S4 Premiere: A Journey Into The Dark Forest | Hasu

Transcript

all right everyone welcome back to another episode of bell curve before we jump in quick disclaimer the views expressed by my co-hosts today are their personal views and they do not represent the views of any organization with which the co-hosts are associated with uh nothing in the episode is construed or relied upon as Financial technical tax legal or other advice you know the deal now let's jump into the episode hey everyone before we get into it today just want to give a quick shout out to this season's sponsor Rook close to a billion dollars worth of Mev has been taken out of users pockets and that's just on ethereum and that number is only getting larger unfortunately Rook thinks that it's time for a change and they've

built a solution which is going to automatically redirect that Mev back to where it belongs into your the user's pocket so you're going to be hearing all about them later in the show I'm a huge fan of this team and what they're building so stay tuned to find out more welcome to the intro episode the kickoff episode to season four I'm very excited to be joined by haasu we're going to be taking a journey into the dark force and exploring Mev haasu welcome to the show man yeah it's good to be back uh to do yeah speaking about Mev and doing some podcasting I'm very excited to do this with you yeah me as well me as well um so let's just get into it and folks who have listened to previous seasons of

bell curve will know that the point of this episode is basically to lay out some of the ideas that we're going to be exploring in depth later in the season and you know I think it would be a very cool outcome if both of us you know we're going to lay some of these things out and we're going to kind of you know speak through and understand what our opinions are going into the season and I would love it if we changed our mind about a couple of things you know after we've discussed these topics with our guests I think that'd be a great outcome oh yeah um I I when we designed the guest list um there's a bunch of people on there who have have not talked to so far so there will be a lot of new for me in these conversations and frankly that's uh

that's a big reason also why I enjoy podcasting and why I enjoy um doing this season on Mev because it allows me to meet new people and get New Perspectives on on Mev which is just a fear that is constantly evolving yeah I agree so I want to start um you know maybe to introduce the first big topic that we're going to talk about um which is what does equitable distribution of Mev look like across the Mev value chain there's a lot to unpack there so maybe hosu could I Tee It Up to you and could you kind of break down this concept right that many people be familiar with but what the Mev value chain is who are the actors and then

what do we mean when we talk about what is equitable distribution of value look like across that chain most people know Mev as the money that can extract from reordering inserting censoring Etc the the list of transactions that they put into a block right so I would actually go and generalize that to say it is any money that or any value that a privileged actor can extract in a in any system as a result of their privileged position and so if you use that definition then

um all of a sudden you start to see Mev everywhere for example central banks you know going out of control printing money that leads to hyperinflations you see um you know that's that's Mev right Google cheating in the ad auction um that's Mev there's there's just so many forms of Mev everywhere and um my bull thesis for crypto is actually that people want crypto in order to Escape Mev in the real world right they want systems that are more fair

that are more decentralized where there are fewer privileged actors in them that can result that can extract less Mev and so um yeah I think um I like this kind of this Arc um that uh the smev and crypto and maybe outside of crypto and yeah so um I think we need to make sure that the that the systems that we built in in crypto RS resilient to mevs possible because otherwise store thesis doesn't work out hmm and I I think there's you know to borrow a a phrase that um Stefan right has has used is there's kind of this idea of Mev dystopia and

let's say that let's go into this season assuming that Mev is an inevitable outcome when you have different systems like this and to your point that could be a crypto economic Network that could be a central bank that could be Google's ad Network let's say there will always be privileged people in these sorts of systems and there's always going to be a way to extract value but there is probably systems that have structures that enable better more fair outcomes and then they're probably more black box opaque systems where maybe the outcomes are good maybe they're not so good so maybe let's even back up for a second there and when you're designing these systems right because flashbots in many ways is involved in designing these economic systems what are some of the

principles that you sort of use or think about oh that's such a great question what are the principles um I would say one of the one of the main principles in reducing Mev is the idea of competition because when you when you have competition for example uh in the Mev supply chain we try to have competition at as many um layers of the stack as possible when you have competition between uh different block builders in order to buy a block from the next validator then the result is that they start to compete on

price and they start to build up bit up what they are willing to pay and the value that makes more money if you go down one layer then if you have more competition between searchers you know for the same Mev opportunity they start to build up the value of that opportunity and give more money to to the Builder one layer before that let's say if you have more competition between wallets or applications to minimize the amount of any whether the user exposes then there will be less money lost and so I think competition competition to do better to be more Innovative and to find better Solutions that's one of the main principles

