The LRT Episode

Bankless Nation, welcome to the LRT episode. Today we have a different kind of podcast than what you would normally expect out of the Bankless Podcast feed. Today we are talking with six different teams at different times, not all at once, all in the LRT space. Five different LRT teams, and then Chunda McCain from Ion Protocol, who can give us a little bit more of a meta perspective as to how to view the entire LRT landscape. I'm making this episode because there are so many questions that we have about the restaking and LRT ecosystem. It is developing at a mile a minute, and the number of questions that we have are growing faster than the answers that I think we have for them. And so I'm hoping to help answer some of those questions that we have in the LRT space. What are the risks of LRTs? What are the different strategies of LRTs? What are the different directions that different LRTs are going in? I've always thought that the LRT game is won by maximizing exposure and minimizing risk. And each one of the LRT teams that you're going to hear from in the episode has different strategies of getting that done and different priorities, different orders of operations. In this episode, you're going to hear from Etherphi, Puffer, Kelp, Swell, and Renzo in that order, all about their different LRT strategies, what they are doing, and what their roadmap is. Some of them have explicitly teased their token generation event, but all of them will have a token, I bet. I'm guessing, there's a prediction, in the year 2024 and probably in the first half. But diving headfirst into the LRTs might be kind of a lot. So first we're going to talk to Chunda from Ion Protocol. An Ion Protocol is an aggregator of LRTs. Kind of like it's kind of like a compound. So LRT and LSTs go in one side, and then vanilla ether goes out on the other side. Where all of the different LRTs are producing different flavors of ETH, Ion uses a compound style market maker application to re-aggregate and re-homogenize all of the ETH that comes out of Eigenlayer so we can more efficiently use capital in DeFi. And because of this position that Ion protocol has in the space.

They have to be extra cautious and extra informed about all the different choices that LRTs are making. So that is why Chunda comes first in this interview. He is going to provide a framework for understanding how to evaluate LRTs at a meta level. And then we will go into the individual LRT projects. Every single interview is 15 minutes or less. I think the Chunda interview is maybe 18 minutes, but this is meant to really give you broad exposure to the entire LRT landscape and kind of download you very quickly about all the different players here so that we can all be informed about what's coming inside of the eigenlayer ecosystem. So let's start with our interview with Chunda from Ion Protocol first, and then we will get into the LRTs after that. We got Chunda McCain of Ion Protocol. Chunda, welcome to the show.

Thank you for having me, David. It's a pleasure to be here.

I want to peek peek into your brain about how you think about the LRT ecosystem. Before we do that, I think we need to explain to listeners what Ion Protocol is and what vantage point it provides for viewing all of the different LRTs. Because you're kind of at the meta level, you're at the bird's eye view, satellite uh level view of the whole LRT landscape. Maybe you could explain a little bit about Ion Protocol and what it does and how it informs your thinking about the LRT space.

Yeah, 100%. So uh I guess I'm gonna give a little bit of context about myself and kind of how the idea of Ion kind of came about. Because I think I'll provide a good kind of foundational starting point to kind of understand a little bit more on how we're approaching the problems of addressing LRT risk and specifically how do we want to approach financial underwriting of LRT risks.

So a little bit about myself, been in the crypto space. I'd say I first learned about it back in 2015, but really got into it in the birth of DeFi 1718 era.

And just started writing white papers on like peer-to-peer lending market design, and then starting doing research on AI-based credit risk analysis frameworks, not for DeFi, but actually for in the more traffic context, thinking about how we can actually build systems for people who don't have access to like FICO scores or extensive credit history, and trying to get build an approach from more like heuristic

data to say, like, okay, uh, how can we assure that these un and underbanked people in America gain access to the traditional financial rules that other people have?

And so, you know, thinking about those types of things in crypto, uh kind of emerging, like DeFi emerging onto the CN around the same time, really sets the stage for me to like it just immediately clicked. And so for me, um, I spent, you know, kind of the next, well, I guess, uh until now seven years, um, kind of getting engaged in the space, working from, you know, being a software engineer, like a smart contract engineer at like a Neo Bank, building out crypto infrastructure there, uh, to working at uh DeFi startups to even spending a little bit of time in venture.

