Ethereum’s Staking Endgame

Transcript

[Music] welcome to Uncommon Core who we explore the Big Ideas in crypto from first principles this show is hosted by John chano co-founder in general partner of DBA and me hu strategy lead at flash BS and advisor to the light of Da cool welcome to the fourth episode of Uncommon Core I'm here again with my co-host John chabo founding partner of DBA and kaspar schwat schilling and Anar dri to uh research us at the etherum foundation hi

guys hi aad to be here so you guys recently um co-authored two proposals that made quite the wave uh to put it uh lightly in the etherum community and um they both explore the yeah what is basically um the endgame for uh staking for how staking Works in ethereum right and like um the economics around that um one of them introduces the idea of uh targeting a specific uh staking participation rate in the protocol uh including uh where rates can

even go negative um so if for example we wanted to Target you know 25% or 50% staked and we were at 75 then the rates would go negative until that for example that Target ratio could be achieved so that's a rather New Concept um and then the other one was um uh a kind of more concrete proposal to uh lower the the issuance curve um for the upcoming Electra fork or uh possibly the fork after that and um so why did you decide uh to to research this area kind of why are we talking

about this right now right um yeah good question I mean um touching the issuance policy is obviously always tricky and for good reason so question is why why are we talking about it now um and I think so basically we just looked at the um economics of staing today and kind of analyzed where we're headed and kind of looking at it um we felt that it's not really sustainable and um I think one thing to remember is also that this issuance curve is from kind of the pre- beacon chain launch days so I think that's 2020 and so at the time liquid

staking was not really a thing um reaking was not really thing and so we've learned a whole lot of things uh since then and um um I think it's important to kind of get to a point where we don't actually have to change the issuance policy anymore kind of the endgame of staking economics if you will and um this is really the intention here and so we wrote These two posts one is kind of kind of a possible Vision to get to a point where we don't actually have to change the ISU policy anymore um weer we refer to it as a targeting policy and then there's the second proposal which is kind of a concrete um step to get in in a concrete step to in the right

direction um and so yeah I would say um the goal is precisely kind of to not having to make changes to the issuance policy anymore and to kind of future prove it in a way that um I think we're now in a good position to do and that's actually really exciting to kind of really see the final shape of ethereum um for me and maybe to briefly jump in there right like um I I because I think that last point is really really important uh in general if you look at the the world map for ethereum of the last few years there were there have been a lot of moving paths for a long time and we are I think over the last year or two years we've we've gotten to the point where we start to really settle a lot of things and and start to

really have pieces into place that that we think are reliable for long term say the road Centric road map the modular approach um it it we're really starting to see basically the final shape or something that gets closer and closer to the final shape of ethereum and and so for the first time we were like hey instead of just talking about you know what gets us to the next few years now would be the time to actually start thinking about what what actually would be places that we could for example now here with the stake economics that we could leave the protocol in for several decades and and could trust that it would actually be in a good place so that's kind of where we're coming from John can I maybe call on you to kind of paint us a bit of a backdrop so

uh where are we right now in in the kind of the staking landscape and so kaspa you you already mentioned some things right so the the current curve is basically from you know before the beacon chain even went live and um I if I remember correctly like there was like an element going into it where we knew we needed some critical threshold of participation in staking so that the chain could be secure but then there wasn't really like a whole lot uh I don't want to say like a whole lot of thinking went into it beyond that because I'm sure it did but like a lot of things happened since then that maybe uh weren't anticipated so John um maybe to start with you where are we right now

and kind of what are some of the things that that are kind of moving um the saking red yeah so yeah so the high level landscape of it is so the the ocean ons have been touched a few times um back in proof of work days um but nothing has changed since um proof of stake happened um and basically just over that period of time what you've seen is just like the amount of e deposited um for staking um is basically just like an a steady kind of lineup basically since it was initiated from then through now um driven by I mean a mix of overtime like the market getting more comfortable the merge happening and now a lot of it lately is also potentially like reaking driven that there's more incentive because of that and people just like

generally getting more comfortable lower risk normal to do it over time um so and as You' mentioned the the curve basically it is like highly opinionated in the front end of okay we want to like be really aggressive with rewards early on to make sure there's like some reasonable number of amounts sted um but then to interrupt you like really quickly so in case our our listeners don't really know what we mean when we say curve yeah so in in what s like how does how does the idea of like a curve work yeah yeah so um and this one will probably pop up on this screen just to give you like a a general idea yeah cuz this is what most people are talking about when they talk about the issuance curve is just this like really simple graph of like x-axis is like the amount

of eth staked going from zero to the maximum amount of eth which is like Paul Park 120 million eth um today and then the y- AIS is just the the yield that you would earn on staking so like the general idea is as you're like close to zero we want a super heavily incentivized staking because we obviously don't want zero e staked that would kind of be a problem um so there's like effect effectively goes to like infinite yield um at zero and then it just kind of like lowers off over time such that like hey if we have 99% of e steak we don't like really need to incentivize the next guy that much the steak like we don't really care um the way that the curve works though is like it's really super aggressive in the front and then it tapers often how

aggressive it goes from like that kind of midpoint of where we are now towards the end um so the yield of the like incremental Staker goes down but over that point the amount of eth being issued is still increasing all the way up until the end just at a diminishing rate um so the incentive to keep staking always does exist and does keep growing in that context of like there is always added incentive um in that of like the more stake the amount the more e that the protocol will issue ultimately just at at a diminishing rate but is always more on net um so where we are today is um a little over a quarter be Stak so like a little over 30 million um and of that it's ballpark I think like a bit over

40% of that is liquid Stak um and of that liquid stake call it a little over 70% um is Lio so Lio is ballpark a little over 30% of like total eth Stak um and within that liquid stake um Lio has been going down in share kind of pretty consistently I would say um over the past call a year or so particularly as you have more new stuff coming up lately of um a number of the Sexes having uh liquid saking products that are growing like CB binance's eth um and then also a number of like the reaking platforms um are going to start growing over time um so that's kind of where we are today um historically it has been that kind of like simple mix and now we're at this stage where I would say

there's like a number of bigger wild cards coming up that could just like radically shift like okay not not just one the total staking rate but also the distribution within okay the total amount of e Stak like what does that composition look like um cuz as of right now after Lio the second biggest is coinbase is a bit over 14% 14 15% range um and so there's a big question of how much does this gravitate towards being institutional staking um particularly as someone like coinbase is likely to be presuming that there are ETFs at some point like they're probably going to be the custodian for these um there are going to be other ETFs like these will presumably stake over time um like other sexors are going to be growing their staking product so like how

institutionalized this become um and then also in qu staking there's finally going to be much more of a diversity as what like early signs seem to look like um at least in the short to medium term particularly with all of the reaking launching um there's like at at present I mean there over 12 billion in igen layer and like probably going to increase a number of that will end up in liquid reaking tokens and so this is going to become very clearly like a very large part of the market um basically goes to say there's a lot of wild cards um and this is also part of what like gives me pause of the so I'm I'm not like fundamentally opposed to changing issuance at any point ever like I'm definitely not I think there's just a a very high bar to do it now if

like you want to be really confident that like this is the thing which is like we're pretty much all agreed on this thing it can last a long time um and there are still just like a lot of uncertainties in my mind and things are probably going to keep changing pretty drastically over the next year or two um with a mix of all the custodial reaking like I'm probably going to change the landscape a lot um so things are changing pretty quickly which makes it exciting um but it's about where we are today and just just to just to jump in right um I think I think that's actually why we we thought now would be a really good time to kind of start that conversation um as as you said in the beginning right we have these two kind of post one of them is really focused

and that's the main one and the longer much longer one of people look them up right that is the one that just explores like right now with the information we have now and that's still incomplete but it's at least you know we we are now we're starting to see the shape again that's where like we we're trying to to to figure out what direction we think makes the most sense there is also like a more short-term proposal but I think we can talk May more towards kind of second half of the podcast a little more like what that what we're trying to do there and why we think there's some urgency but I really as you're saying kind of like for the for the main like the the the discussion around what should be the endgame State that's definitely discussion where we think that's not something we can we can

decide now or even in a month or two months or 3 months that's more something that like over the next year or so right basically we would we would hope that we are on a good good path to to really kind of make that assessment um and and one outcome here could be that we decided hey you know we believe the path we're on is good enough or even maybe the optimal one um we should stay on it that's totally fine it's more about basically really being intentional about it so that it's long-term um reliable yeah I think that point about intentionality is is really crucial here um like the current issuance curve it's it's a given thing but not changing it is also is Al also a decision and so kind of really thinking about um the

endgame Vision here is is crucial and so whatever the outcome I think it should be an intentional outcome okay this I find this very interesting because there's this uh there's this kind of very famous saying I think from Warren Buffett uh where he he says like if you look at your portfolio like you should just like imagine that every asset that you hold like you you should you know be willing to buy it every day right and there's like some some parall thing here right where you have like a certain configuration of rules in ethereum like you should be able to like create them from scratch every day and I think this is one line of thinking but I wonder if it's if it's like a good line of

thinking as it comes to kind of decentralized systems right because your portfolio you should be very loosely held to it and that's the main idea right you shouldn't kind of develop a kind of uh bias around things that you already own versus the things you don't own and and here it's really that you should have a very high bar to change something that is not clearly broken and so um that's why I think like those two things actually diverge and so my question to you would be okay why do you think I mean a like do you agree like we should only like it we should have like a certain bar in a decentralized system to change things that are not clearly

broken um um and if if yes then like why do you think uh the current situation warrants that yeah um so I I totally agree I actually think and especially something like like staking right and that's that's also when we when we published the articles there was definitely some some reactions that were like hey first first instinct instinctive reaction hey we should be careful here and totally 100% agree and and that's why again like the the idea is we want to get to a place where really we can rely that that this can can can remain unchanged um for for a very long time um we basically we just looked at the cont trajectory we are on and we do believe that not today

like today's situation is actually great I think if we could just freeze the count staking situation in place like we would immediately re press that button in a heartbeat uh I think it's more about basically us saying that hey we are not actually in a count equilibrium right as you as John as you said in the beginning we on this very constant trajectory of of inflow and actually the rate of INF is increasing if you look at like how long did it take us to get to the first 10 million State versus the second 10 million versus the third 10 million it is an increasing rate um and um and we'll I assume we'll in a second get into kind of our specific thinking around why we think there's strong reasons to believe this will continue all the way to at least a very high rate

possibly approaching 100% um and then specifically we we we think we have very strong reasons to to to to make the assessment that this 100 close to 100% stake case would be um very kind of um um problematic for for for ethereum um and so it's not about today's situation again it's it's really about saying like if we do nothing we on a trajectory kind of making an uh like an opinionated uh assessment about where the possible equilibria on the trory are and then um um assessing those and that's the a general problem with with ethereum research right like we just like um at least at the time of recording two days ago we had the Dan hard fork and we

started work on E 444 well over two years ago right so basically we always you always have to live in the future one or two years in the future basically with these things so our job is not to look at today and make propos changes right today it's about living one two years in the future and and figuring out where where are we headed if we do nothing and yeah is that an okay place do you have any tips for us like what meme coins are doing well in the future um um Casper uh maybe kind of to to kick it to you so uh I want to unpack uh um what Ana just just said so uh why do you

think the staking rate uh will keep increasing kind of what are the Dynamics that that kind of keep pushing it up versus would also like pushed it down yeah that's a great question um so essentially this question is about the supply of Stak right so what is at what um Point are stakers willing to supply their stake um and this supply curve has significantly changed over the time right we've mentioned solo staking and and the introduction of lsts and um I think it might make sense to kind of quickly unpack how this kind of influences the willingness to supply

stake um because it's I think very core to the whole argument um solo staking is costly you need 30 to East so high barrier to entrance um and it's technically not trivial so a bunch of people would need to sit down and read a bunch of block posts and trying to figure out how the hell um staking staking from home uh would work and so the pool of possible solo stakers um at current rates is relatively limited and the point here is that speaking in economics terms the supply th curve of solo stakers is quite steep what that means is that to get the last possible person to solo stake would require an insanely high amount of uh yield to

convince them to do that now liquid staking protocols really changed kind of that game um as as as you probably know much better um and so essentially LSS are this kind of like oneclick solution um where you click a button and you have access to that St taking yield um it comes obviously at some kind of trade-off you have a bunch of um kind of um kind of trust assumptions that come with it but um it's this kind of oneclick convenient uh solution giving you access to yield and the point here is that this significantly changes how the shape of the supply curve looks like and so um essentially now you can reach

