See all Bell Curve transcripts on Applepodcasts

All right, everyone, welcome back to another episode of bell curve. Before we jump in quick disclaimer, the views expressed by my co-hosts today are their personal views, and they do not represent the views of any organization with which the co-hosts are associated with. Nothing in the episode

is construed or relied upon as financial, technical, tax, legal, or other advice, you know the deal. Now let's jump into the episode.

All right, everyone, welcome back to another episode of bell curve. This is the first interview episode of season 5, the end game for liquid staking. Today we're going to be joined by Hasu. Hasu is a special guest on Bellcurve at this point. It's a fun callback for those of you who watched the MEV season where Hasu was my co-host.

Today we're going to get him on the other side of the mic talking about 1 of his other favorite topics outside of MEV, which is liquid staking. Hasu's been acting as a strategic advisor for Lido for a period of time, and not only has great sort of high-level framings on the space for liquid staking, but in this conversation, we got to get into some of the nitty gritty with him. So in the kickoff episode of this season, Miles and I introduced some of the big themes that we want to get a sense of this season, which is how is the market structure for liquid staking going to develop? What are Hasu's thoughts around the self-limiting debate, which is this idea that if a protocol like Lido gets a very dominant percentage of the market share, should it actually cap itself? We also talked about the impact of restaking and how that intersects in this sort of co-op-editive, co-op-etition relationship in between protocols like Lido and protocols like Eigenlayer.

Now, not only did we get some great high-level frameworks for how we should think about this and also went deep into the history of Lido and how it developed from a first principle standpoint, we also got into some of the more nitty-gritty sort of tactical ways that these protocols actually develop and think about things. In particular, I want you guys to pay attention to the section where we discuss the role of decentralization and how it actually acts as an offensive tool for protocols like Lido. So this was a really, really great conversation. I hope you guys enjoy it as much as Miles and I did. All right, everyone, welcome back to another episode of Bell Curve.

You got me and my co-host Miles here. And today we're joined by Hasu. It's a special guest, I think, of Bell Curve at this point. Got him back on the other side of the mic. Hasu, welcome back.

Hey, Mike, it's good to be back. It's been a while.

So this is gonna be a ton of fun, Hasu. We're going to be, last season we talked about MEV, which is a subject that I know is near and dear to your heart. And then I think we're going to be going down the road here on another subject that's near and dear to your heart, which is liquid staking. So we're recording this on July 6th. We're deep in the heart of a bear market.

And it seems like liquid staking is 1 of the few areas in crypto, which is really taking off. As of today, I'm looking at the TVL of Lido, which has just eclipsed 14 billion. That's more than Maker and Aave, the next 2 largest by TVL on Ethereum combined. So I think we could just start at a high level for this talk and just what is behind the rapid growth and adoption of liquid staking? And then maybe we can get into kind of a state of the union and how to carve out What are the major buckets of different types of providers?

Sure. So I think the reason that liquid staking protocols have risen to the top of the TVL leaderboard, if you will, I think it's in the past, lending markets used to occupy that position because it's a place where people park their assets in order to get some additional rewards on the assets from lending them, for example. But we are seeing in the bear market not very much demand to borrow. And so not very much interest that you can earn from putting your assets into lending markets, frankly. And so That's why I think overall the TVL of lending markets has been on the decline.

We are seeing that in MakerDAO in particular. So DAI outstanding has gone down from 10 billion to 4.something billion. The TVL inside MakerDAO has declined even more. So now a lot more of the die outs sending is backed by stable coins or other forms of stable collateral such as treasury bids that are held in off-chain trusts. And on the other hand, we have seen the rise of proof of stake.

So obviously, Ethereum went from proof of work to proof of stake, and that allowed users to now secure the network by staking their ETH. And we are seeing an increasing demand to participate in that. So for 1, users can earn inflation rewards from Ethereum, but they can also earn MEV and transaction fees. And in spite of the bear market, these rewards have actually been quite consistent, much more so than what you can get from lending your coins. The other thing that I would point out is that withdrawals have now been enabled from the beacon chain.

So proof of stake in general has been de-risked quite a lot. So my thesis has been for a while that we would see more and more ETH being staked as a result of that, because it's as close to, I don't want to say risk-free, because it's not risk-free, but it's as close as it gets basically to the kind of risk-free rate of crypto, which is just staking your ETH, which is kind of the most secure and most decentralized asset on the protocol layer in order to secure the network. And if you run your own node and you do a good job, or if you delegate that to a staking pool or a staking protocol, then yeah, you generally have, that's a pretty good kind of, it's a low return, kind of low risk way of using that asset.

1 thing I'd be curious to get your perspective on, Hasu, is this sort of recursive leverage trade that people run using a liquid staking token like Steeth. So the idea being that you stake your ETH, you get Steeth in return, you turn around to a borrow lending protocol like Aave, you borrow more ETH and you kind of rinse and repeat and in that way sort of lever yourself up. I guess there's kind of a spectrum of healthy sort of capital efficiency and wanting to not, you know, take, lose too much opportunity cost by staking. Maybe that's the really healthy way of doing it. And then you've got this extremely recursive leverage trade that you can run being on the not so healthy way of doing it.

So I guess how much is part, like how much of the market is driven by that more degen sort of activity? And where do you sort of fall on the spectrum of, you know, not to, you know, put some sort of value on trading, But is this a particularly good use of liquid staking versus maybe, you know, it's excess speculation kind of.

So the first thing that I would say is that this kind of looping is not no longer possible today because you can no longer borrow ETH on lending markets for cheaper than you can stake it for. And that's what you would expect in an efficient market. So there's kind of 0 arbitrage anymore. That used to be different in the past, in particular, before it was possible to withdraw actually ETH from the Bitnichain. And so someone who was holding stake ETH, or some other liquid staking token, they were incurring basically like a risk on, you know, the price of that moving against their ETH in some form.

And they could get liquidated on their position, right? If they if they were using the liquid staking token as a collateral to borrow ETH and then the liquid staking token falls in value, then obviously you can be liquidated. And because people were like looping a lot of times, the leverage was very high. And so there was definitely a concern of kind of these cascading liquidations there. You were asking whether this is a useful feature of liquid staking tokens.

I mean, I would definitely say so. We can go into how liquid staking is different from other forms of staking in a moment, but the ability for users to get liquidity on their tokens and to use that liquidity for whatever they want is definitely a feature of the system. And likewise, I think it's good for people who put their ETH into lending markets, because as a result of this, the rate that they could earn on their ETH was driven up to the staking rate basically of what they could have earned from staking as well. And so you could say people who are not comfortable maybe staking their ETH, but only putting it in a lending market. So they were participating just the same from the staking yield in that way, which it basically further democratized access to staking through lending markets.

So let's get into, I'd love to get your perspective on what is a state of the union of liquid staking providers look like? And maybe we could actually, we talked quite a bit about Lido in the first episode, actually, but I realized that we never really gave quite an overview of the history of the protocol and how it functions today. And I know you act as a strategic advisor to the protocol. So maybe we could actually start with sort of Lido and get your sense of the overview of maybe wherever you want to start, either the history of the protocol or what you're sort of focused on today. And then maybe we can go and sort of carve out the existing state of the market structure and then try to poke at guesses at how it's going to develop?

Yeah, sure. So smallnit, I'm an advisor to the DAO, which is the union of the token holders that basically vote on decisions for what the protocol should do. So Lido, the protocol was conceived a few years ago, before the beacon chain was even live. And Back then, there was only 1 form of staking, I would say, really known from other blockchains, which is you delegate your, either you stake yourself or you delegate your coins to a professional node operator. Some blockchains had in protocol forms of delegation that kind of made it cheaper, more accessible and also more secure because the node operator didn't have custody of the coins.

But Ethereum chose a different route. In particular, they were going with this kind of isolated blockchain that would run in parallel, which was called the beacon chain, right? And that kind of ran its own, you could call it an incentivized testnet for multiple years. Before it eventually the, you know, the proof of work kind of consensus was then stripped out for this beacon chain consensus layer. And that's still how it works.

