Restaking Revolution: The New Frontier in Hyperscaling - TOKEN2049 Dubai 2024

all right uh I was called a scarecrow back in high school didn't think it would happen again um so thank you guys so much for joining us to talk about Igan lay and reaking uh I know you just got a light introduction before going on but I'd love it if we could do uh one that's a little bit deeper that covers who you are what you do uh and a little bit about your company um I'll start and then we'll go that way so my name is Victor buan I work on protocols at coinbase uh so I spent a lot of time working on staking we operate uh what was previously called coinbase Cloud which is a largest staking operator and I work on base and governance and policy issues and and things of that nature everything has relat to protocols hey uh I'm tun I uh founder of Gauntlet we do a lot of risk management within Defi and uh I spent a lot of time thinking about different types of risks within avss and reaking hey guys uh I'm vs uh one of the early team members at Igan ler and uh primarily lead uh strategy in BD for ecosystem growth um and yeah pleasure to be here hey guys um I'm Dorothy sorry for my voice I lost my voice a bit I swear to God I haven't gone to any party yet um that's a bit unfortunate um I'm had of pro at Aller at Aller we built reate rollups that combines reaking mechanism from I layer with um rollups we help decentralized rollups um including the sequences and validators and to provide restate rollup services to our partners thank you amazing thank you um so we're going to dive right into Igan lay and we're going to start with more introductory material in case folks aren't familiar uh and then we're going to get spicier as the session goes on and so vs could you start us off with uh an introduction to Igan lay uh and explaining how reaking works yeah definitely um igen lay is a Economic Security and validation Services protocol that um is designed to bring Innovation back to ethereum um you know in the last couple of years we saw plenty of projects launch their own tokens which determined their entire Security based on their market cap and we all observed plenty of debt spirals that occurred and so um in addition to solving the bootstrapping problem for these new protocols and while they can retain their tokens for governance and incentivizing users um we wanted to enable a way for all of this extractive value that was leaving the ethereum ecosystem to come back and um be secured by the ethereum staking and operator Community amazing um and Dorothy uh from all layer which is one of the ABS I'd love it if you tell us a little bit about abss what are they uh and why did all layer uh decide to become one yeah so ABS stands for actively validation service which doesn't mean much um if we don't um look at what it really does um they're really um leveraging IG layers rethinking service to help new blockchains new projects to uh bring leverage or rent the ethereum security um just by using other ethereum nodes and rake each assets to safeguard all these new projects so when you um in the past when people start a new layer one or Oracle or uh any other type of pro blockchain project they normally need to build their validator program from scratch they need to attract a lot of valid datas to join their ecosystem and help them build up all the nodes for to maintain and operate the blockchains um but with Ian lay they don't need to worry about it anymore they can just rent the security offered by ion layer and have all the ethereum node validators to run all these projects on nodes um so it's easier a lot more cost saving and a lot more um easier for them to boost trp yeah amazing um and the really like two second version of that is that you know ethereum validators can do a lot more work and they can do more work for more pay uh and that is incredibly valuable and useful not only to the validators themselves but also to projects such as Al layer they want to take advantage of that and so tun can you talk to us a little bit about so you know we heard about the validators and abss and the network uh talk to us about the Dynamics of validator operators and more explicitly um how should these uh operators be thinking about the risks and rewards of participating on IG layer yeah uh for sure so one thing to think about layer ones is they generally are in a lot of ways extremely over collateralized and how much security they're providing if you look at the net transaction flow in most blockchains the amount that's staked is significantly higher than the sort of Maximum attack surface you could have and that's something that's always necessary in networks when you're bootstrapping and you need to actually you know build the network effect build the network effect for apps but also for validators um and reaking a lot effectively you could think of it as a way of reusing that stake uh for these kind of new productive um Services these Services obviously range from things that are sequencing rollups doing things like running a decentralized sequencer or shared sequencer to something that is you know like a zero knowledge proof generator proving Network to something that's like an oracle now just like applications have different sort of risk and rewards in terms of you know applications like being a unop liquidity provider you kind of generate fees continuously on every block uh but you don't have these kind of like large loss over a single block type rewards versus other things like a A perpetuals or lending protocol where you might lose all your money in a a liquidation when the market moves you have the same sort of properties within IG L Services some of the services will