State of the MEV Union - Jon Charbonneau

foreign [Applause] [Music] welcome to the meth economics uh we've got a ton of great research topics set up here at the entry section of Mev ethereum all things economics um first shout out to each Global and flashbots for putting all this together especially team and Sarah um and thank you again Tina for informing me that I volunteered to be roastmaster for this just a quick scheduling note all the presentations here we'll have some q a at the end um so please submit any questions you've got um if you're on the live stream we will be checking them I'll prioritize the spicy ones so please know softball questions with that I will get into the first presentation thank you there we go well I'm set this will just be a very broad overview to kind of set the stage for all of the other talks today and this is basically the State of Affairs for everything Mudd right now them basically as everyone kind of knows uh merge happened at the end of the last year completely changed with block production supply chain this is broadly what it looks like right now um this picture comes from Barnaby who's obviously working on a lot of the stuff from theorem Foundation um so broad transaction supply chain users submit Transit actions um Searchers will take a lot of those transactions make specific bundles they send them to builders those go through relays and eventually validators will take those blocks um that they'll just make for them and propose those to the network um this kind of separates that specialized task of actually building blocks um so the validators can have permissionless access to Meb and not have decentralizing pressures of needing to be sophisticated actors who need to do that more complicated tasks so this is what the Builder Market has looked like since the merge um as you can see it has Diversified a lot um since the start of it or right around the time of the merge um it was rather centralized in the beginning where there just weren't a whole lot of parties there yet um since then it has broken up very very nicely um such that you have a lot of different participants who are all taking you know 10 to 15 20 market share and then the long tail of a much smaller ones as well um so as you can see kind of in this graph below at the bottom of the table below at the bottom here some of these Builders are able to take a nice margin on that uh for some of the bigger ones these are the profits that you know some of them have picked up over the last week for some of the big Builders um some of them do run in a profit like as you can see with some of the bigger ones like Beaver here arcing uh 0x69 and others just run at flat where they pass everything through and which is the illuminate democratized distribute that one slashbox that's what they do one of the other exciting things with Builders as well is we'll start to see going forward hopefully is their ability in their kind of role here to add a lot of additional features um going forward which I would recommend you to see Alex stokes's talks on those topics because that is a very long topic and it ended up itself um so the big question here is that everything looks pretty good um from like the start of the merge I think most people would agree that we're pretty happy with the way that the building Market is shaped out that there's a good amount of decentralization now in that there's a lot of different participants in there who are acting pretty competitively um so I would say it's a good start um but there's obviously concern that there are very natural very strong centralizing tendencies in this over time um so that's where it comes in trying to build an actually fully decentralized block building Network um that kind of plugs in here and is able to have many many participants who actually contribute to even building a single block um and that is exactly what Suave is as I'm sure everyone here has heard it by now found here is just a very quick overview of it uh so moving on down to the next step of the slide chain um relays is basically kind of just the the pipes in the middle between the builders and the validators who are sending those blocks along um again you see the very similar Trend where at the beginning it was very centralized it was basically everyone was using flashbots for the most part um and since those first couple of months we have seen a lot of diversification on there which has been really nice um and the biggest one now is actually not flashbots um and it is I believe last I ejected is the ultrasound relay um especially after they launched very recently I believe a week ago they're optimistic relays so that is kind of a New Concept where you start to realize that there's a meaningful latency Advantage if you cut out that kind of Step of not requiring about uh not requiring the relays to actually validate the block that they are sending along um to the proposers that cuts off you know 100 200 milliseconds of latency and that does add a meaningful benefit um so it does come along with some risks because now they are no longer checking the blocks that Builders are sending them um to the full extent um but they're also collateralized by these Builders so Builders who want to be able to get that extra latency Edge and give their block a little higher chance of getting accepted what they'll effectively do is they will just put up some form of collateral and you know hey if my payment that I said that I was going to give the proposer doesn't go through for some reason if something's screwed up then you could take my collateral basically give it to them