Will Shannon, Lido: The Power of DVT for Liquid Staking Protocols | ETHSofia

[Music] thanks very much everyone uh so as mentioned we're going to talk a little bit about DVT distributed validator technology uh we're also going to talk a bit about like where Lio as a liquid staking protocol has been over the last year uh and really where it's going now over the next few months uh so I'm will I'm a contributor to the node operator mechanisms workstream this is the workstream of the Dow that's responsible for um helping the Dow basically interface with node operators traditionally this is the curated node operator set but more recently this is also taken form with many solo stakers and Community stakers as well so for those of you that don't know Lido is a liquid staking protocol on ethereum uh currently with over 9.7 million e Stakes via the protocol uh Lio is a middleware so it connects users that want to stake with node operators uh it's completely software driven uh and it's governed fully on Chain by the Lio da so all code uh changes basically go through an anch and Aragon vote all right so to to kick this off I think it's really really important to talk about where Lio was a year ago so you have the context on how DVT is really change things um so if any of you were on crypto Twitter a year ago you would have had uh plenty of drama around Lio right there was only 29 node operators uh that were on boarded via these periodic application windows so I've been a contributor to the Dow now for about two and a half years and from my earliest conversations with uh like Dow uh members uh contributors to the Dow was very very clear there's always been this unrelenting focus on decentralization from the earliest days it was clear that having solo stakers participate in the protocol was the goal about a year ago that wasn't the case there was 29 node operators now I will say I think at this time there's no question in my mind that Lio is and was the most decentralized at scale staking provider that there was it had better consensus layer client diversity than any other protocol uh and basically than the entire ethereum network uh there was no country with more than 20% of Lial related validators excluding Germany and that's improved quite a bit and in May of last year there was a a very important change that happened so uh there was the introduction of What's called the staking router and the staking router changed the architecture of the protocol from something that's monolithic uh to more modular and that is really what's uh like driven a lot of the change and Improvement that we're going to talk about here one thing I I want to call out because this is always sort of a a point of discussion but node operators themselves are completely independent entities the Lio da is not run a single validator for the protocol itself uh as someone who basically is talking to these node operators every day uh I can tell you it takes quite a bit of effort just to get a form filled most of the time uh there's no ability to tell these people to do anything with their infrastructure and actually many of the ethereum client teams are actually Li node operators as well one of the things I've always really appreciated about Lio is transparency and I think this really stems from the fact that Lido is a fully on chain and and uh like mechanism uh so there's a few things I want to touch on uh first first is the Lio validator node operator uh metrics report and so this is a a quarterly report uh that basically covers everything related to the infrastructure so the data is sourced from the node operators themselves uh this covers things like their consensus layer an execution layer client diversity it covers things like where they're running infrastructure so uh are people running out of home nodes are they running in the cloud what cloud provider are they running in bare metal what countries Etc and so all of this is reported and shown to the community one thing I think that's also really cool and this is very different than many of the centralized exchanges for example is that all validator keys are publicly mapped so if you want to figure out if a validator belongs to Lio you can do that in about 30 seconds uh you can do it on chain but you can also just look at a bunch of the interfaces and popular resources like Beacon chain have all Lido validators mapped I touched on this but all of the policies and like proposals to the Dow are community-driven so there's a really vibrant discussion taking place on the Lial research forums uh where like all changes all suggestions to the protocol are discussed there's also something here that's called the LI to on ethereum scorecard and I think this is really really cool and quite different than you know uh not even just like most Web Two or web 3 things but Web Two also so basically there's like a very small little screenshot there in the I guess uh bottom right uh but this is basically something that the contributors put together with feedback from the community on areas uh like basically that are scored so um for example like how is Lio doing in terms of node operator decentralization what I think is really cool about this there's a bunch of aspects that are like actually we're not doing very well here there's a lot of issues here this could use some improvement or uh this isn't looking so great but like we have a path to Improvement um so I think this is quite differentiated I mentioned this but all governance happens on chain either via snapshot first from more like qualitative policies but all code changes go through uh onchain vote and then there's also something called easy track which is used for more routine governance operations so really like bringing this home to where Lio was a year ago uh at this time there