a another one would be privacy that's very big for asset flashbots because if a transaction is private then it's much harder to extract Mev from it because you don't know what the transaction is doing you can't simulate it you can't say Let me let me front run this transaction for example because why the transaction is is private it could do like a million different things and so whatever you do is purely speculative right um so what else I think giving control uh over their transactions to users I think that's a big one so a lot of a big

reason why we are talking about Mev is such a big topic is that users and wallets and applications for a very long period of time did not take control over their transactions they have the they have the power to do it right they they could for example use a DEX aggregator instead of using unit swap or they could use the wallet that that protects them from Mev that routes their transactions through a private channel to minus or they could change their RPC endpoint in metamask to to do the same thing right um there it's not like options are not available but this is a complicated

topic and yeah it's just very hard for people I think to see the differences and see um to to Really compare different options and um and no one even hit them but I think one of the peniculous sides of fmev is that it's often invisible to uh to the eye right you can look at your transaction but even if you look at your transaction in in uh in ETA scan you won't necessarily see that you've gotten sandwiched for example but because the the sandwich would be then in in different transactions that come before before and after your own transaction right and so it's um it's many things it's it's a lack of um awareness lack of

visibility a lack of sophistication and yeah so it's it's really a multi-pronged problem that we are slowly chipping away at all right so competition privacy redistributing web to users those are you know kind of philosophies and ideologies right so I want to kind of map that across the value chain as it exists today and my question to you is so if we kind of lay out right there's kind of users in wallets um then there's Searchers Builders relays validators and when you're you're designing how Mev should flow through that entire value chain those principles that you laid out I totally agree with my question is what do you do when the principles that you're trying to design

a system around bump up against the natural economics that organize that system and what I mean by that is there's every incentive for actors in that value chain to uh vertically integrate or create these little deals in between each other and consolidate so that it looks less and less like the design ideologies how do you deal with that bumping up uh yeah basically you have to include that in your design so you always have to assume I think in in distributed systems design you always have to assume that people act on their own best self-interest and nobody is um

yeah nobody does something for someone else right every Everybody Just does what's best for them and um and you need to understand what these strategies are these self-motivated strategies and then you need to design design the system that it's robust and it gives you the guarantees that you want even when people just do what's best for them hmm and so I think we can so we we could go into an example right so for example why I would say so some people say that they have different schools in Mev which I actually don't agree with and I'll tell you why in a second so there's this is a straw man but this

there's there's the school of uh democratizing access to Mev which which just says that um well we can't it's uh it's like it's very difficult to minimize Mev so let's make sure that the the access to the best block is available to all validators so validators can stay simple and decentralized and we we isolate the the centralization and complexity of block building into this new role in the block Builder right and then we just we try to make this role very competitive but we isolate it first and so this has already done a lot in order to curtail the the level of vertical integration that that you were talking about right

and then the other alleged School says we have to minimize Med instead of democratizing it right democratizing is the wrong path and so um I think these are not um competitive at all I think we should do both all the time right you have to minimize what you can and then you have to democratize whatever is left because if you don't then you will get centralization a divided layer which we don't want but back to the original point so there are many uh Mev designs that have been proposed in the past that are not incentive compatible for example if you just ask validators to run an

alternative uh client that orders the transactions in a way that doesn't maximize Revenue to that validator but that has some kind of that is somehow opinionated about what Mev can be extracted or can't right then um whoever validator will use that um yes they will have less Mev in their block and they will arguably cause cause less harm to users but using that client is not incentive compatible and um it just means that most valeras will not do it and um and those spell letters who who do extract Med they will over time outperform and crowd out the value

that as who uh who ignore the Meb right and so um this shows that it's it's not even so even if you're willing to sacrifice something uh in order to protect users from any vs available it's not even in it's not even it's not in your own best self-interest but it's not even in this in the best set interest of the system itself because all of the solutions must be coherent at the system level there's no individual player here who can solve this except for the user and so that's why we are so focused on educating the user to use solutions that that expose less Mev to begin with

yeah so let's let's talk about because an enormous amount of discussion thus far in Mev has been kind of dedicated to the validator level right in trying to enshrine systems where we don't have centralization at the and that's kind of the proposer Builder separation right which is on the roadmap for ethereum but there's kind of this new Force which I think is going to be emerging in the Meb value chain which is a little bit higher in the you know in the structure of the of the value chain which is kind of the user and wallet layer and that so we had seen you know previously kind of block space auctions happening at the the validator level and now there's talk about order flow auctions happening closer to the wallet and user layer can you talk a little bit about that Force

yeah I'd love to um so the idea behind order flow auctions is that users don't just send their transactions to the public mempool or even to directly to a block Builder because they by doing that by sending it it straight away they just they give away the Mev value of their transactions if if we assume that you have a transaction to make mic that has let's say 100 of Mev attached to it for example because your transaction takes an inefficient route through D5 it trades on instead of trading on all exchanges equally it only trades on one

