On The Brink with Castle Island - Matt Cutler (Blocknative) on block building and the economics of MEV (EP.401)

Episode Date: February 27, 2023

We host Matt Cutler, CEO and co-founder of Blocknative, in this episode to discuss the intricacies of block building and his thoughts on the future of MEV. We cover: The phenomenon of certain block builders subsidizing transactions The requirements for operating a competitive block builder Margins required for sustainable block building MEV redistribution and balancing interests of users vs. validators.  You can follow Matt @mcutler on Twitter and learn more about Blocknative here. 

Welcome back to another episode of On the Brink. In this episode, I hosted Matt Cutler, who's the CEO and co-founder of Block Native, which is an infrastructure provider that's deeply involved in MEV through its operation of a builder and relay. We cover some interesting topics like the requirements for operating a competitive builder, exclusive order flow, the economics needed to justify block building, redistributing MEP to users and transaction originators and the role that privacy can play in reducing the negative externalities of MIV. Matt shares a wealth of knowledge that helped fill some of the gaps that I had in my

understanding of the landscape. So with that, here's my conversation with Matt. Welcome back to On the Brink. This is Ria, and today I'm excited to be talking to Matt Cutler, who's the CEO of Block native about his thoughts on the risks and opportunities posed by MEV. Matt, it's great to have you back on the show. I urge listeners to check out the December 2021 episode with you as a good primer for this conversation.

But before we delve into how you're thinking about MEV, could you give us an update on, you know, yourself, just give us a quick background and also give us an update on Block Native and how the company has evolved since you were last year. Sure, of course. Thanks for having me back, Ria. It feels like a lifetime ago, December 2021, that I was on the show. I am founder and CEO of Block Native. We are specialists in core infrastructure for the Ethereum ecosystem, specifically the

pre-chain layer. So everything to do with the transaction lifecycle from the moment a transaction is as intent to go all the way through. getting on chain. We've been building our infrastructure for almost five years now, and we are lucky to work with hundreds of the top projects in the space, including many of the household names in the category. And as part of the merge, which was September 15th of last year, we took all of our infrastructure and technology and capability and evolved it into block building. So we, today are one of the major players in the block building ecosystem. We hover anywhere from three to

10% share of the network. At times, we've been actually north of 45%, which is pretty exciting. And so we live and breathe this stuff every single day. About me, I'm a glutton for punishment. I like to say I'm a serial entrepreneur. I've done seven or eight startups in different categories, depending on how you count. I've bought them, sold them, split them, shut them down, taking them public, all of the above. My very first startup was the first ever web analytics business. It was called NetGenesis. That was a nine-year overnight success, zero to IPO and the Web10 boom. And my most recent startup was a mobile collaboration platform called Collaborate.com that we wound up selling to Cisco.

And that's why I live out here on the West Coast. I actually am from the Boston area originally. And so it's no mistake that I jumped into Web3. I'm pretty early in sort of emerging technologies and on the sort of infrastructure and data side. So it's no accident that we at Block Native do what we do. do because that's sort of the stuff that I know how to do. And it's been quite a journey to put it mildly. Yeah, I can imagine you've seen the space evolve. And like you said, it feels like an eternity between the end of 2021 and now and how significantly the entire ecosystem,

but especially Ethereum, has changed in that period of time. You mentioned a term that I think might be helpful for listeners, if you could expand on. You said you block native operates in the pre-chain world. Could you just expand on what pre-chain means? Sure. So pre-chain is sort of the hidden force behind most of what happens on-chain. So everybody has access to on-chain data. They can see which transactions were confirmed and when and what the value was.

But the pre-chain layers, all the stuff that happens before the network reaches consensus. It's typically referred to as the mempool, and it is the peer-to-peer layer where transactions that are candidates for inclusion get circulated across participants and ultimately compete in real time for one of the few slots that are available or a few spots to be confirmed in the next block. And it's interesting because there is no truth in the mempool because there is no one mempool, right? That every node on the network has its own unique set of transactions. they gossip them across each other. And typically with low network congestions, most nodes see most of the same transactions with certain parameters. But certainly during periods of heavy network usage and heavy congestion,

you can see quite different sets of transactions depending on where you are in the world and depending on who your peers are. And so it turns out to be a pretty difficult area to get understanding of. It's a real-time layer, meaning anyone in the world can submit a transaction at any millisecond, and often do. And there are no really great tools for typical folks to get access to and be able to work with this data. And in fact, working with this data requires specialized infrastructure, requires specialized capabilities, and that sort of favors large, deep pocketed actors. So we at Block Native, for the better part of four years now, have been building public Mempool

data infrastructure. So we operate the industry's only Mempool Explorer, where you can go to our website, blocknative.com. You can create a free account. You can type in any transaction hash. You can type in any smart contract address. And you will get real-time updates on transaction state changes across our global network of nodes. It's super easy to then work with an API level, so you can incorporate this into whatever you're building. And quite frankly, many of the major traders and searchers and even block builders rely on our infrastructure to power facets of what they do. And so part of what we do as a business is to make working with Mempool and pre-chain data, easy as working with any other part of Web3 data.

And so that's pretty fun and pretty interesting and something we've committed to for a long time. It was historically viewed as somewhat of a sideshow. Like it was interesting to certain people are interesting in certain contexts, but not really a mainstream thing. With the merge and with block building, the peer-to-peer layer has really come to the four. And so it brings sort of new vitality to the capabilities that we have, makes it more strategic, makes more people dependent on it.