And I'd still say, This is 1 of the most impressive technical feats in the history of technology, the way that the engine of this plane was swapped in mid-flight. I still like this analogy very much because it's so fitting. I mean, Ethereum didn't miss any blocks as a result of this. And it's an absolute technical marvel, you know, the way that this worked. So when it was clear that Ethereum would move to proof of stake, then of course, it was obvious that users would want to stake and they would want to participate in securing the protocol.

And however, you had several problems in Ethereum. 1 was in order to run a value data, you needed 32 ETH. There was the minimum at the time. It had already been lowered from, I don't know what the number was before, maybe in a tune of several hundred ETH. That was when Ethereum was the cheapest, much cheaper than it is today.

And they lowered it to 32 ETH. But even at the time, you know, when the beacon chain came out, that that was maybe, you know, $60, 000 or something. So it was a very substantial amount and not many people had. So needing to stake in increments of 32ETH was 1 big problem. The second big problem was that people had to lock up their stake on the beacon chain and the beacon chain was expected to run for at least 2 to 3 years, maybe longer, before withdrawals would eventually be enabled.

And as we saw that estimate turned out to be pretty accurate. So I think it took around 3 years between launch of the beacon chain and withdrawals going live. So I'm not

100%

sure. But it was also clear that people would want liquidity on their stake. So either 2 outcomes, either people could not get liquidity and then the number of people who would be willing to stake would be much smaller. Basically, only people who don't would be comfortable locking up the ETH for many years with an uncertain outcome, if they will ever get it back. Or people can receive, we sometimes call it a voucher, you know, for that represents this, if that they stake on the beacon chain and then they can find another seller, sorry, they can find another buyer for that and they can sell it.

Yeah, and we now call it a liquid staking token. And the third thing, the third problem was that that Ethereum did not have this native ability that we talked about to delegate to to a node operator. So you could find Ethereum did have 1 feature, it had separate validator key and withdraw key. So the party who could withdraw the tokens could be different from the party who was using the coins with their validator on the beacon chain. But it wasn't trustless.

Basically, you needed to find someone, you know, sign a legal contract, needed to do kind of your own sourcing and so on. And there wasn't any way to, yeah, to just like, select someone from the list of not operators or say, I want to withdraw my tokens now. And so you had these multiple problems. And I think it became pretty clear that there would be a new form, like that a protocol that would solve all of these problems would be quite dominant over anything that existed at the time, right? So when I look now where the kind of differentiate, differentiation lines, let's call them like that, are drawn, then you still have this, what we now call liquid staking versus non-liquid staking.

So liquid staking is just when you, when you receive a voucher that represents this ETH that is being staked by an auto trader. And you can take it and you can use it somewhere. You can trade out of it, you can use it in Aave and Meka and so on. And the non-liquid ones are just if you give, for example, your money to an exchange and they stake it for you and they put rewards in your wallet every month. This is another very common form.

I think Coinbase has a double digit market share using that model in staking right now, because it's very easy. So that's the other big category. Then I think custodial versus non-custodial is a pretty big 1. I think there's a pretty big overlap, I would say, between like liquid and non-custodial, non-liquid and custodial, because they fit very well. And then finally, there's retail versus institutional.

So it's more about which customer segments are you speaking to, right? So I think that Lido, the protocol, Initially, I think its value proposition, it spoke particularly to kind of Ethereum native users, right, who wanted to self custody because it was a non custodial protocol, who wanted to use their coins in DeFi because it was liquid, right? And who weren't institutions, right? Because institutions weren't in DeFi at the time. So I think That's where Lido got its start.

But we have over time also seen other protocols come around that focus more on, for example, ease of use. So like the whole custodial liquid staking, like its big benefit is that you just don't have to self custody. Right? So if I'm a user, most users in crypto never leave their centralized exchange account, right? They never withdraw a coin to Ethereum still.

And so there's a huge number of users who you only reach if you have a custodial offering and you make it like 1 click from the existing exchange wallet. So it was very natural, I think, for Coinbase, for Kraken, for Binance and these other platforms to roll out their own staking offerings in a custodial way. And I think I still remember that there was actually a, I had a huge concern personally that these exchanges would capture most of the stake by the time that withdrawals would even go live. Why? Because it's like a custodial staking is so much simpler than non custodial staking for users.

But the main reason was kind of the liquidity. Because if you're on an exchange, then you can very easily make staking liquid in the sense that you allow users to trade out of these staked validators. And I really expected exchanges to go very heavy on liquid staking because they could make it, they could allow people to trade out of it, they could allow it to be used as collateral. And it didn't really happen that way, which kind of made me very, very surprised. So Binance had a staking relative coinbase for the longest time didn't, Kraken didn't.

I would still have to talk to an exchange executive and understand why they didn't do it. What was the exact concern? But it would have been, I mean, I expected them to do it and I think it would have been the right play. And had they done it, then I don't know where we would be today, frankly, because it would have been such a compelling offering. I think that, yeah, it basically becomes a very big risk that eventually kind of

60 70 80%

of Ethereum validators are run by like 2 to 3 large exchanges. So I think into that concern, then LIDAR was born to present basically a counterweight, like a non-custodial staking protocol that has some of the same benefits, which is it's very easy to use and it's liquid so that people don't have to stake with exchanges.

I think that makes sense. And I think you kind of pointing out 3 different user personas here. You know, the first being, you know, maybe an institutional user or somebody on an exchange that either wants the convenience of 1 click from an exchange, or they want, you know, the compliance regular requirements of, you know, an institution. Then you have basically everybody in the middle that uses Lido. And then on the other end of the spectrum, you might have self-stakers that mint Rocket Pool ETH, right?

And how do you view the growth of those, you know, I guess the first bucket and the third bucket relative to Lido's core users right now? And then how do you think about extending support for those types of users, you know, from your seat in Lido?

That's a very good question. I mean, I think, where is, do you know where exactly they are in terms of market share right now in terms of broad size, kind of the more institutional focused segment?

Yeah. So, you know, CBE, which is interesting, because it is retail only, you know, has grown, I think, to roughly about 10% market share of liquid staking. I believe Rocketbull is kind of around the same range. And CB ETH, it's notable that they've done this in the face of 2.5x the fee rate that Lido charges. So definitely tells you something about the users there and their preferences.

I'm just curious to know how you guys are thinking about positioning Lido also to capture some of that market. Maybe this is users that were previously sticking with Kraken, right, traditionally, and now have withdrawn, or maybe they're users that were just waiting for withdrawals, right, to come off the sidelines. And I'm curious to just hear how you think about the growth prospects of either side of the market.

Yeah, so I think and I, if I'm totally honest, I don't strongly understand who uses CBE. Because as you were saying, it's so much more expensive, but you still have to self custody it if you want to use it. Right. So I don't know, like, do what are all the assets on chain? Like, I

don't actually have to self custody it. I think you can you can rent it and then hold it in your Coinbase account.

Okay, so people just hold it as a stand in for like regular Coinbase staking. So it was basically okay. So it's almost like they cannibalized the existing products in a way and just called it CB. Okay, that would make sense. Because if all of the CVE for sell on chain, then I would actually argue that Lido is already kind of a super set of that functionality.

But if if you have to custody it, then I would have to say, I think the way to access this market would be by offering alternative custodial products. A, by existing exchanges, listing StakeEth, so people can buy StakeEth and hold it on an exchange or a wallet. I think that would be 1. And the second would be through more traditional, you know, asset management channels. So if I can hold Staked Eve in my Fidelity account and my Vanguard account and so on, then that is probably a competitive channel kind of to these existing exchanges that speaks to a very similar, maybe even a little less sophisticated group of users.

But that is, again, bigger than the previous. So yeah, that's what I would say about that. In terms of Rocketpool, I think that solo stakers, so individuals who run their own nodes, and let's say who kind of run fewer than I don't know, 100 validators, I think that they are really important backbone of Ethereum decentralization. And Rocketpool, to their credit has been around much longer than Lido as well. So they were, they launched later than Lido.