Um, and it was actually my time in venture. I was working on the blockchain capital team back when we were looking at the Eigenlayer deal. I think this was late 2022,

um, that I first I got my first glance into what Eigenlayer was as a concept.

Um, and it just, I mean, it was one of those, like I'd say it's probably like the second most important life hold moment of my time in crypto, because the idea of extending uh the core concept of crypto crypto economic security of Ether as an asset, and which is basically the whole value proposition why we

All use ETH as kind of the current status quo settlement layer in the crypto ecosystem,

extending that kind of power that Ethereum had to other networks and other services just immediately made sense.

And I realized there there was kind of a big problem that existed within this context in DeFi in particular, which is that

We don't really have primitives that internalize all this infrastructure risk that exists in restaking platforms. This idea that instead of underwriting like traditional financial leverage or these terms that we're more familiar with in DeFi, I think you guys kind of mentioned this recently, there is no inherent leverage in restaking. The risk that you're taking on effectively is like

there's a bit of game theoretical risk about like node operator activity. There's a bit of infrastructure based risk about like ABS setups and consensus network setups. And then you also have a bit of risk around like the balancing the principal agent problem and like decentralization from node operator to network perspective, right?

And so for us, we were like, all right,

in order to tackle this problem, we need to first uh ask the questions of like, what is needed

um in the restaking space? And then, like, all right, how do we build this infrastructure? So, in terms of what was needed, we're built ION effectively, uh, is a lending platform meant to underwrite all of these different types of risks to allow people to financialize their restaking positions as well as staking positions um without having to uh

Kind of

depend on the traditional means of financial underwriting that exists with like, you know, counterparty AMM liquidity and and really deep like Dex liquidity and also all these like price oracles and kind of taking a step away from that infrastructure and thinking about all right, what are the specific things about this market that are important for us to address?

That was a super useful explanation. I think one of the very interesting things about um Eigenlayer and liquid restaking tokens is that at first glance, everyone sees Eigenlayer and they see the concept of you you take your ETH, you stake it, and then you stake it again, and then again, and then again. And then everyone is like, oh, but that's like we're like stacking risk on risk on risk, which is true. Uh and so like then people's like heckles start to like rise up a little bit about just like, okay, what are the risks of Eigenlayer? It's a very natural response. And then like you layer on LRTs on top of that, and LRTs are all playing the same game of like managing that risk. Uh and then there's your game, which is managing the net aggregate of all the LRT risks. Yeah. And so it's this simultaneous like yield seeking, risk seeking, um, inherent nature about Iconlay, Eigenlayer matched with like, okay, but like let's contain all of the risks too. Because at ultimately, at the end of the day, Eigenlayer, like you said, is going to reach out.

To far beyond the horizons of crypto, leveraging crypto economic security, providing new products to like TradFi, to the SaaS model, to like San Francisco, Silicon Valley. And so it actually measuring and containing risk is really, really important. So like it's this kind of like two-faced nature of like restaking, which is like more risk, more yield, more upside,

but we're gonna manage it all. And that's kind of how like that, that's why I think this space is so interesting. And your perspective is not just like a user. We have a lot of users listening to this episode who are trying to get their um just get educated about the risks of LRTs. You are coming from a protocol perspective because you are building a protocol. And so you have to think about it a little bit differently. Can you illustrate some of the inputs that go into the risk assessment? Like what are the different things that Ion protocol assesses when it does its underwriting of the LRTs? Like, what are the inputs that are actually being evaluated?

Yeah, that's a really good question. Um

I think a good thing to remind everyone as well is uh this concept that LRTs, at first, before they're ever LRTs,

they're LSTs. A lot of them, anyone that supports native staking, you're building an LST protocol

first, and then you're building an LRT on top of that. Right. So our first approach is to actually um tackle that um and say, like, all right, we need to evaluate all of these different providers. Um, so I call them providers because effectively that

they they act in a very similar kind of function.