And what's fascinating is this element of the peer-to-peer layer is actually growing in importance and complexity as we look ahead down the roadmap. In particular, there's all sorts of new stuff happening with EIP 4337, which is account abstraction, as well as with EIP 4844, Proto-Dank charting. So it's tough stuff, and it requires sort of big real-time infrastructure to maintain and manage, and that's what we've been doing. So we're really fortunate to be specialists in this area and really excited to work with. so many members of the ecosystem and leveraging this data moving forward. You mentioned block building, and we've covered the MEV pipeline on Ethereum post-merge in-depth in previous episodes, you know, the first with John Charbonneau and with Robert Miller as well. But just so that listeners who may not have heard these convos or may have forgotten are on the same page,

could you just summarize the pipeline for us? And then, you know, this is something that we were discussing before we started recording, but also talk a little bit about who within this pipeline captures most of the value from MEV today. Sure. So, by the way, both excellent episodes. I highly encourage listeners to go back and listen to both of them. Great, great content-filled, dense episodes. So under proof of work, you have.

had miners who competed to solve computational problems for the right to propose a block to the network. What isn't very well understood is that, or wasn't well understood at the time, was that though there were many computers that were mining and therefore securing the network, they were largely controlled by about four or five mining pool operators. So basically they were sort of clustered together. And these four or five entities had this very unique job of building the block, of deciding which transactions would be included in what order, and then trying to solve the math problem associated with that. So as decentralized as we like to think, Ethereum was at that time,

there was a small cabal, literally less than five private entities, most of them located in domiciles that are not the U.S., who basically determined which transactions got included in what order. Okay. And so this was highly centralized, and also this notion of MEV is known as minor extractable value, or maximum extractable value, and it is in any transaction system, which are all ordered, you can't have a transaction system without transaction ordering.

Whomever is in the privileged position of controlling that ordering can extract value from doing so, because there's value of which sequence transactions go in, there's value in including specific transactions in specific places, there's value in excluding transactions overall. And so fast forward a little bit. there was a research project known as flashbots that started to detect that there were some backroom deals happening in games occurring, you know, private relationships that were formed that would provide certain actors with preferential inclusion and preferential sort of ordering, if you will, and that this was perceived as both damaging for the fairness of the network

and also potentially centralizing because there's a lot of value here. They created a project called MEV Geff that basically outsourced block building to flashbots itself and some of the systems that they built in and basically created a off-chain marketplace for folks who cared about this stuff to bid for an inclusion. And this had a bunch of positive consequences in terms of removing negative externalities or some of the things like gas price wars and PGAs that were occurring pretty frequently in the network. at that point. But again, looking back, it was not the greatest situation because you had relatively small sets of folks who were in charge of really important areas of the network. So enter the merge. Ethereum goes through this long plan upgrade where it moves from proof of work consensus to proof of state consensus. But these factors of MEV potentially get amplified under the situation. And there's a bunch of factors why. But one of the most concerning

is just like you had mining pool operators that sort of pooled computing power, you know how it's, you know, have staking pool operators that that pool stake. And a bunch of big staker, sticking pools start to get formed. And they have quite deep resources. And it turns out that if they get good at extracting MEV, because now they're going to be the ones who can propose blocks to the network, they'll receive outside rewards. And they'll receive larger ETH for that. And they'll be able to offer higher rates of API, higher yield for their stakers, and then ultimately everyone will start staking with those parties because they just earn more money. And this becomes an irreversible centralizing force in the network because the stake itself will go to these deep pocketed actors

who have better control over the network than others. So in response, a new approach was developed known as PBS, Proposer Builder Separation, and in particular, Proto PBS, Proposer Builder separation splits out the roles of proposing block to the network from building the block itself. So Proposer is the validator who's the tip of the chain. The builder is the one who determines which transactions get included in what order. This is achieved via an out-of-protacle mechanism called MEV Boost, also developed by FlashBots, as somewhat of a stop gap. There are certain trust assumptions associated with it

that were perceived as sort of the best available tradeoff given some of the state of the research. I think there's desire to do in protocol or enshrined PBS, but it just wasn't ready at the merge. And the perception was the tradeoffs associated with the MEBoo sidecar were preferential to the risks or preferred to the risks of centralized state. So at the merge, the network fundamentally changed in that a proposer, i.e. the tip of the chain, could outsource block production to an open marketplace, that whereby anyone could compete

to build a block, bid against other block builders. And if they had the highest bid, then the proposer would select their block and that builder would earn the right effectively to determine which transactions got included in what order. And by the way, that's what Block Native does, where one of the competitive block builders on the market. And so what's interesting now is there's a couple of new steps in the process. So you have users or bots that conduct transactions. They get submitted to the public mempool.

There are groups called searchers who basically inspect transactions in public mempool, try to evaluate them for MEV. Now, searches are an interesting term because searchers feel like people, They're folks who are out there, you know, methodically searching around. And that's generally not the case. They're just bots. They're automated systems that very high speed evaluate transactions for MEV, like if they could get ordered in a certain way. Or as a result of a transaction, there's going to be an arbitrage opportunity over here.

So if I can put a transaction immediately behind this transaction, I can capture that arbitrage opportunity. And what searchers do is they try to figure out what the value of those. opportunities are and then they bid against each other right and so I see an arm you see an arm and we the arm might be worth $100 I might say well I'll pay $50 to make 100 and you might say you'd bid $70 to make 100 and you would get that opportunity because there's only one searchers do their thing they're evaluating they submit bundles so these are bundles of transactions to builders builders aggregate transactions from the public mempool from searcher bundles and from

private transactions if they allow that. And they put that together into something into an order with a block. And they assign a value to that block. Now, what's interesting about this is you could pass all the value onto the proposer. So you create a block that let's say has one-eath the value. And you could say, hey, proposer, if you pick my block, I'll pay one-eath. And so you just pass it all the way through. Or you can say, hey, I'll bid 0.9th. And there's a gap in there of 0.1-eth. And if the proposer selects your block, then you can collect that revenue straight from the protocol. It's pretty magical. And so there's an opportunity for what's called builder margin, but there's also opportunity

for builder subsidy. So you say, well, there's one ether value in this block, but I have some side agreements over here. So I'll actually pay the proposer 1.1. I'll have negative margin on the block, which interestingly has emerged as a much more common behavior than I think anybody anticipated when they were conceiving of this. So you have builders that actually subsidizing the network, which I think is interesting and potentially concerning in certain regards.