Lido was the first protocol to market by I think a decent margin. But the, you know, Rocket Pool as an idea and kind of the team behind it had been at work for a long time. And I think the reason that they were launching later was because they were set on a very specific model where users can basically like solo notarizos can divide their stake and you know, they can put 50% of their own stake and they can borrow 50% of a of a user stake and then they can stake that together and they secure that in addition with an R Rocket Pool token bond or kind of a collateral I think is a better term, right? And this has the benefit that it's actually trustless. Like, it's at least more trustless.

So there are some caveats that we don't have to go into. But it is definitely more trustless than other protocols that exist today. And so it's very important for Ethereum that the solo, it's almost like leveraging kind of this backbone in order to, you know, give them more stake, in a sense, right. And this makes definitely Ethereum more decentralized. And this is, you know, the same values that that LIDO has as well.

I think the 2 protocols have different, yeah, different priorities, if you will, right, because LIDO focused on this, or the Lido DAO focused on what they saw as a huge and possibly existential risk for Ethereum, which is, you know, a small number of exchanges capturing a majority of the Ethereum stake by the time the proof of stake would go live. And I think that was a very real and very big risk. And it is still a big risk today. Even though many people don't want to talk about it. I was looking actually last year and we will get into this, I think it was last year, when there was the self-limiting debate about Lido and there was a lot of, Lido had a 1 month where there was very bad, like a series of very bad kind of consecutive press, where there was the 3 arrows, kind of for selling their, you know, stack of state Eve, and then Celsius.

And there was Terra Luna. And at the same time, there were kind of people on social media arguing Lido is getting too big, and they should think about self limiting, right. And at that time, Lido had almost no new inflow of stake. Also, because the stake ETH was trading at a discount. It was a bit cheaper than ETH, right?

And so somebody wanted to stake, they were just buying from the market instead of staking. And you could see that over 90% of people who were staking, people who were still staking, but over 90% of it was going to Coinbase and Binance, you know, into the custodial pockets, you know, and it was a kind of, and these were entities that were not transparent at all, you know, about how much share they have in staking and what is their infrastructure like, where is it located, And, you know, they don't care about client diversity or, you know, decentralization and so on. So I thought that was interesting. Anyway, I think the reason why I talked about why we got sidetracked to this, so Lido and Rocketpool, I think are ideologically very, very similar protocols that are starting from slightly different points. And I think that they are moving towards the same point, which is in the case of Lido, there was recently a version 2 of LIDO that included the idea of a staking router, which is a more modular architecture for adding basically different staking modules that represent kind of different lists of node operators or different configurations of node operators.

And what this allows is for Lido to not just distribute the stake to a permissioned list of node operators, but for anyone to write new modules and send them to the DAO to consider. And then the DAO approves, same way that MakerDAO or Aave, for example, approves a new form of collateral, then LidoDAO can approve a new staking module. And that can and will include things like permissionless staking that has a collateral attached to it, for example. Or staking that uses distributed validator technology. Where, for example, professional node operators and individual node operators are working side by side, you know, and have this kind of consensus protocol overlay and are working together to make blocks on the beaten chain.

And so I think I would expect in the next year, LIDO to move from currently I think 35 node operators to hundreds, if not thousands of node operators as a result of this infrastructure. And at the same time you're seeing Rocketpool has long kind of hit the scaling limit of their approach, which is if you need all of the notabitas to over collateralize any customer ETH that they take, both with their own ETH, but also with an additional and I think it used to be something and I don't want to say something wrong in the range of like 10% additional collateral in the form of the RPL token, then it is very expensive, you know, and it doesn't scale. And so I think they hit the scaling limit. And what they are doing now is they are trying to reduce this collateral. And they are also whitelisting entities that is allowed to stake without a collateral.

And so, such as coinbase, by the way, which is interesting. And so I think that over the next year or 2, I think you can expect these protocols to move more towards, you know, look more and more alike, in that sense. So I think a rocket pool, it will have more permission not operators who are professional and to have a rep off chain reputation at stake and so on. And that will allow them to meet more of the user demand. And Lido at the same time, it will have more ways also to support permissionless staking.

And that is the group you were talking about, right? The solo operator who's really passionate about running their own hardware and who just wants to get a form of leverage basically on that, or the user who wants to stake, but they want to stake specifically with a solo operator and not with a professional node operator.

I, cause I want to dig deeply into this with you Hasu, but before we do, I just want to bookend the conversation about market structure and maybe add 1 tiny bit of context there and then ask you a specific question before we get into this, which is, I thought you made a really great point about at the height, it was May of last year, actually, when this debate started taking off and Vasily made his post in the forum and actually Vitalik, right? So it's not just random people on Twitter, This was Vitalik himself proposing some sort of limit to Lido's growth. And the reason why, you know, the, the, uh, the sort of paradox that you pointed out was very interesting that at the same time this debate was kicking off, it was actually Coinbase and Binance that were getting inflows. And originally part of the reason why Lido was so supported was because I think like you, many were concerned that the centralized custodial based offerings would take up an enormous amount of the stake and Ethereum would lose some of its sovereignty, so to speak. But I think the reason why people are a little bit concerned about LIDO today, and we can get into whether or not that's valid, is because of this perceived dynamic of a winner take all market for liquid staking or staking in general.

And I think I want to kind of just ask you the question directly about how you see this market structure evolving. Do you agree that it's a winner take all or winner take most market structure? And maybe if you could get into some of the dynamics that make this liquid staking in particular, you know, what are some of the returns to scale? We talked a little bit about this offline, but if you could explicitly outline what those are and then say whether or not you agree with that, and then we can kind of get into this self-limiting debate.

Yeah. So I would say I'm definitely in the camp who thinks that liquid staking, and by extension, staking in general, has a very large network effect and return to scale. And so it's very likely, in my opinion, that the outcome will be very centralized or I think centralized is the wrong word but it will be more it will be concentrated. Why is that the case? So first of all I think liquid staking is a much much superior solution to regular staking because it's basically a superset of that.

So you can have, you know, liquid staking is like regular staking, but you also have liquidity, which is purely beneficial. And then it is, in addition to that, also non-custodial. And that allows it to be held in different ways. So a non-custodial asset you can hold in custody, but a custodial asset you cannot hold in self-custody. So in many ways, it's purely better in every dimension, right?

Because you can turn a non-custodial asset into custodial 1 if you want. And for that reason, I think that the market share of liquid staking over regular staking will approach 100% over a long enough period of time. And so we only have to think about liquid staking. So liquid staking in general, I mean, What is so special about liquid staking? It is that you get this voucher that represents the ETH that is being staked.

And you can use it. You get liquidity. And as we know, liquidity begets liquidity. And people want to use, or they get the most value out of using the staking token that has already the most acceptance that they can use in the most places, such as lending markets, exchanges, that has the deepest liquidity and that has the highest branch recognition and so on, right? It is really like to the degree that a token is used as money, it behaves like money.

It has the same network effects as money. And so if we are talking about a liquid staking protocol for a token that's not used as money, then these effects will be much lower, right? Because then, if people hold it anyway, and they don't use the liquidity effect, then it doesn't matter. But for a token like ETH that people use very actively, and many people treat it as a form of internet money, then the network effect is, as a result, very big. So that would be the first point.

And 1 thing maybe to add, back last year, people were actually saying, look at stablecoins as an example, for why this network effect is not actually that big. And look what happened 1 year later. All of this, like Tether is exploding in size and all of the other stable coins are collapsing. I think this is a good example, I think, why people shouldn't judge prematurely kind of what the outcome will be of a particular market. Because at the end of the day, network effect almost always wins.

There can be path dependency and it can take a few years. But I think that when it comes to money, the network effect is just super strong. And in the end, it will succeed. So that's 1, network effect. I think brand, and we touched on this slightly, I think brand is just extremely important.