and I'd get kicked off of that relay um so there's some element of trust in there but there is also this fallback option to make sure that proposers don't just completely get rugged on it and recommend checking out the methus community calls there's been a couple of them now where a lot of this stuff gets discussed so the general trend on this I'm sure everyone has seen this chart by now um and this goes kind of hand in hand with a lot of that diversification that we saw throughout the relay and Builder Market is early on um the large majority of relay and Builder adoption that we saw was primarily coming from ones who filter out ofac transactions which obviously got a lot of people worried in the earlier parts of the Year particularly if that started to continue to Trend up you'd potentially have to wait a rather long time if you were one of those sanctioned listed addresses thankfully a lot of the really diversification and build a diversification has come from parties who are not engaging in that ones like ultrasound relay who are not Sheltering those types of transactions out so the the big question kind of here becomes is the ethereum protocol is very kind of simple today and most of these types of actors are sitting outside of the purview of the core protocol itself so the perfect example of that is something like imp uh in protocol proposer Builder separation so as of right now we need those relays because there isn't this notion of that market structure and allocation mechanism for for PDFs actually built into the protocol itself um so we do rely on those other actors and similarly with Concepts like restaking is it possible to bring those kind of commitments into the protocol such that the protocol is actually aware of those things and that becomes a very big question of what is it actually the ethereum protocol's job to guarantee um should it be bringing all of these things in if this is if we see this as a really important and potentially dangerous role that is sitting outside protocol that has a lot of power or something like Builders should we bring that into the protocol itself um something like him from Pro in protocol PBS the nice thing about that is it removes the need for relays so you get rid of one of these vectors for censorship in the middle you remove another party you make that kind of Engagement between the proposal and the Builder more trustless that you know that they will get paid um but it does also mean that you are now enshrining additional complexity into the protocol and there's a possibility that you know you find out a year or two later oh maybe there's this thing that we haven't thought of on the incentive compatibility on the economics of this structure um so there has to be a pretty high bar to like actually enter on something and it probably looks like something it like um and Shrine and TBS will take some time if it is going to happen um so this is kind of the core idea between uh behind a lot of what Barnaby has spoken about in the past year this concept of Pepsi which he has uh mentioned before which is effectively making the protocol aware uh it's it's an idea it's not a specific proposal or anything that's like an EIP or anything um but it's an exploratory idea at least of what if we just make the protocol more flexible in understanding these kind of external commitments um as opposed to trying to enshrine everything uh into the protocol itself like a very specific uh proposer Builder separation model um and this kind of relates to restaking a lot um because we're starting to have this idea where proposers can potentially opt into external commitments outside of the protocol that hey I'm saying I'm going to secure this other application but the protocol is not aware of that at its core um so there are obviously a lot of risks that come along with restaking that people are concerned about but the reality is at zero and we just need to understand how to build with it responsibly and one potential uh kind of answer to that is the vegetable is making the protocol more aware of it and being actively flexible as opposed to enshrining very very specific and complicated mechanisms into the ethereum protocol so looking at the validator market here um I'm sure a lot of people have seen this before but you see the similar trend of we kind of realized that a handful of large operators do control a very very large amount at stake um and while we have high Economic Security we also would like to much more decentralized the actual State behind that such that there are many operators behind it and one of the things there is obviously liquid staking providers they have a number of operators who are underneath the hood of something like Lido something like rocket pool um and they obviously offer a lot of nice benefits to users they're effectively a backdoor way of warrants for users it gives them a liquid token um so there's going to be a lot of development here on things like as Lido talked about their staking router being able to bring in things like distributed validator technology to hopefully decentralize kind of uh The Operators under the hood they're even more than they are today so one of the big trends that I we will definitely hear about later today um and probably a lot for the next year is that users have mostly been not acknowledged a whole lot in a lot of the stuff that has been built um in regards to Med over the past year or two um there aren't a lot of options to really give users back a ton of value today for the most part you're going to send something to the mempool hopefully you set your slippage right