exchange and so it creates an Arbitrage opportunity um then what the autoflow auction does is its auctions off the right to execute your transaction or interact with it in some way that allows someone else to plug the Mev that is attached to that transaction and if you combine that with the principle of competition then what you get is that you have different searches competing for the ride to pluck the Mev from your transaction and you get the same effect that happens when Searchers compete for the same Mev opportunity with the block Builder

what you get is they start to bid up the value of the opportunity they start to bid up the right to take your transaction but instead to instead of paying it to the block bill or the validator they pay to you as the user and that allows you to reclaim the Mev value of your own transactions so you know what is a little bit exciting to me when I think about this and we'll talk about the regulation is going to play a role in these systems that we're talking about we'll get to that later but you know the the analogy has widely been used when describing me via payment for order flow right and that's uh you know to use the way this works in traditional tradfy kind of markets is brokerages

sell their order flow to internalizers like a citadel Securities right and that's why Robin Hood has been able to offer zero trading fees to its largely retail user base which is kind of dumb dumb order flow is the is the it's the word that these people use and if you actually take that to its extreme right Citadel Securities did something like some incredible number like 18 billion dollars or something like that in in profit last year so what's kind of cool what you're talking about Mev redistribution to the user is the next logical step to that is actually to say I'm going to compensate you and pay you actually to trade so not only can you trade for nothing but we're gonna return some of that the fees that we got to you

which is an exciting concept to me I think it's very interesting yeah it's uh it is very interesting I I would say that I think on a common sense basis you can never get paid to trade and so it's a very tricky idea because you you are paying something otherwise they wouldn't give you that service right it's it's just that you're paying in an invisible way and I think so there's a lot of things um oh I think payment for order flow has a very bad reputation um I I think the concept itself is sound but um I think what I dislike the most about it is basically the in transparency of

uh of how much you pay right what what's actually the cost of um getting a particular trade executed and I think that's that's where we can do um a lot better in defy yeah you know something that's funny about payment Florida flow for me to because you're absolutely right right it but to me payment for order flow is almost a an example of fee preference so you know when they're when you have financial markets right people need to uh there's infrastructure that supports that and the people that create the infrastructure need to extract fees in some way so it used to be you know brokerages would charge you it was actually a very honest business model in a way but people didn't like it because

there's this very seen noticeable fee up front right in the form of it used to be much higher but even with Fidelity it used to be you know five dollars or something and now it's now at zero instead people have opted for these invisible fees that you're kind of alluding to and you know what Citadel does is when they get your order flow from Robin Hood a bunch of people buying GameStop they don't trade directly against you that would be illegal but they understand and this is statistical Arbitrage which is happening in crypto and I want to get your thoughts on is they understand that when people are buying GameStop AMC is about to move so the GameStop people don't get run over but in a sense the AMC people the people that are about to buy AMC are going to get it at a worse price so is that front

running I don't know that's a weird philosophical question right now I think the regulatory apparatus in the US says it isn't but you know it's an interesting question to pose it is a very interesting question um I wouldn't consider that I mean I'm not sure if I would consider that front running I mean in a sense you are using the information of one user to um impact another user but I I'm also on the other hand not sure if if there's such a thing as the right to buy AMC at a stale price um this is this reminds me of the The Flash Boys

um debate right where all of the um the old trading firms were complaining that um they were now no longer able kind of to to buy like still Market maker quotes on exchanges because they were starting to update so fast and so on yeah so I'm I'm not sure it's it's a very interesting topic for sure I agree all right I want to keep moving through some of the ideas that that you and I want to explore so one other big question that I have for you and this is you know I know that flashbots is working on this in the form of suave but even that kind of Mev value chain that we laid out there's actually kind of another dimension to it because ethereum has adopted a modular scaling roadmap

right so not now not only do we have um the main chain of ethereum but we have layer twos right in form of arbitrum or optimism or base and then the idea is we might have you know either app specific um kind of layer threes that sit on top of those those two so each one of those systems mevs being generated so one thing that I think we're pretty interested in exploring and I'd love to get your thoughts on now is how is Mev going to work in that world of uh modular chains built on top of one another so I think the first thing that so first of all I completely agree so um the roll-up Centric uh roadmap of ethereum and not even not even just that