be like giving you a small amount of extra income every block and have sort of lower risk some will have higher amounts of potential risk but hopefully we'll be generating enough fees to compensate for that and validators have to take a portfolio approach to this where they look at the set of avss decide how much to allocate to them based on what they they view the risk word profile is so there is this sense in which the validators also have to do a lot more work but obviously they're getting compensated for it yeah yeah yeah amazing well thank you for that thorough explanation so we're done with the introductory part now we're uh going to get spicier and so you know there's almost uh 5 million e being restake via AG L right now and one of the concerns that people brought up is what happens with consensus overload um and so I'm hoping you guys can talk to us about you know what is consensus overload you know is it actually a risk is it a bad thing um and what is IG lar doing about it do you want to Kickstart the yeah sure so um you know I think there's a the the canonical reference for this is a vitalic blog post called don't overload ethereum's consensus and the idea is that you know consensus has to do a lot of stuff right it has to validate sequence of of transactions you have to also get different validators voting on certain things you have to be able to deal with sort of adversarial environments where some of the validators might go down some of the validators might take 100 times as long to reply to a request and the idea that you're giving validators even more work that's outside of the initial sort of form of consensus means that there's more of a likelihood for some type of failure that happens in the application you're providing to Cascade into the Bas layer which would be bad for security and so the goal of of trying to not overload consensus is really about making sure that the types of applications that people are using within reaking don't burden the active consensus maybe they happen at a lower frequency maybe they can happen sort of in a separate process that doesn't have to interact with the main chain as frequently um but another thing people sort of have to think about is how much stake restak does an application really need right you know an application that maybe is only generating thousands of dollars of value Val in extra yield might only really need a million dollars Max of security extra security and so being diligent about understanding how much how you're kind of redistributing the security and how you're sharing it across abss is actually quite important to making sure that you don't overload this and such that you know the valer has distributed over the different avss yeah uh he hit the nail on the head but um you know there's there's some complexity around ensuring that um this shared security pool doesn't get fragmented right and we don't want to enable these um you know as a platform we T re- retain a level of credible neutrality when it comes to um what operators and what abss these actively validated Services want to do but the reality is um you know as I mentioned we want to bring Innovation back to ethereum and we want to bring on these more uh complex more throughput heavy more um High block space heavy use cases and so there's going to be a uh necessity for a lack of stake centralization across both the operator and abs network but also in terms of ensuring that while we bring on these use cases we're also building the infrastructure required to ensure that as much computational scaling is offloaded and we keep the essential burden on the ethereum settlement layer fairly minimal yeah and uh Dorothy I'm wondering if you could just double click on something that tun said from the perspective of you know you guys run on an ABS like you have a certain amount of restak e how do you think about the relationship between the amount of e that's restak to the services that you guys are providing for the rups that you're helping operate to the inflation rewards that are potentially being generated to the transaction fees that are being generated and like how do you kind of like square that Circle and think about that relationship between uh Economic Security and services rendered yeah so um right now I think um a lot of people are kind of pre amping the ecosystem before it like matures out or plays out so right now the the whole um IG L ecosystem I would say it's quite chaotic it's in its early stage where people um are expecting a lot of return a lot of rewards um without understanding um what really are these abss doing and what um are these different projects might be able to provide in return for instance we help rooll LS to decentralize and we as all these roller projects to provide token rewards to the stakers or to the node operators um but in reality a lot of these projects might not be billion dollar worth of businesses there might be only1 10 million worth of business um in reality but right now just a lot of front running and pre-m actions going on so the community and other not operators are just blindly um staking and offering their service to every single ABS they can find um which is not bad for the ecosystem to grow from the beginning but once I I think once we have a few months of more avss to come in once we see a lot more various type of businesses to join this staking Service Marketplace then we can see um who are the real business that can provide real yield and uh what kind of operators or LRT projects are really good at identifying the highest risk reward and manage the portfolio prop properly for their uses so so right now it's very tough to tell you what's the expected return