly impaired. Right. And that's not nothing. It's just more abstract. It just means that their future business is much less likely to yield.

David Hoffman:
[33:06] It's like brand equity damage.

Adrian Cachinero:
[33:08] Yeah. This can be existent. This can go to zero, right? Like there are curators that have completely shut, that have gone from like leading the pack to completely shutting down. So you can say this is minor, but yes, you know, businesses do go like do switch off for making mistakes. So it's not, it's not nothing. It will give a little comfort to somebody who has actually lost money in the vault because it's not like they can take the, you know, lost brand equity and recover their principle. So I, I sympathize with this. I do think a cryptographic mechanism may be better.

Adrian Cachinero:
[33:44] Again, it will depend. Like, I think it does actually depend. Like, if you have a vault with complete discretion on behalf of the manager, no restrictions, they can add and remove whatever policies they want into the vault, there should probably be at a minimum some kind of cryptographic enforcement of a buffer or a surplus or insurance or supervision from a third party, something to that effect. For something like a prime vault lending against Bitcoin and ETH that just lends against Bitcoin and ETH. Is that really necessary? I would argue probably not. You know, it's not the most efficient market mechanism. And then there's a range of scopes in between. And so it's worth, you know, being very specific. But we prefer cryptographic mechanisms to social ones. Social ones have value, but they are imperfect coin compared to a cryptographic one. So I would rather be, have a way to like cert definitely get slashed and benefit the user rather than have to just promise it. But we are, there are many, you know, proposals out there that offer solutions around this.

David Hoffman:
[34:58] One of the reasons why talking about

David Hoffman:
[35:00] vaults is hard is it goes back to the open-endedness of the vault word. And there's like a complete open-ended spectrum of what a vault is. Like on one end of the spectrum is like you guys with just very high quality collateral, you know, the crypto native monies, Bitcoin and ETH. And then on the other end of the spectrum is like, well, we're actually allocating.

David Hoffman:
[35:21] Any depositors money in this vault to an EOA that can take it into their bank account and invest in stuff. Like complete managerial discretion, complete levels of trust, and vaults fit anywhere along that spectrum. And so I guess one question worth talking about is vaults are going to grow. This industry is going to get very big. Like institutions, Wall Street, TradFi are looking at vaults and seeing the value here. So we're expecting this industry to grow hundreds of billions of TVL into the short to medium term. And so I guess the question is like, on the spectrum of risk, where do we expect the growth to occur? Do we expect it to occur on the very safe side of the spectrum with over collateralized vaults with near perfect levels of risk? Do we expect it to be somewhere in the middle of risk taking or do we expect the vaults to be kind of like the trusted asset manager end of the spectrum? Or do we expect the growth to kind of happen equally across the spectrum? Do you have an opinion about where and what types of faults are going to grow over the next like five years?

Adrian Cachinero:
[36:31] This is not going to be super exciting, but we expect the more boring vaults to grow more and faster. You get more benefits to scale by using vault infrastructure. You have a weird situation where you have an on-chain regulated hedge fund that operates using a vault infrastructure is at the extreme end of the risk and is an example of what exists today. But we don't have a very efficient overnight repo market for U.S. treasuries on chain. So there is kind of a weird asymmetry where you get the little pockets of high risk first when you don't have the foundational layer in place yet. And that foundational layer, we are talking about trillions and trillions of dollars worth of activity that takes place on repo markets today. What we're the most excited about is making those markets as efficient as they are today we want to make those markets more efficient by bringing them on chain, making it more transparent and making them more accessible and composable with other parts of the finance industry worldwide. So that's kind of the direction that we're the most excited about. It tends more towards the more boring academic parts of finance that excite us the most, to be honest.

David Hoffman:
[37:49] So there are some fundamental questions about the vault industry and that has especially emerged with Hester Peirce's comments recently. And she kind of just illustrated a few areas of concern that she has with the vault space. Like one is like are curators asset managers? Are curators engaging in you know securities like activities? And like one of the questions she asked is like if you're a vault curator you should ask yourself if like these the behaviors that you're engaging in are like that of like an investment company or an asset manager. From your answer just now about how we expect the vault space to grow on the safer end of the spectrum, it kind of sounds like the things that Hester is talking about is not the bulk of the vault industry. It's more on the margins. It's more like the degen risk-taking margins of the

David Hoffman:
[38:42] vault space rather than like the body, the bulk of the vault space. Would you say that's accurate?