I think it converges and you had this great season on kind of the modulus take in Cosmos I think that the world of Cosmos and the world of ethereum they are converging in the middle um I think and I mean this is not a this is not a new opinion but I think the ethereum um Community has been better at executing um and distributing and so I think my my money was was to be on um the ethereum ecosystem out competing the the cosmos ecosystem from here but nonetheless technologically speaking they are converging and I think um in it will be very hard to say in the future if a particular project will be

an ethereum project or a cosmos project just because of how the different layers can be mixed mixed and matched right so what does it mean for Mev so I think intuitively speaking the Mev accrues to the execution layer with the caveat that actually the Mev accrues to like technically speaking the Mev accrues to whoever gets to order the transactions um and so usually the execution layer has some control over the ordering so they they can decide to forego that they can opt into being what is called um a layer one sequence roll-up or a

based roll-up right um or opt into any kind of shared sequencing system and in that case the this alternate system that now sequences their transaction now accrues all of the Mev there might of course because the the the the execution layer has the ultimate right to decide who they want to give the right of sequencing who they want to Outsource that to so they still have a lot of economic power over there and can can hopefully negotiate some form of profit share so that would be my that would be my first uh comment on that uh Mev in the modular stack then

how is it going to look I think one interesting takeaway is all of these layer 2 sequences right now are centralized and all of these layer twos have road maps for decentralizing their sequencer but if you look at what the sequencer actually does it's a Combi it's basically the accumulation of what the block proposal in the block Builder does in ethereum layer one so it's almost like you're taking the you're turning back time right it's like we go back into pre uh pre-muv days on ethereum and so I think that shows you that this

um when when this layer 2 sequences actually complete their um decentralization roadmaps that that they have today their drop on Mev is far from done I think that they will all have to struggle with the the same problems that ethereum uh was struggling with um about uh basically there being very inefficient Med extraction um the uh the Mev introduces a very big skill differential into block building and because many because abstraction in the public mempool is very inefficient

you now have a strong incentive to vertically integrate between Searchers and and welderas and yeah all of these things will be true on layer 2 as well and so long story short I think what we need on all of these layer twos is a form of PBS meaning proposal beta separation meaning we need to um we need to abstract out the the role of the block Builder and yeah I think these are two points that I'd make what the modular stack and yeah let me know if you wanna dig in any Department yeah I have more questions um so some and some of this I know the rules are kind of being written in real time but you know I think it's a good

heuristic kind of rule of thumb that the Mev will accrue to the execution layer that makes intuitive sense to me I followed that here's my question to you right now and I'm going to ask this question within the context of a single sequencer world where with the assumption that we're going to move to a decentralized sequencer set over time but in the single sequencer world right let's use the optimism model for now so you basically get to bid to run the sequencer for a specific period of time and order the transactions and if the profit you can extract ordering that transactions in that period of time exceeds what you have to pay then you're in the money and everything's good but the difference in between you know being having the ability to order transactions

at the base validator level at ethereum is once you order those transactions the block gets you that's the settlement consensus layer but this layer is still one layer up so you could order those transactions but then how do they get sent down you know how do you make sure that the the way you've ordered the transactions will settle in that way so you can extract your profit ah okay so um a sequencer today uh in a layer 2 does four things basically so they um receive transactions from a user they um decide on some ordering of these transactions and then they give the user some receipt of where they have been

ordered so that's what we typically know as a pre-confirmation and then number four then they send the ordered batch of transactions onto the data availability layer God and that's what basically creates finality uh from the perspective of the layer two so from from the perspective of the layer one these um you basically still need a validity proof of this batch uh that it leads to a valid State transition um or you need some fraud proof window to pass that basically proves the same thing right so it's this is but this is about convincing the the layer one uh Bridge so from the perspective of the layer two once the um

once the the sequence has posted the data availability batch or the batch of transactions onto the data availability layer the transaction is is practically finer my next question for you here and this is the last on this topic and then I want to move on to some other topics is you know the amount of fees that validators can extract from from Mev and then there are other Revenue sources that's the security budget they're securing the ethereum network right so I see a couple of different competitive forces here that might impact the security budget of ethereum one you have the emergence of order flow auctions on main chain and then you also have these sequencers right which extract their own

Mev and those I see is am I correct in seeing that that's kind of uh taking away from the the revenue that validators can extract and should that be a cause for concern for the security budget for ethereum um yes so I think in the broadest sense if we if we could if we look at a um a a build is if you look kind of a sequence signed a builder as one of the in the same then yes any money that is extracted by any other participant in the supply chain that is not a a validator that lowers the security budget of ethereum but I would say that Mev should not really be counted towards the security

budget at least that is how the ethereum core developers think about it so I think their CME vs kind of an add-on but it's not something that they want to enshrine it's more something they it's definitely something they want to minimize over time and um and so Justin I think it was Justin Drake who a couple years ago and should introduce this idea of the minimum viable security budget or it may have been vitalik I'm not sure but the idea is that the the minimum the minimum necessary amount of security should always be paid from inflation um nowadays we it could also be paid

