# Maximum Viable Security (MVS) – a new issuance framework

- Channel: [Devcon](https://streameth.org/devcon)
- Date: 2025-10-07
- Duration: 26:38
- Watch: https://streameth.org/watch/yt-6k3S_pEQgBM
- YouTube: https://www.youtube.com/watch?v=6k3S_pEQgBM

## Description

We derive a new framework for analyzing Ethereum Issuance, based on Ethereum's core values: security and neutrality. Upon discussing various attacks on Ethereum, we study future growth projections and the importance of diverse validator set, and conclude that Ethereum's defendability is the key factor for issuance policy evaluation. Via MVS, we show how the current issuance reduction proposal is dangerous, based on the future staked ETH concentration with CEXs & impact on solo stakers.

## Transcript

[Music] all right uh hi everyone happy to be here my name is uh AR kilki and I'm going to be talking about this paper of Maximum viable security and it's really a framework for how to think about ethereum misss and it's a work in collaboration with adcv Danny mcnut and Kim from steakhous maybe I'll give just a little bit of background because I'm I'm kind of newish to crypto research scene um I have been mathematician for about 10 years studied topology and geometry pure mathematics and yeah over like past six years I've been slowly sort of transitioning to crypto research and uh I work at Cyber fund right and cyber fund is an investing firm and it's founder Le and most important thing about it is that we um we try to accelerate transition of um our economy to cyber economy and by cyber economy I mean an open uh efficient uh cyber economy where everyone on the planet can collaborate right um and the reason why we're so focused on crypto is because crypto and blockchains is a value substrate on top of which this cyber economy can be built quick disclaimer cyber fund holds eth ldo and is investor in P2P so what are we trying to do in general uh about ethereum issuance policy um I think pretty much the call is clear we want to develop a a stable sustainable issuance policy um and I wanted to take a step back a little bit and and think okay what could that really mean and I came up with this four simple steps um and the first one is set the Northstar and have a guiding framework second is collect the necessary data third is analyze scenarios individually and fourth is analyze scenarios relative to each other um and I I would argue that we largely have been focused on number three um uh research led by ethereum Foundation but I think all other threes one two and four are um also very very important uh so let's end basically those four step is is the plan for my talk so let's start let's talk about North Star and framework so first why do we need a high level framework well the first thing is that it's an incredibly complex topic with many many factors right so we need to kind of find find our way through all these factors and second it's very contentious so we need to Rally people around uh a Direction so um that's basically and it's in my mind it's almost unavoidable able if we want to come to a rough consensus right we need some sort of Northstar framework which will guide us and which will help us agree on something next is how how uh do we find this Northstar what to optimize for this one is Big by the way uh for me I think we need to spend like a lot of time talking about this what to optimize for there's three things I think that are floating around one is Network ethereum right um and is asset eth and a third one stakers interestingly right um and I would argue that uh the framework and proposals that have been suggested by ethereum Foundation mvi Stak cap are largely focused on the asset eth while our framework maximum viable security actually is more focused on network and um this is this is basically the the the starting point of our paper is that actually the network is the most important and it's not to say that eth is not important or stakers are not important but it's important to understand that the the value flows from Network to asset primarily of course there is some value from the asset to network as well but primarily value from from Network to asset and also uh am uh about the stakers this one is uh like um you know we we all really love stakers right um and especially solo stakers because they bring most decentralization but let's think about why we live love stakers and we are stakers by ourselves so we love ourselves as well well it's the reason is because they secure ethereum right it's not just by default we love them right so when arguing about ethereum issues we cannot think about stakers without thinking ethereum Network we always need to have this in mind so we cannot just build something that is good for stakers because stakers are providers of security for the network and uh uh uh I I strongly feel that and I strongly feel a little bit of um confusion in the community that you know we serve stakers we like all our conversation about stakers but they are about stakers because they give us something right so we need to ground ourselves in the root cause which is the network and I'll come to that later of course and so what should be the framework well I would argue and I think most of people agree actually that the Network's uh highest value that it brings to the table ethereum's highest biggest value proposition is security and neutrality um so yeah so now if we take security and neutrality as a framework and I think this there's largely there's a large consensus that that's the Network's biggest differentiator what what should we think about well we need to to think about it attacks and there's three types I you know there's more but these are three main types of attacks if if actor has 34% of network it has D it can double sign uh and this attack is is interesting because it is seen by