DVT as a Catalyst for LSD Development with Ariel Zimroni

foreign [Music] so hey everyone my name is Marco I am a contributor to this city Dao and we are building a DVT solution and basically we are trying to decentralize ethereum staking what I'll be talking to you today is DVT as a catalyst for LSD development so let's get started so first I will be talking about what is the current state of LSD Market with roles how will they change the game what is the hard problem of LSD how DVT addresses it what DVD does not address and how do we solve this LSD issue so the current state of lsds on ethereum uh currently there is around 17K or 17 000 each staked which is 14 of ethereum of the of the outstanding geezer and depending on who you ask like the the whole staking uh pie is expected to grow 100 to 200 percent uh currently there is around 40 percent of this eth is staked in lsds as you can see in the graph and that is expected to grow 50 to 100 percent obviously these numbers are not exact uh but here uh the idea is because of the high or of the much better experience of staking when uh you use the LSD experience it's much better than regular one because you can use it in D5 and we have seen we have seen it in the past so we are expecting that uh staking staking the lsds will make roughly 60 to 80 percent of all eth um so that would be the expected growth of lsds of 25 to 40 to KES now if you look at the graph uh you can see that it's right now kind of a winner take all effect you have 77 has lighter than 16 coinbase and Rocket pool is the third and it only has three percent so there is a huge power law distribution there um now what is what are the barriers of Entry or why does the the market look like it does one is a large Capital requirement and the other one is the need to actually run the serum nodes uh with D5 you can uh how many of you are familiar with ethereum's taking uh how many of you own lsds good not very experienced in LSD but I'll walk you through it so uh the capital requirements and need to run ethereum node uh what do I mean by that uh LSD I will try to persuade you that LSD currently is a stagnant Market not in a sense that that LLCs are not growing but in the sense that the market share is not really changing much and there are large barriers of entry and uh you could have seen it on the craft but the the stake in any protocol is never going down because before withdrawals you can't actually move the stake so uh the the second hard problem why it's very almost impossible for new players to really compete with Lido's sdes it's the inability to hold it back we we saw it even with Lido that is orders of magnitude larger than any other staking providers when when the the Luna collapsed last year uh the the peg went down 10 which basically means if your LSD doesn't hold Peg um you can't really be used as a collateral in defy thus there is no borrowing you can really safely borrow against it and that defeats the purpose of whole ever of the whole LSD experience so there is no leverage uh you can't use it in the D5 Casino so there's that's no fun so yeah that's there's a problem but luckily this problem will be basically solved with withdrawals or assaulted at a major extent Stakes can be moved but more importantly the every time Peg is Disturbed there is an Arbitrage opportunity that can be actually taken so anytime the the peg would falter anyone could buy LSD of the market and uh redeem it for actual eth so after after withdrawals we'll be able to hold back better uh and even with much smaller lsds that means borrowing against your LSD that means you could do leverage staking and uh yeah play your favorite D5 Casino uh now let's let's focus on DVT actually so the first problem of large Capital requirements let's say it will be solved with roles but how does the LSD look and how does the DVT fit into the picture so the actors because they get they're a bit confusing so let's define it validator is not a person validator is an ethereum node holding 32 eth and it's defined by its validator key and operator in the example of of taking pools is actually somebody who is running the validators for you and a Staker is somebody who if I want to stake I'm a Staker and I basically give my funds to validator as you can see it in the picture uh no I give my funds to an operator and the operator is running validators so uh now LSD pool 101 how does every LSD pool work what is the basic structure so um you want to get some LSD so you deposit eth and you will get back LSD now the East is in the pool the pool needs to stake the East somehow so and the Builders of the pool they are usually not it's not the same person uh the the Builders of the pool are web3 developers and to operate and basically basically run the infrastructure and operate ethereum validators is a different skill set and a different job and this is delegated from the pool to some other third party so this is a hard problem how to safely delegate this running of taking values so let's look at the main approaches that are used and we can reframe the question basically how to ensure operator good behavior because we are giving or the pool is giving the operators funds and they need to make sure that uh they they will not get slashed and the funds will be safe so Lido is using a trust Bay model basically they have a small set of operators currently it is 29 which are professional operators and