And they then propose it on to the builders sort of suggest these on. And the proposer, i.e. validator, selects among them. There's an intermediary, a trusted intermediary here called the relay. And so relays actually act as sort of the go-between between the builders of which there are many and the validators of which they are many. and they basically ensure that everybody plays fairly. But they're a trusted actor. And so the relays over time will go away.

That's the difference between proto PBS with MABI Boost that we have today and what's known as in protocol or enshrined PBS moving forward. So a bunch of moving parts here. But what it does is allow for specialization at each level. It allows for a more decentralized effort. It allows for more people to participate at various levels of the network. and it would appear to be pretty clearly the network operates a lot more efficiently under this model. Today, of all blocks on Ethereum, 90 to 95 percent are outsourced in this fashion.

And the yield which is generated is generally about double what it would be if it was done in-house and internally. And so this is quite positive. And then finally, this also opens up all sorts of new avenues for programmability and new possibilities for routing value of MED that wasn't possible before. So there's a wave of innovation that's coming around intelligent things to do around managing, mitigating and rerouting MEP. So that's all stuff. I think one thing that's interesting that we haven't talked about in the past is this idea of block builders actually subsidizing transactions and paying more than the value of the

block to the validator. Could you talk a little bit more about that? and what allows different block builders to do that? And I think you alluded to this, but maybe expand on the risks of centralization that that creates on the network because I imagine that not all block builders that are trying to compete,

have the capacity to actually subsidize blocks. Yes, this is an interesting game. And by the way, one of the great things about block building on Ethereum today is that it's very, transparent, that there are many dashboards out there that are widely used by builders and validators to understand what's going on. And there's full transparency and visibility. So the one that is probably most widely used is called relay scan. It's relay scan.io. I believe it's a project of a member of the

flashbots team. I'm not entirely sure about that. But there's, it does two, it does two interesting things is one is it shows share of the network by relay and by builder over various time periods, one hour, 12 hour, 24 hour, and seven day. But then it also shows builder profitability. So there's a little tab that you can flip. And if you click on that, you can see which builders have been the most and least profitable over a given time horizon. And actually very graciously, the relay scan team open source their historic data as well. So there's a CSV of everything that they have since the beginning, since the merge for researchers to do research on. So what that shows is that at any given time, there are certain builders who are capturing margin, i.e. they're bidding less than the

blocks that they're winning and they capture the difference. There are those who are flat. We're among those generally, which are basically have zero margin. And there are certain builders at certain times that are negative margin. And what we're seeing, the, the, the, the, the, the, The block builder market is still pretty new. I mean, the merge was five months ago, give or take. And the economics are still emerging. We think this is an area that requires a lot more attention from the ecosystem, largely because it's in everyone best interest to have an economically sustainable block builder market

with lots of competition. And where you have poor or negative economics, you basically create centralizing forces, whereby only deep-pocketed players or players who are able to, who are, capturing revenue sort of outside the bounds are able to operate, right? And the big fear here is you're going to have an outsource block building market with a single builder. And so to get on chain, you have to go to, you know, Megabilder XYZ. And if, you know, you're not friends with their CEO or you're competing with some of their things, they just say, no, we're not going to include you. And there's sort of no recourse because they control the contents of the block. This is generally known as

MEV dystopia, which the FlashBats team has been pretty vocal about, which is, you know, where you have builders who control the majority of the network and can impose their preferences, which is an interesting term, on the rest of the network. Now, what we're seeing so far is pretty interesting. So first off, one of the general themes of block building is you want to have a decoupled supply chain. So you want to have searchers who are independent from builders so that searchers are competing with each other and that they submit their bundles to builders who are sort of dispassionate about what they include or exclude. But there's very clear evidence that certain builders are also searchers. They do what's called self-searching. And this creates a vertically

integrated stack. And why this is concerning is the objective among builders is to have parity of meaning everybody has access to the public mempool. If everyone's doing their job, you should have access to the same searcher bundles because the same set of actors who are perceived as reputable by the searcher community, there's no economic, there's no reason why not to socialize your bundles. And then there's some private transaction flow, but effectively that each builder has the same ingredients to work with in building a block. And then they compete on, you know, the algorithms and various merging techniques to

most rapidly build the most profitable block, right? And the thing that I think a lot of people don't really understand is all of this magic needs to happen in a 12-second window. And in fact, a less than 12-second window because everybody needs time to collect their stuff incorporated and disseminated and there's latencies that occur and you have global phenomena. So everything we're talking about here with all these moving parts that's somewhat of a Rube Goldberg machine needs to happen in a matter of just a few seconds. And delays of hundreds of milliseconds can be the the difference between success and failure. And failure in many cases is you miss a slot. The proposer doesn't get the information that they need, isn't able to distribute fast enough,

and instead of proposing a block to the network and getting a reward for doing so, they get nothing. And so this is generally problematic, and we can talk about that. Now, getting back to builder economics, there is evidence of searcher builder integration. And so what occurs there is if a builder is self-searching, meaning they're also searching for opportunities, then they can privilege their searcher bundles over competitive searcher bundles. So Ria sees an opportunity. She bids $100 ARB, she bids $70.