protocol protocol can detect it and there is automatic slashing there 51% attacks are trickier because uh you can do short re works you can also censor and these attacks they are not seen by the protocol and gradual coercion by State actors is an attack that we didn't really see but it's very possible we see it and it is also not seen by the protocol and I would argue that uh ethereum's autonomy it's a concept that Gab Shapiro uh came up with ability to resist this these attacks that are not seen by the protocol is a very important thing to think about right because if we don't defend against these attacks uh in some other way not by not in protocol but out of protocol we are vulnerable so what are the defenses well as I said um there's cryptographic slashing for a there is social slashing for A and B uh but it's pretty politically messy to do uh and also like if if we want to automate it there are ways to do it to some extent but it's very complicated and you cannot do it fully right that's why protocol doesn't see it and I would argue that you know the core uh defense that ethereum has against actually all these attacks is the decentralization that comes from distributed validator set um ensuring that no single party controls any of the big thresholds uh of network and this defense is preventive it's apparent it's ethereum's biggest differentiator if you think about other networks uh eum has the most diverse value that they said by far and I would also stress that it is is key to neutrality okay neutrality comes from this fact um so I would argue we need to think about intentionally about this defense mechanism and so the upshot of framework is that we we suggest this mvs framework Maxim valuable security and neutrality and it largely comes from geographically distributed and disconnected validators and if I were to contrast this with the framework of mvi it mvs strives to maximize security of course you cannot forget about asset you cannot about forget about um stakers so without compromising the rest for example scarcity of fees um and mvi is is sort of diff is the opposite IT minimizes issues prioritizes assets without compromising trying to with not compromise security and and maybe a slightly different way of to think about is that mvs prioritize expansion of the network and sustainability right over efficiency and mvi does the opposite okay coming to the second Point um of of the plan is the data collection right and this is what I wanted to focus a little bit on uh because uh I you know there's there's basically a whole stake in economy uh built around ethereum staking and I wanted to uh maybe talk a little bit in detail about it so first what types of stakers uh there are well obviously there are SOL stakers and they have actually about 10% of the market give or take onchain Tech so these are people who are onchain and uh um yeah basically onchain but without really sort of big restriction big regulatory restrictions or something like that and then institutions they're also on chain but they come with their own sort of restrictions and preferences there's retail and then there's ETFs which are not yet not there yet but they're coming and we saw yesterday the that ETF isure actually acquired a test a test and so we know they're coming um so now the next point and this is probably the most important Point uh that I want to con is that the majority of the staking Market 90% Works through delegation there's no way around it that's the that's how the market works and we we need to think about this intentionally and we need to I feel like we need to S almost accept it because most of the talk is about solo stakers and of course we want to support them right but we really need to think how ethere emissions will affect uh delegated stake so what are the types of delegations so first is centralized exchanges uh they have 20% of of the market um and 15% by the way is is is by coinbase and they charge a quarter different there are different fees but roughly they charge actually 25% a really high fee why they uh they charge this High fee is because their uh you their uh users are sticky right they they don't want to go on chain and they just want convenience okay and as a as a result there's very big sort of network effects there economies of scale and it's a not compet it's a really not competitive market and it is fairly concentrated within coinbase and binance next come centralized staking providers so these are F like P2P Kil figment and these are uh you know interesting they are a little they are really it's a different Market segment from centralized exchanges they serve institutions a lot they serve onchain teet and they charge way less fees one to 5% why it's a very competitive market okay there's many many of those and because it's a competitive market it is decentralized so these are two sort of let's say straightforward way to delegate uh state right and then there's a there's one which is uh trickier which we all know is these uh decentralized pools uh liquid staking uh token pools and liquid raking token pools like Lio ethery rocket pool Etc so these are it's important to understand that these do not really run node operators they do not valid validate ethereum they are middleware that take stake primarily from onchain Tech case and they distribute this stake into various validated nodes so uh for example um yeah some some distributed to solo stakers like rocket pool or lier with their uh Community staking module but majority actually is uh distributed into centralized staking providers and the core point is this you know I have these three arrows here three hours here is because it really distributes it uniformly which actually brings a lot of decentralization to ethereum so and that's that's the that's that's sort of uh one reason why exist and the the other reason