was only 29 node operators also 78% of the validator that are utilizing the protocol were running the execution layer client gu this was actually down about 8% year-over-year it was recognized as a as an issue um but like still it was too much also at the time there was no permissionless entry to the protocol and there was no Sol Community stakers I mentioned this but if you were on crypto Twitter a year ago you would have seen this basically every day for weeks and weeks and weeks but I think there's been a really really big change it started uh last fall with uh something that's called hu's uh proposed Goose goals to the L O Dow and there was three three goals the second of which is what we're going to focus on and that's that Lido should attract the best validator set in ethereum that includes key results like permissionless entry to the protocol reaching 5,000 node operators having a performance Effectiveness rating in the 80th percentile and designing the protocol for risk mitigation by Design and having a free market for validation that was quickly followed by two Snapshot votes uh that basically introduced the idea of first off the simple DBT module and then second the community staking module so we'll start with simple DBT uh and if you know me at all uh this has been something I've been very focused on and I can talk to you guys about it all day long and uh you know we're we're lucky here to have a bunch of SSV folks in the crowd so uh I'll be followed by them and you'll hear more about DVT but basically simple DVT was the second module that was added to the protocol and the main differentiator here it actually was uh the exact same codebase with very very minor modifications uh from the cura node operator registry but the difference here is that it uses clusters instead instead of individual node operators that is because it utiliz is uh this technology distributed validator technology which is developed by two main providers Obel and SSV Network all right so for those of you that don't know what DVT is I think the easiest way to understand this is It's like a multisig but instead uh like uh like a wallet right it's for a validator and this adds a lot of different advantages so if you think of a normal ethereum validator you have an execution layer client you have a consensus layer client uh a validator client in your beacon node and maybe you're running this from your house uh you're in Bulgaria you're exposed to your internet provider if the internet goes down you're going to miss mest stations and at a small scale this really isn't a big issue but if you were gone from your home for a month it is kind of an issue so the difference here with DVT is it uses uh threshold cryptography and uh in the case of Lio all of our clusters run with seven people you could also do this with a three out of four a seven out of 10 you could go bigger but those are the most common but in this five of seven threshold right only five of the seven nodes need to be online for the validator to actually attest and that's that's a pretty cool thing because two people that means could be offline they're still receiving rewards and your validators are still online which also really really cool from the perspective of a liquid staking protocol here is there's a ton of benefits to uh resilience uh to decentralization and key security I'm going to double click on key security very quickly because there's a lot of people that claim uh benefits from DVT but they aren't actually using distributed key generation for their uh key security um if anyone tells you they're using DVT ask them if you're using dkg with DVT uh this allows for the creation of the keys to be split across multiple parties so that no one person ever has the full private key to huge Innovation it drastically improves the security of the validator um all right so talking about these like benefits right so we talked about the example of the single validator but now in this DVT like configuration you have five or seven different nodes every single person could be using basically a different execution layer client a different consensus layer client all of them could be in a different country around the world subject to different regulatory aspects for example uh all these people are using different kinds of infrastructure so some are in the cloud some are running from home some are running from bare metal and all of a sudden at this single validator level you have such a distributed and decentralized piece of infrastructure that is incredibly resilient and performance we're going to talk more about that so the module launched in February and since that time it's uh facilitated the addition of over 140 solo and Community stickers to the protocol there's now nodes running on on every single continent except for Antarctica via simple DVT um and so initially simple DVT launched with a 0.5% share limit and if you consider the size of Lio that is fairly significant uh but given the early success the Dow voted about 3 four months ago to increase that Target share to 4% so for perspective 4% of ethereum Stak via Lio is equivalent to over 12,000 validators that's 1.14% of all ethereum validators and that's going to be run by 72 normal clusters so these are clusters uh with seven participants they're made up of professionals solo stakers community stakers community stakers are basically solo sakers that work with their friends and they're going to run at least 80 validators per cluster there's also 10 super clusters and these super clusters followed an evaluation uh round basically by the LI node operator sub governance group that then went to the Dow there was a proposal uh but these clusters are mostly filled with professional node operators but there's actually Solo stakers in these clusters too