Matt sees the same opportunity. He only bids 50, but Matt's a builder and Ria's not. Matt says, well, I'm just going to ignore Ria's bundle and I'm going to prioritize mine because it's in my economic interest to do so. That's problematic, right? That's not great for Ria. It's not great for fairness. And so in general, the notion is you want to have separation here,

but without good economics at the builder layer, you're incentivizing builders to go down the stack and to seek out profit opportunities elsewhere. There's countervailing forces there, of course, because then searchers are saying, well, why would I submit bundles to you if you're just going to steal them for me? So there's games that can be played at that layer. Then there are also very clearly builders who are,

subsidizing blocks. At the beginning of the merge, there were certain entities who were subsidizing blocks, but they were subsidizing pretty small amounts, like literally nickels and dimes, few dollars of value at a time. And that was the difference between winning and losing a block. And so one of the perceptions was in the early days of the network, certain builders would subsidize or to increase their win rate, increase their profile, and drive adoption, which is a perfectly rational thing to do. Generally, again, I think not healthy for the long-term success of the network, but rational for short bursts of time. What's emerged recently are builders who are subsidizing blocks at hundreds of each, like very serious amounts of value. And I think this has a lot of people

sort of scratching their head. These actors are not, these are very sophisticated actors. They're not doing it by accident. They're not doing it by mistake. It's because there's a value that they're able to recognize off-chain somewhere else in the stack that isn't apparent. But what it does is it makes it pretty hard to compete. You and me are competing as block builders and we're trying to eke out, you know, one-eath here, point one-eath there, 0.05-eath there. Somebody else comes in and subsidizes a block with 50th, right? It's just you and I don't have the resources for that. You and I don't have access to those sorts of things. And again, it creates disincentives for participation. And it creates sort of the underlying conditions for centralizing forces. Now, at the end of the day,

like, this is an open marketplace. People can do whatever they want to do. There's no like, you can't pluck your tongue and say bad. But, you know, again, what we want across the board is a healthy market where infrastructure operators like Block Native are economically incented to participate in ways which are transparent and are credibly neutral to the extent that we can be for the entire network, right? And so I think this story is still being written. And I think we're still seeing sort of emergent behaviors in the network and we're still seeing sort of how folks are going to interact. But it's a pretty exciting space to be a part of and every week brings new surprises and new challenges without any question.

So, you know, putting aside vertical integration and the ability to leverage revenue and value captured in other parts of the stack to kind of win the ability to propose a block or, you know, have a validate or choose your block to propose to the network. What is required to build or to create a builder that is competitive? Like, what is the infrastructure, the algorithms, the technical requirements to create and operate a competitive builder? Well, so being a builder requires a pretty significant set of infrastructure. So first off, you need high-speed access to the public mempool.

This generally requires multiple points of observability around the world because you want to have resilience to network congestion. So you're going to have a bunch of nodes located at different regions around the world and configured in such a way and set up in environments that give you pretty rapid access to what's going on, maybe have distinct peering relationships, and are able to move data around the world pretty quickly. We at Block Native have done a whole bunch of work to basically move data faster the peer-to-peer layer in efforts just to give better visibility into the public mempool out of global basis. One.

Two, you need the ability to accept bundles from searchers. And so you need to have RPC endpoints where searchers can submit those to you. You need to have documentation associated with that. You need to build relationships with those searchers so they know about you and they view you as a credible place to share their. bundles with and a certain degree of trust where you're not going to steal their bundles from them. You may or may not have relationships with wallets or other folks to get private transactions. So you have a number of what are known as private RPC endpoints where transactions can be

submitted directly to builders without going to the public mempool for folks who want to protect their transactions. You need to pull all this data together. In that case, would a wallet, would a wallet, it that a builder potentially has a relationship with, send them transactions directly as opposed to, you know, submitting it to the public mempool where searchers can create bundles where they capture MEV from those transactions. Is that the right way of thinking about it? Well, so their private RPC endpoints can be incorporated in various ways. So for instance, in Metamask,

you can add custom RPC endpoints. FlashBots has an RPC endpoint known as FlashBots protect. You can configure your Metamask to submit there, and then every transaction you submit via your Metamask will go directly to FlashBots and subvert the public endpoint. That's up to you as a user. Other wallets may choose to basically build that into the code, so users may or may not be aware of that.

And in fact, we talked about this a little bit earlier, you know, in other systems like, Robin Hood is a stock trading platform that has zero fees. How do they do that? When you trade via Robin Hood, they submit your order flow to folks like Citadel, who then basically try to extract value through MEV-like things there. And that's how the fees get paid effectively, right? And so certainly there are all sorts of experimentation going on right now with sort of

various tie-ups here, some of them above board, some of them less above board. And I think that experimentation will continue to proceed. We are working on an initiative that we'll be sharing more publicly at East Denver in a few weeks called Wallet Boost, which is basically an open network whereby this sort of thing can happen, where wallets can propose or can share transaction flow with searchers in an open and permissionless fashion, and users can potentially receive some of the rewards from the MEV that they participate in. FlashBots has recently shared something along similar lines called MEV shares.

So there's a bunch of innovation happening in this area to make this stuff programmable. But as it relates to private transactions, yeah, there are various operators of private RPC endpoints. We Block Native will be offering one of those who don't offer one of those yet. and it's just a mechanism to protect transactions from observability before they get included. Again, another way to think about it is they act as a bundle of one. It's just sort of a mental model for that. And they prevent anyone from front running or back running or sandwiching whatever transaction you have.

But does that kind of create or does that have other potentially negative consequences? Like you use the example of payment for order. flow in traditional markets where market makers pay brokerages for routing order flow through their systems as opposed to, you know, through like the exchanges. And I think there's controversy around whether that actually results in better or worse outcomes for users. is in general proprietary order flow or i.e. payment for order flow is perceived as a centralizing force. And there's a difference between private order flow and private transactions, right?