is because they provide also liquidity so you you can notice that they charge roughly 10% there's different fees and the reason why it's higher than centralized staking providers is because uh people who stake with LST and LT pools they get liquidity right and so they can get um defy yield on top so that's the market structure right and uh pools have 40% centralized taking PR 25 and centralized exchang is 25 okay um all right so I want to stress one other thing about this whole the these markets I'll speed up a little bit that actually I feel like the market the stake in Market is kind of subtle it's not straightforward it is actually segmented it is segmented into centralized exchanges with very sticky customers who are inelastic to yield essentially I'm not sure I'm not even sure they know the yield they just think that um they know that they're staking and they're earning something but they don't really care like is it two two and a half 1% or something like that right and then on chain segment with csps lsts and lrts and solo stakers where we have sophisticated um participants who actually uh kind of go between each other and I would you know a lot of uh arguments uh in the research they sort of at some point say well it's a competitive market so the margins will compress either way I personally don't think so and I do not and I think we need to try to let's say really segment this Market the reason why is because if the market is not segmented margins compressed we do not have hope for decentralization centralized uh delegated stake staking is way way cheaper and Superior so if it's all high hyper competitive uh there's no chance to decentralize the stake so we really need to find a way how we're going to distribute the stake among many participants and this is I think one natural way as it happens and the natural way is that sticky customers who are willing to pay 25% uh fees with with centralized exchange they stay there and centralized exchanges their optimal solution is to not lower the fees not compete with onchain segment their optimal solution is to charge higher fees you can think of this as Pension funds Pension funds you know uh even though they seem to not charge really high fees but in if you dig if you dig down into the mechanism they actually do charge High fees it's because their their users are extremely sticky okay so I feel like it's an important message of course I mean there's there's a lot a lot of analysis needs to be done here on the costs on the fees on the um on the stake elasticity and but that's how I feel right now and we at Cyber fund uh have a grants program MBI grants program where various teams dig into all this and so watch out for the announcements of various reports that we're going to publish okay so analysis uh of the impact right so from what I told you what do we think will happen if we cut the issues well as I said the sakers are sophisticated with LST LRT and C PS and so as a result you know that uh their yield will lowered will be lowered and they will churn and so the decentralized market segment actually decreases centralized exchange stakers are inelastic and so they largely don't turn and actually that results in centralized exchanges gaining market share and solar stakers will turn um and you know as as ongard and uh kaspar said potentially at a lower rate right because they have very diverse cost and preferences however we also analyze in our paper their cost structure and actually they have really high cost High fixed costs and they're very inflexible so their supply curve is maybe steep marginally but I think it's very low kind of at the lower end of of yields so at some point actually there will be a big Cliff a big jump a big drop of solo stakers so we need to really be careful right about solo stakers uh and by the way the you know the the amount of solar stes will decrease and they will be unhappy that's for sure maybe we can argue that the percentage of solar sters marginally will will increase but will it Mar Will it increase long term I don't believe so and that that's what we show in our paper and then on staking providers and middleware this is the more subtle thing what's the impact well um here lsts and lrts they have high costs and the reason why they have high cost is because they they distribute the stake among like 40 uh node operators so some of those noers will drop and so what the result of this is will be just loss of decentralization or or red reduction decentralization centralized exchanges preserve customers and maintain margins because they just have really really low cost and csps lose customers but they actually maintain margins probably due to flexible low cost so so they cut on their expenses on their personnel and Etc and we show in our paper that it is possible so they will likely maintain their margins okay so what's the upshot well uh centralized Exchange Market segment largely untouched the decentralized market segment of LRT LS LST and CSP feel the pressure so the issuance cut hits this Market segment and so that some stakers and some csps leave uh and as a result centralized exchange market share grows and we end up with less decentralization undermining security and uh important thing is that we actually try to model this scenario in the paper and right now coinbase has 15% share huge unstake reserves and and ETFs assets are coming coinbase is the main custodian for most of the ETF issures and so with the issurance cut we show that with all this information combined there is a very real scenario that of coinbase achieving 34% of the network um in particular under 40 million State e cap right so these kind of uh these alarm that I'm alarms that I'm that I'm making that uh market share can sorry decentralization can be lost are very real um and yeah I encourage everyone to look at our paper and