What Ethereum Can Learn from Cosmos' Liquid Staking Module | Aidan & Felix

Transcript

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 uh 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 [Music] all right everyone welcome back to another episode of bell curve miles today we're going to be getting the cosmos perspective we're going to be talking to Aidan Salzman of stride and Felix lutch Felix as as he we refer to him throughout this episode is sort of The Godfather of liquid staking as the author of the original delegation

vouchers paper that you and I have referred to several times throughout this season from 2019. so this will be a really fun one I'm excited yeah we get kind of a history lesson of really the origins of liquid staking from from Felix um who's had a great perspective because course one's been very active in both Cosmos and Lido on the ethereum side and then Aidan who is really building on The Cutting Edge of liquid sticking protocols today um and can kind of provide a preview of what other liquid staking protocols might be thinking about in a few months or years down the road when some of these ideas inevitably you know translate over to other ecosystems and guys when you're when you're listening to this what I want you to listen for and we'll talk about it in the recap a little bit is this tension input between

flattening out the governance of liquid staking protocols and allowing users to decide when the risk there is that maybe you get sort of an adverse outcome in terms of what your validator set ends up looking like or liquid staking protocols actually being a permanent stakeholder and having an opinion which could lead to decentralization of the validator set so I want you to listen for that tension we're going to talk about it during the actual podcast and then recap I also want you to listen for the parts about fungibility of the protocol that comes in right at the end but that I think is an extremely important part of the episode as well pump to get into it hey everyone we've got a great episode here but before we do I just wanted to give a quick shout out to permissionless this is the biggest and best conference in all of D5

it's the one that we do with bankless who's a great partner for us last year we had almost 7 000 people there in West Palm Beach we are moving this year to Austin Texas from September 11th through the 13th and if you are a listener of bell curve any of these last five seasons this conference is basically custom made for you we're going to be talking about liquid staking the theme of this season we've got a bunch of great panels on Mev if you listen to the app chain thesis we've got a bunch of Cosmos folks out there in full force we're talking about the converging architecture of Solana the roll-up space in eth and Cosmos so I would love to see all of you there and to reward you for being such great listeners to bell curve you get a special 30 off code it's bell

curve 30 that'll get you 30 off tickets click the link in the show notes and then head over to the permissionless site and make sure that you get your ticket today again that is bell curve 30 click the link in the show notes all right everyone welcome back to another episode of bell curve today miles and miles and I are joined by Felix lutch and Aiden Salzman guys welcome to the show thanks for having us this is going to be an exciting episode um where we're going to dive into some of the big questions that we've been asking about liquid staking uh from the perspective of what's going on in the cosmos ecosystem and as per usual as miles and I note in this show oftentimes some of the big questions that are being asked in ethereum uh the cosmos ecosystems tends to be six to 12 months

ahead so hopefully we're actually going to get some answers uh and some early attempts at experimentation for solving some of the big things we've been talking about this season but before we get into that we've referenced several times this season that the first person to post something publicly about liquid sticking was in fact Felix lutch so I don't want to call you the The Godfather Felix of liquid sneaking but uh one might one might describe that title to you so can you kind of take us back to that blog post that you wrote and what you were thinking when you sort of outlined delegation vouchers and kind of the history of liquid staking and then any surprises in terms of How It's evolved up to this point right yeah totally and yeah again thanks for having me on really been a long time coming so

this post I think is at some point in 2019 right or so um we've been kind of circling around liquid staking since like four years now and it really I think got like a lot of traction this year obviously with light already last two years but I think now with you guys like covering it there's there's a lot more eyes on it and a lot of people like interested a lot more um you know protocols being built around liquid staking at that time when we wrote the post it was more that proof of stake was basically just emerging and there was the the sort of Delegation model in Cosmos so you had this concept of okay you have the validators and then you have the token holders that like sort of just delegate their votes to to

validators and and this sort of principle existed in Cosmos already but also in other networks like tesos and given we were like one of the first staking providers we were kind of looking at you know how what's like the similarities between these models or how can it can it be evolved and at the same time while we were writing this actually there was basically the D5 summer in ethereum starting to go on and uh being in Berlin I was in in full node with Brian uh CEO of course and we were like kind of observing some of the initial D5 protocols and their designs so like one notable one is like compound which has this idea of the C tokens where

essentially you deposit your tokens and you receive this this C token return that kind of represents your position in the compound protocol and we realized oh this this is very similar to what you could do in in proof of stake with um with the validator delegations essentially and we foresaw that there would be like a sort of merging between D5 and liquid staking at some point since sort of like I guess what we were always thinking is that staking is kind of the base layer of of these networks and it's kind of like a financial activity and then people will want to like hatch their risk or manage their risk in various ways and and to do that you will require like some sort of more flexibility than what the the core protocols often offer and um

as like sort of through that bravestorming we were part of this hackathon in uh hack Adam actually in Berlin like a very cool uh one of the earliest Cosmos hackathons and um worked together with sunny and death from uh back then Sika now osmosis obviously on this concept of Delegation vouchers so basically just taking it one step further from just having the delegation structure in the code but actually turning it into like a token that when you when you delegate you receive like a validator specific liquid staking token let's say uh and that you could then always like convert back um uh according to this mechanism that's here on the screen right now for those that watch the the YouTube video and um yeah that's really where it started and

then I sort of realized oh wow there is like such a big design space there uh in this in this world and we saw like slowly like projects emerge like some of them including like the stakedown Julian's uh idea for taser there was also um stator and some some other projects that are all covered in in the liquid sticking report that I later wrote so I was kind of realizing oh we need like this is a big topic we need we should be making like a working group to kind of talk about this and figure out what the right way is to like Implement liquid staking or where it's going and what the risks are and that ultimately turned into the liquid second report that is then later I think probably like a year later I released

um and so there I go into like sort of the benefits of liquidation but why it's needed sort of like the comparison to staking with centralized exchanges and the powers of centralized exchanges and why liquid staking is sort of potentially Force for decentralization but also the risks in terms of like what happens when you sort of over financialize maybe the the staking positions and and sort of what what are the problems that can emerge there I'm sure we'll we'll go into a lot of these in in the episode yeah um and I wanna I wanna bookmark and return to this idea of liquid staking as a force for decentralization like we're going to spend some time later in the episode talking about that but can you just kind of walk us through Felix at

the time what was the response initially to this idea because now we've all sort of taken it as this given liquid sticking is one of the few things especially over in ethereum that is really working and chugging along you know sort of despite market conditions but um you know I I imagine there was some pushback and we've had Dave on the program before and Dave has some very particular thoughts about liquid staking I would say so what was some of the initial feedback and then one question miles and I were just we're looking at this initial diagram that you drew um between the user the validator pool and the validator and actually we noticed that there was a conversion ratio of two atoms per one voucher so we're wondering if initially the idea for liquid staking was that you would

for instance in the Steve metaphor you deposit to eth you get one's teeth back and if you would actually thought that through just a small small question there we've been wondering yeah so what we were why this number is more just like illustrative but like it was clear that sort of because you earned the rewards the we sort of were thinking in delegations about vouchers of this model that's not rebasing so basically like the conversion ratio of one liquid delegation voucher would um increase like in terms of the pure atom so that's that's what this kind of represents like in in case of like let's say this delegation voucher system was running for a while then it might be that at some point you know one voucher

represents like two atom um so this was just uh for illustrative purposes there and then um I guess the pushback right I think at that time specifically most of the people in proof of stake what I would say were like protocol developers or very like deep engineering um people folks in general and so their thinking is like I I would say like pretty extremely long term around the health of the protocol and like sort of attack scenarios of of uh you know uh 33 or 66 attacks and then how they could come about or how an attacker might abuse the proof of stake system so I think there there is some pushback there that you know liquid staking might

introduce this sort of uh kind of too powerful thing that like all the stake Aggregates in this in the liquids in one specifically second protocol and and that this then sort of has power over the protocol um which is certainly I think like probably the biggest pushback um and then yeah many would probably also just say hey it's not really needed no one really needs this or something like this but I think in terms of like why what's dangerous this was is probably the main point and I think the the thing is that protocol I think at that time many were thinking oh we can sort of restrict this in the protocol and like sort of design the