a whole bunch of uh people that are willing to stake because it's technically easy you need one ease um or even less um and so um the point here is that the willingness to supply stake is much lower is much higher given lower yields and one interesting Dynamic here with lsts in general is that they have these Network effects so a liquid staking token becomes more useful if it's more adopted and more used it's kind of like uh liquidity attracts liquidity D5 protocols will start integrating the LST Etc and so in a way you can think of it even as kind of as um more staking happens via lsts it kind

of attracts even more stake because the LST in itself becomes more useful and then on top become kind like we have these kind of um um how should I say kind of the tail risks might uh kind of be discounted in different ways if we start seeing kind of 50 60 70% um of total e in one LST the the kind of um the the calculation risk calculations around those might change because they might be considered too big to fail um Point here is that as lsts become more adopted they become more attractive um and yeah that kind

of has this like staking inducing more taking effect and um because it's so low barrier to entrance um to get that yield um I think if we if we look at the current issuance curve we have just a little less than 2% um staking yield um at 120 million States so if all of each state you can still earn roughly um less just less than 2% um and so the worry basically is that um yeah given this yield at those High um staking participation levels um could mean that we actually reach those levels it's at least economically feasible and

not uh kind of impossible the 2% you were citing at 100% stake where do they come from from Priority fees and uh uh and me uh it's basically issuance so basically the staking yield you can access is roughly speaking the issuance yield so the new P that gets minted by the protocol and then yes you also have that me yield on top of that okay but uh at 100% stake the the issuance you can ignore basically right because it's uh like it's you're paying it in your in your own inflation basically well exactly so I mean I think

we have a whole section on this um in our write up and I think it might make sense to dive deep into the kind of differences between ninal yield and real yield we actually made quite a big fuss about it um in our write up but um yeah essentially like if everyone Stakes the it kind of becomes meaningless right so if everyone earns 2% you're actually not earning anything in real terms because you're also getting uh inflated by 2% however the incentive to stake is uh still relevant from the issuance you don't earn any issu but you earn fees and you earn m and you earn other rewards maybe that may exist so sorry for um but but but I I I think it's important to to point out

that as as KAS was just saying even at say 100% stake yes your balance goes up 2% everyone else's balance goes up 2% it's just a nominal thing like basically if you account for inflation B actually at the end of the year you're at the same level you were before um so you don't earn any money from issurance you do earn you still you still have your me income stream you still have the just the normal 1559 burn income stream um but for your for your decision to stake you compare of course your yield if you stake versus your yield if you don't stake and if you don't stake you now get get basically diluted by 2% so so the gap between eth holders and stakers that is this kind of nominal curve that that you're that you're familiar with so basically even at 100 100% stake that

the gap between not staking and staking is still 2% so so that's where the incentive to stake come from comes from yeah and it one little thing to point out on two is like I mean this isn't technically something you necessarily build a decentralized system around but like one of the reasons that it's arguably like really bad for the system as a whole too is like nobody is actually gaining anything um to your point from this issuance when you're at 100% of like yeah you make 2% but you're also duded by 2% like everyone kind of equally together um the really inefficient part about that like for the real world in practice is but you're probably going to pay taxes on that depending on your jurisdiction so now you are inflating and diluting everyone and then potentially making them pay taxes on half of that so they didn't

make any money and now they have to pay taxes on 50% of like whatever they made there it's almost even worse because that I think it's basically almost unavoidable in most jurisdiction for solo stakers because they actually you just have your e Frozen and it normally stays the same but actually just loses in value so maybe you get like some deduction in you know like income investment losses but at the same time you generate paper income that you have to tax whereas for for lsts and especially of course the centralized lsts it's it is actually relatively easy to structure that in a way where at least I don't again not every jurisdiction but but in most jurisdic that is shielded from that income so it's it's basically an effect that very heavily disad is disadvantageous for solo stakers and which is one of like basically you know we kind of in our

list we that's it's it's probably not the most important aspect but it is it is yet another aspect in which high staking is is just really bad for Sol staking yeah yeah in in practice it does make a big difference I mean like that that is a real reason that again this depends on jurisdiction and blah blah blah and all that but like that is a real reason why a lot of people will hold for example like wrapped teeth or very taking derivatives um is because they're the ones where the price goes up as opposed to like giving you the income um and so you get to defer as capital gains like that matters for a lot of people this is actually one of the best Arguments for the burn and reducing issue into my mind in practice is this so there like super tax and efficient on

the system and just to mention because I think actually this is there's a there's a bigger really important Point here that like sometimes a lot of these mechanisms they all equivalent but they're equivalent in not obvious ways and it's really important to build up the right intuition so for example right right now we pay for security we pay for staking with new issuance you could imagine an equivalent mechanism where you just take it directly out of eth holders balances and you just redistribute it to stakers right and if you if you do it that way it's it's completely equivalent right but but you would immediately understand that hey as we approach 100% basically there's fewer and fewer eth holders that you can take it out of the balance for so basically you get less and less paid out to people right so so basically this kind of

nominal real distinction which is kind of not obvious and weird to think about in the issuance case once you if you had an equivalent redistribution model much easier to think about it's it's it's a little analogous to like the whole dividends versus stock BuyBacks kind of case case with stocks um yeah so so so I think it's really important to kind of like really kind of look at it from our perspectives to fully understand all all aspects right and I think yeah one point here to to to add to this kind of nominal versus real yield discussion we've kind of briefly touched upon it already is but I think it's very important as a takeaway is that the incentive to stake is the nominal yield you protecting yourself from dilution is just as efficient as real yield in terms

of incentiv incentivizing you to stake that's very important um but then um the real yield is um important to look at when we compare different kind of long-term equilibria so and that's basically what we're doing when we compare different issuance curves so today today's issuance curve um might give us um a long-term equilibrium X and a different um issuance curve a different might give you a different long-term equilibrium and the composition of the net incentive to yield between these two different equilibria um is very much relevant because as a sticker you would be living

you would prefer to live in a world where in that equilibrium you actually earn more real yield so if you could choose between World A and B you would choose the world that has more real yield and that um is obviously not a choice that you have as a Staker has you don't have it as a Staker but um that's basically what we can do with issuance gr we can try and design um a curve such that we actually end up in a world where stakers actually have real yield left at the end of the day that's just yeah a quick point that I wanted to add to this uh discussion um we already talked about the supply curve and I think you painted this

really nice picture so we had this what we thought is the supply curve uh basically is like the relationship between like at what price are stakers willing to supply how much Eve correct and so we had in our minds like what the supply curve would look like and so we designed the issuance curve basically around that but then several things happened in the market um mainly I mean one was the big der risking of people seeing hey proof of stake actually works you know it's like it's not going to burn all my money and so um of that basically made people more comfortable with the system and shifted the supply curve um another big thing was lsts it it really dramatically lowered the cost

of staking uh and uh and that of course again uh made it made it basically possible for more people to stake at at this lower price but then uh I would say things also happened on the demand side right because when in the past uh folks were only earning um basically like issuance and they maybe were earning like some amount of me and and transaction fees uh it seems to me that right now um there's a lot of kind of exogenous rewards for being a Staker right and that that kind of pushes a lot of folks into staking so to give two examples I I think IG layer um kind of you know the anticipated air drop that

they have I mean we mentioned it earlier I think it's like 11 or 12 billion dollars of uh of e effectively in the protocol right now right some of that um in various like lsts and others in kind of uh you know native stake uh directly on the beacon chain and so um what this did effectively was like change the demand curve right um for St taking uh then you look at kind of others you look at like you know ethena or you look at blast Etc it's like various places where you can kind of get additional yield on on Stak Eve right and so um how how how are you kind of thinking about the

demand curve yeah so you're touching uh uh on something really important there right so um maybe the way we think about this at a very fundamental level is basically you have two you you you said just right basically like there are two types of fundamentally two different types of income that you get as a Staker you get um the the the internal or like that the protocol gets maybe in the first place that is the just the internal kind of rewards that that's basically just issuance that's just a redistribution of existing money basically of existing value between different um entities in in the protocol so in our case from non- staking eth holders to stake stakers and then there's external revenue streams and um the easiest example of that is of course 50 like um transaction fees and

um with how with after 1559 the the the majority of transaction fees now get routed to all eth holders and a small portion still goes directly to to proposers only um basically the way we think about it yes right now basically we have some existing sources like me um of external yield that that already kind of goes only to to to stakers and in the future with reaking We have basically another big category um of of of this external um yield as we approach um 100% stake basically this distinction between external Revenue going to all eth holders and external Revenue only going to stakers kind of goes away right basically because everyone becomes a Staker so basically in the end game all both 1559 burn and um me and everything

everything goes to to everyone basically everyone that holds e um in a way it is we already we already have proposals in the work um um there but basically the end game would be that we we we Direct all of these external income streams to all eth holders directly via some sort of extraction smoothing burning mechanism right so basically like how we did that with transaction fees um we we we also we working on meev burn or some some basically there this Testa proposal separation ideas that that would have the same effect um incidentally they would also do the same thing for at least some sub portion of of of reaking um income we can we can maybe talk about like the different categories of of of of reaking yield and then to what extent you can how easy it is to extract them and smooth them and

whatnot um but basically the end game is to have that all be smoothed away um and that that has two advantages one basically it it removes this spikiness of of um of Revenue esec especially for solo stakers so solo stakers because they only say for example any Revenue that you that you get as part of being a proposer which you only do you know if you only have a single validator it's like less than once a month or so that you propose a block right so basically right now with me there's the source of of of of variance in your rewards that you can get lucky and you can have like a big M block or you just don't get any M for an entire year right um and this of course will get even worse with something like grease staking and whatnot with with

with basically by by by extracting that on the protocol side we now make yields and incentives very predictable right so we basically bring them under control so basically end game here is that that all of this kind of goes instead of basically like it goes to the protocol first and then gets redirected to to all eth holders and that way we basically we only have this kind of very predictable and equally distributed um uh staking yield that that that is that that is what what directly incentivizes um staking so so that's but but but but I think it's it's really important um that we end up you know designing designing an end state that is robust against all these different potential ways this can play out yeah so the the the reason why I

wanted to bring up the demand curve so correct me if I'm wrong here my mental model is when you actually do this kind of targeting then you actually have a mechanism for controlling the demand curve including uh like all of the external factors kind of Weighing on it right because you basically have an oracle that that tells you in a way what the demand curve is uh which you read it basically from like how much EA staked uh and and so you you kind of basically have a uh like the one level which you control which might be for example the issuance rate you can say you know uh if I lower this you know let

let me just like lower this until like I I hit like the the kind of Target that I want and this allows me to kind of compensate for any external uh rewards that might exist outside the system we say that's like a fair meal model right again like I think the the the end goal has to be to always kind of have these external uh revenue streams directly captured by the by by by the protocol um because otherwise you kind of left with this kind of spikiness which basically makes home staking over time unviable and by the way no matter whether we do Target like this kind of this trying to Target a specific participation range or we go to 100% same problem right as we talked about at 100% there is no real yield anymore in terms of from from the issuance side so

all you're left with are these external revenue streams so so if they they they are spiky that's that's even a a worse problem um but but but yeah so basically the the the the idea is specifically to to basically have have a mechanism that in a way is kind of works no matter the external Factor so so say and maybe maybe now it's a good time to like look a into kind of the mechanism we propose for for achieving something like targeting would be basically just an issuance curve that that starts out just as the count one does so basically pays very high rewards as we go close to 0% but then has some some sort of flat part where it you know pays pays out similar to today like at at at at kind of today's level or something and then goes very very quickly like towards the zero

and then even potentially Beyond zero if we were to go to exceed that but by by by um a large amount and of course almost by definition that means that we end up never going into the range where it's it's strongly negative um at least on in net right net with taking taking external um sources into account but there could be other mechanisms right you could imagine more like a 1559 style feedback controller that that would also work just fine same same properties so why would 100% State Eve be bad so I think I heard previously okay like we are right now on the path to 100% okay so I mean yes also some time has passed maybe we learned some things maybe we just like shipped some