read the analysis um and yeah so I also wanted to sort of uh try to apply our framework this Maxim valuable security framework to some of the AR arguments that researchers have been discussing so the first is real versus nominal yield so one thing that you know as a mathematician I really uh you know appreciate the beauty of that uh whole argument and construction but from the Practical standpoint I feel like it's uh concerns with stakers and not the network right because we're talking about the stakers actually earning less than nominal yield right the real yield is is less and but in some sense network doesn't really care about that Network cares about how many of those stakers we have so I feel like that argument is is a little bit not in the root in the root goal of of what we're doing right and then current issuance is costly it's another argument and we in our paper estimate and actually Anders also in one of his papers estimates that actually the safe cost from cutting issuance are in single billions right so it's about four but give or take and if you think about the market cap of eth the Vol the trading volume of eth it's apparent that this price is relatively low for security assurance that you get from it so the the saving on costs uh is like on balance doesn't doesn't is is not as valuable as the security and neutrality that ethereum gets from issuance and so the last one is boosting monetary hardness of eth well there's several points here so first is Bitcoin won that market right the there's a there's a growing consensus Bitcoin is a store of value it's very hard to uh sort of compete with that but even more importantly in my mind if money doesn't really come from hardness of its in you know issuance rate it actually may decrease if we change the monetary policy because we uh undermine The credibility the sustainability of our um issurance policy and also security losses like if if security or neutrality sentiment is is suffered we you know e Mone will suffer for sure because the network and eth are very very connected um and then yeah and then I would I would even argue that if we want to like if we want to think about e and if we want to think about ethereum contributors and how we really around coordinate and rally the whole Community right uh actually bolstering the network might be the best way to go yeah I know the time's up I I'll I'll speed up in just one minute um yeah and then there's an attempt like really quickly on relative analysis right here we didn't do uh there's there's almost no analysis out there I would say that you know there are arguments uh Pro Cut there are arguments against cut but uh on balance it feels like AG aggregating State eth on the transparent and smart contract l right that boost decentralization is better than aggregated in on the opaque uh tread file layer with centralized exchanges that undermines neutrality and I would say that this is this is also a very important thing that we we haven't really uh done too much research on but the both arguments are security aligned right the LST dominance argument and the centralization uh centralized exchange State concentration they're both very important under mvs okay I'll speed up a little bit so we have lots to do let's collect elaborate um let's accelerate what we want fearlessly it's an infinite game there's no end game thank you thank thank you so much Aro people have questions let me start with the first one solo stakers have paid most cost up front Hardware time to set up Etc maintenance is very cheap but then why do you think staking via sex centrala exchanges is more inelastic right so well first of all uh I would say that you know there's a misconception that stakers have uh costs up front right and they're fixed costs they're actually variable costs and they need to upgrade their uh setup over time so one needs to take this into account second you know solo stakers are people who are very engaged in the community right and they are very engaged on chain they know their options right and that's the reason why they are more more elastic is because centralized exchange users they don't think about ethereum they think about their life they think about their Investments that's why they're in elastic solo sers sure I mean I I I I admit they have uh diverse uh cost preferences so likely they're their you know their their supply cses is is rather steep at least marginally but if you think about this fixed costs you also realize that actually at some point of the yield it it becomes unsustainable so it's a very important point to dig further and we did it in our paper encourage everyone to read thank you this one more time for one question as more e is St real profit margins go down even if you doubled issuance why do you think staking margins will not compress either way yeah so I answered this question on my fourth or fifth slide is because the market is segmented right that's our uh let's say main hope that decentralization there is a decentral there's a we can sort of pay for decentralization is because the other users other stakers other Capital allocators uh will be uh will not let's say will be willing to pay higher price with centralized exchange is the segmentation of the market uh and it really by the way it's a fact that it is like that right now if just take a look at the fees 25% fee with coinbase and you know 1 to 10% uh fees with uh pools and csps so Market is segmented right now question will it be segmented in the long future I don't know I would argue that there is a very good chance that it will be segmented for example investment the whole investment market right there's Pension funds and then there's professional investment firms it's a segmented Market as well thank you so much uh time is up arum and thank you for the audience for the questions if you have any