MEV in xDai with Flashbots and Nethermind

Transcript

great hi everyone so this will be the talk about mev and x i with flashbacks and nevermind uh testing preface take i may be ahead of ethereum 2 happening so my name is tomas kettenstein um great to be invited here to speak at that com um so maybe start as a reminder what's mev amiv is a minor extractable value and in theorem 2 it becomes a kind of validator extractable value and it's all about reordering censoring and injecting transactions in the mampo while constructing a block so all that value that can be extracted by executing practically those three operations and

while we say it's minor or validator extractable because they are the ultimate acceptors of the block in whatever form it is so even if that meb is extracted by someone else it has to be approved by minor validator and in the past this would be something that they will be available only to miners without any democratization and nowadays we're talking about mvv in the context of access for for arbitragers for cooperating with miners uh using various types of tools and uh what's x die exactly is proof of stake based with 20 validators but open to more ethereum sidechain uh has faster blocks it has arbitrary message

breach a stable coin is a native currency it's packed to mainnet so you can move the die to x die between there and back and it has quite a lot of defy activity uh it's much smaller in this regard than magnet but it's a great playground and a growing network that's cool to uh to play with mev and mov itself in ethereum one maintenance nowadays uh already struck by explorer.net already amounted to around 700 million dollars and it grows significantly during the periods of increased volatility

uh it does use the opportunities that arise from big price movements because you have temporary misallocations of assets temporary decentralized exchanges showing different prices and also very often if those price movements are rapid then you can have multiple liquidations of various d5 protocols that are available for whoever is able to access those via mev and currency you might be aware that flashbots relay is available to deliver the bundles uh so-called bundles of transactions extracting the meb uh to the mev clients like mev geth

and this allows multiple searchers so anyone in dependency trading and analyzing the chain to interact with relay and deliver that so you don't have to work directly with the miner you can use that relay in between that ensures that you have many miners to work with now in ethereum too what's particularly important is we will have after the merge mev for validators and it will significantly significantly increase uh the rewards to validators on top of the existing reward for securing the network you also can expect that mmv will be important part of the cash flows and generally

mev will become vev if we can say it or simply will keep using the same term um and it will be important to know that validators are known slightly ahead so it changes a bit how we interact between the searchers and miners or whoever block builders and miners we deliver information early to very specific validator we can decide with which validators to work with um with some of the mev that is available on the longer term of time it opens a few other things we'll discuss later and generally this research of ethereum to mev can be tracked at flashbacks frp14 flashbacks research proposal uh where we

discuss um a lot of those different scenarios like how exactly movie behaves in ethereum too and when we discuss those scenarios you can think of multiple levels if okay you simply run it validator or maybe you go a bit further it starts extracting mvv um sounds like a must but then maybe you start speaking cash flows and trade and decide whether you want more risky or less risky cash flows whether you want to maybe sell your mev cash flow and then go further and start asking questions about multi-block mev a bit unexplored but potentially very powerful tool how things interchange inter layer mev when you start thinking about layer 2 and side chains and main net interactions

uh so for the mvp post merge there are a few proposals or maybe many of them but few that uh we really look at a bit more deeply uh so for sure we built examples of mubs gx in ethereum 2 on top of the existing flashbots mbbs gx so more like hardware based ensuring ensuring privacy of the bundles delivered to the validators we to work with tools like blog builder proposer splits that were suggested by vitalik so it's more like consensus a level idea where you ensure that the rules of the fork choice and the rules of the accepting the block

are built into uh how ethereum to operate we experiment with carrier transactions that are more like a way of delivering those transactions and ev bundles using the devp network directly to validators because suddenly when you know up front who the validator is you can send those messages much earlier and encode them encrypt them in this method for specific validator and because of all of the splits between black builder block proposer the validator the um searcher and so on so we start to see that uh there are multiple possible dynamics between all of those participants so the stakers validators and so on may define how

exactly the split mev rewards and it all depends on their power uh inside the construction the consensus construction in etm2 after the merge who exactly has a final say of how the block looks who exactly can choose from multiple different parties and choose the best option and how they can compete with each other and so on how they discover and how they also value the partners that are smaller or bigger or more trusted and so on we want to remove that trust requirement and make sure that these small validators and small block builders and so on are equally important in the system otherwise it may create some uh difficulties to the how to secure the

network efficiently and when we explore the media on site checks later too very often we start asking the same questions so um we start having in many of the solutions we have validators it's more like proof of stake solutions uh xdia being one example but we also have energy web we have polygon we have layer two solutions which have more like uh sequencers for the transactions that are delivered to mainnet so start net or beach from hermes optimus casing and so on and so on all of those competing ltf solutions uh they sooner or later they'll start asking questions about how how to efficiently use an evidence and we ask questions about the intro chain versus inter interchangeable visa whether we stay inside the side chain or

or try to ask maybe is maybe available when you start being a validator sequencer on both networks can you can you do something with that and um so using this imev with nethermind uh knowing that the mind is now practically a major client supporting uh xdi hormones the only one after open etherium is no longer supported and many validators already switched the nethermind and probably around about 50 percent of the validator network for x-die is operating on nethermine we could build the mev delivery like bundle delivery tool for x-die and

it uses simplified flashlights relay architecture um as we were building that in cooperation with flashbots at nanomite and it uses another mind mev plugin architecture for adding all of this bundle support on the validator side so it's similar to mevcaf construction uh initially it started with one out of 20 validators and soon you can expect that uh majority of the network of validators will be accepting bundles in this format and then you can think of examples of where we can play experiment with maybe it can be dark forest perpetual protocol which is a major one of the major d5 projects on xdi nowadays but also lots of other amms d5 projects um and i can start exploring that uh similarly as you

would slowly learn about using anything on main and yxi is particularly important when why did i start talking about ethereum too because buff x die entity and two are preface tech networks and when we start asking those questions about how to solve the problems uh it means that we can experiment with those on on proof of stake networks side chains like xdi we can use them a bit as a calorie network uh for the value chain the body in the change it already has hundreds of millions of dollars body locked uh for testing the concepts like multi-block mev interactions between searchers

and the block producers like the block publishers and one of the things that we also want to test is how the mev cash flows can be split so uh can you look at in the future for the ethereum to cash flows you think that okay first of all let's fix return from the current number of validators the reward from ethereum two right the the validator return and then you have a bit of a dynamic component if there is a fluctuation of number of validators you deposit you have new deposits you have withdrawals and then there are transaction fees and mev and these are very different risks and rewards scenarios and you may start

trading those cash flows and as a validator as an operator of many validators and you can use some of the existing or future protocols euler yield protocol and integrate those maybe a slider rocket pool i think we can see some interactions of those pools in the future so thank you so much and please contact me if you want to explore further some of those concepts and you can find myself or nevermind at twitter also write to me at flashbots nevermind thanks a lot and i hope that you start using xdi in the future you use it for experimenting how it will work in ethereum too to run mg