Adrian Cachinero:
[38:47] Well, it might be. But I do think like what Hester, I think Hester's memo was very thoughtful. You know we'll certainly take her up on the invitation and go and have a chat like we we definitely think about these issues a lot in the way that we configure the vaults and you know like what's the point of a vault like why would anyone use a vault versus something else or an etf it is a type of asset management sure we call it like internet native asset management it's similar but it's new and meaningful ways and so it's worth a second look like there are some things, like she said, it's not worth, you know, you shouldn't do what was like somersaults and fist bumps and like rolly-collies. Like you shouldn't contort yourself to evade an existing law. That seems quite commonsensical. And there are elements of like vaults and repo markets on chain and stuff that are genuinely new and feature mechanisms that protect the, you know, user or the lender in novel ways that can't really be captured in existing frameworks.

Adrian Cachinero:
[39:50] So even in those like low risk, so-called low risk, just because something is low risk doesn't mean that it Doesn't that mean that you disclaim managerial discretion over it or it can't be done in a way that's violating or, you know, it can be done in ways that don't violate security laws and it can be done in ways that do violate security laws if not done correctly. And so we do really want to be very thoughtful about how we configure it. Because our feeling is that if we do align with the way she is describing the vault space, it will be a better vault for users because it will have more protections, more distance between the curator and the vault user and so forth. And for us, that's kind of the point of a vault in the first place.

Adrian Cachinero:
[40:35] We already have mutual funds, hedge funds, ETFs. These already exist. We don't need to reinvent them on the blockchain for the sake of it. You would do a vault if you could do something in a meaningfully improved way for the user or the investor.

David Hoffman:
[40:53] Is there something that needs to get figured out in the vault space when it comes to engagement with the SEC and who has legal liability? Or is it something as simple as certain vault managers just need to register as hedge funds and or investment managers or whatever the technical term is, and they need to take on legal liability? Or is there like a new category that needs to account for the quirks of on-chain stuff? Like are the questions answered or do we need to like as a vault industry go sit down with has to burst at the sec and kind of hash this stuff out.

Adrian Cachinero:
[41:29] Yeah probably the latter like because vault is such a broad term um it like it almost certainly captures like there are almost certainly players that are doing activities that are already covered under securities and so those those already exist and so there are already frameworks that people can register for. We've been, like Stegas has been particularly, or tried to be thoughtful about this. We don't believe that's the case for us, but I mean, I would rather have a chat with the SEC first and confirm that. And if there are areas that we need to register, we will just register. That's fine.

Adrian Cachinero:
[42:06] And then for the most part, I think yeah, our view is that there are sufficient um, And interesting elements of the vault space that merit thoughtful consideration. Like legal liability is one aspect. Insurance is another. Disclosure regimes is another. Like these are important investor protections that feature in regulation today that have no parallel in the pure vault space. Do all the vaults need to meet the same standards? In our view, probably not. But there is almost certainly some level of like standardization and requirement that would benefit the industry as a whole without hampering the ability for somebody to just whip up a vault if they feel like it. Like to the example that we put on at the beginning, this one of the cool features and innovative features of this industry is the barrier to entry is very small. So if you as an individual user want to make your own vault like you just make your own vault you know and you can be the only user and that's fine and it would be a shame if the laws prevented this for example so we do believe in you know more competition and more openness as a way of facilitating innovation which hopefully you know the the regulator will recognize this is like a part of the the ecosystem as well yeah

David Hoffman:
[43:34] Yeah this has been my third episode doing vaults I did I recorded with Paul from Morpho yesterday and then Andrew from Heard a while ago which kind of kicked off this whole thing, I'm kind of circling around two main quest lines for vaults that have like, all of the interesting stuff in them. One is what you're talking about, which is all of the lower risk vaults that can scale massively in TVL because Apollo, BlackRock, they come in, it's pristine collateral, it's the repo markets. We basically rebuild the repo markets. Maybe that's a good way to describe it. It's like we're rebuilding repo markets on chain more efficiently with stable coins and pristine collateral like Bitcoin and ETH. And that has the opportunity to scale towards trillions of dollars on the longest of time horizons. And so I guess the core question there is like, in immense levels of scale.

David Hoffman:
[44:38] What is emergent out of that immense level of scale on the crypto side of things? How does that impact the economics of the on-chain economy? How does that impact ETH and Bitcoin? Just because of the immense scale of these stable coins and players that come in to rebuild the repo market. It's the true basement of finance. That's one quest line. The other quest line is on the other side of the risk spectrum, how do we establish.