So a private order flow is there's a certain wallet, wallet X, and they're only going to submit their transactions to builder Y. Okay. Nobody else gets them. And what happens there is builder Y can build blocks that nobody else can build because they have transactions that nobody else has. And so, and again, this is part of the reason why builder economics make a lot of sense. If there is so important, if there's not good economics, then builders are incentivized and try to create these types, right? Of course, in the perspective of the wallet, then you have this critical dependency on the single builder.

And, well, if the builder is not a dominant builder, then your users are going to wait longer for inclusion. And so, well, if I'm a wallet who wants to do this, then maybe I, I, instead of doing it with builder Y, who's small, I'll give it to Builder Z. And then what happens is everybody else has to give their proprietary order flow to Builder Z too, because that's the only way to get included. And now Builder Z becomes the dominant builder overall. And nobody can't even approach them because they see all these transactions and nobody else does.

Private transactions are something different. Private transactions are a user expressing the preference to say, I just don't want this transaction to hit the public mempool, okay? either for my own purposes, I just don't like it there, or I'm worried about someone trying to do something nefarious with this and I want to protect it. And there are various providers out there who enable that. What will typically happen for private transactions is they will submit this to all of those providers simultaneously. So there might be 10 or 12 providers of private RPC endpoints. They're all relatively trusted, i.e. they're actually going to keep it private.

you don't know which one of them is going to be most attributable for inclusion or which one's going to win the block. So it's in everybody's best interest that you socialize the private transaction to all of these entities. Okay. So that has less of a centralizing force and is not, again, equivalent to proprietary order flow. But again, these are, you know, these are relatively sophisticated capabilities. I mean, imagine, you know, you're at the supermarket and you say, well, How do you want your credit card transaction processed? You're like, well, based on the time of day, based on the day a week, based on what's in your shopping basket, based on how crowded it is, based on who makes this little box, based on what card do you have, you have preferences that you can express that may have different consequences for settlement. You're looking at, I just want to buy my groceries, right?

I don't want to deal with all of this, right? And so generally, this sort of stuff is only used by sophisticated actors on the network who have a different level of understanding and are. probably doing a different level of transaction than quote unquote, a normal everyday user, which I find a little bit pejorative to use that concept. But we don't want to force all this complexity onto the user to navigate what really is sort of inside basebally stuff. Ultimately, we at BlockNative believe that this is best addressed by MEV-aware infrastructure, where nobody needs to really worry about it. This is an infrastructure issue and an infrastructure problem.

And as long as you have MEVAWR infrastructure, that everybody up the stack sort of deals with it, but you know, can leverage this. But that's probably another story altogether. And so these are capabilities that exist on the network today. These are things that are emerging. They're still, you know, relatively esoteric. But I think we'll become less and less esoteric as MEV moves to the four, as it has post-merge. And as everybody, as all the builders, people who are building stuff in Web3 get more savvy and sophisticated.

about managing and mitigating MEPE on behalf of their users. But to zoom back out a little bit, we're talking about what does it take to be a competitive builder, right? So you need all these data sources. You need to move the data around very quickly. You need to have a fair amount of compute in order to basically both incorporate all this data in real time, because remember, transactions are streaming in constantly, and to basically experiment with various configurations to extract the most value in the block, right? And so, for instance, generally what builders have done historically is the bundles are perceived

as more valuable than public mempool. And so what would happen would be all the bundles that you get would go at the top of the block and then whatever is left over for the gas limit, you just fill with public mempool blocks, public mempool transactions. But that's not always the case that when there's a big spike in gas prices, like for instance, big NFT drop, that in fact, many of the transactions in public mempool might be more valuable than transactions than searcher bundles. And so, you know, searchers, the builders start to get more sophisticated and start to build various strategies, start to compare them against each other,

do so very quickly because you only have 12 seconds per slot overall. And they need to broadcast whatever they have out to the relays. Oh, you also need to worry about bit optimization because if you're trying to extract profit, you have a certain amount of value that's in your block, then you're trying to guess what everybody else has. Trying to figure out how much margin is there? Because obviously, the more margin you take, you'll affect your win rate.

So if you bid too high, then you're giving more of the value to the validator. If you've been too low, then you don't win the block, right? And so it's a relatively resource-intensive activity that requires expertise across mobile. multiple domains requires, you know, real operating expenses as far as your compute infrastructure, your networking infrastructure, your data layer, and real, you know, coding capability. These are not sort of off the shelf, you know, offshore engineers. These are highly specialized people who obviously

demand, you know, competitive salaries. And so a block building is not for the faint of heart. and every day there, every 12 seconds, there's a real-time competition. It's one of the things we love about it is you can make changes to your infrastructure, changes to your building algorithm, changes to how you do stuff. And within minutes, you can start to measure the impact of that, if any. And also, sometimes you have problems and things get mucked up and your wind rate drops right away and the alarm bells start going off. So it's a pretty interesting ecosystem to participate in.

And it's one, again, that's just, you know, starting up. The total addressable market was zero on September 15th. Depending on what time frame you look at, it's, you know, tens of millions of dollars right now, but it seems to be growing pretty rapidly. And so, you know, we very much hope that compelling economics emerge in the space and the market grows as the Ethereum ecosystem grows and that there are more and more participants who are incentivized to join in because they perceive opportunity there. We really don't want to see, you know, deep pocketed players who have sort of other objectives or other ways that they can sort of, you know, see value, introduce artificial economics that basically discourage competition.