because it just varies yeah no I'm I'm so glad you brought that up and I think that the point around you know a little bit of the chaos in the ecosystem is is both a good and a bad thing because it's so good for grows and so good for people to be excited um you know but I think to your point right now because people are optimizing you know for airdrops and they're optimizing for like being early and they're optimizing for like all kinds of stuff and at the same time there's no slashing conditions uh enabled for most of these Services yet and so you basically have like all reward very minimal risk and so everybody's doing everything and there's not that deep thought necessarily around you know doing the proper evaluation and being really diligent um and that's something that's again great for growth but at the same time like has consequences down the line you know when the rubber meets the road and then you know as you mentioned like people have to see who's actually generating yield and really what that means is who is providing a service that end users whether it's rollups or somebody else really cares about like who's contributing value to the world uh and I'm super glad well cares and pays for cares and pays for that's right they show their care via payments yeah I think to expand on that just a little bit um I don't think the market has fully uh understood or has been informed from our perspective um from our side as well on how reaking yield separates itself from defi yield and I think the reality is the majority of people that have been um you know for lack of a better word aping into IG layer are assuming that it's just like every other D5 protocol that they've interacted with in the last few years and um I think a far more um a far more you know robust mental model to understand igen layer is we're trying to build um the equivalent of like Cloud infrastructure for web 2 but bringing that to web 3 in a more decentralized Manner and so thinking of us more almost like a decentralized SAS company where people are paying for the security as a service in like fixed yield um in proportion to the usage is more accurate than looking at us like a dii protocol and so there's definitely some you know complexity around how we communicate that but also you know the incubation of this ecosystem has led to um you know rapid growth and Rapid speculation and the reality is like we don't have slashing Liv today we don't have payments live today we're actively building both of those things um coming this year and you know it could be interesting um to look at uh almost like these operator ABS positions similar to how catastrophe bonds uh view very local systemized risk and you know these liquid reaking tokens that have emerged which is like this amazing new entrance that we didn't even expect into the market as almost like catastrophe Bond ETFs that aggregate these systemized localized risks of these abs operator positions so yeah there's uh there's a lot of complexity and there's a lot of information asymmetry right now but as we you know deploy payments deploy slashing and we start to see a more um robust Open Marketplace where abss have to compete with each other for the yield that they pay out to stakers and operators based on the amount of block space that's required we're going to start to see far more interesting Dynamics yeah I I totally agree um and I'm really glad you guys are thinking about it from that way so another uh spicy spicy one that I think other people here will really enjoy is so you know we were talking about how validators could potentially earn more uh from participating in IG lay in addition to the staking rewards that they earn by participating in ethereum and so I'm really Keen if you can talk to me about like what does this mean for ethereum and so specifically you know what does it mean for ethereum staking rate and inflationary schedule is it a good thing for ethereum staking rate to continue Rising because of these increased rewards should ethereum considering lowering its issuance rate because validators have access to increased rewards from IG layer uh you know would a lower issuance make eth the better money like how do you think that that all works out yeah um so you know if you look at and this is just one mental model that's like not fully complete but if you look at ethereum almost like an app store we try to think of ourselves to some extent as also ethereum's like xcode playground or like test flight where the you know some of the protocol features that could be inherited into the core protocol could be first beta tested on this subset of stake within igen layer and so the question becomes as you inherit more features into the protocol um why wouldn't you necessarily pay out a higher staking yield right um and it really depends uh on the ecosystem and the demand for those features and things like that but um it's going to be interesting to see how the market reacts to you know the the stabilization to some extent of all this like speculative yield becoming more specific yield that we will be able to see paid out and from there we can make a more accurate assessment on whether we need to adjust the staking rate of return yeah t do you want to comment yeah so I think adjusting the inflation curve for ethereum is you know a topic dour nowadays uh but I would say that you know one of the the rationale that people have for doing this is that it it makes it less profitable to have these liquid staking vehicles that aggregate a lot of stake because their economies of scale only really make start making sense at like a particular yield number I think the problem with that for something if if you know you