protocol in a way to make it very hard to do this potentially but given we have like smart contract capabilities we have centralized exchanges we have so much infrastructure and the market generally I guess is my opinion and and sort of is proven over time that like kind of strives for this Capital efficiency there will be ways that can kind of circumvent these restrictions and those um will be very successful so I think like my personal opinion and I guess we might also get back to this is that in the end the protocol if it's like successful enough will have to deal with these uh issues and and might be better off like sort of enshrining it like as you say or like putting it native into the protocol um these liquid staking uh kind of

systems to to avoid having captured by some external party too much um and uh yeah I'm sure that's like probably like one of the quarters we're going to talk about so I'm really excited yeah and I think we can definitely get into a little bit of that later especially with the liquid staking module coming um which is essentially a version of what you just described but um maybe just to stay back in the air the 2019-2020 era um when he wrote this paper ethereum was still proof of work I think you know the beacon chain was on its way but it was not yet live um and so I guess to ask directly why do you think you know this concept took off on ethereum when it originated in Cosmos

and you know liquid staking is now alive and thriving and Cosmos thanks to folks like Aiden here um but it is taking a while um so yeah I would just love to you know as somebody that was deeply involved in both the cosmo side as well as one of the early validators of Lino um you know I think you have a great perspective to to kind of look back and do a retrospect here yeah but also almost like surprised me I think we were like I guess in Cosmos the main or many of the people in Cosmos are validators or like sort of have a deep understanding of proof of stake and like infrastructure versus on the ethereum where the actual like Community was more D5 maybe an nfts and and all these applications or smart

contract developers and I think we were in a way Cosmos was thinking earlier about this concert because many of the folks were like on that layer but then sort of the adoption from the applications is missing in Cosmos and I think like liquid staking relies a lot on having that to to be useful and and also that the difference being on ethereum that you actually don't have a delegation model so it's very hard to stay here if you're if you just have eat and you're like a normal person you don't want to run the infrastructure there's very few ways to actually engage and um like it's essentially light or like who's taking brought this model to ethereum in the first place and then obviously also the other point being that um given their the merge hasn't happened so like this kind of gradual rollout of

of staking of proof's sake and ethereum meant that um there's a long period where you if you stake normally like um you couldn't actually get liquidity on your staked assets and Lido uh was able to kind of provide some some liquidity for people that might want to withdraw through the curve pool and like incentivizing that heavily to like uh maintain the pack between SD and eth for the first year until the margin and and sort of that I think is a very big value proposition that just wasn't there in Cosmos and maybe still is less there uh also like now in ethereum it's less there but I guess now Lido already has like such a market share there's not there's other

fact or like other forces at play I think that keep it uh very interesting in ethereum also like again the the just greater amount of liquidity and more defy activity going on that so push this liquid sticking a bit more right so it kind of sounds like the perfect storm where you have a combination of you know higher barriers to entry to actually participate in validating um you have this you know lack of liquidity even Beyond a 21 day on bonding period it was you know at earliest a couple of years and then you know you actually have higher amounts of utility through you know in early D5 you know ecosystem where you could replicate some of the benefits of what you'd be able to do on a centralized exchange

whereas that in Cosmos just took a little bit longer to develop is that kind of a fair summary yeah yeah I think maybe also to add like literally from the like experience as a node operator trying to get started on ethereum you know we were lucky enough to like be connected with the light of folks early and sort of participating in that but in in terms of like offering uh staking as a service for ethereum outside of these protocols it's actually quite a hard task you need to build a lot of infrastructure like API or some sort of product that allows people to stake even just 32 each with you which is a very compressive activity and it's basically like an entire team that needs to build something like this and this is also something that why

there aren't that many staking products on if you're even from node operators right you need basically a team of like 20 plus to to be able to support such a such a product and um in in in Cosmos it's very different right you just register your validator you can be one dude in the in the basement and and you can receive delegation so that's uh much more um sort of easy yeah lower barrier to entry like you like you mentioned I think that totally makes sense and maybe just shifting gears a little bit towards I think we want to get your perspective on just the relationship between the liquid staking protocols and you know professional validators um because I think it's it's very interesting that it seems like you know Lido is developed by P2P um of course

one was very early in in getting involved there and you know you've uh were very involved in the development of Quicksilver on the cosmo side and it seems like you know as a you from the seat of a validator you can look at liquid staking protocols as you know at the time probably an existential threat um but also you know one of the best distribution channels that you could you could imagine right um it's just they take care of all the BD for you um and you have plenty of you know ways to use your these liquid staking tokens and you can just kind of get into the set and receive deposits but of course that's a much lower share that if you were to receive a hundred percent of those deposits

um like you would in traditional staking and so you know I'll actually pose this to both of you guys but you know how do you guys view the relationship between validators and LST protocols um today and and what was it like back then when some of these were you know just getting started all started in maybe I can add like his view from the other side I think my yeah our perspective I think what's interesting is right there is like this sort of conflict like you said when you're on the Node operator side where you sort of understand deeply the staking mechanism and I guess that led many to sort of get involved in liquid staking in some way uh also developing protocols there but then once they are there you have this sort of conflict that okay now I have to basically or spread my stake across these

different providers like in most of the cases right um and that sort of makes it less attractive you you would rather have someone like stake directly with you I do think that though given this this depends also on the size of the note operator for for many I think liquid staking is actually or like it's it's one of the best yeah distribution methods and also like force for decentralization where before that you basically have the foundations which basically bootstrap the the validators often so there is a foundation of a certain protocol they have maybe like 10 20 of the tokens and they would like pick the node operators that provide value to the ecosystem and sort of distribute the stake across them I think that was like the first entity where

which many um like smaller operators that don't have like Big Marketing budgets or BD rely on that that sort of um existed and then liquid saving protocols is sort of a extension of that where now it's also it's very focused on the specific stake in use case the foundation generally has like a lot of other um kind of things to do and it's also like not profit oriented so the liquid's taking protocol is more like I guess a product um that has its users or like the the yield and the safety of the program more in mind and it's actually I think doing often a better job than the foundation at like sort of picking the best operators and so uh from the perspective of the network I think that's very

that's actually desirable and then as a node operator um yeah again for many it's actually a very very nice um yeah uh way to get some stake and I think many like desire to be in in one of these sets um and um but then the very big ones they have like sort of this problem that oh now I actually would rather have people stick directly with me and there are a few um ways where how this can be achieved like for example stake wise V3 is interesting model that allows a little bit to the node operators to retain this ability by having these walls that you basically stick directly with an

operator but you can get still get a liquid staking asset by essentially collateralizing these validator specific positions and and that's I think an interesting model that that works for large node operators uh and still kind of gives their clients the ability to still like opt into liquid staking if they would want to so again I think yeah again General the liquid second design space is quite quite large and um overall I would say yeah very interesting for relatives but yeah yeah I think it's a really interesting question maybe one of the most interesting between uh on the relationship between liquid seeking protocols and violators um so when I think about liquid seeking my assumption is I I'm obviously very

biased as a founder of a liquid saving protocol but my assumption is that liquid staking will go to like 80 to 100 percent um there's so I think Felix and Miles you both brought up some really good points around why it took longer for liquid sinking to take off in Cosmos so there's this cold start problem with defy you know eth was very liquid you could use it um so D5 was bootstrapped and then steath demand was uh was there because D5 existed um like you said it's also easier to stay in Cosmos so it's uh if if you're a bit more crypto native maybe you just go and delegate your state to a validator um whereas on ethereum it's it's pretty

challenging to run a node um and there are also some technical reasons why it took a bit longer on Cosmos so the cosmos sub didn't have Smart contracts uh but there were actually places in Cosmos that liquid sinking did take off so in the uh Terra ecosystem which was a cosmos SC kitchen um I think B Luna launched uh it was a Lido LSD and it launched sometime in 2020 or 2021 but my assumption is that in spite of all of these these challenges in Cosmos having liquidity and yield is such an attractive value proposition for end users that liquid seeking probably will go to 80 to 100 percent uh so therefore it is it is a pretty important question for validators