David Hoffman:
[45:05] Legal liability and consumer protections so that that side of the spectrum can also scale immensely? And so like that part is yet to be figured out. And I'll take your point and I'll concede to your point that like the repo market end of the spectrum is far bigger. But the hedge fund and more risky and, more managerial side of the spectrum is also very interesting and like probably made me to the DGEN seeking more yield, seeking more upside, seeking more risk, more interesting to them. So the question for that quest line is like, how do we establish a capital stack with legal liability insurance and everything that you've said on it with the regulation involved to make that side of the risk spectrum also grow into immense scale as well? And so I think those are the two fundamental paths forward that the vault industry needs to take and will take both simultaneously. But I think that's just like kind of mapping out the future of vaults. I think those are kind of the two arenas.

Adrian Cachinero:
[46:08] I think so. Yeah, I think that's a reasonable way of looking at it. I mean, not to rehash the old

Adrian Cachinero:
[46:15] Talking point, but You know, the reason you would use blockchain is because you want to be able to rely on a crypto guarantee instead of having to rely on a social guarantee or a trust me guarantee or even a cord enforceable guarantee, which is just a flavor of social guarantee. Like the benefit of a network like Ethereum is the crypto guarantees that it offers are total and final and immediately enforced. And this comes with risks, but also benefits, because it means that you can build a regime where you can aggregate liquidity and strongly constrain the managerial discretion of the operator. If the vault is correctly configured in a hedge fund scenario, let's say, you can guarantee that the operator doesn't stray from their mandate, for example, using crypto guarantees or having policies that only a nav administrator can update or something. So we do view vaults as, for the same reasons that they can deliver benefits to scale in repo markets, they can also deliver meaningful improvements in things like investor protection. Obviously you have to do it yeah the trade-off of the crypto guarantee uh you know coin is that when you transfer stable coins into a black box like stream finance then it's gone and you've you've completely surrendered the managerial discretion of your stable coins to some to some guy

Adrian Cachinero:
[47:45] The flip side of that is that you can also severely constrain him and make it make it safer even than a hedge fund, right? I mean, the graveyard of TradFi is littered with dead hedge funds that strayed from their mandate or took more risk. And it's not like we're, you know, inventing risk from nothing. Right.

David Hoffman:
[48:06] Is what I'm hearing you say is like, as we grow the, safe side, the very scalable, safe repo market side of the vault spectrum, that kind of gives us the tools and the optionality to push into, the higher risk, more managerial side of the spectrum, but without having to need to leverage social contract and courts, because we can do it with building out the tools, the network effects of the safe side. We can take the power of the safe side and apply it further and further into the margins of the risky side while still being cypherpunk. Is that kind of what you're saying?

Adrian Cachinero:
[48:46] Yeah, for sure. And it's not to, it's not to suggest that you can disclaim liability or evade regulation to Hester's point about the somersaults and the backflips it means what i what we're trying to say is that you can actually enforce these regulations better and quicker and faster and in a more like secure way you can improve you can take that same framework for investor protection and use vault technology to enforce it to like actually enforce it and make it a stronger regime than just uh you know whatever disclosure regime for example yeah

David Hoffman:
[49:21] I see i see and that's built on top of a very robust scaled foundational side of the the vault market and kind of kind of move out from there and we'll always still kind of have like the legal liability side, it will just kind of get pushed further and further out into the margins as we kind of figure out the safer more cypherpunk more permissionless trustless side of things.

Adrian Cachinero:
[49:42] Like the trustlessness and permissionlessness are a way of improving market efficiency this is like this benefits everyone through lower search costs, lower frictions, lower spreads. And Volts allow you to find equilibriums of market efficiency faster by aggregating liquidity. And they also happen to do it in mechanisms that allow for a very high degree of investor protection. The legal liability doesn't magically disappear. But it can be codified or enforced in a clearer way with a vault versus waiting for two years for a whole court settlement, as an example.

David Hoffman:
[50:24] And potentially also marginalized, right?

Adrian Cachinero:
[50:27] Hopefully mitigated, yes. So in the ideal, you would mitigate the total amount of credit risk by constraining the operator as well, right? Yes, for sure. But in the event that you did need to enforce an action against an operator, a vault would allow you the tools to do that without having to go through an arbitration court. Right. Cool.