want to lower the inflation curve because of something like Ian layer is that the actual yield per AVS is going to vary wildly right like I said there there might be some avss where you know say you're sequencing a rollup you're getting a small amount of fees on every block but you're not getting like a huge jump it's very consistent kind of low variance almost like a a bond and you might have other avss that maybe it's like a ZK proving Network and like hey you don't have to generate proofs every block but once you know once a a week you have to generate a ton of proofs in response to some type of event in that type of thing you have a payoff that's very you know infrequent and so I think you know if you think about inflation Curves in blockchains there are a continuous process right every block has some amount of inflation uh there and some of it is deflation in the case of when you have a burning mechanism like ethereum does and doing this match matching between the the kind of durations of of yield you get ver where you have this continuous yield coming from the base protocol versus the variable yield coming from you know things like me things like I layer it's it's actually quite a delicate balance to strike and I don't think that that you're going to just be able to kind of come up with an answer uh a priori for how it should look it's definitely a type of thing where you need to really see how you could homogenize those two before you kind of justify a larger decrease yeah I I think it's a really great Point um especially on the on the lumpiness because you know one of the things that we think about is predictability of cash flows um and so to the extent that validator operations continue to get professionalized institutionalized um you know there's going to be more and more of a focus on really understanding you know what a security budget look like over time so another uh topic toour as you mentioned but also a spicy one is there's been a slew of reaking protocols that have launched or or been launched or in the process of launching uh that are you know in some ways but not others necessarily competing with I layer um and so one question is do these make sense right like you know is there a need for multiple reaking protocols uh is reaking particularly optimized for ethereum based on the way that the ethereum validator set is um I'd love for you all to chat about that yeah um I think there's a few different uh points here right I think one um we at Ian layer you know one of our uh our you know dominant narrative right now is um and you will see this all over our Twitter is like infinite sum games but the reality is like we don't believe in zero some games there can be multiple players in this market that will be able to drive value um it might be a long head short tail we don't know what that Dynamic is going to look like but um I think it's very positive for us because like you know you need the optimism to your arbitrum you need the competition in the space to nurture The Innovation so I think that's really healthy um so again super excited for the protocols that are building reaking whether it's on ethereum or in other ecosystems but I think there's a uh a very high bar for diligence and security and risk management that's required to execute this effectively so because people's assets are locked up right people are reaking their assets and you know calamitous slashing events could have really negative circumstances the second part of the question which is you know in other ecosystems and whether ethereum is like specially designed for this um is you know a conversation that we were having earlier which is you know in salana you have um the the majority of this like initial half of this bull run is happening on salana because um one you know Traders love the ability to execute transactions quickly there's a lot of liquidity to enter and exit positions but you know the reality is um Solana's market cap uh to stake ratio looks much different than the ethereum stake to market cap ratio we're dealing with uh nearly $70 billion um and my stats might be outdated but nearly $70 billion of ethereum stakes when it comes to somewhere around like two to four billion dollars of Salon State whereas you know the market caps are completely different and so I think these ecosystems value different things whether it comes to stability versus agility and so um ethereum as a community has valued stability in terms of their decision making and you know protocol upgrades and things like that where you can see um where and also has a far more like vared diverse very minimal footprint operator set so you can really enable people to participate in this ecosystem and like bring in more Innovation back but on salana you're starting to see some new designs that like share similar uh you know mental models but um it's going to be very interesting to see how they execute considering a leverage ratio of stake to market cap so I'll take the other side of that um I actually think there's a lot of things that people are doing on salana that are opt-in extra slashing rules so um for instance in jto uh there are uh for me payments there's sort of a a slashing mechanism that's used you only get your your your me payment provided that you don't miss a certain number of slots with the Neo and you get slashed otherwise and it's an extra opt-in slashing rule that you immediately subscribe to by running the GTO Sol validator a second thing which is the uh s SW qos upgrade in salana which is basically uh you have to to stake in order to have a certain number of channels um in in RPC node and a lot of the apps say like drift are running their own validators