but when I think about how validators and liquids taking protocols interact like what's taking protocols I think are just larger entities that validators can attract stake from so they can go and attract stake from uh little fish like you or I or they could go after a big well or they could go after a liquid-saking protocol and there are different strategies for validators to attract stake from each of these entities so although liquid syncing protocols might take a cut of the yields they can also in some cases attract so much more stake from a liquid-saking protocol that I think it outweighs that um but for uh for large validators in particular I think it's a bit of an interesting question because in

delegated proof of stake especially stake is skewed to the top of the set so I think uh in a previous podcast you and hasu were discussing this where validators can only really differentiate on a couple of Dimensions uh the main ones being brand and yield so being at the top of the set is a really strong indicator that the validator is good and most most users don't go and stake to the bottom of the set I mean we see this empirically like 90 of sake is with uh distributed across the top 10 validators in most delegated proof of chains um so in that sense liquid seeking protocols might take some stake away from the top validators and redistribute

it across the set um but if liquid saving Protocols are well designed then I think they can be distribution channels for good validators so it might be a little bit of a reframing from taking away stake from the largest validators to Distributing stake across validators that are really good because some validators at the top of the set are really good but there are also really good validators at the bottom of the set and then defining this question of what is good and rating validators on a scale is a real Challenge and I'm sure we'll get into this more later um but I would reframe it slightly as as a redistribution to validators that the liquid-saking

protocol has defined as good maybe that's decentralization um uh maybe it's geographical diversity there's all kinds of attributes you can think about I actually do want to push into that a little bit right now Aiden actually because I feel like this is a very important conversation that's happening again over in ethereum land with Lido and the staking router and that's what we're trying to move towards here right which is a more decentralized validator set that's what everyone sort of wants to see so can we actually get into that idea of you know how that the presence of a liquid staking provider like a stride might actually counteract some of those sort of Natural Market forces that

concentrates stake in the very top sort of echelon of validators yeah so I start from the assumption that on most blockchain networks ninety percent roughly of Network Power occurs to maybe the top 10 entities and I think this is roughly true for most proof of sake and proof of work chains if you look at Bitcoin there's large mining pools if you look at proof of shake stake chains uh specifically delegated proof of stake chains most stake is a cruise to the top 10 entities and maybe there's like a little bit of uncertainty here maybe a chain can really push on decentralization and they can get this stake distributed across 15

entities or maybe they care a little bit less and it's only distributed across five but for the most part uh steak does follow this sort of power law distribution to the top of the set um and I think this is where liquid staking protocols can really help decentralize the network so it is I think it's a big reframing as LSPs as a decentralizing force typically when we've talked about liquid sinking protocols we focus on all of the risks of liquid staking protocols and there are some centralization risks so for example if a liquid's taking protocol is hacked that would be really bad for the

network this particular argument has never uh really resonated with me given lots of things can be hacked in networks for example the core staking logic in a chain can also be hacked but there are definitely some centralizing forces and risks but if you look just empirically at how liquid staking protocols or good liquid-saking protocols distribute their stake across sets the distribution is typically more decentralized and then you know there's a question of how you actually Define decentralization but the sake is more distributed across a broader set of validators that hopefully are differentiated on things like

clients geographical diversity jurisdictions all of these things I think uh actually John um Carbono had this presentation at modular Summit where he was looking at uh Lido's stake distribution versus ethereums and like any normal Observer can look at this and tell which one is more distributed uh so I think there is uh there's a lot of potential for liquid sinking protocols to distribute steak much more broadly um across sets I think that's really important

um because the I think the Assumption or at least the narrative last year um and this starting to change is all about the self-limiting debate right like this is okay Lino's become too big at this point um but a lot of people just there's two ways to look at Lido you can look at it as one entity um because governance right is just one entity and it controls a lot there or you could look at it as you know a distribution of what is it now like 30 30ish right um entities that are actually you know operating the underlying stake and are Diversified geographically different clients right and I think that that does you know pose like a pretty Stark

picture when you look at Lido compared to the rest of ethereum um and on the cosmo side it's a little bit different because you actually have the ability right to delegate your stake to the validator that you want um versus ethereum the assumption is that the next validator will be some you know solo Staker that has their own 32 eth and that every validator you know every net new stake that enters the system will further decentralize the system which we know is actually not true um but yeah how would you how do you view kind of like this the difference between a delegated proof of stake Network and you know what we have on ethereum and maybe the role of of lsts um in in comparing you know the role of

lsts on both systems because I think that they kind of address different problems um because with the delegated proof of stake you give the users your choice but you know how that actually ends up looking um and it's very very transparent how it looks right but with on ethereum I think it's a little bit less transparent it's a little bit more opaque how these same forces actually do translate and extend over to ethereum but yeah I'll kind of post that question to the group um you know basically how do we shift this narrative on on both both ecosystems what I found interesting and I think is also on the modular summary actually or like I think somewhere was that you know some some people are like arguing essentially also for the

decentralized so sequencing right is also like a topic now where they think about okay we're going from a single sequencer to like decentralizing which essentially is like again pretty much the same if you look at it at a high level then any proof of stake Network and then I guess proof of stake is sort of the easiest kind of solution that achieves some sort of distribution because you're just saying hey let the market do its thing here it's like a financial thing and then you have on the other side of the spectrum maybe like a system where the governance decided and now whatever governance means I guess it can be that you just have a proof of authority chain where the the project Founders say Hey you five people run it or it's like some

token voting within that Dao of that project or the way it is in light or where all the existing operators have like one vote to like uh do new onboardings of operators or potentially even you know some algorithmic solution that tries to distribute the stake across uh operators and I think you know I guess for my side um yeah all these different models should somehow exist and I guess the the users will decide what's best I think it's also like kind of fazili's core argument last year on on HTC where Lido had like a lot of backlash from the ethereum community sort of oh you're a danger to the

ecosystem but then it's like hey you know each holders decide what's what's best you know and then I think that's I I definitely subscribe to that too so you know that these Solutions compete and um then sort of the best went out I think if you like kind of forcing one solution on on the system that will probably emerge like some some like sort of protocol that will try to kind of introduce differences and and I think that's actually healthy uh for the ecosystem at large so um for people that think you know our lighter we need to like shut it down I guess that's that would be my sort of framing that you know um it just because with Lido goes away

it wouldn't be that like suddenly like 30 homesteakers appear and and stake their youth but it would probably be some other entity that potentially is worse uh because like glider was actually you know everything's sort of Open Source and very um contributing so I think having that those um these forces is very important I have a I have a question because this is something I've been curious about this entire season which is so the Vasily core argument of the etholder should just decide right the you know uh we had sriram on an early episode of this season and he mentioned something very similar the demand side of middleware operators that want to opt into that system they should just decide the issue though is that at the current

stage of where we are at in crypto is I think if you just blanket allowed users to decide you'd end up with some pretty negative outcomes so I think what we're saying on this podcast is that actually liquid staking tokens part of the value prop that they can do is they can exert some sort of opinion on the validators right they're this really valuable distribution channel the state of you know per John charbonos uh discussion at modular Summit the validator set that Lido if you look at lydo's validator set versus Ethan General it's actually a sort of a source for good because it's more decentralized but in that you're implicitly admitting the sort of power that liquid staking token networks have

right and I think that is really the core Crux of the question and it's you know we're referencing John charb here he came out with another recent piece we're recording this on Monday August August 7th so this piece came out about an hour ago so we haven't had the full time to digest it but it's just very interesting again the I think it's very like a lot of the arguments that are happening specifically right now in roll-up land I think the correct mental model actually is Cosmos chains because they're going through a lot of this very similar sort of experimentation um and my question is what happens so maybe Aiden I'll direct this to you as the operator of a liquid staking protocol what happens if the decentralization

that you would want to see in your validator set runs counter to Market forces and what I mean by that is what if you actually have a market of users that want just the top five validators and no matter how many times you try to cycle through this idea of decentralization or promoting other smaller validators what your customers are telling you is I really just want to be with with the big guys here I mean how do you balance if Market forces go against what we all want to see in terms of decentralization yeah that's that's a really a really good question I don't think I have uh an answer to it but I'll give some thoughts so