Um

And we need to assess, all right, specifically,

how do you look as a provider first? Because that's your core, that's the core dependency that exists right now, especially without Iconlayer being live. We need to answer a question and make sure that

everyone in from a provider infrastructure standpoint is being safe. So for those things we we look at, we take into account uh node operator setups, the idea of uh kind of two things. One, it's like a technical implementation standpoint.

Are you requiring bonds? Who has access to the node operator keys? Is it the user? Is it the underlying node operator that has all the control? Who decides what node operators get included? Is it permissioned, permissionless?

How exactly are you choosing these node operators? Who are these node operators? Are they docs? Are they undocked? All these questions that you would traditionally ask in a liquid staking provider context, we're also asking for these LRTs.

So that's the system wide kind of perspective. You also have technical implementation details that are important as well.

How much are you using DVT? Are you using DVT at all? Do you plan on integrating TEEs to abstract away key ownership?

Um, how what's your client diversity looking like? Right. All of these uh kind of more infrastructure based questions are also things that we're looking at internally when it comes to assessing these various different providers on the staking level.

Now, after the staking level, then the approach is okay,

uh, you kind of have to balance two things when it comes to um

uh avs delegation which which creates this is actually interesting thing which uh I think we you know we we wrote a bit of a paper on it um with with some uh folks uh I'm sure

we can link that later but

um we what we call it is is the efficient frontier of uh restaking risk where

You know, even if you're delegating capital to what we seem as like the most safe AVSs, you still end up centralizing your risk vectors. And so there's actually this interesting balance that you have to take where you want to decentralize your delegation, your internal delegation, to multiple AVs, thereby technically earning more yield. But uh also finding that balance where you don't over allocate to AVSs that take on additional risk.

Um and so this idea of like creating silos of where you're uh allocating and delegating your risk and diversifying that risk across multiple different AVS delegations is something that we're considering.

A lot. And the one big thing right now, and this goes a little bit more on kind of like infrastructure based underwriting, um, is this idea of like

how common is this type of consensus model? Proof of stake and BFT based consensus protocols have existed for a long time. We understand their risks really well.

Um, and you know, as we've seen in throughout the history of Ethereum, we can mitigate those risks

pretty decently well so far.

But one of the big concerns is when we start innovating on novel consensus models, even DA-based consensus models have some interesting things that make things like anti slashers unviable for DA,

which is an interesting kind of constraint there. You then we kind of start ending up walking into uncharted territory where we're not exactly sure what

black swan events look like in these situations, what can cause them, and even what does the status quo of that network look like.

And so for us, I think we're going to be heavily indexing. And I think users should also think about heavily indexing on what kind of consensus models and what kind of underlying infrastructure are people building that we're already used to underwriting and we're already used to kind of assessing within the context of Ethereum and the consensus protocols that exist today.

One of the roles that Ion uh protocol will will do is it it it'll have a algorithm, I'm assuming, uh to actually like judge parameters. Just to make it super clear, Ion protocol like compound, uh but LRTs go in one side and just normal ETH goes out the other. Or

Yeah.

Or LST. LST is on yes, right. Yeah. Any sort of ETH derivative on one side and that's bait with a protocol underneath. And then an vanilla ETH on the on the outside. Which means that you need to have an algorithm to like produce parameters, risk parameters about the collateral assets. This will, of course, ingest some of the inputs that I asked you about earlier. But like also, how do you even design the algorithm? Because where do you get the data to even make weighting decisions? Right. Like this is we've never done this before. You're doing it first. How the hell are you doing it? What do you know? Like, how do you actually like build this thing?

Yeah. No, that's a very good question. And it actually goes to the thing, like the two things that concern me the most with Eigenlayer, like holistically, right? The two things that concern me the most about Eigenlayer has nothing to has nothing actually to do with Eigenlayer, the product, but instead on how we are able to observe the product.

Those two things are opacity on data, uh, basically the limitations on being able to see, access, and view data uh in real time when it comes to all of these networks.

And then, two, is like the ability to then ingest and analyze that data. I feel like there