and I think this is very differentiated versus many of the other staking protocols which are using DBT they basically have a pool of 10 node operators they're all professional running all their validators these superclusters actually have Solo stakers in them so simple DBT in 2024 is expected to facilitate the addition of 300 plus net new stakers in total to the Lio protocol and 2024 is almost done uh and we're not far off there's another thing that's basically related to this that I want to touch on and this is uh a vault that was launched by uh Melo who's a member of the Lio Alliance and and this is a a kind of interesting thing because uh it basically directs new net new stake to the simple DBT module this is doing uh multiple things but like the most important thing it's doing is really furthering the decentralization of the protocol users put stake uh to this Vault it Stakes it to the protocol and that stake goes to the simple DBT module this is allowing solo stakers and Community stakers to run validators faster it's allowing for users to participate in running validators and decentralizing Lio which is decentralizing ethereum through this uh much quicker and also there's a proposal right now that's currently live uh on Snapshot vote to also incorporate the community staking module we're going to talk about that from a user perspective the other thing that's pretty cool here uh so users receive points I know I know terrible terrible from SSV from obal and from mow but I think there's like one kind of interesting thing here so SSV already exists you can go find that token look at it it exists it's not fluff Obel has many many validators currently running uh in the simple DBT module and and over the the next 6 months obal and SSV will have an equal amount but obal currently has more and obal points are generated by the number of validators running uh because there's so much more stake in validators running in the simple DVT module there's currently a 9x Point boost just based on the math of the amount of validators in the protocol versus The Vault so it's interesting to look at uh stakers I just want to call this out should not unstake stake deth or wrap stake deth on the base chain on ethereum Main net um if they want to receive these points because the goal of this is to add net new validators to the protocol that are allowing community stakers that are allowing solo stakers to come in run validators and decentralize Lio okay so we talked about where we were let's talk about the impact of DVT so today there are now 235 independent node operators using the Lio protocol to run validators that's up 710 per year-over-year it's huge huge huge Improvement right we went from 29 to 235 and it's not stopping that number will be over 300 in the next two months my favorite stat on this slide is actually that 53% of that 235 are Solo and Community stakers running validators via Lio I think that's incredible incredible progress uh there is a test net that just finished with Obel two months ago the proposal to onboard cohort 3 is uh just been accepted so there's 36 net new node operators that are on boarding next week and there's an SSV test net that's ongoing with another 90 spots open so these numbers are very achievable to go over 300 this year and there's also a slide at the bottom left I mentioned gu and how the utilization of gu was over 78% a year ago it's now actually uh it's actually less than 40% and last quarter was flipped by nethermind for the first time it's a really really significant Improvement I think that Lio has had uh from a protocol perspective a ton of progress and it's very very hard to argue that there's any other liquid staking protocol or staking protocol in general at scale that is more decentralized but simple DBT is really cool it's kind of what's happened over the last last year now what I really want to talk to you about is what it's I think like incredibly exciting and it's it's the story of Lio over the next few years that's permissionless staking with Lio and this is going to be facilitated by something called the community staking module but this is going to allow for anyone in the entire world to run a validator via Lio without ever having to talk to a single person it's completely permissionless and I think from a user experience perspective it's an incredible incredible product so uh I talked about this it's permissionless and this is important right it's true to the ethos if you run an ethereum validator as a solo Staker which really is the gold standard right you're running a node at home you have the full 32 e it's your validator you're not touching a liquid staking protocol or any other staking protocol it's completely permissionless and this was not an option uh with Lio even with simple DBT you had to go through a test net and it was quite easy to do as long as you basically paid attention uh but it wasn't permissionless and so this is a big Improvement there is a bond and the bond is basically just eth that you keep on the side this is to prevent me stealing or if your validator was to go offline for a very long time uh if you go to the lto docs you can understand more about this but basically this is quite similar to the mechanism that exists with rocket pool so there's two different uh categories here uh so the module had a test net we're going to talk about that uh and basically there's an early access period for people to participate if you uh have Early Access which also would be uh you'd also be included if you're running a mainnet validator today in any other liquid staking protocol any other uh basically like by yourself uh this is data collected by rated and by state cat as well uh but BAS basically then it's 1.5 e to participate for your first validator your second validator is only 1.