through the base fee so right now the base fee is basically set by the most recent degree of congestion in the network so the network was very congested and the base fee goes up and the network was not very congested it basically goes down but you you could change this this base fee based on other factors as well so for example if uh if the recent security budget was too low then you could increase the fee and so on right um it's just not guaranteed that this will lead to more demand to transact and so it's It's Tricky um but I don't think ethereum will have the problem or the need to do that um because if ethereum has a lot of holders and um I think that um

ethereum can pay its security budget through inflation um and yeah that should be enough it should not have to rely on Mev yeah and there is one topic that I know our first episode we're going to be talking at tarun chitra and Justin Drake and I know he wants to talk about Mev burn as well um so we'll we'll pick his brain on that hey guys quick break from the show here I want you to imagine something for me imagine swapping two stable coins on chain paying zero dollars in gas and instead getting a rebate of two thousand dollars this is something that's actually happened on chain to understand how I want to introduce and thank the season sponsor Rook zooming out for a second

the current state of affairs at Med is billions of dollars so far have been extracted from users Pockets using Mev Rook is coming in and saying enough is enough blockchain should drive value to their users and the applications they use it is time to leave the hobbyist era behind us if we want to move forward we want to get this right that's why Rook has built a custom blockchain settlement Network and it's one that gives you full control over the entire transaction life cycle today you can connect to an open source rough node The Rook protocol will automatically match bundle and auction your orders and transactions in seconds with zero gas overhead also any Mev that's discoverable along the way will be returned to you the user created as a collaboration between the industry's top mechanism designers and Mev Engineers

Rook was built from the ground up to be scalable safe and programmable you can get your own mempool choose searches and Builders and link your mempool with others to discover even more Mev you can Define how the Meb is shared and delivered as well and what could basically process anything from transactions to meta transactions and more this is the way that blockchains basically should have been from day one so if you're a user listening to this here's what I want you to do I want you to go to your wallets go to your favorite app your node provider and say Hey I want you to be working with these guys rock I want the Mev that I create to be redistributed back to me if you're a developer and you want to stay ahead of the game the best way to do that is to follow them on

Twitter they are at Rook or even better yet slide into their DMS they are lighting responsive they'll get you set up today and if you do slide into those DMS as always please tell them that I sent you I I want to move on to you know I think many people are familiar at sort of a high level with the idea of what a Searcher is and what a builder is um and you did actually a great episode on on Common Core a number of years ago I think called interview with a Searcher and we're going to be doing kind of a 2.0 version of that with a couple of Searchers in the season as well but I would love it if you could just give for people like help formalize like what does the search or landscape kind of look like today and maybe if you could

kind of touch on like Searchers that optimize for long tail Mev versus short tail Meb and if we could also touch on this idea of statistical Arbitrage kind of sex to Dex Arbitrage as well that'd be great yeah so I think that there's a couple of ways to slice that pie one of them is to look at the different forms of Mev that exist um the biggest one by far is arbitrage so Arbitrage meaning there's some trade opportunity and if somebody makes that trade then the opportunity disappears for example an asset is underpriced relative to some other asset or the same asset on a different exchange then

if you buy that asset and sell it at the same time then it kind of evens out in the middle right and um so Arbitrage I believe represents about 80 percent um of the Mev and I want to caveat that by saying that Mev in general is very very hard to track a lot of it is invisible a lot of is happening between ethereum and the centralized exchange or ethereum and some other exchange and a lot of it is not easily recognized as Mev basically and um that said so there's Arbitrage arbit charge is the biggest ticket no matter how you slice it then you have sandwich attacks um and that's when you uh that that's

when you basically buy an asset before someone else buys it and then you sell it right after them they buy it at the at the higher price and and you sell it back at higher price which makes you win something at the expense of the the sandwiched person and uh the last one um the the last one is is liquidation so sometimes lending markets have to liquidate collateral um of someone and then they just auction that off have some mechanism basically for selling that and it's typically Searchers who who do liquidate that and so this is one of the examples where um applications actually rely on searches in order to do an essential job for that application which is to recapitalize

um their balance sheet um so of these um I think that um so Savages have gone down over time a lot actually um that's for a couple reasons I think for one users became more sophisticated um there's better tooling so more users understand how to use Dex aggregator more users understand how to use um a private RPC endpoint like flashbots protect that allows you to send a transaction directly to a block Builder without it ever touching the public member but arguably the biggest is just a existence of Unison P3