I think as a user, most people are inclined to stake with what is already the largest provider and the oldest provider, by the way. And so if you're both, then that's a pretty big position, pretty good position to be in, because staking is something that requires a lot of trust. You are basically putting your money into a box, and you are leaving it there for years, if not decades. And This requires basically a lot of trust in the operator. And that's exactly what a brand gives you.

A brand is basically like it reduces uncertainty about the future. And so that's so important. And then there's this idea that running or getting a, when there's already some established liquid staking protocols that have built this, these integrations and this liquidity. So that's almost like a minimum that you need to get to, a minimum size that you need to get to, to be even anywhere near in the conversation. And so in the beginning, I think you can think about these in terms of fixed costs.

So if you wanted to break into liquid staking today, you have to spend multiple billions of dollars, probably in that range. And so I think it is it is just very hard to spend that amount of money as a as a smaller protocol. And so the bigger you are, the less kind of the effect of the, you know, fixed costs is hurting the revenue. And there's 1 1 final thing that I would say. Actually, this is a thought that's kind of more new in my mind.

I haven't really verbalized it before. But all of the criticism, not all of the criticism, but many of the criticism against LIDL is completely correct. Like the concern that, you know, things can go wrong if most of the ETH would be in 1 protocol, And that protocol is not part of the Ethereum core protocol. And it has some governance layer, for example. And Lido is taking aggressive steps to decentralize the protocol and the governance.

And 1 of them is this idea of dual governance, where, so Danny Ryan, we will get to this, he has basically voiced a concern, yes, Lido is not 1 entity. It consists of many entities, but under extreme conditions, the incentives could play out in a way that these many entities behave as 1 entity. And then they do something that's bad for the users of Ethereum. So that is his concern, right? And what LIDO can do or is doing about that is introducing the idea of dual governance, which basically says for anything that LIDO wants to change, that LIDO now wants to change about LIDO the protocol by voting on it, the stake e-folders can actually just veto the change.

And so you get this filter where bad changes can no longer pass through because they would not be in the interest of the staked e-folders. And I think the bigger the protocol, the more stake e-folders basically have to be convinced that a particular change is good for the protocol so that it can pass. And so you actually get this herd immunity from being in a bigger protocol where there's more people watching the governance and more people have to vote, have to kind of be in favor of any changes for these changes to come into effect. And so for me, this is definitely an argument for why I would use a protocol where there's more checks and balances and The bigger, the more people use it, the more people can spot anything that goes wrong.

It's a really good point. And I agree with you on all of that stuff. And the reason why I just wanted to interject that is I think many people agree with that point of view that it's a winner take all protocol. So people get a little bit worried about the amount of influence that a protocol like Lido has, because the general thinking is this protocol seems pretty unstoppable at this point. And just to get more specific, You know, to go back, if we rewind the clock to May of last year, I kind of mentioned that it wasn't just some random pushback on Twitter.

It was originally kicked off by SuperFizz, who openly wondered about the first staking provider to publicly commit themselves to not limiting their, basically keeping their stake to lower than 22% of validators on chain. But Talek retweeted that and said, speculative take, we should legitimize price gouging by top stake pool providers. Like if a stake pool controls greater than 15%, it should be accepted and even expected for the pool to keep increasing its fee rate until it goes back down below 50%, uh, 15%. That all led to an actual, uh, governance proposal in Lido, but it was initiated by vastly 1 of the co-founders basically voicing the idea. Should we, should we effectively self limit?

And eventually it was voted that they shouldn't do that. But I would love to get your perspective on the self-limiting debate anyway. It's kind of an interesting philosophical question, which is maybe a cousin of the should billionaires exist question, which is getting bandied around outside of our corners. But is it a good thing to be from frankly, just a capitalist, encouraging entrepreneur standpoint is a good thing to, to limit the growth of some of these protocols and then maybe we can get into some of the specific ways that Lido is choosing to limit itself. What you were just alluding to the staking router and dual governance.

A really good question. And I think last year it was a very pivotal time for the staking market in general and for Lido as well. So I think, first of all, I have to say that many of the concerns that people like SuperFizz and Danny and Vitalik have had and may still hold today about 1 staking protocol growing to a particular size, that these are definitely warranted. And so, for example, you know, Lido, the Lido protocol, the smart contracts are still upgradable by governance, for example. And so governance could vote, for example, to, you know, change something about the smart contracts to the point that it would be detrimental for stakers or for node operators or for Ethereum itself.

And so that is definitely a concern. There could also be bugs in 1 of these smart contracts that could have a negative effect. For example, ETH could be frozen. So I don't know how it would happen. I'm not, you know, super technical.

But I think that in general, when someone says, well, if there's a huge amount of TVL in 1 protocol, and it's like doing 1 of the core functions of Ethereum, which is moving the beacon chain forward, then, you know, this could have a spillover effect on the security of the whole system. I think they are completely right. So that's the 1 perspective. The other perspective is, what do you think is the structural outcome of competition in the liquid staking market? And I am also, I mean, I'm starting from the position that, you know, size is dangerous and has responsibility in order to manage all the risks that come with it.

But I'm also coming from the perspective that exchanges are, were and still are in the best position to capture a lot of stake and become very influential players in the staking market. And that the dynamics of liquid staking drive towards a winner-take-most outcome. And when you combine these things, then you kind of get into the mindset that you can't really, you know, choose for there to be 10 staking protocols that all have

10%.

I mean, you can dream about it, but it's not going to happen. And if you, and like, you have to be pragmatic. And what being pragmatic means is you try to make the best liquid staking protocol win. And you make that most, you know, as decentralized and as aligned with the underlying base protocol as it can possibly be. And so I would say that is the summary of my position.

I started from the point that I thought, you know, this is the liquid staking industry, This is very clearly going to have 1 big winner. And there's nothing that can be done about this. You can tweet as much as you want. You won't change that outcome. I didn't even know all of the players involved.

And I just thought, well, what is... What can we do from here? Like, how can we salvage the situation? And I saw Lido being, I saw Lido as a protocol that was in a good position to fight the exchanges. And that I also saw as like culturally, very aligned with Ethereum, having much of the same values and goals, and having a very clear commitment to transparency and to decentralization and to constant self improvement.

And I thought, yeah, this is something that I want to support with my time and, you know, with my, yeah, with my money as well. And that was the position that I was in. And when the self-limiting debate came up, which I mean, I thought, yeah, this is definitely going to be kind of a problem at some point. So it wasn't exactly a surprise. Then Lido has a grants committee that's called Lego.

And so it has a budget and people can give grants. And the way that Lido protocol works is it does the DAO, you know, and the DAO basically decides the direction and the strategy of the protocol, right. And so it turns out that, you know, whether Lido should self limit or not, is a pretty big part of what the strategy should be. And so it was very clear that this decision cannot be made without asking the DAO for their opinion. And so Vassili did what I think is exactly the right thing.

So as a member of LEGO, he looked for someone who can write what we call a Swiss booklet, because in Switzerland they have these ideas of the referendum and people can vote. It's kind of a form of direct democracy. People can vote And they publish these small booklets that that basically say, if you believe in, like, if you believe in these things, then you should be in favor. And if you believe in like these things, then you should be in favor of the other option. And they are written by an independent committee usually of experts on the topic.

And they are basically a form of voter education, and in lowering the cost for them to come to a decision that would be good for for everyone, right. And so I, and there was someone in the Cosmos system, we had in the past written such booklets for governance decisions in Cosmos. And I believe Vasilis thought that this would be a great thing to also try in LIDL. And then he reached out and he asked, you know, Sasha is the name of the author of that post. And he, yeah, he agreed.

And he didn't hold any IDO at the time. I don't know if he does now. But yeah, he wrote this unbiased analysis. And it had, I thought actually, it was a fan, it was it almost reminded me of Vitalik's writing style, because Vitalik is like incredibly good at representing both views of an argument. For example, when he wrote this post about, you know, the Bitcoin maximalist argument that he published on April 1st.