important upgrades so like some time to breathe right now and maybe think a bit further ahead so that's like one side of the argument which is hey let's figure out hey what should the end game look like but I also heard maybe some urgency uh kind of coming from what you said um around we don't really want 100 100% of e to be staked so why why is that the case what what kind of risks do you see from 100% being sted yeah um maybe I can take it first stab at it in that answer you can jump in um so I think there there's a couple of kind of negative externalities um there we see um one of them probably one

of the more fundamental ones is that um in a world where we have 100% State just let's use that as a kind of meme value um in that World um given some sort of Winner Takes most Dynamics um of L of the LST Market I think um you had a whole kind of episode on on why you think LST um is a Winner Takes most Market um essentially I think the tldr you can probably give a much better one um the tldd has essentially Network effects um more liquidity Etc um and so if we assume that there's this one dominating LST um and all of e is sted then

essentially that LST replaces raw e as the de facto money of ethereum and um that's a very kind of fundamental change to how I at least I think about ethereum and um so I think it it really warrants like us as a community asking the question do we want to live in a world in which eth is replaced by some LST um that inherently carries some assumptions around slashing risk governance and maybe some some smart contract risk um and I think that kind [Music] of the ideal money um should be maximally trustless maximally credibly

neutral and um I I feel that um an LST is in a way too opinionated to be truly scalable in that sense um and and in today's world I think we have basically the following problem holding holding raw e is expensive because of the opportunity cost um of of uh that issuance yield and um the alternative is basically then holding an LST um at least on that kind of 100% State Range it's feasible to expect everyone to solo stake as we know um and so essentially we we live in this world where it's expensive to hold uh e and it's risky to

hold lsts at least um more risky than holding raw e or than State Le and so money should be cheap and it should be maximally um uh uh trustless and so I think um that's I think at least personally for me a big reason um maybe I give it Anar the opportunity to jump in here if sure yeah I mean I I feel like in my mind I I ially I I have four main reasons of course you you primarily mentioned the the the money one I would even maybe on that point I would we it a little stronger I think basically if we

decide to replace e with um an like an LST or some basket of lsts or whatever you know like at 100% it's not guaranteed that there would be Winner Takes most but doesn't really matter because raw e will by definition no longer be be like a a a a a big portion of of of of eth basically on the Chain um I think in that world basically it's it's it's uh almost impossible to imagine that eth would still or or whatever kind of you know e equivalent we have would still be in the competition for as a crypto cryptocurrency right like we it might still be like a good kind of utility token investment token you you get you get your Revenue share of of whatever kind of Revenue is the ethereum BL chain produces you can still send transactions uh you know by maybe unwrapping just in

time or something like that um that all is fine but basically at that point we would basically just say hey you know if you if you want if you actually want like a new type of currency go to bitcoin U and if you just want to basically invest into kind of ethereum's Revenue stream hold hold the basket of lsts and that that you know like again all of these things are not to say that that's not a choice we can make but I I personally you know like part of what excites me about ethereum is ether the asset and I basically think just like you know like we we keep talking about things like credible neutrality there is just no such thing as a credibly neutral um LST that that has the same type of of um of Aura to it as as as as the native

asset right so so basically on on the point of of Mone I think basically it's just like giving up I'm curious what is it that you think that like that that you feel that any kind of different version of it is giving up on the money because like analogies are never perfect but I I mean like same thing in the real world of like the reality is that most of for example the dollars that are out there are not the like they're not like physical dollars or any form of Base money like they're generally like credit issue by some Bank there some like various M2 version of the dollar that's like credited by someone else that like yeah it's healthy your bank or whatever um in practice I expect the vast majority of e to look like that

regardless um and this is one thing that I've gone back and forth with people on I I especially think that like taking a rollup Centric map and like an L2 Centric view I think just like fully accelerated and accepted all of that for me um just in that world of realistically even if everything is not like even if it's not steth or CB eth or whatever like all the eth that you're using on an L2 is always going to be fundamentally this like synthetic issued version that's issued by whatever this L2 contract is um and this is why we see people like blast and any other native yield whatever taking advantage of that they realize that they are effectively a bank in some sense they are an asset sure and they realize that we can earn yield on our assets that we're holding um so I think that there's always going

to be this like issued representation if we expect that the majority of eth is going to be used on these al2s in the long term um the the only counterargument that I think that exists to that is if we think that most people are going to end up al2 and we take that as the assumption is that like you think that there will be a meaningful Market of these like enshrine ZK evm rollups like with no governance or upgradeability or anything like then I think you can make a plausible argument like okay okay like there is a credible base eth alternative um in practice I just don't expect that to be the majority of like eth usage in l2s I think like we see that it's very very likely the majority of eth that's going to be used on chain is going to be used on l2s that are in

some form upgradeable using potentially Alda like all these different things like it's never going to be this perfect one so I like I I've already kind of like accepted that it's never going to be quite the like purest version of it um and then like anything beyond that is like it it's an issed version of it that is like effectively very close um well right and I mean maybe I'm I might be the most you know like libertarian one here right but like like um I I personally still very much like kind of this the distinction like crypto with where this like Fiat where you know with with Fiat there are intermediaries like Banks and you can choose your bank and it's kind of hard to use to live life completely without banks but you know at least at least you have a choice with

the government you don't like if if if I if I live in the United States I have to use the US dollar and I cannot choose my intermediary there it's it's it's the United States right whatever Choice the choices they make or the fat basically like whatever choices they make in terms of how much new dollars to create like how the prices should behave all that kind of stuff I just have to basically I'm exposed to that I I I can't help that right um and I think that is the m main appeal of of of of cryptocurrencies as currencies again there's a lot of things about crypto that is super exciting that has nothing to do with the currency side of things but as a currency right if if we want e to be an interesting currency I think that very much kind of hinges on on on on

basically no one ultimately being in control and you can you can have that thought experiment right like say you have some LST winning out um that LST that charges a fee right and that charges a fee and as as it becomes Like a Winner Takes most it charges more more of like a monopoly fee and it can basically just choose like hey next year I'm raising our the our fee level by 1% right um and basically now now all my kind of my my my currency starts um being inflated away at at a slightly higher or like B taxed away at a slightly higher rate right I I don't want that I don't I don't want to basically have that intermediation and I think not only do I not want that I think basically people will look at this and they're like hey why should I trust

the you know like say Li but again that's not about calling them out anyone like the light of the coinbase governance over the US US government or the fed and actually after all the FED is a very competent institution right I trust the FED quite a lot I just I'd rather trust no one than the FED but I'd rather trust the FED than any LST governments and you can always make the governance kind of immutable you can say hey the fees fee is set and you can't change it anymore in the future but the moment you have that locked in right you basically you have all the stake in one one and um um um Network you can always basically like that always has governance power the facto you could always F off you could always kind of like say Hey you know we just deploy a

new one and we all switch over or something something like that right the moment you have then those networks effect you just for better worse you have to follow what they do and and and on your layer two point um well I think basically two things right one we were we working super super hard on making Layer Two experience as trustless as possible right that there's no governance in between that can do anything to my e no Bridge kind of risk like of course my contact rist is always always there but like basically minimizing this to to the extent possible and in a way it's like I would almost say it illustrates the opposite point right out of all of the problems that ethereum faced to like get to the current place and hopefully to get to like the the future that we want to get

to only scaling ever was like hard enough challenge that it required us to to add this one layer of inter intermediation right like out of all the possible like like aspects and challenges like this one was was hard enough that we needed this this indirection level um and now we're working super hard on trying to basically minimize the negative impacts of that um and so just basically now saying hey let's add a second one just capriciously just because we feel like it I mean of course you know that's yeah but you know making the point here right like Bally adding a second one that that bar for that should be incredibly High just because it is such a such a basically problematic thing to do maybe one thing worth adding here is also that these risks um compound so raw

e like e on an L2 is less risky than an LST on an L2 um and these lsts fundamentally carry slashing risk and that slashing risk I mean we can maybe get into penalties capping and trying to come up with a construction where trustless lsts are possible I don't think as a Cav there the Silver Bullet um that people make them to be but um definitely worth exploring but Point here is that um yeah I think the whole point of the RO rollup Centric road map is that L2 e is as close as possible to Raw e in terms of trust assumptions um and yeah I think what Ana was saying

scaling is so hard that we kind of couldn't do it in in a way that didn't require rollups um so kind of now kind of uh just accepting any kind of risk as a as an artifact of that seems kind of like there should still be a kind of Battle For What risks should we accept versus not okay so Casper I I hear you say that just because we have layer twos doesn't mean we should like willy-nilly kind of be accepting of all kinds of risks right so and uh there is I think the the general notion that uh while if

that's issued on layer twos is a derivative um it is stored in this like special Bridge construction that we work really hard to try to make as trustless as possible um I I I would say that I think in in practice there's probably going to be a fair bit of trust uh involved with rollups for a long period of time um then you have the element that there's a lot of derivative Eve uh outside of rollups as well so you have uh I mean e on coinbase you have e on binance you have Eve in in all of these ETFs that are coming and that number is is kind of

only going up right so I think um the amount of this kind of purest Eve that I mean even if we include like the most trust this kind of rollup in the definition of purest Eve I think that's that that that was your argument right that that roll up e really should be countered as long as it kind of is uh below some bar of like just assumptions that we would accept I still think that the amount of purest e will over time Trend towards like very very very low amount um and that's probably that that's probably like very little that that we can do at the same time I would uh I think there's a a good case to be made why you can build uh decentralized staking pools

that uh probably fall like in a similar range as I think like the most trustless uh kinds of rollups that you can make um I think you can you can at the very least make it fully non-custodial um you can fully kind of constrain the amounts of fees that can be charged I think where we all agree that that probably like some amount of governance and maybe it can be extremely slow moving just like some kind of defining some kind of objective function and like constraints for what the node operator set should look like right uh and that you know you could even have that changing you know maybe on like a six month for like a 12mon basis or something but you you need some kind of input uh and like some kind of

reaction I think to to what's happening so I think um well the the nice thing about LST like at least I personally believe that you can get them pretty pretty trustless uh to to a point there where I would I would personally be comfortable uh counting them towards the purest Eve so so two points on this right um one is just that I think like we don't need purist e right it's totally fine if your e is in a makeer contract or something because you chose to put it there it it's really more about that's why I tried to make the distinction between the US government and the banks because it by de facto you probably also trust the

bank but you chose to trust the bank you didn't choose to trust the the government right and so same thing here basically we just want to avoid the things that you have no choice but to but to trust and then you can put it wherever you want and with with with whatever trust assumptions you want and the problem with with with lsts is that but just by by by their very nature you can never make them like fully trustless specifically because there is always the slashing R like it's almost a catch 22 otherwise staking would be useless right if there's no that's the the whole point of stake at least in the theorum model with slashing is that there is something at stake and so inherently it has to be risky and I almost that's part of like you I think very in the beginning of the

podcast you or John in the intro said that like you know with ETFs coming online and whatnot I think actually it's really important that we keep reminding people it's not just free yield the free yield on ethereum you get is just the income streams that that that the protocol chooses to directly direct to you right like the V burn once we have it that is free yield the 1559 burn that is free yield staking is not free yield staking is a specialist activity that you can choose to do yourself or by delegation but it's actually something where you have to understand the risks you take and it's not it's it's it's it's very much not free and specifically um sorry for but specifically um to maybe make that point a little more um one thing we are worried about very much

and that's almost kind of like already the second you know we talking like the Mone is like the first thing wored we worried about the second one that I would I would say is kind of the the outsized influence on on ethereum governance and so you could create the scenario where um we you know people always talk about censorship right and um in principle you could have a dominant LST that at some point decides you know because of regulatory pressure maybe coordinated regulatory pressure across multiple different jurisdictions that all the operators will from tomorrow on just F out anyone who includes transactions that you know on some Black List Right right not not just themselves no longer include them into blocks but actually Fork out anyone um who uh who who would do that and and that's specifically like in the past

we've always said that's a red line we draw like you can do in your own blocks you can do whatever you want and I mean in the future maybe we we add things that constrain your behavior but for now it's not it's your block you can do you you're not allowed to just reok someone out because you don't like their block right um so but in effect what would happen if if if people started doing that well we would basically have to say to to have this instrument of of of of social slashing right where we basically like hey you know you either stop doing this or we come in we take away your e right and that that is right now I mean in the future maybe we can automate that somewhat with with with there there are all these fancy ideas around detecting