3 if you are not on that early access list then it's 2.4 e for your first validator and then it's 1.3 e for each validator after that so this has brought the accessibility of running ethereum validators to an incredibly different level than any other option on the market uh so we'll talk more about when you can start but it's really soon third is rewards and this is important because you need to buy Hardware or you need to pay for a cloud provider to run this right this is really really interesting so the APR multiplier and the APR multiplier is basically considering the staking rewards rate that you get for running a validator is up to 2.37 times because the 32 eth required uh for your validator um is uh fully delegated to the validator and your bond sits on the side but it's actually also held in stake deth there's execution layer and consensus layer reward smoothing you're benefiting especially on the El side from being a member of the largest validator set on ethereum and here's the the really interesting I think for people that aren't super experienced uh there's this performance threshold and the performance threshold basically means that if you were to go offline for a day or so you would actually still receive full rewards if within the 28 day snapshot periods you were staying online the whole time uh this is a pretty cool mechanism and you know that means if your power goes out you don't have to freak out that you've uh just basically wasted a month of work number four is that the user experience for using the community staking module I think is hard to argue that is not the best in staking Lio has benefited here by being one of the latter to launch a product uh that has uh or like an ability for people to access permissionless staking and this is really really meant that anyone uh like that looks at this basically says the UI is great but more importantly than that the user experience is great so we have this test night I mentioned it we're not going to spend a ton of time here uh there was over 2,600 node operator entries that participated uh there was quite a few sibles that were farming something uh 300 unique entities at least identified within that almost 5,000 validators and that was all just Within 3 months um cool main net November 2024 the onchain Aragon vote to deploy this is happening this coming week uh that'll start the Early Access period if the Dow does vote to approve it uh though they have approved multiple snapshot votes before this uh and and basically this is incredibly exciting fully permissionless q1 2025 but next week if you're already running a validator on Main net as a solo Staker you can participate fully permissionless after that anyone in the world in q1 probably January February will be running validators permission sleeve via Lio okay so wrapping this up uh I talked about this path to 5,000 independent node operators uh this is going to happen with three main things first is the launch and growth of the community staking module this is the most accessible product for anyone that is interested in utilizing ethereum to run validators and especially using Lio to run validators there is an incredible user experience here I'm telling you if you compare it to all of the other options in the market it's taking the best features of those there's been years of research and consideration uh and I really think that it's hard to argue that there's going to be something better here number two is the continued expansion of the simple DVT module so there's more cohorts on boarding there's more net new participants coming to the protocol to run validators uh and I think this is something that has brought Lio from a very professional uh only org uh participating from the curated module to something now where there's hundreds of people participating there's solo stakers any single person in this room could participate right I'm not a technical person but the benefit of DVT are very very numerous and what's also really cool is that with CSM you can also run DVT so you can go find your friends you can set up a cluster via SSV via Obel um and you can also use different tools like dap node or stereum or sge to participate third and finally on this uh there's also the potential for modules in the future now these will have to go through the Dow governance process there will be proposals there will be long discussions but I think it's it's really really uh like easy to think about something that combines the best of simple and the best of community staking module imagine being able to go to SSV or obal uh without having to talk to a single person set up a cluster and run validators via Lido uh this is I think uh like a huge Improvement in terms of the protocols decentralization I think it's truly good for ethereum and the best part is it's not ending there's going to be thousands of node operators running Pro uh running validators via the protocol over the next couple years with the launch of CSM with the scaling of simple DVT and with the potential for other modules so thanks very much [Music]