which led to more concentrated liquidity around the current market price um and uh and that basically made it possible to set tighter slippage limits and and generally have have less price impact in pools and it made um sandwiching just a lot less um profitable Arbitrage is is an is an interesting one because uh a lot of Arbitrage used to be atomic and I think this is a theme that I wonder if the Searchers we talked to will confirm this but this is this has been on my Mev bingo card basically for a couple of years

um which is uh which is basically that Atomic transactions will disappear except for when the markets are going really crazy uh why then because then basically statistical Arbitrage players um turn off their models because they are afraid of getting blown up and so only the the most riskless trades basically get executed um and that's when kind of time time for for Atomic habitats to shine but so the reason why uh statistical Arbitrage and that means statistic laboratory means it's um it's a form of trading where you take balance sheet risk so you buy an

asset and and there's a small period of time or there's some period of time it may be very very short where you have to hold the asset before you can sell it so that is um more risky of course because the asset can go down while you hold it right and so it's not Atomic it's not riskless and um but the reason why this form of trading would dominate nonetheless is because uh if you don't have to do it atomically if you can do it statistically if you can internalize a trade you can generally do the same trade cheaper you can use less gas you can bid a higher price and you can ultimately win

uh the the right to execute that Arbitrage with the block Builder and in any case so any kind of trading like that will happen at the Block Builder level anyway so it's already today and this may be also interesting to mention um many of the top traders in D5 are also block builders so they don't even go the step of the search anymore and submit their their trades as a search or two different block Builders they may do that also because they want to maximize their inclusion guarantees so they want to maximize the chance of being included on chain when they have a profitable trade but they also want to be a block bidder

because that gives them much greater control over their trading strategy basically they can make the trade at the very last micro second before they have to submit a block to a relay they don't have to send it seconds earlier to a block Builder we still has to simulate it and validate it and whatever right they can they also they know the position of everybody else's transactions in the block they can cancel their transactions they can make new transactions they can update them they can shuffle them around it basically gives them much greater control over their over how they execute their trading strategies and so I think if I had to guess then um these are my three kind of observations about the current market

structure so Arbitrage is King uh sandwiches were disappear over time as kind of we get more better privacy Etc and and and uh and better dexes and then um statistical arbit charge will completely crowd out Atomic Arbitrage in any form of atomic trading and then block Builders and big trading firms I think this role will emerge and this is something that we've already seen yeah tons to dig into there um I agree with all of that and one thing that I I wanted to get at with you is this I I think that idea of the lag right this idea of warehousing risk even for a very tiny number of seconds is very key you know because before this

season and started doing a bunch of research you kind of heard a lot about this idea of cross domain Meb sounds very exciting right but you know when you kind of dig in that's not really existing today and I my understanding is part of the reason outside of sex to Dex Arbitrage my understanding of the reason is if you look at like the block time for ethereum is 12 seconds if you look at something like polygon that block time is two seconds so even if you had an Arbitrage opened up right let's say you are three seconds into an ethereum Block in one second into a polygon block you could lock in the polygon side of that Arbitrage but then you still have nine seconds right on the ethereum Block and who's to say

whether or not the validator executes on that second leg of the Arbitrage and the spreads are wider on both of those chains than something like a sex anyway so it's just not economical today to execute those arbitrages um although baby Suave fixes this in in some way um and you're going to school me half halfway through the season about why that might ultimately become possible but I think that's an important idea to concept to highlight and then the last thing that I wanted to get at I think you're getting another important point about latency and the importance of latency as well because these large sophisticated trading firms do not want to Warehouse risk for more you know

they're not even operating on a second basis they're operating on a micro second basis than they need to and you you talked about Flash Boys before a little bit and we know in tradpy the role of latency is extremely important and that's starting to rear its head again in crypto so with that is kind of the context could you dig into like how do you think about latency what is the role that it plays both in the view of the mempool but then also in the strategy of these large sophisticated builders yeah the role of latency that's a very interesting one so from the perspective of the Searcher and the block beta I'm going to treat them here it's kind of the same bucket so the more time you have the more you can

slow down time basically the more valuable the block that you can build for example if you can get more information about the word faster so if you can read the chain and any other quote-unquote chain so the external world you can read all of these prices and price movements and and whatnot and and everything that's going on the faster you can feed that into your model the more time you have to kind of compute and search that's where the name Searcher comes from the more you can search for optimal or like or profitable trades um that you can do against the real world right um

and uh so and then of course you want to you want to maximize the efficiency of your algorithms itself So the faster you can run them the more you can parallelize and so on um the faster you can compute and search and then you also have the lack of getting uh whatever trades you want to make to the block Builder or if you're a block beta then to the relay through the relay to the proposal and so on and you would like to do that as late as possible because that again that maximizes your time and it minimizes the risk of something going wrong and that also hints to um this going wrong also hints to kind of the value of cancellations so