David Hoffman:
[50:48] One last conversation before I let you go, Adrian. The balance of power between distributors and curators. And so distributors are going to be like people like Robinhood, Coinbase, Kraken, like consumer front ends that aggregate a bunch of consumer deposits. They have the relationship with the consumer, which is a very lucrative, powerful thing to have. And then they can take a pretty fat cut, I think. And then there's the curators on the back end who kind of get marginalized in this position. Is there a tension here between the distributors and the curators in terms of the actual cut out of the retail product that gets earned? And if we do, does that kind of... Threaten the vault business model for the curators?

Adrian Cachinero:
[51:38] There's always a pressure. In every value chain, there's always a pressure to go in either direction. Every time there is an accumulation of excess profits, one party will have an incentive to go in one direction versus the other. You see this with Aave, the biggest curator. They have a pressure to make their own savings up instead of, depending on, you know, the Robinhoods and Coinbases of the world. They're like, no, we want to own the full customer end-to-end relationship and actually compete with Robinhood and Coinbase to some degree in...

Adrian Cachinero:
[52:12] It doesn't always have to be that way. You can have scenarios where specialization can deliver benefits to the overall product. We would view our role as a curator in that light as kind of specialists. It's a sufficiently narrow activity and it's sufficiently difficult or not trivial that it's often not worth the time to invest to capture the remaining margin for a platform like Coinbase. And for a smaller player, they might not have the skills or desire to push into that space anyway. All they want to do is just have an earned product for their users and focus on the customer experience. And so there's a nice balance where I think the curator, and at least in our experience with Robinhood and Coinbase and all the other partners that we integrate with, it's more constructive and collaborative. There may be a time where excess margin accumulation will lead. I mean, Stakehouse is never going to try and compete with Robinhood or Coinbase. So it's more likely to go in the other direction. But we view the specialization and the role as sufficiently distinct as to be valuable for the foreseeable future.

David Hoffman:
[53:23] What's the likelihood that Stakehouse or somebody like Stakehouse simply just gets acquired by one of these very large players and everything kind of just gets integrated and verticalized?

Adrian Cachinero:
[53:34] Possible. So, Stakehouse has the odd feature that we're not venture-backed. Unlike many businesses in crypto, we're a family business. And so, we like to think in centuries. So, our horizons are much longer than venture capital.

David Hoffman:
[53:52] Our plans are... What's that mean from Dune?

Adrian Cachinero:
[53:55] Yes, our plans are measured in centuries. Yes, exactly.

David Hoffman:
[53:59] Well, what does that mean? How are you thinking in centuries? What happens in a century for Steakhouse?

Adrian Cachinero:
[54:05] No, we're not looking to make an exit for investors. We're looking to make the biggest possible business, deliver the maximum amount of value to users.

David Hoffman:
[54:15] Okay, so because you're not VC backed, because you're not looking for an exit, you get to play the long game because you're here to play the game.

Adrian Cachinero:
[54:24] The price of the brick is just higher for Steakhouse because we have that luxury. And we, yeah, our goal is really just to do the best thing for our users, you know, make the market as efficient as possible, build the most invested protections as possible, and just get bigger doing that and delivering value.

David Hoffman:
[54:43] Well, couldn't you see somebody like Robinhood, Coinbase, Kraken on the crypto side of things, but also, I don't know, BlackRock, Apollo, I don't know, Vanguard on the Tradify side of things, It's like, well, okay, Steakhouse doesn't want to sell to us. We'll just hire somebody in-house or we'll acquire their competitor. I guess I'm just asking about the tendency for mergers to verticalize the vault business.

Adrian Cachinero:
[55:09] For sure. Like there's a strong pressure in any industry. Well, I don't know that we've reached that stage in the vault management industry yet because it is growing and growing industries have a tendency to delay consolidation because there's just more opportunity for more people. So I do think that there will be a good amount of years while the opportunity set expands for everybody to kind of come in and take a slice. You know, like Aave is the biggest curator, but Steakhouse has staked a significant claim next to Aave. Like we are comparable in size to Aave and we didn't exist two years ago. The market will continue to grow. So there will be other entrants. There'll be niches and verticals where we don't play, Ava doesn't play, nobody else plays and people find a little thing and that thing then just grows. Consolidation will eventually come and you know we're not against it you know there are large acquirers you know i like you know big mansions and yachts like i'm not going to say no to a lucrative exit but our horizon is is very long and we want to do right for users yeah cool cool uh

David Hoffman:
[56:21] One question about ai uh does ai get involved in the vault business uh like you could imagine just, more data more intelligence being managed by an ai and ai kind of like offloading a lot of your manual labor what's the future of vaults plus ai.