and they are actually putting up stake for to get a higher quality of service those types of things are extra slashing rules effectively and I think what end up happening is that in these markets you may find that people have opt in extra slashing rules but in different places because of the architectural changes and also differences in the economics that are being talked about so in salana the meev revenue as a function of the stake is actually much higher than ethereum nowadays and because of that you actually have to have a very high amount of security on the on people uh kind of missing slots but taking payments and griefing people and so the the place to look is actually at where you're seeing fee acral in different parts of the system and where you actually need to have penalties for people for sort of missing some of their tasks um with regards to the first question about um the the sort of how the market will have shape out market share wise I think um you know in a lot of ways reaking kind of is in some ways harder than both lsts and defi in defi you can have a lot more sort of like multi competitive markets that you know might be oligopolistic but generally there there there tends to be market share distributions of you know sort of like 60% 30% and than everyone else in lsts you have something that's a little more competitive but often times the the competitive Advantage for lsts is a how much more yield can they provide due to to me so that's a variable Yi compon component that can be used to differentiate them and also how how much better of a liquidity Advantage does an LST have liquidity advantages come in two forms one is in terms of Dex liquidity for trading the the staking derivative for the underlying so that you know you have the sort of synchronization of of what the the true premium of holding the the derivative is the second is withdrawal liquidity and so this is important to node operators like coinbase which is how how easily what's their opportunity cost of capital for being in the withdrawal Que how hard is it to get my money out of the withdrawal Que in reaking you have both of these problems combined and in a lot of ways this this this plus the security features make me at least personally think it's a much more concentrated Market than either Dy or uh liquid saking yeah I I think I think that's a great point and you know something you kind of touched on is that um you know not only is are these pieces getting more complex but like staking in general is becoming more complex you know if you think about previously it was just you know individuals or enthusiasts running individual validators uh and then Lio came along and started offering this like node operator set to which you can stake and it just you know automatically delegate and got your rewards and all that stuff and now we're taking things to the next level from there and that now you have to manage not only like the liquid reaking token right which is kind of like similar to LSC where that's a regular um liquid saking token so you have that you have the ABS selection right you have the not operator selection and all of these things are are fundamentally you know they they have attention there um they have to manage like really really actively that adds complexity and then on top of that you're layering the the defi complexity of how these things are actually being being utilized um but yeah so we have uh a little over two minutes left what I'd love to do is I'd love to give everybody an opportunity to kind of like wrap up with a call to action uh for the audience um you know where can they learn about your project or your company what action can they take uh whether it's to participate in something that you're doing to subscribe and and follow to visit your website to redent what would you like people to do um okay um I think you can easily find um on the Ian page there are six a seven abss that are live right now one of them being agda and the rest um the six others are third party avss and we operate two of them uh one is AER M ala which provides fast finality service for um optim optimism and opum and the second one is for exterior um our partner that's a major gaming studio that launches um number of games with millions of players in the crypto space um and we're going to launch a l more avss in partnership with s in the future so you can check out um our homepage Ian Lair homepage to all these information uh yeah please check us out on Twitter or our website um or any of the abss building on top of us um and I would say the call to action for the crowd would be two things one um let's get more informed as a uh community in terms of you know as I mentioned how reaking Guild is different from defi yeld and um for the Builders out there uh you know I think we've built a culture of deploying with you know a lot of constraints when it comes to block space and the reality is like you know Web Two devs don't think about scaling don't because Cloud just scales so the question becomes you know what would you build if block space was no longer a constraint in the ethereum ecosystem and um if you have some ideas please find me hunt me down um shoot a message on telegram would love to chat with anyone and uh I I've written a bunch about L LRT risk and I have a couple other papers coming we have a couple papers and research articles coming out soon on different risk will be on Twitter so amazing and then for me uh coinbase is one of the node operators on the uh Ian layer uh reaking Network and so please come and uh come and stake with us so thank you panelist so very much that was lovely I really appreciate it and thank you to the audience [Applause]