you may submit your transaction but then the price moves rapidly and you kind of you say oh I don't actually want to make this trade anymore and so you have to cancel it and that happens all the time and so that shows you why many trading firms actually opt to become block Builders themselves because then any cancellation is instant because they're doing the competition and the block building locally they they don't have to go through around half the word to to some other block beta who has to well you know the the cancellation has to travel and they and then they have to execute it and then they have to send you back a receipt and whatnot read and so um yeah the more this has the models can be compressed of course the better for

you and you can see why Searchers and Builders and proposers all have a very strong incentive to lower the latency between themselves as low as possible because that just maximizes their Collective payout and that's a very that's very very dangerous basically because um if this trend is or if if this incentive is is sort of allowed to run wild then what you get is at the limit you get Searchers who are co-located with block Builders and you get Block Builders who are co-located with validators all in a very small geographical area in the case of treadfy

even within the same data center right inside um inside New Jersey or whatever right watch Chicago um and uh and and this might be okay in threadfy because threadfi is built on top of um local jurisdictions or local law law um and so on anyway but the ambition of crypto is that it's a system that is global and that is nutrient and when you have systems that rely on low latency where it's better to be if you have a lower latency towards other actors than to have a higher latency

and so you can never completely avoid this but where it just pays a lot to have a low latency to other actors then you will always erode um this geographical decentralization and this neutrality over time because you basically disincentivize being somewhere in the world that's not close to everybody else and this is something that I predict um would be become a big up topic uh in the next year of talking about Mev and as usual day and um so Easter is the one of the co-founders of flashbots who wrote the original flashpoints 2.0 Paper

that um that that first talked about Mev he recently put out a a post about latency and Geographic decentralization where he laid out this argument uh in much greater details so I would recommend everybody to to check that out because it's really important um what this means basically is that in the systems that we built we need to make sure that they are not sensitive to latency otherwise we would get this we would just recreate tread file we will recreate the same latency games and we will not we would basically fail in Building decentralized blockchain Systems instead we have to build systems that

are more tolerant to to higher latency and who to actors who who have a higher latency and don't punish these actors for being at a different part of the world um because otherwise you will get what you wish for and that would be not good for crypto hmm I've got one one question for you here before we move on to our next topic here which is I totally I I generally agree um I agree with you on that front but I think also you know this this kind of hits into it there's Fierce debate right across where you would call kind of like the Mev maximalization corner of crypto and then the fair ordering uh order of uh corner of crypto and there's no sort of Shangri-La you know when it comes to

this sort of thing so if you remove latency as a vector of competition for how Mev is extracted what is the other side of that coin if you had this steel man do you want me to Steven or do you want me to see him in the other side what I heard from that argument basically is like we we can't be up we can't allow uh latency to be the tradition the determining factor in how Mev gets extracted so what is the you know if you push down the latency competition Vector what is the other competition Vector that rises up so what competition Vector Rises up when you minimize latency is price and so because you turn you turn basically a latency auction into a into

a price auction and you force the parties to compete on on who pays the most in order to fill a certain opportunity and price is more compatible with geographical decentralization than latency and so that is one big reason um there are other reasons as well so auctions are generally more fair and more welfare maximizing um and they also uh with a maximizing meaning so the the party wants the item the most I'm more likely to get it and then they are also more efficient in the sense that they maximize revenue for the seller of block space and so in in in a market where the the value data

or the sequencer can pick their own ordering then they would always prefer to do an auction because that actually maximizes the revenue that they can get I would assume and this may be where we build a bridge to the modular stack again so I can totally understand the appeal of um a fair ordering protocols um they just feel more fair they are it's sort of very natural that as a human you you think well I was like my transaction should be ordered in the order that it was received right and um it's only at the limit that it starts to

break down um when Searchers start to compete and basically leads to this geographical decentralization but that's the same argument as why users love Robin Hood it's basically the good is very visible that the bad it's very invisible to them it's a kind of systemic effect that doesn't affect them personally it only affects the guarantees of the system as a whole but that said there are some nice properties that filtering protocols have that that options don't have and I would say the biggest one the biggest one is it's much easier to make a fair ordering protocol private than to make an auction private so if you have a list of encrypted

transactions then you can only reveal um the timestamp when the transaction was received and you can come to consensus on that in a decentralized network um and and then you can order these transactions and you can later decrypt them and then you have your ordering but it is a much harder problem to take the same list of encrypted transactions and compute the output of an efficient auction where you say the transaction that pays the most goes at the top but you also want to minimize any conflicts between them right because in in this type of system a lot of transactions were