I think one one the first thing I'll say is I think that is mostly correct I think n sakers actually don't really care about which validator they're delegating to except for these two things they want brand which I think you can just swap out for safety and then they want yield so really they just want like the best risk adjusted return across validators that they can get um and the the simple way to do this is to assume that the top 10 validators are the safest and then just you know like sort them based on one or two things such as uh such as yields

um but I think it might actually not be totally correct that the top five or ten validators really are the uh the safest um or the least prone to sloshing things like that so I think uh one way that you can sort of combat this is by introducing a new layer of abstraction between users and validators and there's there's different ways that you could approach this so one simple way that this is done today is through liquid saving you know when you mental liquid staking token when you deposit an atom or a theory or an eth into a liquid saving protocol you just get an st atom

or an st eth back and uh there are some liquid syncing protocols where you do choose a validator at delegation but for the most part you you don't um there are other ways where uh so I think osmosis has been working on a protocol where you can have validator set preferences so rather than saying like I want this validator you can express a preference that says I want a validator that is geographically distributed uh um or I want my stake to be distributed across validators that have never been slashed so there are preferences that that you can express so I want to get into the the liquid staking module here

but one one idea another question that Miles and I have been sort of we're currently asking throughout this season is I've are there parts of the protocol that you might be able to uh if this is a winner take all Market or a winner take most Market which which miles and I have sort of been asking and assuming that that's correct are there parts of the protocol that you could actually ossify so the way this gets talked about sort of over in Eastland is dual governance right and allowing uh steath holders actually some amount of of say over core decisions or the ability to veto essentially but but something else I found myself wondering evening this seems very technically complex so I don't know how you would actually do

this but is there a way that you could give some sort of guarantee that and ossify this directly into either the protocol or the liquid staking module the the design of it that you actually say you know we can maybe this is a good time to get into the liquid sticking module because there's a cap uh you know from my understanding um on the amount of staked atom that you can have but maybe you could actually bake in some sort of distribution and set of logic or rules that you know we will not allow the validator set to be you know above X concentration right is that something that you could actually design into these protocols so that you wouldn't have to necessarily trust Aiden the operator of stride but I could trust

the way that the product is constructed yeah I think it is something that you can bake in although it gets close to something that I think you can't bake in uh or or ossify so the only or the the most challenging thing to ossify and liquid sticking protocols is in my opinion the validator selection mechanism basically everything else can be fully ossified but validator selection is uh there's a very human element to validator selection one trivial way to think about it is as validator sets to change over time someone has to tell the liquid sacing protocol how the validator set has changed so let's say you know today we

say like validator a is in Europe validator B is in the U.S and then we have some notion of how to distribute steak across these two entities um in a few years if validators a and b are long gone and now we have validator C and D there has to be some human in the loop that is telling the liquids taking protocol what the attributes of these validators are so then I think you can sort of break it up into a delegation strategy so thinking about um you know how a liquid sticking protocol distributes stake across validators with different different attributes and then there's a second

question which is how do you um how do you assign these attributes to different validators and I think they're they're challenging in in different ways um the second question hasn't really been broached yet I think for the most part liquid sinking protocols assume that validator operators are honest in some ways like if they say they're in a just in a in a certain geography maybe they run some checks but you kind of trust them that they're that they're doing what they say um but I do think this will become more challenging Through Time and then the the first part of this which is can a liquid seeking protocol just ossify the delegation strategy I think it could but I don't know if it would be a great

idea um yeah I mean one reason is uh like the world changes through time so uh like for that simple reason you'd probably want your liquid saving protocol to be flexible enough to update delegation strategies based on real things that are happening in the world um and there's there's a spectrum of ways that you could update this it could be on the very conservative end it could be super infrequent or on the more aggressive ends um you know maybe it's maybe it's more frequent but even uh yeah so I I don't think it's it's possible to fully fully ossify a validator suction although pretty much everything else you probably can I think

I agree with that I would say maybe just to push back the only I guess path to ossifying that is that I see would be just to give that uh ability to the user and then find a way to you know wrap these non-fungible positions into a larger pool and then don't touch it um and I think that this gets into really you know part of what the liquid staking module is is trying to do um and it's probably a good time to shift there so I know Felix you you alluded to it earlier but you know there's always been this question of can we bake some sort of you know way to Mint liquid staking tokens into the chain itself um in a way that the chain is aware that

they are being minted um and aware of you know or maybe opinionated about how users can delegate that state um and the liquid staking module is getting at this and and I think you know just to give kind of a a preview of why we think it's important we've been talking a lot about you know disability on ethereum and with Lido if this could ever be you know an option for sneakers to delegate to a staking module of their choice that is more or less uh you know a version of this is being implemented right now in in Cosmos and so you know Aiden I know you've been very involved in the the development of the liquid staking module um and I would just love if you could

kind of walk us through what it is um you know what the what the flow looks like for a user and then where stride comes in I guess you know after the initial minting of this position and and I'm actually going to ask you to go for for folks who might be listening even aren't exactly sure what these modules are sort of in Cosmos Lane could you even just describe what a module is relative to Costco and then we can into the liquid staking one definitely so starting with this idea of a module in many general purpose smart contract chains the base layer is abstracted from most end application developers so you have this clear delineation of infrastructure code and application code

for example ethereum Solana any other smart contract platform there's one set of developers that works on infrastructure and then typically there's a mostly uh disjoint set of developers that builds applications on top of that and in Cosmos the lines are much more blurry so because uh there's this design philosophy of creating application specific blockchains where the infrastructure and the application are interwoven to make the application better um the technical architecture of uh of these chains is a little bit different so the idea of a module is just it's one specific piece of either infrastructure or application code that that lives on

the Chain so you might have a module for staking so when you stick to a validator some logic needs to happen under the hood and in fact there is a cosmos module which just handles sticking there's another module which handles slashing so if validators misbehave this code will uh we'll deal with slashing validators and stickers um osmosis has a module that deals with swaps so it has all of the amm logic embedded within that module so a module is just like one piece of infrastructure and application code that is bundled um and lives at the actual protocol

layer so it's not built on top of the protocol as a smart contract it's actually baked into the protocol itself and it it lets uh developers have a little bit more flexibility in how they build things and sort of a nice uh abstraction that's really helpful explanation there so that's just sort of helpful context now for setting up specifically the liquid staking modules then sort of like just miles question there could you just describe what that is and then maybe we can get into the user flow of like where the liquid staking module ends and where stride begins sort of so the liquid sticking module is one of these infrastructure application modules and the idea is that the liquid safety module will be built into the cosmos Hub itself

um you could build it into any Cosmos blockchain but the cosmos Hub probably be the first so I think we've talked a lot about why lsts matter in Cosmos um maybe one thing that we haven't touched on yet Cosmos has very low liquidity but there is lots of stake and stake rates are also typically much uh or staking yield is typically much higher on Cosmos so D5 is extra challenging not only is there the cold start problem but no one wants to use a token that has like a 20 or 30 yield as collateral so for this reason uh I think lsds can Kickstart defy and Cosmos but as Felix pointed out

there are lots of risks with lsts and especially for Cosmos which I think is one of the first um it was one of the first ecosystems to really engage with um with proof of stake um and specifically DFT systems on a really deep level I think there's a lot of skepticism around liquid staking tokens because it's it's taking something that is really core to the to the ecosystem overall so LSM or the liquid safety module attempts to solve two problems uh and it starts from a couple of assumptions so the first is that the risks associated with liquid sinking aren't very well understood um there are some logical arguments why

liquid sinking is good or bad but it's been pretty hard to quantify some of these risks I think there's you know Felix did a lot of work on this in the initial report but I haven't seen uh really in-depth rigorous analysis on this since then and it's been years um I think this is starting back up again um you know for example with this podcast lots of people are are talking about liquid seeking and trying to quantify these risks but today the fact of the matter is it's uh it's just not very well understood um and the the second thing that is a bit unique about the structure of Cosmos is once there's a lot of NATO's sake liquid staking is hard because stakers have to