that in protocol but because it's kind of hard to detect in protocol right now we would have to do that manually now imagine if that is done by the dominant LST right and then now all of a sudden 80% of all Ethan existence is up for slashing right good luck trying to come to allods and and and and propose that but it's completely impossible right right so in in that scenario basically now now you basically just de facto kind of overrule like basically now ethereum governance is just that LST governance right or to the extent that we can still resist that now that e again is at risk so you know what wasn't trustless in the first place so basically you know this is kind of this inherent tension you can never have a situation where the state e is not

fundamentally atrisk and it has to be a risk otherwise the entire system no longer works so one thing is I I think you can pull apart I mean and like I think there actually is a valid argument for this of honestly removing slashing penalties um for stuff like double signing um one I I mean the penalty is quite frankly not that high and like you can get away with things that are like very clearly profitable even though you are double signing um in practice that like penalty is effectively negligible on like any SC and like I mean this is part of why I mean people are very comfortable with Ste there any other staking providers because in practice a normal course there's like effectively no slashing

risk like it's incredibly rare that anyone would ever have any kind of penalty from this um fundamentally like you have the stake there such that like it can be socially slashed or forked um is like realistically where like any kind of meaningful penalty is or if it's like an activity leaked out like you ignore this chain Etc um so I like the stake is like fundamentally kind of there in either case but yeah I I mean it's never going to be like technically zero risk if you are um putting your money there it it's just like at the end of the day to me like what a lot of it comes down to is fundamentally is like what are we optimizing for is the in practice like what do we think that 99% of users are going to go do is it going to be go do this thing that like has

some form of trust tradeoff whether that's using an upgradeable L2 or wanting the staking yield or all these things or to be optimize for what we view as like the ideal case but is like in practice Pro like very possibly the 1% um like it reminds me a little bit of like a bit different but like the idea of like not having delegation and protocol of like it is like a little bit purist in the sense of like we like we don't want everyone to go delegate and so like we optimize for the system of like we like we just don't build around that as a model and then in practice like most people just go do that thing anyway and then like it was like kind of did it help optimizing for that thing in the first place um the other the other main thing for me is just like I it's

not forced in any of these cases to like go use an LST um it's just that currently the incentive to do so like continues to increase as there's more stake um which is why I think the idea of like kind of flattening off the curve and the electro one is like a more reasonable thing I I don't think that you need to go to the idea of like doing a negative issuance curve to do that like any form that like effectively caps the amount of delution like achieves exactly that um anything beyond like that's part of the conversation around this is like there's actually like multiple distinct goals in my mind between these two different proposals that are kind of like embedded underneath like it is very different to cap the amount of eth that's going to be issued versus actually reducing the

issuance curve even those are coupled together or like making the curve go negative um and a lot of it is also like it's targeted at we're trying to make sure that like lsts like don't go to 100% basically um which is different also than making sure that the amount of staked eth doesn't go to 100% um and like that is one of the things that like I think about a lot in this conversation is in practice you you can limit the amount of e Stak to some number but does that mean that lsts won't still become 100% in practice of like what people use um for the simple example of like everyone could put all of their money in a Black Rock e ETF that is like staking and even if like rewards level off at some point like Black Rock can still

just like extract the maximum number of awards and like socialize them to everyone in there um and that's like rational for everyone to do it even though you're not getting like more rewards out of the total like um structure so well almost almost by definition that can't be the case though right because like if we if we if you go to a targeting mechanism we say like for example you know people always hung up on the 25% but just it's a good mental model as a Target right it could be that we end up choosing 10% 50% 70% I mean you know whatever but like just saying 25 just you know like for good intuition um by definition that means that the majority of eth is is RO and sure you could basically have 100% pulled and then you just give everyone only quarter of the of the of the revenue right but

but but basically the whole point is that in that level because the demand is so you know for staking is so kind of unchanging is that you you just come down to to basically zero reward and at that point why why bother you know going going through the the the black Road ETF versus holding it holding it yourself right like the the whole point would be that if if there's if there's so much pooling demand then you know you that that means there is no there's no yield anymore right that the because there is still just the fundamental like safety Network effects desired to hold this thing like people are going to put their money in ETFs with no staking yield um if like like that is like a very plausible like the first ETFs that are proved as like very possibly ones that like they just don't allow them to stake

in the US and like you could very plausibly see just's like a ton of people put them in there and then like staking being added after the fact is like kind of like a bonus to that um and same thing with steth of like if like at the limit of like if there's anything positive like I'm just going to hold it anyway um and so like that that is still the optimal thing for them to do at the end of the day um it's it's a diminishing return but like it is still always just going to be possible so it's extremely I mean very unlikely that there will be like some like that if ETFs for example in state EFS so I think this I mean what you see in traditional stocks is hey like phys like stocks that like ETFs that have all of the stocks there like

in physical backing they used to be more derivative based ETFs now they are mostly physic back but all of those still lend out the they lend out the the stocks as well right it's like natural competition they try to generate some yield on it so they reh hypothecate it and they compete on like the total expense ratio to investors and so I mean the same thing will happen with Eve right they may like lend out some Eve they may stake some Eve um and if if they think that like staking Eve is not that risky which it's not um I think it's like there's no reason why it should be right I think I don't think staking Eve has to be risky um I think the the like all that all that it does is basically you just like making some

economic commitment to like the correct Fork of the chain and identifying that fork of the chain if you're not a malicious actor is a super trivial exercise and the only times when that fails is if you you make like an operational mistake but operational mistakes happen everywhere right like Banks uh you know fat finger like and send money to to to do the wrong counterpart right I uh and then they can't get it back it's like this this stuff basically happens everywhere right and so yeah I think the competition from ETFs is uh is going to be kind of a very real one um whether like no matter kind of where the the issuance curve is and it will not make a difference between like regular e and and and state e just

to spell spell that out right because basically um in a way it that that that's not conflicting at all with what with what we're saying right because basically the point is right now we on this trajectory we 100% And then if you think again about this kind of this this model of like external Y and internal yield basically it means internal yield would by default go away and you only have the external yield left and it's spread equally across everyone now basically what you're saying is Well you know even if we were to do targeting we would end up with the same situation right because they would basically the the the external yield would be spread equally across everyone you know and then and then the internal one basically again would basically come come come down to zero because kind of under targeting there's there's no such there

would not no longer be this two separate classes of the stakers and the not stakers because they would all be managed under the same entity right so basically you're just saying it's hopeless to recover any real yield for staking right and that might be the case I mean I hope it's not right but at least in that scenario staking would also like like targeting targeting would more be like physically like a a hopeless act but it wouldn't actually change anything it would not make things worse so basically what we're saying is hey the end game inevitable all the external yield will be distributed AC all all eth that is inevit in either scenario and now we just trying to say let's try to at least let's do tar things so that at least plausibly there's still some real yield that goes to stake us and that might fail for the

reasons you just you just outlined I personally I think actually the competitive Dynamics would be that there would always be some entrance to to basically to to make it so that it would not be sustainable to just to to just um um smooth over uh basically your yield across across the the entire kind of e set but you know we can let that play out in any way like if this situation would not be worse right it's it's basically you're just saying oh it's completely inevitable to maybe get into some uh uh kind of the other side I think uh of of the of the argument so I mean if if we kind of map out kind of what what the arguments uh roughly are so I think um the idea of targeting in general I think

is a kind of very sound one uh because for all of the the benefits mentioned right it kind of internalizes some of the like external rewards into the protocol and kind of smooth them out and kind of I think those are all like very very sound arguments uh I think in the it it gets more interesting I think if we talk about like what is the number to Target right because obviously like when you target 100% uh that that doesn't that's like meaningless right if if you target 10% versus 50% ver 25% you can have kind of very very strongly varying outcomes and so I would say um one of the one of the counterarguments that

I've heard um and this is maybe where we get into some like concrete uh numbers to discuss so what uh um if we say kind of 25% uh Target uh Target stake rate uh how very concretely like speaking what that look today like where do you think you would need to put the the kind of um the ne how negative would kind of the what the fees uh the issuance have to go in order to to Target that in your mind today well I think I think kind of because you also said negative in the intro actually that's I'm I'm not quite sure that seems to me to be more

misunderstanding like basically the the the um the point would be that of course the nominal yield would indeed be lower than than it is by because at the nomal like if would get to 100% at 2% 100% St at 2% nominal then the ninal Le would have to come down to prevent us from getting there but it would still if you if you if you if you if you talk about the aggregate yield including me income including staking income reaking income that has that that would have to by definition still be positive otherwise you just unstake right oh yeah yeah yeah I'm saying like how how how low does the issuance have to go kind of keep uh like the real Yi at a level where the marginal person is no more

likely to stake at 25% versus unstake at 25% right so basically um when we have like example kind of Supply curves we can put them on the screen of like um where we we just you know totally just like qualitatively give some examples where basically you could plausibly end up in a scenario where the um your real yield would be um slightly positive right it would it would it would definitely be lower than today I mean the whole point is that unfortunately today we still have inflow pressure which we would have to get rid of but you could plausibly end up with still 1% half a percent of course depends on the the exact circumstance of of real yield and that is basically yield that you would not have if we go to 100% so it would still be be be by definition positive if the if the total yield is

negative you just unstake so um but it would be lower than today all we can say is it's lower low lower than today higher than zero somewhere in between would be why do you say you wouldn't have the real yield when you at 100% issu yield I suppose oh but I mean in order to actually stabilize now the protocol at 25% the noral yield would have to be negative already right it like it the you need to pay negative you need to charge saers negative issuance right so I think in general for this kind of targeting endgame Vision I think it's useful mental model is think we have me burn we have a form of capturing that

exogenous yield because I I genuinely think there's kind of No Way Around It And the reason is we can't allow reward variability to be too high and otherwise solo staking becomes unfeasible um so the point is that if issuance yield comes too close um if the real issu yield comes too close to zero then the exogenous yield of M V makes makes your reward too too spiky and so um that is a problem in both the targeting world and if we reach 100% State and um that's also an interesting observation it's actually a problem that

exists in the long run under the current policy and it would exist in the targeting scenario and so I do think that um and I mean there's so many kind of ideas around this execution tickets me burn it makes sense on on on all sorts of Dimensions um so I do think kind of as a mental model we should just kind of work with like some M burn mechanism in place um and yeah I I personally I I see this as a prerequisite um at this point and then of course um issuance can't actually go negative because then you would unstake there's no you're not going toake okay so the yeah what I

didn't know is that you already assume that we have me burn in place okay yeah which we don't today it's it's not not not your fault um but I do think I think we listed it in our open questions like how to deal with this reward variability yeah um but yeah in this case it's very clear like if you have me burn then the nominal yield can never go negative and if you have M burn H sorry if you can't if you let me start over so to summarize if you don't have me burn then the nominal yield will have to go negative in order to probably fix uh the the the staking rate somewhere like where it's now or like it's somewhere where we want

to Target it right um and if we do have me burn then uh uh then the normally can never go negative because otherwise everybody would unstake so because then the nominal yield becomes the real yield in a sense okay so uh where I wanted to to take this is is basically what is the relationship between different uh staking ratios I.E 10% stake 25% Stak 50% Stak 100% Stak and the quality of the node operator set and kind of how those two things interact so previously we talked about the supply curve and for example how

lsts Chang the supply curve by making staking a lot cheaper right because um a they kind of remove the liquidity penalty but they also brought professional operators into the mix who can like an economies of scale and so an LST can stake the same funds for uh a lot cheaper than a solo Staker can and so now in my mind what happens if you have uh a Target ratio that is too small uh let's say we put it at 10% okay and we have the following distribution of note operators we have 10 % someone someone who Stakes

for uh you know. 2% uh real yield and then we have someone who does it you know another 10% at like point4 and another one at like 6 and those numbers are completely made up right but that's what it costs just in operating costs basically to to run those operators so the outcome that that occurs is like only the first node operator remains and they stake uh at at basically their uh they here like because they the the the cheapest right to to operate and the other to drop out um and the the kind of the protocol makes sure of that and so

um that is why I think um with targeting in general like this is like this this very kind of underexplored area right now like how low is too low to the point where you kind of end up with a really really cheap and centralized note operator set yeah right I mean that's a great point and I feel like one of the main concerns um that has been kind of put forward um and so I think there's like couple things that um are worth mentioning here so the first point is we're heading to a low issuance environment under the status quo already um and so that that's like the