And all of the maximalists were saying, you know, this is an endorsement of Bitcoin maximalism. So It almost reminded me of that in a sense, because it really had a full list of arguments for why you should be in favor of self-limiting and why you should be against self-limiting. And I could read this and say, yeah, this is like why we shouldn't self-limit. And somebody who was in favor can read this and say, yeah, this is exactly why we should set limit because exactly like it isolates the assumptions that different people have and the different lenses through which they view reality that then lead them to, you know, come to voting a or b at the end of the day, right? And so I thought that this was actually a pivotal moment for Lido governance because it illustrates how the DAO should work.

When there's a major decision, then you should ask token holders for what to do, and you should try to make the decision for them as easy as possible. And yeah, so LIDO token holders did end up voting against self-limiting. Because I think at the end of the day, I mean, of course, there's a self-interest, but also, I think they were swayed more by the arguments against because I think people working on LIDO or investing in LIDO, I think they are kind of, you know, in a sense, agreeing with the idea that the staking market has these tendencies of being monarchic most and that it is very important for a decentralized protocol to win and so on. And yeah, so that's why we saw, I think, the outcome that we did. And that is the yeah, that is the track that the DAO has been on since then.

Yeah, I think that's a great overview. And, you know, it strikes me that I think a lot of folks on either side of the argument are actually very aligned at different stages, but a lot of this comes down to sequencing, right? And there were, I think everybody was more or less aligned that Lido needed training wheels in the early days in order to create an offering that was competitive to the centralized exchanges. And I think everybody can somewhat agree, or most people can agree that if there is a winner take all, winner take all winner, liquid staking protocol, then it should be more or less as hard to change as Ethereum itself. And the ability to stake with it should be as permissionless as staking with native staking, right?

And I think the disagreement here really came down to like where Lido should be at its stage of growth. And it almost had, in the eyes of some folks in the Ethereum community, had already mitigated this risk of centralized exchange capture, but yet hadn't gone far enough to take off the training wheels. And I'm curious to just hear if you agree with that sort of overview and also curious to hear where you think Lido is today, with things like dual governance coming in, the staking router coming in and early conversations about, you know, ossifying parts of the protocol so that governance cannot upgrade them.

Yeah, I think this is a really good, I think, observation. I can't speak for everybody who's on the side of, you know, Lido should self-limit, but I think a fair amount of people would say that if Lido had, you know, like was much even more decentralized than it is today, then they would feel comfortable with a higher market share, for example. And so they can point to particular things and say, you know, that's why I'm in favor of self-limiting today. But I may not be in the future. I was I was missing like a bit of nuance on that topic.

I think also like it was very, it felt very tribal, which is never good. Because then it's it felt more like I'm supporting like rival protocol X. And that's why I'm actually like making the stink against LIDL. But anyway, so you're totally right. Like LIDL in the past year had even more training wheels on than it does today.

So in particular, not, I believe at the time, so I don't know when it happened, but I believe at the time, at least like part of the validators was still controlled by the original multi-sig committee. And that was simply for, you know, because it wasn't like back when LIDO started, it wasn't possible for a smart contract to control Valdetta on the beacon chain. And so you had to control it with an EOA and the way that you did this was with a multisig. That was the only way to do it. And when it became possible, Lido switched to smart contract custody.

The smart contract was controlled by Aragon governance. So when it was possible, then Lido switched, they rotated, you know, all the EOA controlled keys, and then the Vultrasec has been retired since then. So they have made, you know, further steps towards decentralization. However, the degree to which, we think we call it ossification, right, to the degree to which this is possible, is not just limited by the amount of, you know, like how much Lido DAO wants to work on this, or like how much resources it can allocate, but it's also severely constrained by the underlying base layer. And so, for example, while the Ethereum, for example, you know, withdrawals wasn't finalized from the beacon chain, it was actually impossible for LIDO to make any ossifying steps on in the protocol side.

So I think these are definitely like these timing concerns. I think what I'm driving at is this is very, very nuanced. Like you, you need to unlock certain ossification on the layer below. So you can unlock ossification on the layer above, which is the staking layer. And at the same time, what you cannot control is the demand for staking that comes in, which is actually like unbound from this.

And when you say that, for example, a decentralized protocol, And I use like decentralized now here as it's not where it wants to be, but it's way more decentralized than, you know, let's say Coinbase and Binance. And you say, well, it can't grow above, let's say 20% because it doesn't have feature XYZ. And then you look at, you know, the top 2 exchanges, and they don't have any of the same constraint. It just doesn't make any sense. And so yes, I think that decentralization is incredibly important.

A lot of the roadmap that the Laibo DAO has laid out for the future of the protocol is centered around decentralization. And it has been now for a long time. And that is because decentralization for many systems, it can be seen as a form of sacrifice. Right? Like, in Uniswap, for example, you know, you you sacrifice a lot by having to run your system on Ethereum, in a smart contract, you know, very inefficient kind of curve design.

And like the cost of liquidity is very high. And if you if you could do, you know, essentially, a mid order book, then you could support professional market makers and would be way better. Right. So something around these slides. Yeah.

But in in staking protocols, this does not apply, you know, the thing that it does is very simple. And the more decentralized you can make it, the better it gets. Because it's so much about neutrality and trust. And this is why for, for LIDO, decentralization is not a sacrifice of anything. It's extremely good.

And it's, for that is the number 1 priority in the sense that it's not a defensive move. Like decentralization for Lido is offense, if you will. So that's why it is the number 1 priority. And so I am planning to post some suggestions over the next month, so that that lays out some of my thinking and my role as strategic advisor for what the DAO should do. And that includes 2 things in particular.

1 is taking steps to decentralize the governance, very specific steps. So I personally see dual governance shipping that I see that as the number 1 priority, because that already de-risks so much about what can go wrong in governance. And the second thing is getting the staking router going. So basically increasing the node operator set. Because when you have a permissionless staking module and you have a DVT module and you can get the node operator set from 35 to

350.

I think that also alleviates, I think, a lot of concerns. And it's not just that these 2 things will make an impact on their own. It's very similar to, you know, how in Ethereum, if you go back a few years, there was a general attitude, you know, they will never ship proof of stake, they will never ship scaling, and so on, right? It's just like a project that just, you know, makes eternal promises and doesn't deliver anything. And with Lido, it's not at all like that.

But I think there's always an element of the more of your roadmap you have already shipped, the higher the confidence and the momentum that like the rest of the roadmap will be delivered as well. And I think this trust is just very important. And so, yeah, I think for that reason, I think that's where Lido stands right now in terms of the decentralization roadmap. And I think those are the things that, you know, the DAO should do next.

Yeah, that's a great overview. And I think, so, with the staking router and DVT, you're moving towards permissionless validator sets, right? With dual governance, you're basically have an interim solution to mitigate governance attack risks before these contracts can actually be made immutable or ossified. I am curious to get your opinion on the hypothetical end state where this is the ideal end state or the happy path for Lido, where it is winner take all market structure and you have executed on this decentralization roadmap. Which functions will be left, will be impossible for the doubt, not to ossify completely.

And I think about whether it's how to distribute deposits across these different validator sets. I think about fees as well as it's going to be a tough decision to ossify a contract like withdrawals, where if you, you know, maybe ossify too soon, you could have a catastrophic situation. And so I'm curious to hear in that end state, you know, what will be left for the DAO to really control? And, and in particular, will, will they get will there be pressure to ossify fees?

Aha, Yeah. So I think the question like what is going to be left is a little similar in spirit to a question that I sometimes also hear, like, in order for like, what should Ethereum do about liquid staking? And should they make like an in protocol liquid staking solution? And like, if you squint, the 2 questions are actually the same. Because we're also asking, well, how can Lido be ossified so much that it basically becomes a part of the Ethereum protocol.

And you realize that there is actually like, let's say, Ethereum would put liquid staking into the protocol today, you know, and there's 2 problems still remaining. 1 is the stake is not fungible. So every validator would create its own liquid staking token. And you know, you don't have any of the liquidity aggregation effect that actually makes liquid staking so effective. And the second thing is, if you want to have the liquidity aggregation, then somebody needs to pick who what node operators are receiving the stake.