unbond so it's going to take to to onboard lots of liquid stick everyone's going to have to unbond their tokens and forego weeks of yield and this is a really painful process so given these two assumptions uh or these are these are sort of the foundational assumptions that led to LSM um so maybe diving into how to best think about LSM I think LSM is best understood as a form of regulation on liquid seeking providers it creates a safety framework and there are lots of governance control parameters to modulate the safety framework through time but it creates a safety framework to help regulate the

adoption of liquid seeking um so this sort of addresses the first problem of the risks of liquid seeking or just not well understood LSM or liquid zaking module creates a framework to regulate liquid saving providers um and then second it lets stakers liquid stake instantly um and this might sound a bit uh sort of counter-intuitive like why would stride um be supporting something that uh is sounds like it's in direct competition with the protocol but the way that liquid the liquid safety module lets users liquid sake is it gives them a liquid sinking token that is an nft that is specific to the validator that they

delegated to so it does let users unbond their sake instantly but in return they get this nft that is not composable or really useful in D5 so if you think about the liquid seeking module I would say it's closer to infrastructure than it is application code it is built into the base layer of Cosmos and it does technically let users or stakers liquid stake but in return because you get this nft there's not a whole lot that you can go do with it if you take this this nft that you get you can't go to a lending protocol or osmosis to use those collateral or swap it because it's not fungible with with other liquid saving tokens

um so that's that's sort of a high level summary of uh why LSM was built and how I think about it and a little bit of what it is under the hood maybe actually what I'm wondering is is the Artie's nfts per validator fungible or like yeah it's a good question um sort of uh so the slashing risk is fungible um if if you and I both hold um a liquid staking token nft for chorus one and course one is slashed both of our tokens will be slashed um but they are nfts so uh they're not

totally fungible like the yields for my tokens goes to my address the old for your tokens goes to your address um but this I think this sort of illustrates the point that these things are really challenging to explain and they were designed with liquid sinking protocols as the end consumers in mind rather than users like you and I that's super interesting and maybe just before we jump to I guess where stride comes in um I'm curious if you could just speak to some of these risk parameters that you mentioned uh that are you know customizable through governance um you know I think going back to the self-limiting debate uh it sounds like that could actually you know something like that could be baked into the chains logic itself

um as well as you know making sure that some uh validators are not getting all of the delegations right that is they're minted through this liquid staking module and so I would just love to hear like kind of what are these what is the regulation and the parameters around it that that you're referring to here specifically like I mentioned there's really two components to the liquid second module the first is acceleration of liquid seeking adoption so I think we'll talk about this later this is where stride and the liquid safety module sort of meet um on the and then the risk mitigation is the second part of the liquid sake module and when the liquid sticking

module was first proposed there were a bunch of different types of risk that were talked about like I said none of these were perfectly Quantified but we along with occlusion Zaki the osmosis team lots of Cosmos ecosystem participants we're trying to think from first principles about what different types of risks liquids taking could introduce and there were three or four main categories that we thought about and the liquid seeking module doesn't try to address all of them but it does address some of them so two of them are governance risk as well as vote power concentration risk so governance risk is the idea that if a

liquid's taking provider gets more than say uh one third of the total stake Supply on a chain that liquid seeking protocol could actually halt the chain because Cosmos relies on um two-thirds of validator signing blocks for liveness so if you have more than one third that stop signing blocks then the changes halts and this is of course uh of course quite bad um so the way that this is mitigated is there's actually a limit on how much Network stake can be liquid currently I think the the limit in the signaling proposal was something like 25 and the

thinking was if 25 of all stake is liquid you would have to have 25 of total stake on Bonds in order for the liquid seeking protocol to have one third of the network so you'd have to have basically like a huge unbonding uh 25 of all stake would have to unbond for a liquid saving protocol to a mass one-third because so there's sort of this safety the safety buffer built in uh I I don't view this as like a long-term solution uh like I said I think the the incentives around liquid staking will drive it to maybe 80 or 100 adoption but it is maybe a good check on liquid shaking well we explore what the

risks are and how exactly we can quantify them um so I it's more of a I think a speed bump than a wall no it makes sense it's a number that both sides can agree upon right and and it's beneficial for for the liquid staking protocols because you get past this hump of or this barrier which is the unbonding period and it's you know something that the other side that is maybe more wary of the risks of liquid staking can get behind because you know it kind of limits what this you know worst case scenario could look like um and that's super interesting and so maybe just to but of course that could change over time right it's it's once people get comfortable with you know the way that you guys handle the delegation

of stake see that it's healthy maybe we can raise that cap Beyond 20 in you know maybe even just lift it completely um yeah so it's just a parameter um and because consumers chains have governance built in someone could just put up a governance proposal to raise it and then after the two-week up period it would it would be raised um this the second uh there's one other safety measure that's built in um so that the second one so the first one tries to address this governance risk of what if a liquid seeking protocol um is hacked or tries to attack the chain uh how would that be handled the second one addresses not liquid staking protocols overall rather IT addresses the individual validators that make up

the liquid second protocol um and specifically it tries to address the principal agent problem so the idea is that when there's one party of the agent that takes actions on behalf of some owner principle then sometimes these parties can get misaligned and this sort of already exists in delegated proof of stake systems but some people think it's maybe a little bit worse in liquid second protocols um so the liquidity or the liquid sticking module introduces this thing called a validator bond Factor which forces validators to self-stake some amount if they want to participate in liquid sinking So currently in Cosmos if you want to accept delegations you don't

have to put any of your own stake at risk you can have no skin in the game uh and there are like off-chain reasons why maybe this isn't totally true like uh you know validators uh if they attacked a network there might be legal repercussions it might be hard to offer him funds um but it is a fact today that most validators and Cosmos don't have any uh skin in the game on chain and this thing called the validator bonds will force validators to put skin in the game if they want to take part in liquid saving and it's proportional so if they want more stake delegated to them from liquid second providers they'll have to put more skin in the game yeah the that's that's a fascinating idea Aiden we we honestly haven't gone that much into

this topic of bonding and unbonding in this in this season to be honest miles and I may be a little bit guilty of focusing a bit too much on Lido but they're just kind of the dominant uh liquid staking provider over on ethereum and for the original whitelisted set they do not have they do not require their node operators to post any sort of bond or collateral I think going forward in the future for other modules in the staking router they are going to it's a fascinating concept that you're getting at here because the principal agent problem has also been a big sort of theme that we've discussed throughout the course of the season and one one of the things that I want to highlight in just hearing you describe you know what you're describing is sort of de-risking major Black Swan events so I I'm sort of reminded actually of the

the recent um merge and Lido's the ability to initiate withdrawals is kind of a de-risking event for liquid staking over in ethereum and I'm wondering if the liquid staking module could ultimately be something relatively similar um I don't know if you have any any thoughts there but I want to get into this idea of principal agent problem mitigation then I actually want to ask you about nfts because fungibility is another big topic of this season and I find the nft approach extremely interesting actually I think it is definitely a de-risking event there are these two very explicit risks that are being mitigated so there's the governance risk the risk that a liquid seeking protocol decides to Halt a chain and then there's this validator corruption risk and the validator

corruption risk quite honestly is a little bit subjective uh like I said no one has really done research into how much skin validators should have in the game in order to mitigate the principal agent problem we thought through some sort of basic scenarios and uh we proposed some numbers but I think there there needs to be a lot more research done in this area um that said there are there are still open questions um so two other risks that are not being addressed in the liquid staking module and I think maybe liquid taking module V2 or some future iteration of this we'll try to address these uh one is vote power concentration risk so this is the idea that

um liquid syncing protocols just decide to delegate to the top five validators because maybe that's what users want um and the second is this idea of cascading Liquidation on bondings this is one that uh Dave from osmosis has talked about the idea here is that if you have a liquid staking token and it's used as collateral throughout the ecosystem if you start to have cascading liquidations the token might depeg and when liquids taking tokens depeg arbitragers will come in and they'll unbond them with the liquid seeking protocol in order to get the underlying value of that token so if you can buy like one St atom at a 50 discount on the open market you can go back to stride and redeem it and then get a full atom