first kind of important Point like just because you're at 3% or 4% um staking yields today doesn't mean that's going to be the case in the kind of long-term equilibrium so we're sliding to that world anyhow then burn or just because the number of stakers goes up or both no because yeah exactly like kind of assuming that we're sliding into this kind of uh world where 100% the Stak or 80 90% um as people are getting more and more comfortable eventually maybe the LST flips e as the de facto money it kind of becomes the default you don't even know what ether is anymore and and so basically um in that world the issuance

yield is not magically going to stay at the current levels so that's kind of the first important um Point here um then the second point is that the underlying com relative competitiveness between different staking types be it Solo stakers versus a liquid staking pool or versus kind of a centralized coinbase staking solution for example these underlying relative competive competitive the the underlying relative competitiveness exists no matter what the issuance curve is that's fundamentally true and so um in a way you can kind of think of uh adjusting the issuance curve as a sort of

speedrunning the compet the competition between different staking types but they're not changing anything that's under fundamentally underlying there um the one caveat here is actually um that the kind of network effects in economies of scale that more centralized um staking Solutions can kind of benefit from are actually higher in high staking environments so if you kind of take this idealized model where a centralized staking provider has one huge fixed cost up front and that's it and then maybe like a very low yeah basically one big fixed cost um then uh you're going to benefit from economies of scale more

than someone that has also variable cost um and so that this is kind of actually something that works in the favor of targeting a lower stake participation you kind of reduce the effect of these economies um of of scale and then um maybe kind of one solo staking specific kind of uh comment here is that this um kind of if you think of the of solo stakers um it kind of I think it it it's relatively fair to say that the majority vast majority of um existing solo stakers um the vast majority of solo stakers um

are the ones existing today meaning that if we slide to 100% State it's presumably mostly via lsts and so and these solo stakers have basically put up the fixed cost already they bought their Hardware um they acquired the technical KN how to stake and now just have a bit of Maintenance and they are actually more inelastic to changes in issuant and the reason is that they've paid already their cost and now they they have basically pure yield coming in as opposed to um a liquid staking token holder because they still have to pay their staking service provider fee um as they earn the staking yield so in that

sense I think actually liquid uh solo stakers um benefit um more and I think another Point here is maybe that um if you think about the kind of relative importance of solo stakers would you think that solo staking is kind of the share of Solo stakers in terms of percentage do you think that's higher at 2% issuance yield with 40 million each St or 2% issuance with say 100 million e of state and I would argue probably the former can you repeat that do you think the share of solo stakers is higher at a 2% issuance yield with 40 million e state or % with 100

million each state point being like do you think the relative share of solo Stak is is higher in the low or high yeah the relative share is obviously higher like the less you have stti right and and just to to make the one point a bit more explicit right like because you talked about um basically more more a more harsher yield environment maybe like if we if we do the targeting like our point is kind of the exact opposite right like we we say that if we do nothing you know every every month now basically the environment gets has and hash and hasher until the end point where there's literally no intrinsic yield anymore it's all just at least no no real one right like you only have this kind of counting thing where everyone's balance goes up by 2% but no one makes any real

money and we even talked about the tax implications of how that's act actively really bad for for solo stakers um you you do you do get your share of the external yeld right but the whole point is with with me capture and these kind of things you will get that share No Matter What right the basically the point being that if you own 1% of all Ethan existence you will get 1% of the external yield that basically is fixed and then by doing targeting we're trying to on top of that recover some real issuance yield for you in like basically on top right so so basically the the incentive to stake will go down but the real yield will go up so basically you'll make more money as a Staker under targeting than without targeting and we have like these these graphs that kind of really illustrates that so so

basically you there'll be more leeway to do kind of the that maybe slightly less efficient staking under targeting than this without targeting in the super harsh 100% state competitive environment to to make this point a bit maybe to illustrate that again if we assume to get to 100% the issuance yield um the ninal issuance yield is basically worthless everyone's getting 2% so you have 0% real issuance yield if we however end up in Via targeting or whatnot um in a in a world where less less e estate then the composition of the incentive to stake might be composed of real yield again or

not might is and so the point here is that um in that world you actually have real staking yield as opposed to in the 100% State world that's I mean that's just I still don't get it you know like that's assuming at 100% there's no real yet which that doesn't seem why would that be the case bur again imagine we're in a world with burn there is no kind of exogenous yield um let's just simplify okay yeah so I mean basically you're arguing let's say

there's like 20% staked 80% not staked then it's it's like the real yeld comes from the 80% paying the 20% is is what you're saying and that that makes weird yeah and I think maybe one kind of point because we kind of left off the kind of negative externalities at some unfinished point I think another thing to note here is that um staking is a service and the protocol demands stakers to secure itself and that's why it pays for it and um I think kind of this whole notion of 100% State we've talked about it a bunch

um staking kind of becomes this kind of useless yeah kind of activity from a financial perspective as a Staker um and instead I think as an ethereum user you should come to ethereum and not have to concern yourself with these considerations should I buy this LST or that LST should I solo stake or not um and instead I think staking should be treated as what it is it's supposed to secure the chain in uh and I think it makes sense that this is more of a sort of let's say specialized task it should be someone that is able to make these risk assessments um and kind of understand what they're getting into um

whereas that's not really feasible in a world where everyone just holds um some some LST effectively right it's it's a very basic kind of division of labor argument right so basically like in any case the eth holders are the ones that pay for security this was true in proof of work right via kind of delusion with issuance this was is true right now with proof of stake this will be true even at 100% the difference is how you do that right and in the 100% World everyone basically contributes to security because they have to either stake themselves or they have to delegate via lsts so basically you are now part of the labor you know you everyone has to be part of you know has to be part of like participating and securing the chain and you know it's it's plausible to set it up that way but

the alternative I think is much more elegant that we have today is is a scenario where the majority of people say hey I'm not specialist in this I just basically pay via inflation via issuance for the Specialists to do that job and that does not include only solar Staker that solo stakers and people that understand the risk profile and that delegate V lsts but that's that's the The Specialist minority and they get paid by the rest so I personally think it makes much more sense to basically leave it to the Specialists have a division of labor here instead of yield going away and everyone doing it as a duty if they want to have part of the upside of of being in ethereum it feels I mean I see why kind of this line of thinking is compelling uh it feels still like a little bit

constructed to me in in the sense that when I look in the real world um then I don't see I mean I don't see people holding physical coins and I mean you can go and say oh there should be this division of labor lending is like so incredibly dangerous and we have like example ABC of like Banks blowing up because uh uh there was a a crisis or a depression or or whatever but nonetheless like everybody it's like it's riskless enough it's low risk enough and like it's the the the risk reward is there still for 100% of money in the economy to be staked in a sense that's that's what you see in the real

world right like we all uh none of us hold like uh base money like we all hold credit right and especially if you're saying hey we have this vision for ethereum to become this better money uh and we want people to use it as money Etc then to me like we talk here about like a staking end game but like the end end game kind of beyond that is is once again e be just becoming kind of abstracted uh away and like people trading in like some form of like e based credit you know um where maybe they hold uh like an e deposit but the actual e like on the back end is land out right um that's that that's what you've seen with with basically every

currency uh over time so kind of I wonder if we not trying to Anchor to some kind of idealized version that maybe doesn't hold uh up to reality right but I think actually this this this comparison exactly makes the points because it's not it's not the I think the right comparison because if you put your your money into the bank you actually that is external yield that you're generating it actually is put to good this is like taking your e and putting it into defi right in defi actually you land it out to other people it actually generates value whereas um by staking you basically you only contribute to kind of this common duty of the security it's like almost saying the division of labor between you you

yourself having to fight in the army versus paying taxes and the government paying a specialist to fight for you for example the restating you know you could argue hey like okay we expand staking more and more to add for example uh economic commitments to other services you know like a testing like Oracle updates that come on chain like holding data and like in Da y y I mean there's probably a bunch of things um that that are possible right and like I think you can also we now talk about like based sequencing we talk about like pre-confirmation Services right I think the range of like useful services that um that like e staking can unlock uh and like offer to the kind of

ethereum Community uh I think we've only see that number uh go up right but that's I think you know in the beginning we talked briefly about reaking and we said that there's different types of reaking so maybe now would be a good time to briefly dive into that because there's there's a big difference here right like there's some forms of of reaking that have nothing to do with your specific role as an ethereum validator it's just like say Oracle updates or something that anyone could do and that in principle you could you could always do with just having native eth or any other collateral locked in the in the restating protocol as well right there's nothing special about State e as a collateral there and then and so basically that in a way it's just again that that's like putting your money into Defi and and generating real

productive external yield and that's perect fine you know like you everyone should go ahead and do that if they if they like the risk profile um and then there is kind of types of reaking that very specifically rely on your role as an ethereum validator and so right now we kind of have this proposing and validation and basically testing as two parts of the same role in the future with proposed a tester separation we'll probably have those be split out so you'll have some like just the the general task of all validators will just be testing to blocks and then you basically have some sort of bidding mechanism where you become the proposer but it's a specialist role and then there might be some avss on like reaking that basically specifically give you extra duties if you're the proposer there might be even some with special

duties as as an tester but again those then are much more basically those types of of fre activities you should think of as identical to me right you can you can do the same thing that do with me you can smooth it out everyone you know like that that's just protocol income that that's basically that we just haven't captured yet and that we will capture and we will distribute to everyone so if you hold e no matter whether you stake or not you will get access to that yeah the whole idea of like uh making all the Val us equal is is is a kind of really interesting Revit hold to go down right because in in in like some sense it's it's saying hey like the eum government is like taxing all economic activity at 100% right um I mean and in practice it won't be 100% % it will just

be like whatever is kind of the maximum that or like the the minimum price that people will do the activity at and like all of the rest is basically captured by ethereum but it's not all economic just to be clear right it's not all economic activity if you want to build some project on top of layer two on top of mean in terms of validator services like things that that that validat can do yeah exactly so basically instead of in a way it's almost weird right basically that you have this this you already even in today's world you basically immed in a way you have this this kind of indirect um extraction because basically the more attractive it is to validate to to be a validator the more validators come in the lower the yield of for the validator becomes so so but but it's a

very indirect mechanism and um and all we're doing is we make it more more more direct and by making it more direct we can smooth it and so we can we can remove this this disadvantage that solo stakers have so I think I think basically if we want to have solar sta survive in any of these worlds we have to find a way ways to smooth internal yield otherwise otherwise it will become completely inom uncompetitive to be a solo Staker when you say smooth uh I I hear that like used a lot but I never hear weirdly like okay so it's charged in the form of like me burn but then it's like pulled and paid out over time to like the same parties I guess that's not what you mean right it's it's really it's it's it's just taken and

like given to all the e folders right and not the actual Solo sters or like any like the sters in general the the key Point here is to have some in protocol mechanism to capture me if you then burn it or redistribute it to everyone doesn't really matter um but the it's more about being able to in protocol capture me um right those are equivalent right basically like because especially in a targeting if you whether you smooth it and then everyone gets 1% extra out of that mechanism then you can pay 1% Less on the base and that would up to the level or you directly burn it and then you can p p out 1% higher at the base kind of equivalent mechanisms um it just

it's it's it's cleaner to to to burn it but yeah in the targeting world I would definitely agree if you don't have targeting and you just use M burn then you just like lower the the yield basically for everybody but then again same same thing just because the mechanism is different but like same thing because now everyone stakes and so everyone gets access to it anyway so basically that our point is basically either we don't Target and then we go to 100% And then everyone gets access to that yield or we do Target and then we smooth and then or burn and then everyone gets access to that yield so basically it's inevitable that what however much external yield there is coming like via this validator role that will be distributed across all ethereum like eth holders at the same time those eth holders will then pay whatever

incentiv is necessary to have sufficient staking right but but that is an inevitable outcome no matter whether we do targeting or not it's yeah awesome okay so what are the next steps from here what's next for you with this research right and I feel like because so far we've we've been exclusively kind of talking about about targeting and I think that's good because that's really our main intention was to really kick off this kind of this this debate and if we take a step back it's not even you know you you hear like I at least I hope we were we were trying to be very transparent about kind of where we still have uncertainties as well right like a lot of this is like us just being hey there hasn't really been that much work on this there hasn't been a lot of people writing about this so here's how

we see the world how we see kind of the future for ethereum please disagree with us basically and so it's great I think on this in this conversation already we had like some productive disagreements at least that's that's you know how I perceived things well I'll go back now and I'll think about wait you know what if an ETF smooth it like Stakes stakes and smooth it out across everyone these kind of things right and and that's that's exactly kind of what we wanted to we wanted to basically say we already it's it's a little overdue already that we kind of take this serious as a topic and there's no the the point here is not to make any Rush decisions um it's it's really more about like basically making sure we figure out what is a stable end game is 100% stable like would we like