And if you want to be very simple about this, then you can say, well, just let the stakeholders decide. And, you know, then there's a list and, you know, people sort it by the name that they trust, or the amount of revenue that they can generate. Doing that, I think then basically what happens is the competition between node operators, it becomes concentrated in 1, at most 2 dimensions. And that is actually extremely centralizing for the underlying node operator set. Because basically, the stake goes to the parties that have the best brand.

It goes to the parties that take the most risk, or it goes to the parties that have the lowest cost of capital. And, and or kind of where the stake is secured in like the safest jurisdiction. And proof of work is a good example. What happens if you let the kind of free market, quote unquote, for validation play out? It used to be, mining used to be entirely in China, because that's where the hardware makers were and the energy was the cheapest.

And now it's almost entirely in the US because all of the mining companies are now public because that's what gives you the lowest cost of capital, you know. So you would get basically a very similar outcome in staking if you just follow the incentives. And that is why actually, I think 1 of the huge values that Lido has or the DAO has for Ethereum today is exerting some amount of governance over what a good node operator set should look like. So for example, LIDO has quotas for geographical decentralization. For what amount of stake is allowed to be on any particular continent, in any particular jurisdiction?

How much is allowed to be controlled by 1 legal entity? How much is allowed to be run with a particular combination of different clients in order to drive client diversity? How much is allowed to be run in a cloud hosted setup versus an on-premise setup? And different things like that. And so due to this governance over the node operator set, Ethereum today has actually a much, in many ways, better node operator set than it would have otherwise.

And so I think what this answer kind of hints to is that some amount of governance over the node operator set will always be necessary, because you need to match the demand for delegating stake with the demand for staking, you know, actually like running these validators. And you need to do this in a way that satisfies the objectives of the underlying network, which is Ethereum. And so at the very minimum, I think the DAO has to vote and it can do this very infrequently. But it has to vote on the objective functions of the node operator set. And I think Beyond that, you can, you know, ossify almost everything, but I think this particular part will probably always require some amount of fine tuning.

It can be infrequent, and it should have the full buy-in of the stakers. And so I think this is a realistic place for Lido to get to.

I want to just draw a line under the point that you made before about decentralization, not anting antithetically to the product, but actually being offense. It's a core part of the value proposition of the product. And I would sort of draw a comparison in between Lido and a couple, Ethereum is very picky about what it takes into the actual protocol, right? It doesn't like to have an opinion about a lot of stuff and it outsources actually very critical functionality of the protocol. Lido could be an example of that.

Flashbots could actually be an example of that, right? Ethereum outsources its block building. And I think actually, Eigen layer or restaking protocols have the potential to be that as well. And I think it just bears repeating that that hadn't really made that connection until you just, until you just said what you said, but those are very interesting, those are very interesting group of protocols that do very core functionality for Ethereum. And the example might be, Miles, this is what we talked about in our first episode, but it could be something like utilities.

It could be something like the banking system or health, things that are very close to the metal or the social contract or infrastructure of life. And what I want to make sure that we've at least got time to touch on a little bit is this idea of restaking. And I know you have a little bit mixed feelings on restaking. We're going to be talking to Sriram later in the season as well and sort of hear from the horse's mouth. But 1 of the ideas that I wanted to, because I've heard you talk about the principal agent problem and that definitely plays when it comes to liquid staking in general.

And I think you might add on an additional layer or potential risk of principal agent problem when it comes to restaking. So can you just give us the high level of how you think about the interaction or intersection in between restaking and liquid staking that we can poke out of the bed?

So restaking is the idea that Ethereum's validator set is an important asset, or like a valuable asset in itself. And you can leverage the set to perform other activities by having these validators opt into additional slashing conditions. Rather, they opt into performing additional activities and they can then be held accountable by opting also into additional slashing conditions. And so a few use cases for that have been put forward. I think they go from you can run kind of protocol like data valid, like if you invalidate us also providing data availability, For example, also sequencing layer 2 rollups.

They could provide oracles, you know, bring prices on chain. I think these are just 3 examples of what you could do with this. I think it definitely has its place. In terms of narrative, it feels just a little overheated right now. And I think you can basically see this in when EigenLayer launched, you know, the contracts filled up a stake right away, but there's no 1 on the other side who actually wants to use these services and would pay for it.

And I think we're quite a bit away from actually seeing any revenue generating activities from this. And yeah, I think When it does, of course, it will become important. But I think, yeah, we will have to see. And on the other side of that, we talked about the competition between staking protocols. So definitely all staking protocols compete on providing rewards to the user.

And you don't have to be the best 1. Like I think you can always boost your rewards by taking more risk, which is not desirable. So it's about having a good ratio of risk and reward. But to the degree that restaking will drive rewards to node operators, then the stakeholders will want to participate in these rewards and a protocol that uses restaking, it will improve its differentiation in the staking market very clearly. So I can tell you restaking is something that all of the staking protocols definitely have on their radar.

Today, there are 2 ways of doing it. I think 1 is Eigenlayer can, you know, they can take what is already kind of a staking protocol, staking token and you know, stake that and then kind of use that or the staking pool itself, they can run, they can implement Eigenlayer or another restaking protocol, right? And so I think this is another strategic questions that, you know, DAOs behind these, or the operating companies behind these liquid staking protocols have to have to think about. In terms of timeline, if I have to, my personal timeline is, it's probably between like 1 and 2 years away of becoming like, a serious differentiator in the staking market. So that's why I'm personally more focused on executing the decentralization roadmap.

I think that's more important today than anything that can boost the rewards for stakers.

So you just laid out a really interesting relationship in between. There's sort of a co-op petition relationship between something like Lido and something like Eigenlayer, right? And I think a big question of that is who is ultimately upstream of whom and who has the relationship with the user. Is that roughly how you view it as well? Or maybe talk a little bit about that dynamic and if they have the potential to be more friendly and complimentary or if it ends up being a struggle to who has the user relationship.

I think that's a good way of thinking about it. I mean, today, they are clearly very complimentary, right? Because the liquid staking protocols want higher rewards. And what a protocol like Eigenlayer wants is access to node operators and users. And this is what the LiquidStacking protocols have.

So it's a perfect match in many ways, assuming that there's demand to use the LiquidStacking protocol. And that's, I think, that's where it hinges today. Yeah. So I think they've been incredibly successful at bootstrapping the meme that there will be a lot of demand, but whether there will also be a real demand to fill that expectation. I think that's to be seen.

To the degree that they can, I think the relationship can be very complementary? I think it is possible that a restaking protocol will eventually also look into, well, how can we integrate vertically and like run our start running our own staking protocol, or a staking protocol might eventually say, well, why don't we internalize some of the restaking? But I'm not so constructive on these moves, personally. And that's for a few reasons. I think 1 is that protocols in general, and this actually goes to what you said about field smiles.

And I realize now that I didn't answer your question directly. Protocols, I think the thing about decentralized protocols is they can get much bigger than anyone thinks they do. And if you think you know how big they can get, they can get 10 times bigger. That's my perspective. And they will probably have very small margins.

Yeah, so and when you have a situation like that, you have a protocol that has very slim margins, there's very little to gain from internalizing that and vertically integrating, you're better off just collaborating. Especially if it's in a market where kind of customization, like you don't need to control the whole user experience. And I think like crypto is actually a pretty good example of that. You can have a pretty good user experience just from composing with other smart contracts. And the second thing is, it is related to the particular strategy that Lido DAO has, or operates and, And that is to become a neutral middleware, i.e.

Kind of a very, very thin protocol. Like actually like the things that Lido DAO should do should be minimized as much as possible. And so the protocol must have an incredibly small, what I would call management surface, you know, so because that surface is what then the DAO ultimately has to manage, right? And that's why I, as an advisor of Lido, I'm very, very against any form of vertical integration. And when somebody says, shouldn't we do XY?

I'm like, no, forget about it. Like, don't do vertical integration. It is not right for LIDO. Like LIDO should be as thin as possible and as neutral as possible. You don't want to manage anything.