um so for there's like a 50 ARB there on a one dollar investment but the challenge or the the problem there is you might have large amounts of stake on bond in the network security deployments so there's like these more exotic risks that haven't totally been addressed yet um but to your point I think like the very basic risks uh liquid saving protocols taking over the chin and um like the principal agent problem just not being addressed at all which is the the current state of affairs I think these very basic risk this is like maybe the 80 20 on liquid seeking risks and these These are being addressed so I think um I do see it as a big de-risking event for for liquid saking and I I hope that

many more stakers will uh liquid sake after they see some of these risks start to be addressed maybe one question actually from me in in this scenario that you need the sales take from the validator is this something that stride validators will have to provide in in a scenario or is stride actually providing that for the validator so it has to come from the same address as the validator itself yeah this is a really good a really good question um it doesn't have so uh uh the I think what you're getting at is you know uh validators need to have skin in the game but uh liquid sinking protocols could also just cover that

skin um for the validators and then maybe the liquids taking protocol in the validator strikes some kind of agreement uh technically anyone can cover this skin in the game anyone can put up the bonds but this was mostly just for the ease of the validator so we've heard from validators that using the key uh that they that they use to sign transactions for other random things like doing a self bond for some of them it's operationally challenging so um technically anyone can can put up the stake but the hope is that it does come from validators and not liquid syncing protocols right and and then this certain cellphone has like a higher sort of slashing

probability or is like a first trench or or how do I yeah yeah that's another great question um in the first version it doesn't have a higher slashing rate um but I think it's a very broad design space and being the first tranche of tokens that are slashed I think is uh is probably a better design um it's just not built into the the first version all right right cool yeah they're very interesting actually also for me I haven't like followed it that closely I think you know what's interesting is that maybe just to comment because I guess we had these self-bond ratios already in especially like tesos I think basically like the first proof of stake protocol which is very renowned to have like a 10 like bonding

requirement for the validators um and and um I think there what was interesting to observe is that validators would like enter into um agreements like off chain with with other Jesus holders like loans to kind of source that South Bond and I think that introduces basically in some ways it's sort of um decreases the chances for like smaller operators to even get in this business and it's sort of like a little bit of centralization driver almost because only you know like a large enough entity kind of maybe knows like large teaser soldiers or can kind of create these agreements so I think there are also like sort of

downsides in in these restrictions uh in terms of like how the impact the centralization which are sometimes maybe overlooked but uh yeah just a comment yeah it's super interesting because I've always thought of the self Bond as like um maybe a higher level of decentralization or risk mitigation um that is has come at the cost of scaling right for something like rocket pole um and now but at the same time higher self Bond you know is super important when you're actually trying to make these validators have skin in the game um and do you see this on ethereum as well where the slashing penalties are often a teeny fraction of you know whatever profit the validator of the

proposer has made off of you know some malicious activity right um so you know each of these decisions comes very interesting trade-offs and maybe those trade-offs are a little bit different between ecosystems based off of like you know choices that the protocol design level like delegated proof of stake versus totally permissionless and you know um all those all those pieces so I think it's really interesting I'm learning a lot as well um and maybe just to get back to kind of the user flow so Aiden you've let's say I've been holding off right for the liquid staking module to finally liquid stake in in Cosmos because I'm very wary of all these risks and so I finally you know it's implemented onto the Hub I

um use the some sort of interface I'm guessing to interact with the LSM and basically shift my you know native stake to a liquid stake um it you know uses some sort of logic potentially to redis redistribute my delegation or maybe I don't have to um and now I get an nft right that represents my uh stake and the validator that it's delegated to so what where does stride come in after that and um or or Quicksilver or another you know LST protocol and how do you basically take these nft and wrap them in a way that makes them fungible again so uh what one minor point I think there are

some uh very ideological thinkers and for those particular thinkers or stickers I think the main driver of them deciding to liquid sake might be this technical de-risking event um but for most people I think the reason that that they would be pushed over the edge to liquid sake rather than natively sick as it just reduces the friction by like an order of magnitude um they don't have to unstake lose all this yields remember to come back three weeks later um they can just go and like try a little uh they can try liquids taking a small amount and experiment with it um so that the way that the user flow works is if you are a Staker on the cosmos hub say your sake to correspond

you can use the liquid safety module to turn any amount of your sake it doesn't have to be the entire amount say you have 100 safe tokens you can turn one of those staked tokens to correspond into an nft that nft says the reward address that earns yield is your address say miles and the validator that the token is stake to is course one the reason that it matters that um the validator is associated with the nft is different validators have different slashing risk so for example if you're holding one of these nfts and it's stake to course one and I'm holding a none of T and it's take two

imperator even if it represents the same amount of underlying atom if my stake is slashed suddenly my nft is worth worth less than yours um so this is sort of step one of the flow um you just turn your your one unit of stake into one unit of nft liquid stick then you IBC transfer it to a liquid sticking protocol currently none of the liquid second Protocols are built on top of the cosmos sub they're all their own chains so you could IVC transfer it to stride or Quicksilver or persistence any of these Cosmos um LSPs and then you deposit the nft into the liquid second protocol the liquid seeking protocol has to have logic that understands how to handle

these nfts um one really challenging thing is the validator could have been slashed so there's this like sort of asynchronous logic where the liquid taking protocol has to query the cosmos Hub and check like hey is this nft actually worth one atom or was this validator slashed now it's only worth 0.9 um so there's a little bit of a little bit of complexity that happens under the hood there but um that's the that's the high level flow and then you just receive an st Adam in return um just as if you had staked one native atom with the protocol and then you don't have to think about it anymore your your user journey is done so you you get your nft you bring it over you're like would sake it you're out um and that's that's the user flow Aiden

I have a I have a question for you about this this idea of nfts and uh the fungibility argument that you're getting at here I think is a really critical one and it's an idea that Miles and I have poked out a little bit when in some of our more detailed discussions about the staking router in Lido and one of the questions I think they're wrestling with is you know how do you do we eventually want to in my sense is that they the the folks at Lido are trying to design this in so that delegators can actually pick their specific set of validators um the issue is that that kind of breaks fungibility so there's sort of this technical discussion as well as should we do this kind of discussion about sort of uh because you're messing with this

idea of fungibility it seems like this nft solution that that you have here sort of addresses that am I am I in am I right in thinking that or do you just like redouble down and explain why these things have to be these positions have to be nfts um and if there's a yeah can you just re-double down on how where nfts fit into this this whole discussion yeah totally um I think I think it's a good question and there's a lot to unpack there um so one question is why uh why build nft liquid seeking infrastructure into the core protocol rather than some more opinionated enshrined liquid second protocol right a second one is what are different valid designs of liquid

sticking protocols um and uh you know on the second one like a valid design of a liquid sticking protocol is just a basket of these uh these nfts um so for example like someone could just build a very simple liquid-saking protocol that doesn't do any redelegations under the hood and if you deposit your correspond nft into this protocol it's never redelegated so you're determining the stake distribution of the LSP with stride the design actually does redelegate so if you deposit an nft of a validator that is not in the set stride will redelegate or redistribute that stake to the Target weight distribution of the set so there's sort of like a very rich

design space there um but uh in terms of why this is an nft and not some fungible token um the first thing I'll say is I think fungible lsts are uh absolutely required so composability is I think one of the most important um uh things one of the most important design principles in Define there's already so much complexity and if you break composability I think it just makes uh running infrastructure and making D5 work much more challenging it makes like