That, we think, is net destructive for the ecosystem. And so that story is still unwinding as we speak. Something that you've alluded to is the margin that building. earn on their activities. Could you provide some more data and insight around that? And maybe talk about what kind of margin is necessary for block building to be a sustainable activity? Well, it's interesting. So, you know, I'm looking at Relay Scan right now. I've gone to builder profitability over the past seven days. The most profitable builder, at this moment, which today is February 17, 2023,

is a builder known as Builder Zero X69, an anonymous group. Over this period, they have won 12,328 blocks. They're one of the most successful. They have 2,265 blocks with profit, but 10,000 blocks with subsidy. Okay. So literally 80% of their blocks are so. they're subsidizing 20% of the blocks, they extract profit room. And over this period of time,

they have extracted 45.65 eth of value. Okay. So 45.65, and we'll just use $1,600 USD eth. So over the past seven days, through block building alone, they have recognized, they have realized $73,000 of profit. Okay. Doesn't sound very sustainable. Okay. It's not a big number. They've subsidized, by the way, three-eaths. So even though they've subsidized a lot of blocks, it's been a fairly small amount there. But the question is, is that a is, how do we feel about that? Do we think that given all the work and given the important role that that builders play for the network, that the most profitable, you know, realizing that's, is that something that we think is sustainable long term?

It's just an interesting question. And again, how do we feel about do we want, you know, what level of transparency do we want from these builders? What level of sort of governance and oversight do we have? Because you have, you know, anonymous groups who, you know, you just don't know who they are, who are quite common. And you have corporations like us who have investors and venture backed and things like that. And again, they're just sort of different classes of actors and entities. Now, of course, I'm not a dispassioned observer here. We're trying to build a business out of this. And we think that long term there will be very interesting economics out of block building and

and things that are built upon it, but so far those largely haven't emerged. Now, at the same time, okay, there's another builder who has their overall profit over the past seven days is negative 445. Okay. So 445 times 1600, oops, 1600, that's negative $712,000. in the past seven days. It's $100,000 a day on average. This builder has subsidized the network. That's insane. Curious, right? I mean, again, if you're a participant on the network, you go, wait, you know, there's someone, some private entity out there that's doing, like, why and to what aims and what are their objectives?

And is it aligned with us? Honestly, we don't think that there's anything nefarious going on. But just from a pure external optics perspective, probably we want to have a, a, a, a, a more sustainable, more transparent, more accountable set of actors in this ecosystem, because that will encourage innovation and competition. But ultimately, we think that's what's going to happen. It's just too early in this market for that to prevail. How do you think we get there? Is it just rallying the ecosystem and saying, hey, this is the ethos of,

of Ethereum, this is against the ethos of Ethereum, we need to be more transparent about disclosing how different entities are able to do this. Like, how do you get to a stage where you do incentivize that kind of transparency and accountability? It's an open question. I mean, again, this is an open marketplace and permissionless. Anyone can participate. Guess what anyone does.

So that's great. We don't want to have this like you got a KYC on your way in the door to be a block builder. That would be that would be bad, right? You know, ultimately we think market forces will prevail. And in particular, you know, what's happening that we see right now is the market is just really small. So you have a lot of sort of external factors that can can influence a small market. But as the market grows, sort of it will it will revert to the mean, if you will. I think also like there are probably actors in the category who are exert.

inserting artificial downward price pressure on it. And over time, I believe that that will not be sustainable or not be viewed as desirable. But again, like the best way forward here is to let market forces prevail because that's ultimately what's going to be most sustainable. Any sort of top-down control, any sort of cabal-like thing or some sort of collusion among builders is ultimately destructive, right? Because again, then you're going to have the in crowd and the hour. crowd. You're going to have the, you know, the folks to do that. So, you know, we just think that the necessary

ingredient here is time. But we do, you know, one of the things we think is important is that this is something that we think everybody in the ecosystem should be aware of and should be paying attention to. Generally, everybody feels like well, infrastructure should be free. And that was probably true in the formative stages of, of the Ethereum ecosystem, but probably is not going to be true moving forward. And if we're ultimately reliant on high performance, high availability, high highly reliable infrastructure, then people need to understand everyone in the ecosystem needs to understand that doesn't come for free, right? That there is very real expertise and very real compute and power and all that sort of stuff that goes into it. And again, we're an infrastructure

provider. So of course, we're we pay the bills in many regard. But that ultimately, you know, having venture capitalist finance the operations of the network is not great, right? It really, the network should be funding itself and should be financing itself. And we think that'll happen over time. Something else that you mentioned is this concept of allowing transaction originators, users to retain more of the value that they create from their transactions. Like you said, FlashBots this week announced MEV share. You also alluded to something that you're working on Wallet Boost.

So maybe could you give us an overview of what those implementations look like in practice? And then, you know, I have a follow-up question around how you balance interests now of users versus validators, both of who are seeking to capture value from these transactions? Yeah, so it's interesting. When the merge was being contemplated and Mavie boost was suggested in the creation of this block building network, one of the key factors was as Ethereum went from proof of work to proof of stake, the overhead cost to operate a node on the network and help secure the network goes way down. Because you don't need specialized computer that consume a lot of power.

You just have a validator that you can run literally on a Raspberry Pi. And so as a result of that, the rewards per block, the rate of issuance went way down. It was called the triple happening. I think it went down by like one eighth. Now, when in that happened, the percentage of the percentage of, value per block that is attributable to MEV went way up. It's not that MEV itself went up. It's that because the denominator went down, right, that MEV as a force became a higher

factor, more influencing force, right? And we said, well, what is the root of MEV? Where does MEV originate from? It's from users conducting transactions. And this was something we were champions of very early, which was saying, hey, there are these entities that we call transatlanticians. transaction originators. So RIA uses her Metamask wallet or her ledger wallet to conduct a uniswap trade via Zapper.

Who originated the transaction? They go, well, there's four parties. There's Ria. There's Ria's wallet, metamasker ledger. There's the DAPRIA used, Zapper. And there is the protocol she used uniswap. And without all four of these things, there's no transaction, and therefore there's no MEV.