You don't want any complexity. And that's how you need to play in the staking market if you want to survive. And it's not even about winning, because if you don't win, then you don't survive. Like that is the nature of the staking market, because I don't think there's an equilibrium where you can have 20% market share, or 30% market share. This is why I'm like so against self-limiting because if LIDO doesn't keep growing, then it first stagnate, and then it will die.

And I think this is true for all liquid staking protocols. And so that's why I'm very pro collaboration and trustless interfaces between these protocols. I'm very anti vertical integration. And that's why I'm also very pro and now we finally get to the fee question. So I would be very I don't I don't think that 1 concern that people have about protocols that get very big or companies that get very big, that have that again, outsized bargaining power over their users or their suppliers.

And then what happens is they start increasing the prices. Like Apple taking 30% fee from the App Store, which you can argue today. It's clearly okay for them to do it because, you know, people can still not leave and get the same value elsewhere for cheaper. But I think 1 promise from decentralized protocols is that they cannot behave in this way, that the value that they can extract that they can capture is actually limited at a much smaller size. So I think definitely fees should be kept.

I'm not even sure that they should be subject to management at all. I think Lido could have an internal fee market the same way that Ethereum has an internal fee market today. Ethereum is not opinionated about how much gas should cost on a given day. Right. It only defines what is the supply of it, for example, that is allowed.

And I think Lido will get to a similar point, where the fee that the users pay and that the node operators receive is simply set by supply and demand. And as long as the node operators adhere to the objective function, that is very, very infrequently set by the DAO for what the notarized asset should look like, then the price discovery between these 2 parties can be left to the free market, basically. And the DAO should have no ability to jack up the prices, and say, this is how much users should pay, and this is how much notepadders should receive. I don't think it should have that ability at all. I think that the fee that they can get, that they can take.

It should be small and it should be fixed. Because this is what gets both users and notepad us and all of the protocols and integrations and so on that are building on top, it gives them all of the confidence that they need to treat this as really kind of long-term, reliable and immutable infrastructure.

Yeah. And I think that makes sense. And I think there's an, you know, an obvious concern about, you know, the demand side fees and, you know, Lido's ability to rack those up to Apple Store like rates once it has a monopoly. But I do think experimentation on the supply side, fee split between the protocol and its validators could actually be interesting to explore in making the protocol healthier. My gut sense is that the largest validators of the Lido set right now, the corest ones of the world, the P2Ps, would actually take a smaller commission than 5% in order to receive larger deposits.

But as you're trying to bootstrap, say these permissionless validator sets via the staking router, they're going to need more than 5% potentially, or maybe whatever that revenue that you, higher rank for the largest validators could be used to bootstrap their collateral. So I am thinking from your strategy seat, how do you view like, because that's an example of maybe not being super thin, right? You can be very active, but towards a, you know, a healthier outcome.

That's why I think a staking module should have the ability to set their own fees, basically, right. And then there can be competition between different modules. So you're totally right. I think, especially like if you imagine from here that the price of Ethereum increases by 10x, it's not impossible, right? And then what is the margin of the average node operator?

Like 97% maybe. So the cost for them of actually running the hardware and infrastructure is very small compared to the fee that they are then getting. And it's very clear that this is not, the market equilibrium nor is it optimal for users or for Lido. And so I think in that case, it's very good for the protocol to have a price discovery on what is actually the cost of providing the service of running validators. And it can also work in the other direction, right?

So if, if, if Lido goes, you know, if Eve were to go again to like a hundred dollars, then Lido must also be secure, right? And it can't all of a sudden be that Node operators are not incentivized anymore to perform their services because there is a fixed cost to running, you know, to running these Node operators. And maybe if Eve is 10%, then you know, the equilibrium for not operators on the beacon chain is actually like 2% of the network. Like, I don't know, it's possible. In either case, it should be possible that not operators also can increase their fees if their cost goes up, or their revenue goes down.

And another example of this would be like a bonded node operator who is permissionless and who puts up some of their own collateral but doesn't have a reputation, at least at first. They should say, I mean, I would participate in Lido, but someone needs to pay me for this cost of capital that I'm incurring because otherwise I couldn't be competitive. This would not be profitable for me. And so they should ask for a higher fee. And so I think it's very important for that reason.

And I think it's part of the vision, also for the staking router, that different modules can set their own fees. And then there's competition between modules. And there's price discovery on what the best supply side fee should be.

Hasu, unfortunately, we could keep talking for, that's the mark of a good conversation that we had about 5 topics we didn't even get to here. And we could keep going on forever here, But you've already been really generous with your time and just really appreciate you having back on Bellcurve here. So Hasi, this has been a ton of fun. You've given us a lot to think about and really teed up a lot of the questions that we wanted to ask for for this entire season.

Yeah, Hopefully not the last time. I'm really looking forward to what's the rest of the season. I think you have a definitely a knack for topic selection. First, you know, AppChains and then MEV and then liquid staking. I think you're definitely at the frontier where it's most interesting.

Thanks, sir. I appreciate that. All right. Well, Miles and I'll try to do the subject justice. And yeah, thanks again for your time.

All right, Miles. What an episode there. So many great thoughts from Hase. It's almost tough to digest and figure out where to start.

I know, I know. I think we covered a lot of what we covered on the first intro pod, but just to click deeper and really interesting to hear the perspective of somebody sitting in a seat as the strategic advisor, the light of doubt.

Yeah, I agree with you. I think 1 of my 1 of my takeaways, the question of market structure was, in my opinion, answered about as directly as you're ever going to get it done. I think Hossie explained a lot of the ways and he was thinking about some of the network effects and returns to scale of a liquid stake in token issuer, similarly to how we were. So I really, I thought the discussion started to get really interesting when we started to talk about, you know, he thought about it exactly the same way that we did, which was he drew out these, these 2 sort of principles, frankly, of DAOs and decentralization that we talked about a couple of seasons ago. But it's kind of these 2 ideas, which is 1, Lido sits in this very interesting camp of DAOs where it's very close to the base metal infrastructure of Ethereum.

It's basically Ethereum deciding to outsource a core function of the Ethereum protocol itself. And so decentralization is, I love the way he phrased it as being offense instead of defense. And then the other thing that is on the roadmap for Lido and sort of indicative of how they think about it is limiting the surface area of what the DAO does. I thought that was fascinating. So then there are a couple of ways to go about this.

Part of it is ossifying core parts of the protocol, but then it's sort of a mentality of limiting itself to whether that be operating on Ethereum or not vertically integrating or whatever it is. It's just a very interesting perspective. I don't know what takeaways you had

from that. I couldn't agree more. I think we kind of wound back the clock before Lido existed. And just thinking from first principles, his conclusion was that there will be a winner take all of this market. And so the best thing we can be doing right now is making sure that whatever the winner is, is as decentralized as possible and the best version, the healthiest version of the network as possible.

And I think the sequencing of the decentralized and the protocol is very important, right? Because those training wheels were necessary to compete with the centralized exchanges. But, you know, to your point now, decentralization is now a growth unlock for them, right? Whereas with more typical protocol DAOs that are building more applications, user-facing applications rather than something that could be considered infrastructure of Ethereum itself, that same sort of decentralization actually is a growth hindrance at times. And so now, you know, it's at the stage of, okay, we have gotten the early lead.

Now it's time to execute on the, you know, decentralization road roadmap as a growth unlock, right? What's going to get Lido from 30% market share to 70% market share? Well, it's going to be getting the entire, all the stakeholders of the network comfortable with their winner take all, you know, market position. And to do that, you need to basically start making the protocol as hard to change as Ethereum itself, but being very strategic in your decisions and maybe even interim solutions like this dual governance dynamic.

Yeah, we didn't get too deep into dual governance and Steeth. Our next episode, we're going to be diving into the nitty gritty of how that all works. I did think 1 core idea to call out was, I've always thought 1 of the most elegant parts of Ethereum's design is how it's tackled fee markets. I think the way that it works in Ethereum is a really elegant solution. You're starting to see Solana where there was a fixed fee potentially move more in the direction of fee markets.