they are fully trusted by Lido if anything happens yeah that's uh but it's not a good thing but the the overall idea it's a trusted Set uh then rocket pool has a trustless set of operators anybody can become an operator but to ensure good behavior uh operator in rocket pool needs to post collateral right now it is 50 of the whole validator which is a huge amount like maybe if you have some ease you can earn some additional ease but it is not really a scalable solution if you wanted to run operators for rocketpool you would need to own huge amounts of eth just to just to spin up a lot of validators so basically this is this is the problem and these are the current Solutions and we are working on a Magic Bullet basically um the dvt-101 what is DVT in the first place or what is the basics defined by its validator key what DVT does we can you hear me okay great so as I was saying you have the validator which is defined by by its validator key and this key is split into multiple partial keys let's have an example let's have four Keys uh the important part is that that each part partial key is controlled by a different operator so and you can't really produce a valid signature just from one key so every duty every signing every attestation needs to be a consensus of of the operators so let's have an example Alice Bob Charlie and Dylan our operators they are running validity together and to perform any validator Duty or to propose a block three of them need to agree that the duty is valid so what are the implications of this so the implication is that this drastically reduces trust in any single operator basically if one of the of the operators is Byzantine or loses his or her key or basically sign something that should not be signed nothing happens and this is because there needs to be a consensus on the on the duty so so one is signing something else the the other the three that we had in an example are signing on the correct data nothing happens so this basically removes not removes but press Eco reduces the trust in any single operator and it minimizes the slashing and dumptense downtime downtime risk and it improves the resiliency of the whole system by allowing basically every validator now can be run with uh with using different data centers or different regions different geographical reasons etc etc different setups you can be running different notes now what does it not solve there is also an important point because DVT on itself is only what I was saying it's splitting the key into multiple parts and we have three important areas that it does not address or does not address fully coordination uh how to coordinate between operators and stakers without introducing single points of failure this is not entirely something new uh because takers and operators need to coordinate even before this even if you are not using DVT but we want to use uh we want to have a trustless system and not have this as a single point of failure and also you have additional coordination that operators that are operating a single validator needs to coordinate with each other so you have these coordination problems then you have a performance what if an operator does not perform or becomes malicious as I was saying before this is not as much of a deal as it is with uh without DVT as there will be no slashing and performance may not be uh may not be affected but if we don't want to I have a trusted set or require collateral we must be able to uh solve on the network level the the non-performance validators the problem of this basically this is a problem that would always arise if you have pressless operators and you don't have any any stick to punish them uh the the last one is collusion what prevents third parties from coercing the whole clusters or compiling operators to collude so uh our solution what we are building is DVT Network and it's basically made of a set of smart contracts that is used for coordination and a multi-cluster operator Network which is dvt-based so coordination um in our design the the stickers and operators are coordinating leveraging ethereum smart contracts so basically what happens practically is um a staking pool or whoever wants to run and run a Staker they just pick The Operators they they need to deposit some SSV so because the payments are also handled by the smart contract and they basically just do a smart content transaction and then I'll go back and then basically uh the network observes the smart contracts uh they see they get the all the information from there and they can pick the duties dynamically create a cluster and start operating the validator now performance as I've said since we don't really have any strong carrot for The Operators we must solve the issue of performance like the naive approach is whenever you have one of the one of the cluster one of the The Operators that is on performance you could after withdrawals basically exit the validator and re-enter it with different operator set but this is not obviously ideal you don't want to all the time just be or not all the time but even if it happens once in a time you don't want to be exiting a validator and then re-entering it because it's costly it's costly based on time you need to pay transaction etc etc so in DVT network uh in our Solutions stickers can change operators dynamically and this is a big deal uh so they don't you don't really need to even if you want to change from