100% do we maybe need to do targeting do we want to do targeting what kind of ranges would be possible for targeting that's basically the base conversation now in the very beginning of the podcast you also mentioned we had a second proposal and I think on in a way that actually that that was the more maybe controversial one because we did we did make a propos proposal for a change in the upcoming Fork that that we will have in roughly a year on ethereum um and maybe like for the for for the last section of the podcast we can we can talk about that as well a little bit so basically the the um the problem we see today is that basically right now we have this continued inflow of stake right we will def we will the next time we can make a change on ethereum is this

upcoming PR Electra Fork that will be in 9 to 12 months roughly you know we never quite know when but like roughly something like that so we will have at least 9 to 12 months of continued inflow that'll get us you know depending on of course will we we we now have like a maximum infow rate that we introduced with with the with the the recent hard so um the maximum inflow now would be something like 20 million additional e stake per year of course we don't have to quite have inflow at that maximum level it could be less than that but you know to give you an idea so right now we at 30 million e Stak within a year we could be at 50 million Stak um if we don't do anything in Electra and basically our thinking is any Electro change would have to be finalized in in

the upcoming months we won't get to a agreement whether targeting is even the right thing to do because again that's perfectly fine that we don't have agreement on that yet and maybe that's actually the wrong thing to do you know for all I know um and the specification everything implemented within the next few months right so we will not do targeting in Electra now the alternative is basically doing nothing and then means like okay what's the next hard for that where we could maybe optimistically have targeting ready for that's like another year in the future so now we're talking 2 years in the future that's up to 40 million e inflow though now now we would be at 70 million right and up to 70 million maybe we need one more hard for to figure out all the details of of targeting then we up to 90 million right

so basically we looked at this and we like by not doing nothing it's almost it feels like the much more committal choice if we do nothing we basically we we we very likely end up with like a lot of e State more more than more than half very plausibly right and we see all these effects already you mentioned blast you these L TW that that that are already thinking about natively just taking taking the eth in the bridges and staking that you have ETFs that might at least consider staking what if we kind of at least plausibly move towards this world of 100% stake then of course all of those are rational choices you know I'm not I'm not criticizing anyone specifically for these choices and but of course that that that kind of creates more and more of this inflow pressure and we might choose to then go to 100%

And then that wasn't a problem then we just keep going on that on that path and then we go to 100% and everyone is happy and great if we choose to actually to end up doing targeting it will be a incredibly painful process to come back down right you basically now not only do all these protocols that just started to enshrine their their kind of native staking now they have to undo all of that but you also have to basically like push them out by actually having like really harsh yield environments for quite a while until they're all out right that is a very very painful process I mean of course you know if that ends up happening that way like look we we'll find a way to make it like as as comfortable for people as possible but it will be painful no matter what

right um so basically what we what we felt like is right we we on a path where we have two doors in the future that we see and know like if we do nothing we make one door really really painful to go through and the other door you know basically almost the default what we are proposing is hey in the next possible moment where we could make a change which is Electra let's just try to slow this process down so that we can decide which door to choose right so let's basically have a onetime small adjustment that's basically we don't propose to do targeting we have like a a colleague of ours and us who for independent reasons he have his has his own kind of independent line of research around minimum viable issuance he came up with a really kind of elegant set of

of of of Curves we picked one of those and we like hey it so happens that this curve would also just slow down the net inflow because it would it would basically just to give you an intution it would just cap the maximum amount of total e issued so as there's more stas it would be distributed across more more stas so that basically the yield would the the yields would come down faster as we SL slide to the right hopefully that would slow us down so that by the time we actually choose which door to go through both of them are still viable options and it's not like one of them is becomes really painful so we we looked this we like hey doing nothing seems really really committal we propose to do something that just leaves our options open so that's kind of that's the

thinking we we came to with a specific proposal just to bring them back together um could you just kind of concretely summarize one more time um just bringing the two proposals together of okay so this one of like the potential electric curve that would basically effectively just caps the number of eth that would be issued and then after that just um like there's no additional eth issued so like the Y goes down that achieves a lot of the goals that you were talking about with targeting um so what concretely would be the additional incentive beyond that to potentially do something like we discussed earlier with targeting where like theity of potentially like it would drop off to negative and people wouldn't actually do that so what would be the added benefit of doing that on top of the hey we've already capped the amount of eth issued

anyway right um yeah I mean basically as you said like the the proposal for Electra already has a lot of great properties and is kind of a stepping stone in a way or if you want to see it that way um and but the thing that it fundamentally can't um get you is any sort of guarantee about the staking ratio that you end up achieving um essentially the curve proposed for Electra still in in its shape looks very very similar to the one that we're accustomed to it just levels off a bit more aggressively because you don't have that con monotonically increasing issuance but it it's basically it Peaks

and then even slightly goes down and so the the whole point about a targeting policy and why we believe that's kind of critical for any kind of viable end game where you really don't have to touch anything anymore is to be able to make a precise kind of guarantee that we're going to be staying in a range of staking participation so say anything between like 20 and 30% is fine um and then after that the reward curve goes down dramatically all the way to negative Infinity um same for very low participation just positive and so you basically have like a guarantee that you're going to be um in this window of

State participation and the fundamental difference is basically having this guarantee versus um just having slightly better properties um or better is I guess opinionated but we believe better properties with uh the electro curve and and the more conceptual difference is just at what point do we say hey we think this is a set and forget solution and we basically start signaling and guar well guaranteeing a strong we like committing to this never changing in the future again right and like for now at least like it could even fall out like it could be this way and I think actually Anders who proposed these curves actually has Arguments for why these could be long-term durable um and if if we if we end up agreeing on this

we might never have to change it again the point is that we we don't propose it with that in mind we don't we don't say hey this would be the one time change that afterwards we don't have to touch it again I think it's really important that we and again maybe this is also why what we didn't do perfectly in the past right we should have we been more proactive in communicating the current issuance might might might not yet be the endgame State and so now basically the point was just like we with this curve we could not yet promise that that would be the end state it and as as KAS was saying I think that I we cband I think there's good Arguments for why this this would not be the case but yeah and and just to specifically sorry I just want to make it very clear

because we talked about a lot of things over the last two hours um just to bring it back together so what specifically are the properties that you want from this we guarantee that we stick to whether that number is 25% 50% 5% whatever because a lot of the properties that we talked about being a benefit of that are achieved by um they are they are properties of just saying like like if we just cap the amount of e that's being issued like then beyond that point like that you aren't being excessively diluted anymore um so like what specifically are the properties just to like make it very clear again sorry of like if we keep it at 25% versus if we just say that that's the amount of e issued right I mean I think fundamentally the question is why are we

staking in the first place right why do we have an issuance curve we have it to secure the protocol um recruiting validators essentially to secure the protocol and so um that's the main reason why we are staking in the first place and so honestly to me it's kind of funny that um any issuance policy that's kind of supposed to be long-term sustainable doesn't take this kind of fundamental staking ratio into account like how much Economic Security do you actually want so right now we have a curve where we honestly can't even like is it going to be 20 was going to be 100% I mean we're above at 25 so I guess we know that but the point here is like why are we staking for economic

security so let's make a precise statement about what kind of security level we actually want to achieve um and so that's why why um targeting is is kind of so important for the end game so that we can finally not touch the issuance curve uh anymore and and just to you know like John you ask kind of which which does not address if we kind of I mean yeah we maybe were a little unstructured in the earlier section around like kind of the the Downs sets we see with 100% if I kind of the way I would summarize them would be four points right like one the kind of the monetary aspects kind of replacing e with the worse money the two being just the kind of the the governance uh risks around ethereum kind of having this this external governance that is too big to

fail and potentially can kind of influence eum governance decisions three would be solar sta viability and four would just be kind of worse yield in in in general for for stakers and it kind of the Anders Cur makes what it makes better is the kind of the the the yield situation right because you you don't have as much of this negative the dilution kind of doesn't get so bad so so more portion of the overall incentive to stay comes from from actual yield versus just kind of um um dilution protection um but a lot of the other ones it leaves untouched right because we can still go to 100% it's still there's still the same problems around replacing the the the dominant money of the of the protocol there's still the same problems around governance uh risks and too big to fail there's still kind

of the the same worries around solar sta viability so again to the extent maybe that people disagree with these downsides maybe then they go away and this this curve is good enough but it very much depends on kind of continuing this research path around an analyzing it and figuring even out under which in which Dimensions to compare these these models yeah yeah that's awful and that's what I wanted to bring it back to because like of those basically like those kind of like four points the yield one is the one that I think is concretely quite well addressed by the curve proposal for Electro like that one is actually like I like I have a much easier time like getting my head around of like I I could see myself getting there um which is also separate from I would note that included in that

was also a reduction in issuance but you don't have to reduce the issuance you can just decide to like cap the amount of e without reducing the curve um which I think is actually more viable but but they putting that aside um and then yeah some of those other um benefits of doing the actual targeting or then further at least to some extent rely on some assumption that okay we keep the taking at 25% and then the synthetic representations of stake don't just like blow beyond that anyway because like if lsts go to 100% it doesn't like it didn't really buy you much um so so that is part of the question too is like okay if we keep it at 25% like hard capped how much does this actually stop the representations of stake on top of it

that's the part where I'm not super convinced um but yeah I mean like that is part of what a lot of the fun discussion will be um but that is I realized one of the other points I mentioned there is something that is worth noting is that um for the electro one there are two different points in there one is effectively like capping the amount of e that would be issued um another part of it which is opinionated is reducing the instruments um which are separate things like you can decide to like set a curve that basically caps the amount of eth without like bringing down where the current amount is um and that is something I view as probably more viable of like basically leveling off the increase in issuance that we haven't actually gotten to yet feels like less of breaking a promise kind of to the community as

opposed to like bringing down the issuance where we are today like if we tell people that like hey that issuance that you would have gotten when we went to 100% like we're getting rid of that but like we're not touching the amount of like issuance that's happening like now or anytime soon I think that is like much more palatable um because I I I think there's like a a gigantic problem SL risk of like if you like actually do something that meaningfully reduces income today um like so many staking businesses so many lsts so many D5 protocols whether it's Athena or whatever um like do get kind of rugged by this and like that is going to be a big problem um I think it was like Frankie who had like a funny tweet like a couple weeks ago it was like broke uh like Athena USD like the risk is like

that the St that that it deegs and it was like woke is like no the risk is like issuance goes negative uh which is very funny which is like like actually like a reasonable summary of this is like it's difficult to do that kind of thing I don't know that that is a palatable thing to do um and this does get into the other like speculative part that you were touching on is um um and this was I feel like probably what will be our last conversation is like so like what is the right amount um this is a completely like speculative I don't think anyone has a great answer to this but like the idea um that you're expressing is like we want to take some view of like what is like the right econ Economic Security um um that is like a sufficient amount like that is

supposedly like kind of the the fundamental reason for this um I like I I have expressed a number of times like I I actually don't think that the economic security is like I I think this number is like sort of made up and I don't know how representative it is um like the basic two components here are like what like what is the security budget that we are spending on a given year of like this is the amount that's like being paid out and then like for that what is the like Economic Security that we like get for that of like dollars of e state um and uh the kind of analogy in like the proof of work world is like um the security budget is again like how much are we like paying out in a given year to Miners and then um the like parallel

to the dollar state is just like all right like what is basically the cost of all the equipment and stuff like this like what's the hash rate all of that um because that's what you're kind of getting rid of like the difficult part for proof sake is like when we're saying these like numbers around Economic Security I like this is a conversation that's been had a bunch of times but like like all of this stake is delegated like do you in practice really feel more like economically secure if like coinbase is running all the validators and they have like hundred billion dollar staked or they have 10 billion do staked there is some reasonable Point once you're beyond like you just don't want someone to be able to like travely go buy up literally a bunch of steak and go take over the network um beyond that the risk is