Adrian Cachinero:
[56:37] Yeah heavily so heavily i think andrew talked a bit about this i think his product is very very ai native it's a significant accelerant to the development i don't know that it's in a i don't know i don't know that i would use a vault managed by an AI entirely. But I know for sure, at least in our case, our work is significantly accelerated in pace thanks to AI. And this is only going to increase. The scalability of teams is only going to increase.

David Hoffman:
[57:12] One thing we learned about the, at the very beginning of AI, or what was discussed at the very beginning of AI is like radiologists are losing their jobs because AI is so much better at radiology than the radiologist is. Like they're so much more precise. Ultimately, we are actually in a supply glut of radiologists. What did we, what did we learn here is that like radiologists is not about being a good radiologist. It's about a human professional being responsible for the answer. And so radiologists are using AI to do their job, but ultimately it's a radiologist that puts their stamp on it. And so I could imagine that like, no, an AI is literally never actually doing the curation, but it is helping you do your job and then Steakhouse puts their stamp on it and they've been like using AI in the background and that's kind of how it grows.

Adrian Cachinero:
[58:02] It's like Jevon's paradox, no?

David Hoffman:
[58:04] Right, yeah, yeah.

Adrian Cachinero:
[58:05] It's more like you see an acceleration of consumption and a faster improvement in capabilities that allows for much more growth with fewer people. And yeah, but people are hungry for vaults. So all it will do is increase the number of vaults.

David Hoffman:
[58:21] Adrian, what's your 18-month plan look like to grow the stake house business? I'm sure it's great to be in the vault business right now. There's a lot of external capital looking at us. There's Tradify looking at us. What's like the biggest opportunity for you and Steakhouse right now that you guys are working on?

Adrian Cachinero:
[58:40] I think it's basically more of the same. Like we've done a very good job in positioning ourselves as product builders for fintechs and for distributors. I think it speaks to our competence over the subject matter. I'm very, very interested in seeing how real assets integrate into this. I'm very happy to move away from point farming and start to move into regular meat and potatoes, carriage rates this is going to change the nature of volts it will increase the amount and degree to which volts differentiate from each other it will increase the complexity uh regulators clearly are are having a look and merit uh you know a breather to just understand the space and do it responsibly and so yeah really more of the same less points point stuff more real more real stuff these are the sort of things that excite us for the for the near term horizon

David Hoffman:
[59:41] What questions keep you up at night or feel difficult or what obstacles are there in your life in the vault business that uh will make your life hard.

Adrian Cachinero:
[59:53] I mean, credit is like credit risk events are the ones that are an insomnia inducer. Even when they don't happen, this is just a constant. So, you know, every day we're monitoring vaults, we're checking the liquidity, we're looking at the price articles. It's difficult to avoid. It's a very stressful position to be in. I don't know that I would wish it on many people.

David Hoffman:
[1:00:23] When you wake up in the morning and you regain consciousness from sleep, is that the first thing that you think is like, is there enough liquidity?

Adrian Cachinero:
[1:00:31] Check the vaults, check the collaterals, check the... It's not just me doing it.

David Hoffman:
[1:00:36] Oh my God, I do not want that for anyone.

Adrian Cachinero:
[1:00:38] Stakehouse is a worldwide business. So we have people in Australia and Africa and Hong Kong and the US and Europe.

David Hoffman:
[1:00:45] And you guys like pass the baton as who's observing the vaults as the sun travels around the world?

Adrian Cachinero:
[1:00:49] Exactly, yes. a lot of it is obviously automated but you know it's not a there's still a human in the loop so we're always watching

David Hoffman:
[1:01:02] Adrian I've learned a lot thanks for coming on the show and teaching me about vaults.

Adrian Cachinero:
[1:01:05] I hope it was useful yeah it was a pleasure chat

David Hoffman:
[1:01:08] Bankless Nation y'all know the deal crypto is risky but it's not risky enough it can get even riskier the institutions are coming here and so we are going even more west this is the frontier it's not for everyone but we are glad you're with us on the Bankless Journey. Thanks a lot.
---

*This article is brought to you by [NEAR](https://www.bankless.com/es/sponsor/near-1785257427?ref=podcast/whats-next-for-vaults-steakhouse-co-founder-adrian-cachinero)*