However, in the current scheme, RIA is not made aware, nor does she consent to her transaction being subject to M.EV. That feels not particularly equitable. And so we basically very early said, we think this is something that, one, is a problem under the current system. And two, the realities of block building presents pathways to address quite constructively. And we started a meme called Share the Meth, right? And you can find that out there. We've spoken about that quite a bit, which is that fundamentally what should happen is those transactions which are the source of MEV should have the same opportunity to express preference about how it's handled as other steps in the network, right? And that actually the idea that users are ignorant or naive and that wallets are these sleeping giants that are blind to it and that there are these actors out there who just can do whatever they want. And then, oh, by the way, the value they split at the end of the value chain between themselves

and the miners of the values. And so we work very hard on this notion of MED recirculation whereby the value could be recirculated back to the beginning and could be distributed throughout. Now, of course, that means that everybody, it's the same pie that just gets divided into smaller pieces so that all members sort of, get a different slice. And there's certainly a lot of room to argue about, well, how much should

go to which party and why. And the good news is, I don't think we need to solve for that. We can let market forces sort of determine that. There are various strategies. So there's a couple of entities out there which are beginning to do work here. But most of this is sort of research phase right now. So FlashBots proposed their thing called MBV share, which really seems to focus, and I'm no expert on MEB share on users basically selecting which parts of their transaction they reveal. And then searchers sort of guess based on those pieces. And there's this new entity called, I think, a matchmaker. A matchmaker.

Sorry. There's a lot of roles there. I think it's executor in Swab. Yes, exactly. This is like phase one before Swap. A lot of moving parts here who try to match these things up. And that's just, you know, obviously FlashBats. is a very savvy organization, very interesting.

They've thought about this stuff for a long time. And much to their credit, they're putting this out there. Here's an idea. Let's talk about it, right? We are doing something similar. We're privately circulating while it boost specs right now. We'll probably be publishing those for open comments around the Yth Denver time frame. And our approach is focused on exposing the intent layer.

And so one of the things that we do at Block Native is we work directly with many of the leading DAPs and many of the leading wallets, sort of as an infrastructure provider. So one of the many things that we do for the ecosystem is we provide one of the leading sort of wallet connection libraries called Web3 onboard. And if you've ever used Curve and the thing pops up that says, you know, connect your wallet, that's a block native library. Same thing, Zapper uses the same one.

And there's many, many sites out there that do that. Similarly, we provide real-time APIs directly to wallets for things like gas estimation, for transaction preview and transaction monitoring. Well, all of this infrastructure is becoming media aware. And so what we're focused on is as the user is constructing a transaction, but before they've signed it, there are mechanisms whereby information can be shared, and searchers can bid on the transactions while they're being composed. And one of the things that we think is super interesting about this is it doesn't put any extra burden on the end user to understand new stuff. And this value that they could receive can be compartmentalized into an

existing part of the UI of gas price. And so imagine conducting a transaction and your gas price starts to roll backwards as bids come in. And maybe even your transaction starts to get a negative gas price, a gas rebate. Right. And that rebate is just a mechanism of expressing, hey, searchers are willing to pay more than the price of gas in order to get access to your transaction, right? And there's a whole bunch of really interesting nuances to how the handshaking would happen. I don't leak information and things like that. But then ultimately what you could do is buying the transaction or the searcher, so everybody knows what they're going to get. And you can have a very smooth value capture, value recirculation experience that doesn't sort of pollute the user

experience in a way that now everybody participates. Okay. And again, these are research concepts that we're putting out there and that we want to develop. develop in an open and permissionless fashion so that anyone can participate both as a supplier of transactions as a transaction originator or as a servicer of those transactions and facilitating various parts of that. And the good news about all of this is it's a pretty small ecosystem. We all know each other. We all, you know, can debate the merits of each other's approaches. And, you know, the market can decide and maybe there's multiple solutions or maybe it gravitates towards one or the other. But anyways, this is sort of where we are in this, I like to say,

rapidly emerging category of block building. I really appreciate kind of the background around that. And I think it's important also to maybe pause here and talk about how this concept of recirculating value and at least having this debate around, you know, how do we, number one, like who should be able to capture this value and how do we allow them to capture this value. That is something that is very distinct from how things work in traditional markets. And it's necessary to have these conversations to build systems that are more fair and equitable and transparent compared to traditional markets. And, you know, I think earlier we talked about how MEV as a phenomenon is similar to

strategies that take place in traditional markets, but this is one thing. This is one area of research that is very distinct. Yeah. So M-AV is a fact of life, right? Ordered transaction systems have this, right? There's this sort of, there's certain research about how to get rid of it, but most would argue that it's, you can't get rid of it, just as economic force. And the question is what you do about it, right? It turns out the more you ignore it, the worse it gets, because then it just sort of accrues to privileged actors who have deep pockets, deep resources, special access. And so, like, in traditional financial markets, it becomes this latency game where, you know, who can co-locate the server rack next to the point of origin, who can have the minimum number

of hops, who can have the fastest fiber optic cable, which, again, it just favors the, you know, an elite class of actor in this category with specialized skills and generally to the detriment of all the participants in the network. The amazing thing about Web3, and again, you really do have to credit flashbots is pioneering this is the incredible level of transparency that has occurred around MEV in this ecosystem. So Web3 systems are ordered economic systems, ordered transaction systems, MEV is a fact of life. It started out as sort of backroom deals. It got brought into the light, sunlight is the best disinfectant. And it is getting progressively decentralized, progressively more programmable. Okay. And these are incredibly healthy.

things and incredibly beneficial for the long-term viability of this ecosystem. I believe we are at the very frontier of what's the right way to handle this? What is the ethical way to handle this? What is the correct? And my personal senses, there's many valid POVs here in frames of reference and that there's probably not going to be one ring to rule them all. There's not going to be one right or wrong way to do. There's not going to be one right or wrong approach.