whatever reason the operator because you see he's not performing up to the power you can basically just do a contract call and replace it or replace him or her with another operator now collusion important thing is in SSB Network validator clusters are not known in advance why is this important or why is it important to have a dynamic operator clusters you can you can have a simpler design by requiring operators to coordinate in front and create clusters and and basically but let me give you an example to to Really illustrate this uh let's say we have a smaller Network we have 100 operators in the network but if you are required operators to create clusters they would basically you would have smaller amount of clusters then you have operators in the system so let's say we have 100 validators and they would combine themselves into 20 30 clusters the problem is if the Clusters are formed on the on the basis of operators themselves now uh it's much harder to coerce or to cores or to influence the whole cluster than one operator but it's not impossible and since since you have smaller amount of clusters then you would have individual operators you have larger honey pots that are held with each cluster so it is a big deal for uh stakers to be able to pick additional or not an additional to pick specific operators and this allows and this makes the network much more heterogeneous basically you can see in the network explorer that this cluster has very many validators and you can basically just choose different four operators or just two of the four operators could be different and you are creating a a new cluster or a new cluster is created specifically for you so let me recap how we are solving the hard problem uh Lido is a trust based solution they are using trusted operator set for running the validator rocket pool is collateral based and they require operators to post collateral and the secret LSD cells that we have done is the DVT network based solution and it allows us to use trustless operator set without collateral thank you so let me recap how do we how do we get to the LSD Revolution we have withdrawals this will allow for new players to competing to start competing for for the existing stake they'll also allow small lsds to hold the pack thus be usable in defy and start to be compatible with larger players DVT then reduces trust in single operators because you are splitting and uh no splitting operators you are splitting the validators it also introduces some coordination and collision issues which we then address in the SSV Network we I've explained how we address the the issues that are either not solved or are raised by DVT and basically also SSB allows for trustless to say delegation without collateral this is the important thing it is capital efficient meaning operators do not need to post collateral and it is able to handle non-performing or or malicious operators and it's a support large trustless operator set also anybody can join the operator set earn earn rewards uh earn for their services and they don't need to ask anyone and because it's it's a network the performance of all the operators can be can be shown and can be tracked in time so basically for the after after we launch and after we develops are enabled uh you will be able to start earning as an operator additional rewards and build what is the word build reputation yeah so what does this allow for SSB Network allows devs to focus really on building web 3 applications and not really need to care about how the staking is actually handled and it is handled in a decentralized way you can think of it it's like if you are doing a web 2 website you need some backend so you are using AVS for the back end we are doing kind of the same thing for web3 applications we are basically the back end for your staking application but we are doing in in a trustless and permissionless way so and this this also alive allows D5 protocols to stay stake ease directly so the punchline is SSB Network lets Builders focus on creating the best LSD experience for their customers now we are funding your LSD experience since 2022 with either grants program or ecosystem fund uh I would be actually the person to talk to if you are thinking about building some some kind of a staking application and are looking for funding uh I can help you with with getting a grant or connect you with the right people to get funding via ecosystem Grant and you can see who is in the SS vehicle system you can see it's already already quite vibrant even though we are not on the Main and yet but that will be this year uh we are talking about the staking pool so one of the big players are stakestar anchor or the bill building with us swell network uh also Lido is running pilots and in their V2 they'll also be uh employee employing deploying DVT Solutions and we are working with them and we are also working with savewise so and come to meet us at both here we have a booth on this floor uh right there right at the end okay so that would be everything from my side so enjoy trustless LSD experience with your friends um you can business contact on me you can directly contact me if you are thinking about building some kind of taking gaps and check out our website at ssv.network you can join our Discord Twitter whatever else I have a little bit time for questions if somebody somebody wants to ask something okay so thank you very much [Music]