such that you know you don't get sandwiched um but there's not generally this kind of competitive process where people are bidding you back the actual value of what your order is creating and that's kind of the core concept behind order flow auction such as the flashblocks has proposed with webshare early on there's a number of others who are building very similar Solutions whether that be block native and others um and this is basically that competitive process where Searchers can say hey I realize that I can capture 100 if I have the right to background this trade so I should be able to bid in this auction I'll bid 99 for the right to do that and that kickback can then go to the user returning them some of that value it becomes difficult to do this in a trustless manner and that's where a lot of the kind of programmable privacy Tech that you hear about that is going to need to be layered on over time such that you're removing trusty components like something like the person who's running the auction effectively in medshare early on or not technically running the option but the Matchmaker who's kind of doing this you are trusting them and so removing those trusted parties is going to be a big part of flushing out these kind of systems another trend is going to be Wallops um they're they've been a bit of the sleeping giant for a while here where they're in a very powerful position of aggregating user order flow and sending that through the supply chain like that is effectively where everything comes from so the simplest thing for them to do is you strike private deals you do payment reward of low I hey I'll only send it to this Builder you execute this for me you kick me back some money um the threat of that obviously is it becomes very centralizing um that gets back to the centralizing forces that we talked before with Builders um so you would ideally like to see a much more permissionless in the open system where they're not doing that otherwise we've effectively just created a worse version of tradify um if you kind of play that out at the application layer this will be another Trend that you start to see over the year is most of the general purpose environments like let's say uni swap or some decks or whatever on a general purpose environment most of the value from that is generally going to go to the ethereum validator some Searchers and Builders Etc but realistically applications want to be able to capture that value from for themselves because that's one of the core values that they do create so you see very custom uh ideas around this stuff like osmosis is implementing a module where their validator will just automatically do these arbitrages and capture that revenue for themselves to be distributed as they want it's much easier to do in an app game something like that where you can just actually customize the consensus becomes more difficult to in general purpose environment um and then lastly here we have l2s that you'll hear a lot of this for uh for jasus and through in stock later on um but Mev is obviously going to be a big problem on layer twos it doesn't the problem doesn't just go away once we kick it up there and it already even is a problem today even with the centralized sequencer um so even with out decentralizing at all there are meaningful problems um that do need to be addressed significantly better because the naive implementations of this something like a first come first serve uh you decide up with these latency games and as Troon will explain later uh the idea of fair ordering is a bit questionable um so decentralizing it will make it even harder um and you introduce a lot of new complexities and height how do we try to maintain a private mental anymore now that we have more participants in here stuff like that and the reality is I think it's more important to deal with the mevfx out of something like first gun first serve then it is probably to decentralize the sequencers predicted in the short term because the whole point of Roll-Ups is if they're implemented correctly where you remove the upbringing Keys you have permissionless proving Etc um they're safe anyway and users should be able to exit from them and now you have this whole new idea also popping up of share sequencers there's a number of variations there which we'll talk about later today um one of them is what Justin has proposed called based Roll-Ups which is kind of an idea based on what uh taluka described a couple of years ago as total Anarchy which is basically letting the layer one itself sequence for all of the Roll-Ups that are on top of it comes up with a number of trade-offs kind of on ux at a lot of very different economics which are very interesting and then there are also other chains which are looking to provide similar Services ones like espresso and Metro um who will be offering this kind of shared sequencing layer that many Roll-Ups on top of them can plug into and they will be providing that sequencing service for all of them and that gives you a lot of exciting benefits um it also again raises a lot of very complex questions around the economics and the trust in these systems Etc um and those ones also do give you a bit of an easier route to one of the nicest things that people like about Roll-Ups of the user is you want those really fast pre-conformations that's really difficult to do possibly not practical at all to do um with something like a baseball where you probably are going to fall back to the L1 block times um so these are all very interesting ideas which we're going to hear a lot more about later on today and that is all from it foreign