really the operators themselves like what like what is their loss for like what is coinbase and figment and whatever PTP all these other different companies like what do they lose if they just like say screw it like go Rogue and like well it's like Fork the chain go like go start extracting me Etc and like fundamentally that is like what we're paying them at the end of the day like that is the security budget so I I think at the end of the day the security budget is fundamentally at some level kind of what your Economic Security like actually is is because like the incentive for these validators not to go screw the chain is whatever you are paying them in the future like like that is what they are losing like that is their future business um but yeah I'm just curious like how you think

about like because this is something that we talk about a lot of like we're like overpaying for security at the moment like what like what do we really mean by that and like what feels intuitively kind of in the right area for you guys yeah I think that's a great great question just maybe let me just finish with two quick last comments on the kind of the elector proposal and then I'll speculate um because I just wanted to to make sure people people fully understand that so so one just being that um often times you know when researchers propose a change like often the the ideas it just goes through the governance process largely unchanged is accepted I think people saw this and were like hey wait you can't just like push this very opinionated change through the government's process this was not the intention right like the intention was

more we wanted to be like hey this is one thing we could do please give input right we this this will not by itself just move ahead right like this is very much meant as something we need to get to agreement to with the community um also timeline wise there's still a couple of months before we would have to lck this in um and the other point just be that you're definitely right that basically it's it's almost you can take of it as two changes the the shape of the curve and then the the the the the the kind of the the height the extent to which that also comes with a direct insurance reduction and there's a bit of a trade-off the more of an insurance reduction there is there the more it actually achieves the goal of slowing down inflow right but yeah again

basically design space there we would want to come to agreement with the community it's not like just something that will just on its own just go ahead through the governance process um coming back to how much how much security is enough security how much how much do do we want right H I almost like I think of it I think of it in my my mind much more as like what is I find it it's it's a harder Define range like I think there's Arguments for I say even at 5% Stak we could probably be fine and I'm not even I'm not convinced that 5% would have bad properties I'm more basically I'm I'm but but I assume 5% is something most most people would not like um um and so to me basically it's more of a question at what point do these other kind of

this this down sides of a high staking rate become large enough so that we should stay away from that range so basically more like what's the upper bound what's the upper Bound for what we could what we could reasonably do and in my mind basically it's again this will require a lot of of of further work but some rule of thumb is that to me it feels like you would want to stay below 50% and I did probably not like right at 50% either so I could you know I it's not impossible for me to imagine that 40% is fine or 30% right 25% 20% something like that right the further away from 50 % we are the the more of a safe feeling I would get but basically because a lot of like say for example on the yield side right if it's a small

relatively small set of people being paid and a large set of people being taxed that just fundamentally is economics that work out quite well right because you kind of you have much more real yield than the kind of the the the dilution that's happening um you also have a much bigger kind of portion of unstake ether would still make sure that governance doesn't just take into account the the kind of the the the the the needs of of of state eth and kind of socializing would still be possible and these kind of these kind of things so to me in my mind basically 50% seems like basically some like a boundary where if we stay very comfortably below that seems things seem to be roughly fine rule of thumb wise and above that I would start feeling more more and more

uncomfortable but again this is of course just like just to people understand this is really speculation at this point just yeah I don't know Casper what's what's your what's your feeling Power of Two um but yeah no I mean I guess if we if you want to stick to the powers of two you have 1/ 188 1/4 or one half half I think what you said um basically agreed um then I guess obviously um you need you need that to be high enough for Economic Security right and so then you can start thinking about certain costs of attacks and how much e you would need to be able to pull

them off um essentially you would want it to be hard enough for any kind of one single entity to control enough e to to pull off some sort of 50 51% style attack um although in proof of stake obviously you have this accountability so you actually get slashed unlike in in in proof of work so especially because of that the argument for having an incredibly high security budget yeah is is um tricky I mean I think right now I if I remember correctly from a couple of weeks ago Justin had the this 10x Bitcoin Economic Security flippening tweet um where ether is at roughly 100 billion today and and

Bitcoin at 10 um but yeah I mean then um yeah I think um it's inherently subjective um there's no like d answer but um 1/4 or even 18 sounds sounds like a reasonable Economic Security Budget now is that then obviously still kind of debatable if that's desirable as a kind of Target but in terms of Economic Security um that seems reasonable to me so my I think my general View and I would say at this stage it's still developing so I I I kind of only started

thinking about this uh I think more more recently than like maybe some of you guys but to me the the impact on the value that I set is the like the relationship between kind of the number of stake that that we're targeting and also like the the the kind of demand curf I like how how much money we're paying Val us and like the quality of the Val that us set for me that Still Remains kind of very underexplored because well sets are not the same and like we talk about Economic Security As and we're trying to kind of make the quality of a Val asset really one-dimensional to this number and then we can say oh yeah clearly it levels off at some point so we should Target some some lower amount and if if that was an

accurate representation of reality then I would agree but I I feel like that raw Economic Security is very overrated uh Geo distribution and political distribution of Valas across different jurisdictions and countries and some level of civil resistance on top I actually find very underrated um and so it's not know it's not to me it's like not nuanced enough to say Hey you know just staking is this kind of service you know it's those guys they do something for the protocol and it doesn't really matter who it is because there's actually a huge difference in who it is and kind of different incentive systems produce very different outcomes and to me this like

throws some level of shade on the like like ideas like targeting but also me burn which I see in some ways as very similar mechanisms because they both try to internalize uh external reward slme and like try to basically set like some Target level of real yield for staking when I think like in the limit if you kind of take two two the two cases and I think it was you ano who said this previously that there's no difference between like all the money like all money is paid to e holders and no Mone is paid to stakers versus no money is paid to stakers uh uh all the money is paid to stakers versus no money

is paid to e holders you said like in the limit those are the same if I kind of understood you correctly I think I think the more I think about it the more I think they are not the same uh because if if all of the money is paid to the E folders then there's no no statement about what that does to the valat set I think if you have if you just have some base like amount of real yield to e folders even accepting that that you know that might make the staking ratio go up to 75 to like 100% which it probably will right if you have like some meaningful nonzero real yield then I still think this buys you something very significant which is the ability to

have decentralized taking protocols um and yeah like pools like rocket pool pools like Lio that basically are uh you know just like more decentralized kind of derivatives of of Eve and they exert some very like positive pressure I think on the quality of the Vue that I said well there's several points I would say for one I think in terms of the actual Staker yield like my my intuition is actually because I've heard people say that my intuition is actually the real yield for stakers is probably the highest it's something like 5% Stak and then it just goes lower if

we target like a higher value and I think people people seem to have the the opposite intuition so this is something we just have have to keep kind of diving in and figuring out but my prediction is that it's actually the opposite kind of relationship that actually if you want to maximize yield because that enables for less efficient stakers to still exist that actually I think that happens if we limit the amount of of of of people staking so that's kind of point point one I actually think basically that my intuition there is just the opposite so we have to kind of keep keep kind of diving into that topic but how can this be true I mean if you just limit if you just limit the the if you target 5% Stak I mean right effectively I mean it it almost doesn't matter what

you target but you always end up I mean it's just like you you're basically saying are there like 5% who can do it extremely cheaply and then you Bas basically in Shrine like that 5% but that's not that's so so so right if you actually it's probably I don't know like really really needs a whiteboard but um so right so basically if you think of him it is just income streams like I think it's phasically inevitable that all the transaction fees all the me all the reaking basically ends up going to everyone and everyone basically here just means that if we don't do it via me burn via targeting or something it'll just implicitly go to everyone because everyone will stake right so basically if you have 1% of the assets of the network you will get 1% of that external yield I think that is inevitable like if

you could Point yeah okay right so basically then we can exclude it from from from the analysis right so that that I don't agree with no no no like okay it's like but how how not agree yeah that's because that's what I mean exactly like you're saying it all the money going to eold us and none to the stakers versus all the money to yes but it still produces very different Val ass sets that's I think the distinction between those two because when you have the money paid to the when you have all the money paid to the E folders and none to the stakers you end up with a very poor valed asset and if you have the same money paid to the stakers even if they the exact same set as the eold do

then you end up with a significantly stronger value that set but so then what what you're just saying is basically you you think the only way to have a have a strong Val ass set is at 100% so that's basically you're saying the argument is because otherwise there is no this is not like I mean there will always be like 20% of folks who don't stake uh but um yeah I okay and I basically I disagree just because I think at 100% basically these kind of competitive pressures on solo stakers would be so high that we would not have many solo stickers left and also again that like say if you compare coinbase and li like it's it's it's almost it's a little unfortunate because I think you know in the right after we published this post there were some some people that tweeted things like you know all these E3

Searchers that they must they must hate Li and what not and nothing could be further from the truth I mean look I I also also coinbase you know of course as a company but I would sleep much better if if like you know most of the stake e is via Lio and not via coinbase right and because just Lio is so so clearly more decentralized like has all these great properties like you know nothing and I would sleep even better if it's all solar stakers of course right so so there's like a Cascade of kind of of of of of preferences definitely and but I think basically if we if we go close to 100% it's just that the entity that only have to has to run one big note and then this basically has no no kind of no variable costs on top of that will just

win out that competitive Dynamic that to me is just very very clear whereas if we if we basically limit limit the the kind of the scale of staking there's much more room left and again kind of yeah it's it's worth kind of again taking out a whiteboard and actually diving into this and that's why we we just wanted to start the conversation here but but I really don't understand like to me basically in my mind it's clear like again these external revenue streams whether we want it or not they will over time go to everyone equally like no matter what scenario we are in and to everyone hopefully equally like in a smooth way where there's no variant that could affect things right so so that's kind of point1 the only other income you will have as a Staker is from issuance and issuance yield is only only

meaningful if not everyone gets it and so basically like the the more we have this distinction between the group that gets it and the group that doesn't get it basically the more you actually still have issu and as a meaningful contribution to kind of your staking income and and and and and that and the story so so basically in my mind again I think it's just really important to basically take out a piece of paper and really kind of like draw different plausible endgame scenarios and then compare them and mind definitely we still also at the level where like a lot of this is also still intuition based or like some hypothesis some assumptions we've made that could be wrong and I think some of them will be wrong right so so I I would love to be corrected on this and I I think we have to you know

overtime iterate on our views um I would love if basically I would I would really love like if over the next few months we start seeing like more people being like hey here's my specific counterproposal of the alternative world we could get to and how the incentives look like this and this is our argument for why for these specific groups of people it would be preferable right because there's not only I think by the way one distinction that is important to make is say on the on this on the yield side there's for one there's the yield that goes that's going to stake us but then in terms of say the feasibility of something like lier right that is not only dependent of the pass through yield that they give to stakers but also of the the the revenue that that's that LI or whoever you know

whatever other SSP there is um themselves can can generate right because of course in order to pay the operators The Operators are not the state are not the capital providers right so there's one there's the staking yield itself and then there is how much much revenue can you generate and so those might be different like for example I think the the staking yield will be optimal at a very low ratio but of course the I think the the SSP Revenue will be higher at something maybe like 25 like my my my intuition would be that SSP Revenue Peaks at 25% state or something because then they have the like you know then then there's still a lot of real yield generated so you can actually charge a fee a meaningful fee that you know like adds to your Revenue but then you also have enough assets on management that it's

like in a sufficient amount of time so so maybe like something like that would actually but once you kind of cross into 50 70 80 90% then basically because all the real yield goes away over time the competitive pressures would just erode the the the the the fee that the ssps can charge maybe except for the the dominant Winner Takes most like basically there's one world where one one LST runs away and basically just becomes the dominant one and then that one can charge Monopoly prices so that's one way of keeping an SSP competitive even if it's like less efficient maybe more decentralized but I think it's that one is just such a high price to pay if if the only we can keep B competitive is by basically enshrining it as the replacement money for eth I think that would be a too too high of a price to

pay yeah I think that's fair I think I'm I'm my main hangup I would say in a lot of ways is like with the idea of kind of me burn and and kind of targeting in general I fully acknowledge that uh solo staking is harder without smoothing but I think there are better Solutions than than than what we have seen right now and and my feeling is that I mean rocket pool already does a lot basically to smooth me for solar stakers uh Lio by end of the year wants to have I think a thousand independent note

operators uh so like there are modules kind of being worked on for for solar stakers Etc right and and those might include also m smoothing and I feel like um yeah I I I I just feel like it's uh I have a better feeling about the Val that I set with kind of the the real yield not being internalized and then turned into issuance um but rather kind of going going to the stakers and addressing the volatility problem another way I guess the point though is is that when 100% assume me burn right yeah like um there is no exogenous yield from in under the current issu