I had never really thought of the staking module for Lido introducing the possibility of a local liquid staking sort of fee market. I thought that was such an interesting takeaway and it hadn't even occurred to me. So I thought that was a very cool point.

Yeah. I think it's interesting to hear that the last part of the protocol that probably will always need some sort of DAO involvement is the curation of these of these validator sets and also, you know, how deposits are distributed across them. I thought his point around, okay, well, what if you just let the stakers decide? Wouldn't that be the best possible way for the network? Well, problem with that is that it actually would end up with a less decentralized set because stakers' natural incentive is to go with, you know, the validators that they know will not going to lose their stake, right?

And so, again, this curation is something that, you know, Lido has been criticized for, but turning it on its head going forward, once there is permissionless sets, you know, that curation is going to be necessary in order to actually bootstrap deposits for those new permissionless sets.

Yeah, it was a super interesting point. It kind of got me thinking, Myles. I didn't want to ask... I never like asking about questions like, what's the potential return for something like this. But I'm increasingly starting to see this connection in between to use, let me use a really high level, non-specific metaphor, but then try to get a little bit more concrete with it, is these sort of levels of different corporations in America and the function they do and how critical it is to sort of human fabric and the social contract.

So initially, you know, utilities was the example that we used, But I think you could largely actually lump defense, you could lump banking services, you could lump healthcare, these things that you can't really just leave totally to the free market. There has to be some degree of regulation. And then the impact that has on returns of those various industries. And highly regulated industries, they tend to be more predictable. There's less upside because you couldn't, for instance, for the example that you used, you couldn't implement fee surge pricing for heat, for instance.

That would be an extremely unpopular thing to do. But because it's regulated, it's much more dependable and predictable. And Hasu, the way he described this primitive of Lido as a liquid staking provider, it almost takes it 1 step further. And you're isolating by limiting the surface area and making so much of this just ossified smart contracts, you are limiting the surface area. You're increasing the TAM, but you are decreasing probably the rake that you take there as well.

So all in all, I think what this ends up looking like is kind of an annuity, something like an annuity, right? A very large total addressable market, winner take most with non-punitive fees. So what you're left with is sort of this stream of income. And then I guess it just depends on how much that stream of income is bid up or down. But you could sort of look back at historical periods of time, even in the US, at various times, annuities have been in, they've been out.

But that's kind of the framework that I'm starting to look at a lot of these protocols as. I'm curious what you think.

For me, it's still a we'll see. Because as I was kind of alluding to, I think that if you were to take a super active role in trying to drive revenue to these protocols, you could actually do so in a way that you could increase the rate without necessarily increasing the rate from the perspective of the user, right? So you could, right now, basically, LIDO fees are set at 10% for the user, and that 10% is split between, you know, equally between the DAO and its validators. And I think there's some very interesting, you know, kind of auction dynamics or market dynamics you could introduce to how that 10% is split up among validators in the DAO. And I think that, you know, the protocol could actually increase its rate if it wanted to by taking a higher cut and giving the largest validators a lower cut.

And that is more of a kind of a capitalist sort of approach here, right?

It is. But just not to interrupt, but I mean, how much can you really do that? It's 5% or something right now. You could move that, what? You could move that a couple hundred basis points.

There's no opportunity to go from a 30% to a 50% sort of gross margin. The point I was just trying to make is that there's the opportunity to exert some influence leverage on your supply side, and maybe you can increase your rate a couple hundred basis points, but there's no opportunity to... The sort of flip side of not vertically integrating, not trying to reinvest those gains and trying to be a very small surface area, sort of flat, credibly neutral protocol, is you don't have any opportunity to go from a 30 to 50% gross margin kind of thing or invest in all these new and exciting business lines. I think that's the flip side of that sort of point I was trying to illustrate.

I think that's fair. I do think that the returns could look a lot better than an annuity if we end up with these small margins, but in a winner-take-all end market structure, I mean, it's going to be very, very valuable. Right.

Yeah.

But I in general agree. I think it's definitely a very different, you know, approach than a typical business, given you know, what dynamics are required to kind of, you know, get that winner take all market share.

And actually to that point, I think Hasu is absolutely right in just saying that everyone tends to understate the market size here. If I'm not 100% sold on the idea of Ethereum as money, but it might just be 1 of those things where reasoning by analogy is inherently imperfect. People use Ethereum for money like things today. They probably will in the future. I increasingly think a helpful metaphor to understand the relationship between Ethereum and Steeth, for instance, is dollars and state dollars, which is bonds.

And if you look at the global market size for US dollars versus US treasuries, it actually is about 1 to 1 and the market for treasuries is larger than the market for dollars because there's more demand for yield bearing money compared to non-yield bearing money. So it's a very fair point. We are probably drastically underestimating how big this market's ultimately going to be.

Yeah. Yeah. And despite treasuries not being necessarily fungible, you know, there's plenty of liquidity for those markets and ETFs. You know, maybe the only counterpoint there that I could make is that, you know, Steith, right, is completely fungible, whereas a lot of U.S. Treasuries are not.

And so maybe it's more it's, you know, a stronger network effect there. And it's, you know, maybe less than 1 to 1 and more swings in the way of the yield-bearing asset. But I think you could be right as well.

You never know. It's tough to really gauge this stuff out. And then last thing, we sort of ended the talk talking about restaking. Hasu was not super bullish on that. I think his primary point there was just that there's not an enormous amount of demand on the demand side of a two-sided marketplace like Eigenlayer.

There's actually an enormous amount of people that want to restake their stake-deef, but there's not an enormous amount of node operators that want to run hardware for XYZ random chain. And for that reason, he wasn't super bullish on it. I did think it was interesting to call out the sort of co-opetition dynamic in between Lido and Eigen layer. And I do think it goes back to our original AppChain thesis point of who has the relationship with the customer and who's upstream of who, because there's a natural push to vertically integrate there between those 2 providers.

Yeah, I think that's interesting. I think there's 2 parts of the supply side for restaking. It's the stakers who's, you know, especially liquid stakers who are not running hardware. Then there's, you know, the validators that they've delegated their stake to. And I think the stakers would love this idea of additional yield.

The validators are saying, well, actually, you know, that marginal yield comes at a significant additional cost for me, whereas for the stakers, it's no additional cost. But what will be interesting to see is just how much demand there is from these roll-ups, sequencer sets, oracles, bridges, and at what point this is secure enough to get some of the larger players here and not just the ones that are looking for, you know, a narrative to kind of latch on to, right?

It was it was interesting.

I think we'll,

we'll have to 1, the 1 thing it's almost a philosophical answer and I'm glad some of our responses were more tactically oriented and sort of history of liquid staking. But I think it's a question worth posing of, is it a good thing to limit like Ethereum? And I would at this point be more in the camp of I think there are more creative solutions to implement than some sort of cap. I don't think it's a great message to send to entrepreneurs. There's actually a sort of a rich history of monopolies.

We could do a whole podcast on the histories of monopolies and antitrust and things like that. But starting with Rockefeller, probably even before, there's this rich history of entrepreneurs coming into an early stage, very wily, up and down sort of market and actually helping to tame it with the form of monopoly. And then there are sort of natural undermining effects that lead to more diversification later on. John Rockefeller and what he did with Standard Oil is actually a really good example of that, where it was a boom bust industry that was impossible to make any returns in. He came in, standardized everything, and then the monopoly gave birth to 1 of the largest industries in the world.

And there may be something that we could learn in crypto from doing that.

So I

think so.

All right, Myles, this next episode is going to be really good. We've mentioned this a couple of times now, but we're actually going to be talking to Izzy from the Lido team who's in charge of their node operations. So he's the perfect guy to be talking to about nitty gritty questions around the staking module, dual governance, and how they think about decentralizing themselves. You all can tell today that that's an enormous part, not just of kind of a technicality of trying to align and seem decentralized. It's actually a core part of the value proposition for these liquid staking protocols and it's how they go on offense.

So I think that'll be a great conversation.

Yeah, really looking forward to it. It was a blast chatting with you