Searchers jobs harder it makes protocols other protocols jobs harder because they have to deal with multiple tokens it just it it exponentially raises the complexity of protocols so I would be very very wary about breaking fungibility even slightly like even even breaking LST fungibility into like an Institutional product and a base product um I think that is uh it might be a good idea but I think there are lots of risks associated with that um so I think yeah fungibility is is very important um and then maybe the question is like well if fungibility is so important why why are these all entities like this is

the opposite of fungibility every validator has their own liquid syncing token um and I think the design principle here is rather than enshrining one LLC protocol I think building LST protocols is uh quite challenging and there's a very broad design space um so rather than like one protocol trying to build their own and enshrining it by building this infrastructure it's Let's uh free market participants try and build and experiment with different designs um so that I think that's why the design decision was made to build uh liquid seeking infrastructure rather than enshrining a single protocol and the nfts are Downstream of that

it's not sort of like an implementation detail but it's it's also the reason why fungibility I think is less important for the liquid seeking module uh absolutely guys I I think um you know unfortunately we got to wrap up here soon I guess uh Felix or or Aiden sort of any any parting words that you want to leave folks with um or or just good way to follow you or keep more more up to date with with your work I know Felix uh up until extremely recently you were the col at course one where I'm sure you'll be very missed um maybe if you just want to drop some either concluding words or where to follow your anything like that yeah uh thanks again for for having us on first of all I'm sort of like exploring how to continue my contribution to space might

also get back into like some more research and content around like these sort of topics so if you follow me at Felix LTS on on Twitter you'll you'll be up to date on that and yeah really enjoyed the discussion I do think the liquid staking module generally speaking is like a great Direction like this enshrined idea and sort of not um opinionating too much and letting like the aggregation happen on the liquid staking uh protocol layer is is a very desirable like kind of Direction and I hope sort of that ethereum is also moving in that direction um and learning from Cosmos first of all we have to push it through it Cosmos I guess too but I'm definitely very excited for that and yeah thanks guys for hosting this this season on liquid

staking and having us on yeah thanks thanks so much for for having us um I'm Aiden Xerox on Twitter and I would love to chat about uh more of this stuff if it's interesting well thanks guys it's been really great uh getting the the cosmos perspective from you here Aiden it's an operator and Felix says the Godfather of liquid speaking we're gonna we're gonna make that a nickname for you Felix um guys it's been a real pleasure uh thanks so much for coming on and chatting with me and miles and I'll have to do it again yeah sometime soon all right partner that was a great one a lot of fun with Aiden and Felix there it was really fun because I think we got to dig into really like the origin story of of lsts with Felix and then dig into what I mean

The Godfather you know oh The Godfather right so sorry so sorry our uh definitely not self-dubbed Godfather uh but I'm John I'm trying to do him a favor right he can't ask exactly oh my God I'm trying to wheel this into exactly exactly I think it's totally fair um and deserve and then we got a preview of you know what is I would say like The Cutting Edge of LST you know infrastructure development in the LSM um and potentially a preview of you know what this could look like if it was mapped over to you know ethereum and Lido um because it does touch on a lot of the big questions that we've been talking about this season you know as it relates to delegator stake distribution um you know and how you deal with

fungibility and then just really how opinionated the governance of the LSD protocol should be um and so yeah happy to start wherever you want but lots to dig into yeah let's dig into this idea of liquid staking protocols as a force for decentralization and this kind of comes down to I feel like we're going to sound like broken records here but this tension in between things being permissionless versus the outcome that we want to see so I broadly see this playing out so like let's let's go back to Lido and the staking router and everything that we've been talking about so far this season so in order for uh Lido to mitigate again

the logic train is it's going to be a winner take all Market structure in order to mitigate these calls for self-limiting that Lido has already faced the road that they're going down is to make the governance of Lido as decentralized essentially as ethereum itself and that's what the staking router is is eventually going to do right it's kind of this combination of the staking router and dual governance so instead of just having this permissioned set of node operators you've expanded it to many many different modules DVT Community Etc eventually you can Envision a world with thousands of specific modules so that's one approach the risk that you take with that approach is that what

users ultimately end up wanting is not what if is in the best long-term interest of ethereum AKA having an extremely geographically dispersed and all of those different criteria that comes into having a decentralized validator set so the risk that you're running of essentially decentralizing governance and then leaving it up to the users is the users could make a poor decision right yeah but I would actually say that's pretty likely that that ends up happening so the other solution is what Aiden was describing which is having the liquid staking protocols be a stakeholder that has an opinion about what the validator set should look like

and the opinion being a decentralized validator set is better yeah right and that's an interesting tension I mean that I feel like really got pulled out in this in this episode yeah and I think it's again a little bit different the conversation is a little bit different from Cosmos to ethereum because of this idea that delegated proof of stake is built into the system in Cosmos right and so users today already have their ability to pick a validator and from that we can see just empirically that leads to a very top heavy set right people don't aren't charitable with their delegations right because they think some Grassroots validator is doing a particularly good job in governance or something like that they might care about that but they

don't care about that as much as safety um and and you know yield essentially um and so I think yeah it's very interesting they're kind of threading the needle over in Cosmos between you know finding a much safer and you know almost regulated way to Mint nfts but still you know on the other side to make them useful you have to give them to a protocol like stride and stride will take an opinion on you know what a good validator set looks like and if you know you give them an nft that represents your stake that's been delegated to a validator who is you know already above its quota of stake they will redistribute your stake to a different one that's lower in the set but if you

give them an nft that is you know representative of a delegation from you know let's say like the lowest validator in the set that is not part of strides you know curated set today they will also re-delegate to someone in their own set um and I think it's it's super interesting because yeah I guess to your point you know where Lido could go is Sim basically a delegated proof of stake model where you let the users decide exactly where to delegate you know within a module or a particular validator even um but that might not be in the best interest of of Lido's you know growth um and not in the best interest of ethereum the protocol

um in the same way that stride believes you know it's actually in their best interest to take an active opinion and redistribute some of these the stake here um and yeah I think I think we'll see a lot of iterations and experiments where it's kind of you know trying to do either full on one side or full on the other um and yeah again I think it's why you should pay attention to Cosmos as a little bit of a preview of what could be coming on the ethereum side and I want to continue to draw this connection so in the srivam episode we talked about the site this different approach to what a validator should look like as being a defining difference in terms of how blockchain architectures end up panning

out so the defining difference between ethereum and Cosmos being in ethereum land we want our validators to be our proposers to be dumb and commoditized in Cosmos land we want what can your validator do for you and I want to connect this to we alluded to the the John Charbonneau post that went live uh before you know we didn't really have time to look at it before this recording but I'm gonna I'm gonna share this this particular quote and this comes from dankrad and John tribe was riffing off a quote from Sunny that proof of stake was a mistake and there was a very interesting choice that that he chose to or a quote that he

he chose to include here we should establish a social Norm here that building on ethereum like building smart contract protocols on top we should almost never touch that I mean it'd be a very very rare situation but messing with the staking layer is a different matter for example if lots of validators start censoring we do want the ability to intervene so I think it would be a good signal to send yes we might be opinionated on what you do with the staking layer and that might include messing with your protocol and destroying it like that very phrase right is a very cosmosian sort of phrase to include there and it I see a lot of arguments sort of starting to shift back around around towards the cosmo side of things of expanding the role of the validator of having the protocol actually having opinions on what the

something like what the validator set for liquid staking protocol should ultimately look like and if we do accept liquid staking protocols not as this sort of neutral layer and platform but instead is a permanent stakeholder with an opinion then the governance side of of of of these protocols becomes extremely important that's where you some have something like steath dual governance in the form of Lido but yeah it's just very interesting yeah no I think the other theme that you see being kind of uh more embraced on the ethereum side is this leaning into the powers of governance um you know I think this is a big a big theme here where if you have you know

if you are aligned at the governance level um you can make more opinionated decisions that are you know result in a better outcome or at least a more healthy protocol than if you were to basically let users decide where the stake goes in a delegated proof of stake way or just make it you know more like ethereum l1's proof of stake where it's completely permissionless to join you know you don't really know who these actors are um and yeah we're starting to see a lot of that these ideas kind of translate back in terms of you know what your what Your validators could be doing to be more opinionated as well as leaning into governance as a force of

decentralization whereas you know up until I would say like this year governance at the you know smart contract level has always been viewed as this existential threat right to ethereum the base level layer for for protocols that are super close to the metal like Lido yeah I completely agree with that and actually that's that's what John charbonneau's post is entirely getting into here right we're talking about how governance is ultimately going to work and the role of governance on Roll-Ups and I feel like we've talked about this quite a bit but I think if you want a mental model for how things like governance and staking are going to play out in roll-up land on ethereum the best