poliy or under the Electra curve or under a targeting policy now if you stake 100% there is no real yield all of the issuance everyone gets and so there's no real yield yeah what we're saying is that if we move to a world where less is St say 25% or 18 or whatever there is actually redistribution possible between stakers and E holders that is Meaningful and so that that's where we say there's actually real yield to be gained by by stakers I think I think's disagreement was with the assume meth burn I think the PO point was like What if we don't assume meth burn yeah yeah I'm just not a fan of like the idea I think of me

burn in general but to your point Casper it's uh it's I mean it's like holding Eve in this world is like not also equally ractive right I mean it's uh why not I mean yeah you can say like hey we give this like one party real yield because we have like real inflation on the other side but then the other example like there is real inflation but also real yield like and it kind of cancel each other out right I guess I'm I'm saying is like do you want to be in like the 80% that's kind of being inflated or do you want to be in the 20% that's I think I guess the argument here is that this kind of separation between uh

the kind of specialized parties that are kind of um you can see as an e holder in this 25% State world you can see holding e you you have a tiny amount of inflation that is basically a fair tax in quote unquote if you will um as a kind of this is your kind of share that you pay for getting a secure chain and the whole point is to make e um kind of this cheap to hold money but at the same time making sure it's maximally trustless the money that we want is maximally trustless and very cheap to hold obviously it can't be free because then

we have no security but the whole point is to find something where we think it's secure enough but we're not overpaying for no reason in terms of security and so e is just fundamentally maximally trustless on ethereum and more credibly neutral um and I think just to speak to your point about the kind of stake distribution that different issuance policies might induce I think that's a very important Point um H I guess a question that I would have for you is um the point about basically underlying competitive pressure between different staking types existing under any issuance curve so Lio relative to

coinbase relative to Rocket pool relative to solo staking have fundamentally different cost structures and these cost structures are true for any issuance curve and so I think right now we have a market that's clearly not mature like coinbase charging 25% um I don't know I feel much again also echoing Ana's point like I feel much more comfortable holding something that is on chain like L um Steve or rest or sort of sing but um the point here is that I would would like to better understand how why you think these competitive underlying kind of Dynamics don't play out under the

current issurance curve but wouldn't uh would play out in in a kind of lower issurance curve I think that you basically have to assume what Ana said that um if I think the Dynamics are different if like 80 to 100% are sted um because and there's no me burn right so I think in this world you can assume just like a higher kind of real yield for all the participants of the staking Market that has less competitive pressure right I think like the the the less Revenue there is like to distribute overall I think the the stronger the competitive pressures get if my

intuition is correct and the more Revenue there is in the staking Market overall the kind of the the more kind of safe spaces there are also for I would say decentralized providers nonetheless it's very important that um you're you're shaking your head you you don't agree um I mean I guess my my question um would be if the market was competitive and one staking solution is so much more efficient than another then you can price out the less efficient one even at I don't know we can double issuance the whole point is M um um the competitive the one entity is more competitive and

can offer more competitive yields that's a fundamental kind always the case absolutely absolutely yeah I mean it's the the like the data between like what different providers can offer you know is bigger than you know the the the the perceived difference in a lot of ways is like bigger than like half a per and 1% than like two half% and 3% and I think this is like a real this is a thing that like really happens uh so I think that that's more what I'm saying but again like I I think it's important like at least our our thesis or like what our conclusion is that the harshest un environment out of all the possible ones is the one where we go to 100% that is at least how we understand

the situation with M burn no no matter what because again at 100% I think basically whether we have m b or not becomes completely irrelevant other than for well let's say ideally we still have some sort of smoothing but it doesn't matter whether you pay it out to stakers or to e holders because those are exactly the same entities at 100% so again and and like I think if we take a step back I think what's happening what I see what I think is happening is more that unfortunately proof of stake in a mature system in the end game in an equilibrium state is really really hard to make work in ways that do not collapse into these kind of um monopolist situation and all this kind of stuff that is inherent in in proof stake you know like pro pro work people

always like to point out how proof of work is so much simpler so much easier to reason about and that's not wrong I think the payoff is worth it I think actually you know like I think the all the struggle we're going through right now that that is worth it for proof of stake but it's still it is fundamentally I think just really really hard to make to get to any good place and I think by basically being the first ones that describe a specific place you want to go to it's very easy to point to all the things that are wrong with this place and I agree a lot of these things are especially compared to today clear downsides now the problem is I think today we just can't again I would press the button in a hard to just freeze today we can't like we will not stay in

to today so I think the problem is the moment people start to actually write out the specific counter proposals we will have the same situation of like being like Oh but but this is wrong this these and these and these and these things are wrong with that alternative proposal and and after looking at this the one we propos basically the kind of the targeting at some you know something call it 25% whatever we think out of all the equilibria that is the one that has the least downsides and and we could be wrong with this but at least basically I think the problem is it sounds so bad because it's the first concrete proposal and you compare it to today and it's worse than today which acknowledge right it's just to you know and and so that's that's what I think I think is happening here

but that's why I would I love I really would love to start seeing other people kind of like write out their specific kind of endgame equilibrium state that they think is achievable and maybe just to to add on this like I think intuitively why these proposals seem so kind of counterintuitively or they they just seem bad for stakers right is because you look at the current yield and you're like oh it's going to go down right yeah but that's the mistake you shouldn't you should not compare the status like the the current issuance level we have at 30 million e what we should instead do is compare long-term equilibria and then understand which equilibrium is actually better for our staking set and I agree

the stake distribution should be fundamentally part of the equation and not just Economic Security like I 100% agree um should it turn out that the only way we can keep solo stakers around and more decentralized onchain versions um of staking providers is like high issuance in 100% state I'm happy to be convinced but I I just personally don't um see the arguments for why this is the case and in fact I think um because of the economies of scale um of different staking types actually limiting them and making the yield uh sustainable is actually net benefit for for for

stakers yeah I think you make a lot of really good points I think this is probably like a good good kind of General tone to end it on so don't don't compare the proposals with the status quo because it's it's not meant to replace the status quo it's like meant to paint kind of an endgame vision um and I would love to see more analysis B on like where do we see the current equilibrium end up right so what is the what is the current equilibrium like what is so bad about it that that kind of warrants uh you know this kind of admittedly like pretty dramatic change uh how is it impacted also by proposals like me burn which I think you know when like Justin goes on stage and says like

you all of the e should be burned you know like it sounds very like it sounds very it sounds very good right but it has all of these nuanced effects I think on the staking set that uh that that that maybe haven't been explored in kind of all the all the the the detail that is warranted so yeah I think those are generally to our audience and anybody who's listening who's on E research who cares about this topic um yeah like paint us your uh your kind of end game scenario right like where where do you think what's the kind of path that that you see us on right now what is the path that you want to see for E theum long term right in general like starting from

the valid that are set that we want to see and then reverse engineering kind of what do we need today what do we need to do today in order to to get there I think that's the right way of thinking yeah and maybe two closing comments for me there one I think it really is unfortunate the timing BAS Bally around this is unfortunate but basic ideally we should have this conversation this should be a research conversation we should have it for 2 3 years figure out everything make sure the community is involved and only then even start thinking about how to move there right that that is the proper way of doing these things I think a lot of kind of the push back came from this kind of Electro specific proposal which you know it's like hey let's change the issuance mechanism in six months or like

9 to 12 months yeah which you should as a rule of thumb you should not do right I agree with that right and so it's it really just happens to that I think we we are very unhappy with kind of this this default drift that we are on right now which kind of makes this entire thing on like an on this unfortunate accelerated timeline which I am not happy about and I don't think anyone is right and so that's kind of just unfortunate and then the other thing is yes it's really like we don't think we have figured this out yet we we have like we have some rough understanding we want to see other people's uh positions there are some aspects that we haven't touched on at all right so for example I think some push back came from the DVT people who are like hey you know you

don't mention us at all you're talking about solo stakers is the lowest kind of the most decentralized part of this but like we could even decentralize further than solo stakers what if you have 10 people running one validator right so totally valid right like expand expand the analysis really kind of dive into all of this and it will take time that's why again we we we don't we don't even we don't think it's at all feasible to jump to the end game anytime soon and what do we do in the meantime while we figure this out right like these are topics that I think the more people look look into this and look into this you know hopefully like you know soonish um share their opinions kind of so that we can hash this all out I think I think that would be super super important and

I'm I'm very happy that you had us on the podcast because I think this is a good great venue um and I think yeah you you you guys also push back quite quite productively today um pointed out quite a few things I think we all our thinking is also still still quite quite kind of raw and and not fully reflected um which you know it isn't and I think that's that's super important to acknowledge and to to then take as prompts for further for the work and I think yeah I think this was super productive yeah it was um amazing thank you so much guys um yeah and just like I think one one thing that basically yesterday would have been the best time to talk about endgame staking economics and I guess the second best

time is today um and so here we are unfortunately in in etherum land often things get thrown at the protocol and then we kind of try and figure than not um but um yeah super good to have this conversation and uh productive push back as as Ana said one last quick thing um that's probably worth noting I remembered it when you were talking about the um like you're bringing up the proof for work um differences and I I was thinking a lot about proof for work yesterday when I when I was um like preparing and thinking about this episode and one of the things that I was thinking about and again to the vein of like actually there are plenty of fair criticisms from Pro work of like yes it's a lot simpler to

reason about um the other one that is always a criticism is like oh it's like proof of stake is token holder governance um and while I obviously do not agree with that um it has given me more appreciation thinking of it over the past day of while it is not directly token holder governance there it is very clearly a lot more politically complicated to do this in a proof of stake system and you saw this very clearly and we just reduced minor salaries multiple times and then literally fired them and like it was totally fine and like no one really cared um it's just like all right your employees you're fired like we're cutting your pay it's fine um doing that in a proof of stake system when there are many different uh stakeholders in that system with it's literally

stakeholders the asset holders themselves and then a lot of defi protocols businesses Etc that are very directly built off of these cash flows now um it it is m a much much trickier obviously politically uh political environment um and have that discussion um which is just like just an interesting thing to kind of observe and think about more um which I've definitely appreciated more like thank you about and in the light of that strong shout out to Mike nuda right and his article on set theory which I think exactly makes this point yeah yeah amazing article and pro works great I mean I'm a big fan I feel like these days the eum community is like too way too harsh with it like I mean I'm very happy we moved Beyond it but it's it's still great and uh I hope you know Bitcoin stays on it forever and uh yeah

also not to forget I mean proof of work has also very strong uh centralizing forces um around like in the end game I mean it it kind of it used to be uh you know more around the the price of energy uh but that was always kind of going to change uh as we kind of move to like more and more kind of capex heavy machines and then it becomes all about the cost of capital and it's like no surprise that like every year the amount of like publicly traded hash rate goes up because it just gravitates to markets that have the cheapest cost of capital and uh already we have like 55 60% hash rate in the US and that that's just you know unless the US collapses that's

never going to reverse uh there's no forces that could change this um and so uh like proof of work uh you know while being easy to reason about doesn't make it doesn't make it the better solution well luckily we only have to fix proof of work proof of stake and make it work and uh we can leave that to the Bitcoin as to figure out yeah vision of Labor FTV okay maybe maybe one click one last point to to John's comment on like the differences between kind of Miners and stakers um completely agree right so obviously there's this like notion of e stakers being fundamentally aligned with the the the the underlying protocol um and that's certainly true um

but I think again for me or for us I guess the the takeaway of our proposals is that we're we're not rugging stakers we're actually trying to help them um I think that's like a foundational misconception or like this like again this very natural intuition that hey issuance is going down must be bad for stakers um we're saying it's actually not it's it's actually just um less stakers but at least that portion is earning sustain real years um and then um yeah now I forgot my second point but um right uh anyways probably a good point to

end okay guys thank you so much for doing this debate I think this uh I mean this has by far been the longest episode I think of the four that we did uh so yeah this has been a blast thank you so much likewise thank you so much us thanks for joining us today as always nothing we say here is investment or legal advice the views expressed by the course are their personal views alone please see our podcast description for more disclosures if you enjoyed this episode please feel free to subscribe and share it on Twitter thanks and [Music] [Applause] [Music]

goodbye