There's going to be solutions at every layer of the stack. There's going to be solutions in the protocol, solutions in the application, solution in the wallet, solutions that end users use and solutions in the infrastructure, like the ones we're pioneering. There's going to be various providers who have various incentive schemes around how to do it. You're going to have a wallet that might rebate you 50% of the MEV. There might be another wallet that rebates 20% of the MEV. If this turns out to be decisive, then what you're going to have is users will abandon the wallet that only rebates 20% in favor of the wallet that favors 50% because there's economic

I'm a consent to do so. Well, if we look today, there are credit cards that pay you various degrees of cash back. And you can argue that everybody's just going to gravitate and hold a single credit card, but the market forces would suggest that's probably not the case. And so just like in other markets, there will be many, there are, I anticipate many valid approaches and pros and cons. And it's really, it's just a question of which tradeoffs are made. The current setup, the current equilibrium is probably not sustainable. There's enough going on right now that is centralizing. There's enough going on right now, which feels not as equitable as it could be.

That it strikes me, we can anticipate a fair degree of advancement and evolution in this category. But that ultimately that vector is pro user, pro transaction originator, pro recirculation, that that seems relatively irresistible as what will happen. But, you know, the market will determine how that goes. And, you know, we're lucky to be able to participate in that and be one of those players who proposes ideas and sees what happens. There's a lot more questions I could ask you, Matt. But I think maybe one thing that's top of mind

and is maybe a good way to wrap up the conversation or set us up for another conversation is this concept of programmable privacy and maybe how you can leverage privacy to reduce trust, the need to trust different actors in the MEV supply chain, whether that's trust between users and searchers, whether that's trust between searchers and builders. Where do we stand on that question today? And of some of the solutions that have been discussed, you know, SGX, MPC, homomorphic encryption, you know, what are your thoughts?

Well, so I think, as I like to say, all of this stuff is a journey, not a destination, right? That we're on our way. And so in general, more privacy is going to be better than less privacy, more encryption, less trust, is all things that are pretty core to the Web3 ideals. And we believe, you know, core to creating more equitable economic system moving forward. The challenge is how to get there. And as much as we wish there were magic bullet solutions that you sort of wiggle your nose, and now everything's fully private, we're pretty long way off from that.

I'm no deep expert here, but there are others who have said pretty clearly that the best solution, which is available in the near term, is probably SGX. That's from Intel. It is by no means perfect. There are many limitations to it,

but it's kind of the best thing we have right now. MPC or multi-party computation has been around for a while, but has a bunch of, downsides to it. So the tradeoffs, at least today, are probably not worth it. And I think there's a lot of optimism, to use an overuse term here, around fully homomorphic encryption, but we're not there yet, that it will be sometime before that is ready for prime time. And I just think it's really interesting, like right now at this very moment, we're seeing the consequences of research projects that get released into the wild, the public prematurely,

with these generalized AIs based on LLM's large language model. So you have OpenAI, ChatGBT, GBT, you have the Bing experiment, and all these headlines about these relatively concerning conversations that people are having with it. Meanwhile, I'll have Google saying, we haven't released ours because it's not ready for consumer-grade applications. Maybe Google knows what they're talking about. And so I think in many of these regards,

there's a lot of areas for research on encrypted mempools and others that's moving forward quite rapidly. And we just want to be pragmatic about releasing technology when it's ready to be released and not before then because premature things here can have pretty negative consequences for the long-term health of the network. But at the end of the day, the Ethereum ecosystem, which is what we're really focused on, is a big bet on technological progressivism. And technology means every day, every hour, every minute, it gets faster, it gets smaller, and it gets cheaper.

And so many of the challenges that we face today in the Ethereum ecosystem, the answer is just add time. Okay. This stuff is going to get easier to use. It's going to get faster. It's going to get cheaper. It's going to get smaller. Right. And though it may not be fully ready today, there is such a groundswell of intellectual energy being devoted to these problems.

that it feels pretty likely that we're going to solve them in the near-term horizon. By near-term, I mean three years or less. There are those who would say there's too much change, there's too much experimentation, but this is the Ethereum bet, is that through technological progressivism, through really pushing the boundaries of the possibility, you can create the foundations for the next economy. We very much subscribe to that. We're very much a part of that.

And we feel lucky to be able to even contribute to that in certain small areas. We also recognize that we inherit tremendous progress and tremendous innovation from pure organizations that we're really happy to be working with and inspired by. And so, you know, I think that's the grand experiment that is Ethereum. And that is the, you know, the bet is for all of its limitations today, it is vastly less limited than it was just a couple years ago. And in two years from now, we'll look back at how things are today and go, wow, that was primitive, right? The same thing was true of the internet. I was building core internet infrastructure back in 1994. It was laughable compared to what we know now.

Many of us had the experience of witnessing the launch of the first iPhone. That was 2007, 16 years ago. It used a processor from a DVD player. It was laughably slow. It communicated over 2G. It didn't even have copy and paste. And today we take for granted the supercomputers that we carry out in our pockets, right? What's the difference?

Time. And so, you know, in all of these areas, I'm quite up. optimistic that we will find viable pathways moving forward and and build more and more useful and interesting things on top of this ecosystem. And so it's fun to be a player at the infrastructure layer. It's also challenging. And we're excited to see what happens over the course of the coming year. That's excellent context and an excellent way to to wrap up our conversation. Matt, thank you so much. It's been such a pleasure to absorb the wealth of knowledge that you have on this topic. Really appreciate your time. My pleasure. Thanks for having me on again, Ria.

Happy to come back. You can find me on Twitter at Adam Cutler. You can find us at Block Native. And if you're going to be at Heath Denver, please do come up and say hello. I love to check.