thing that you can do is look at Cosmos and frankly what people were talking about in Cosmos six to 12 months ago because it's going to be an extremely good roadmap for for what you're going to look at and you know this this diagram if you're following along via video is you know John charb is comparing different solutions so delegated proof of stake without liquid staking tokens liquid staking tokens liquid sticking tokens plus dual governance and proof of governance and sort of all of the different merits things like Capital efficiency being permissionless real-time censorship resistance and web protection Etc so these things all exist on a spectrum and what I one of my favorite things about Cosmos in general is that

one of the benefits of frankly not having as much stake or being as large is that they can be more bold in what they experiment with you know this this 20 limit is just such a great example of that right where in ethereum because there's so much more Capital locked up there and the stakes are so much higher there's a lot of discussion that needs to be had first do we even agree with this High idea of a limit and if so what should that limit be and we're going to need months of research in Cosmos like yeah we probably should have one 20 sounds good cool yeah I mean you know it's like well it's because also like the way that just governance works on both sides right like it's a pretty straightforward process in Coswell so you have tokens and you vote with tokens

and you can change that 20 right whereas on ethereum it's much harder to say like okay we're just gonna start with 20 and then we're gonna have to come to social consensus you know all over again if we want to actually change this thing um and so you know any decision you make there because of the way that governance works is just much more permanent in nature um or at least you know that's the implication um so yeah I think it just unpack Sonny's tweet there I think his point is okay you you start with proof of stake you know the Natural Market forces obviously move you know make lsts very very you know attractive in traditional proof of stake okay everybody Stakes with lsts now you basically have a

pseudo like proof of authority sort of system right because if the majority of steak is steak with lsts and the LST governance is in charge of the distribution of that you've basically you know sacrificed like that that ability to express an opinion from the protocol layer um and then you know but connecting that that might not be so bad right and in some cases um if the basically the liquid staking protocol is as aligned as possible with with the base layer chain um and you add in some sort of you know protection like dual governance um but I think you know the point is why don't we from the very beginning start with governance you know electing like

the validator set right um and that's where you could in its most extreme form kind of replace proof of stake with a form of proof of authority that is more controlled through governance um and yeah I think that it's it's interesting that people are finally starting to poke at like these you know almost well held assumptions that this was the best way to do something and and now we're just kind of you know dealing with the consequences of of that choice but the trade-offs were always you know LED that to be the best choice in the first place um starting to look at that a little bit especially at the roll-up level right and say like a cosmos chain yeah well I'm slightly stealing this idea

um from from Sam Hart so I want to give him credit here but I do think there's especially in in the early stages of a protocol you can probably scale governance and the decentralization of that governance with the size and maturity of the protocol which I think to be fair is what you're seeing in roll-up land you know I mean I don't think you could look at any Roll-Ups really today and make a particularly strong case that they're a decentralized I think we're all probably okay with that because there's a road map towards decentralization and all you have to do is talk to these roll-up teams and you will realize very fast that they really do care about that and it's not lip service or anything so

yeah I think there are there are pros and cons to all of these different things and even these buckets here that we're looking at are not necessarily mutually exclusive and you know you you could probably mix and match uh different parts of these things as well I want to make sure that we get to the idea of fungibility because we we talked about that right at the end but I feel like it's a critical point of discussion for this season and Aiden had a very strong perspective there which is you cannot break fungibility even though there's this concept of nfts which inherently acknowledge that not all positions are fungible the fungibility of the stake token itself we do not ever want to throw that into question and that is I

think an open discussion over in Lido land with the staking router of allowing delegators to to connect with certain validator sets and unless I'm just fundamentally misunderstanding something there I just think it like undermines the value prop of liquid staking so much that the trade-off is never worth it um you know the whole reason you would want to use a liquid staking token over you know replicating some pseudo version of that in in C5 is is that you have this you know great defy ecosystem where you can go put these assets to work and if you you know break fungibility then all of that goes away right so I think that that's too big of a risk but the

real question is how can you basically go from an nft that represents your um stake and your choice of Delegation um wrap it without actually you know I guess commingling funds or commingling positions socializing losses of slashing and socializing rewards you want to keep all those things you know basically nft specific but then spit out a fungible liquid staking token on the other side and this you know you could see see in the long term like uh outcome for Lido or or stride this might be a way to just fully pull you know governance out of this delegation you know stake

distribution like decision making process um and I think Aiden's point is that's actually not a great thing um in that if you just pull that out of governance or pull that out of the protocol and give users your choice like we already know what's going to happen we see it on delegated proof of stake networks already like empirically you're going to get a very top heavy set um and so that's going to be a very interesting strategic choice because it's like you know first of all technically can we pull this off like is this a feasible it can we can we have some sort of architecture that makes this work where the nfts can get wrapped and spit something out but then longer term should we you know sure that would allow us to capture institutional users

that that's really the segment what we're kind of talking about or not explicitly but implicitly talking about right those are the people that care about where their stake is is delegated to the most um but you know that could be at odds with what you know the ethereum foundation would consider like a healthy distribution of steak right um and so the choice is really if an LST protocol is going to more or less you know manage the majority of stake for the network should that be a proof of authority system or should that be a delegated proof of stake system um and I think what Aiden and what Lyda would argue with both is that you know we can do a better job

um than users and and as long as you know we stay fully aligned with like the underlying protocol we can distribute this stake in a and create a healthier Network than it would if we just pulled that out of governance completely but then of course you lose that institutional segment and you increase the uh principal agent problem as well right that's right this this removes the principal agent problem but you could see you know it's a very clever way to you know kind of grow if you're or if you're a light out um and and really kind of the counter argument to all these like EF arguments you know about self-limiting is hey sorry we're just giving the bit you know giving that

choice to the users and this is what the users are choosing they're choosing to stick with the you know the largest validators and you know before you were saying hey governance is this huge risk well we've like pulled all of governance's you know abilities to be opinionated out and now you're upset because you know you end up with a very top heavy set so uh right I I could totally see that happening um yeah and and it's like well tough tough tooties you know like this yeah users want but I don't think that we're gonna you know I could totally see it going the other way too the one point that I would just poke at in terms of fungibility is all of these arguments are correct in

terms of pure purely technically what connotes fungibility but weirdly it's a social Choice as well and let me let me let me give you two examples which is one this is one we've already talked about but uh Bond ETFs Bond ETFs are a liquid instrument that should not there's a bunch of non-fungible positions bonds you know that get grouped up into these one this one sort of index that gets that gets traded and everyone sort of knows that the liquidity of the bond ETF doesn't necessarily make sense based on what is actually in there but I think the collective Financial world has said yeah it's good enough and

we're actually just okay with this I agree I think that's the exact model that this would look like right because each of those you know individual I would say like bond agreements are you know nfts right it is correct if you are an agreement an agreement between two entities that is unique versus you know the same version of that which is a different agreement between you know the maybe one entity but a different you know lender right but the risk profile of all of them are roughly in this in line with each other and so you can package it up in a wrapper right so that's what would be the equivalent of like wrapping all these nfts and having like one fungible asset on the outside that represents a share of them uh that's what we're talking about I think

um but where it gets really hard is like what if one of those bonds defaults right um and then you have to socialize that loss across right users um or holders of the of the bond or do you not or do you or can you direct that you know that loss directly at somebody's who's holding the underlying asset so yeah I think there's some parallels there that are interesting to kind of tease out all right miles uh we gotta wrap it here but I want to give listeners a little bit of a preview in terms of what we're going to do next episode so this entire season we have been going off of a very similar set of design principles in terms of what frankly makes a good primitive right which is you know flattening out and decentralizing your your governance in the form of of liquid sticking tokens

and eliminating principal agent problems and not doing anything that resembles vertical integration and next week we're going to get the Counterpoint because you and I are going to be talking to Sam kazamian of frax and Sam has taken a very different approach in terms of of how to conquer this Market um which I frankly I would it's always great to get the Counterpoint and and hear what Sam's doing because he's got a very interesting um way that he's organized fracks so I'm going to be hearing from him next week and get the Counterpoint for a lot of the things that we've been saying this whole season it'll be great yeah because I think we could see more of those models too so excited to get the contrarians view they're great our partner wrap it there is a good one

[Music]