basically conflict with each other right and so this is true in both cases but at the limit it's it is easier to to make a fail during Protocol private than it is to make an auction private and so that's my personal Steel Man of fair uh ordering protocols I think it'll be interesting to see because you know we've got the two maybe biggest brand kind of layer twos on ethereum have very different approaches to Mev between optimism and arbitrum so it'll be fun to see how that kind of plays out in practice um yeah in stock net I don't think has commented at all yet right on on what kind of uh ordering Preference they would prefer at least that I'm aware of

no no me either um and I mean this decision will be made I think in real time there's going to be I mean we're recording this on the day I think ZK Sync has gone live so congratulations but you know there's a polygon is launching their their ZK Network as well so I think this this decision will be made in real time we'll probably see the pros and cons play out and it'll be very interesting um one other uh so a lot of this season is going to be focused on Mev on ethereum but we also have a couple episodes dedicated we've got a Meb on Solana episode and an Mev on Cosmos episode uh now this is the cosmos one's a little near and dear to my heart we the whole

last season was pretty much focused on Cosmos and we did an MVB episode but you know I'd be curious you know as someone who's spent a lot of time like very close to the metal or kind of in the guts of ethereum you know consensus and how Meb Dynamics are playing out on ethereum I'd be curious you know what you're interested in or kind of hoping to get out of the Solana and Cosmos episodes yeah for Solana I think um if I'm not totally mistaken I think in many ways it it can be a case study for why you shouldn't build low latency blockchains um um so it's I I kind I see it almost more as as an anti-example in that sense so

um it I think we will talk about the centralizing effect of of low latency and what this does to a valid asset and what this does to vertical integration between um block Builders and uh and validate us um and it's maybe not a surprise I think that the block Builder said if I'm not totally mistaken at least one of two block builders on Solana so I saw um operating a liquid sticking protocol um because just by running um the extra validators and the Builder in the same place they just have a big and the big big advantage over others right then I think we will talk about uh well free market design which is I

mean that's tangentially related to Mev I guess but it shows that if you don't have a robust um mempool and fee Market design and um then you get uh basically a lot of spam a lot of failed transactions I saw an example uh the other day that 50 or more of Solana transactions actually failed Arbitrage transactions which is so insane if you think about it 58 yeah and then for Cosmos I think Cosmos and Mev is super fun I I listened to all of the episodes that that you did on Mev and Cosmos I know most of the people I'm an investor in um osmosis uh because I think they are doing something very

interesting um I know the guys from skip very well I think their approach to on-chain block building and onshine searching I think that's something new that we haven't seen before and then um I think there's a Noma and penumbra to uh two projects that are at least kind of intellectually quite close to suave in the sense that they have this so they both have privacy as a big part of their project and they have this they want to move users from making regular executable transactions to having kind of intent based uh transaction framework and this is something that we also think

a lot about and that I am personally very bullish on and so I'm I am really looking forward to picking these guys minds and yeah diving into some of the unique Mev ideas they have over in Cosmos me as well yeah it you know it was funny to to hear miles laid out in that season uh an articulation of the perfect Mev solution with two of which are gonna seem they apply perfectly to ethereum but one is different and one is the three prongs of a perfect mbb solution from his perspective was one avoid centralization at the validator level check for ethereum uh minimizes harmful forms of Meb like sandwiches I think check for ethereum but then the third prong is this is a little bit uh osmosis specific which is you have a way to

return Mev to the protocol so osmosis has a proto-ref module right which is a pretty exciting experiment I think there is a difference in ethereum which is uh returning of Mev to the user um so it's there's you know the cool thing about Cosmos that I love doing that season is you know it's very each they're very different opinions there and they kind of think about the world in a different way and like osmosis has a very opinionated approach to Emmy fee and because they're their own Sovereign chain they can do that and it doesn't necessarily perfectly align with the rest of the cosmos ecosystem which is just cool it's a diversity of ideas and opinions yeah and I'm I think something that we should dig into when we talk to them is

how much of this how how much of the opinionated nature can they retain when they actually start Outsourcing um their security needs right because the big drawback in uh in Cosmos right now is all of these chains have their own Val data sets and um so they have this idea of mesh security and they have this idea of uh The Interchange security that you can you can Outsource your sequencing to to the cosmos Hub and as that catches on how can they keep their own uh ordering and mempool policies while also getting the security from another blockchain I think is there a compromise there that would be very interesting for me to know I agree I agree

um so the last topic that I want to get your opinion on and frankly this is one that we might not have an answer to but is is this is the idea of Regulation so right now kind of the big theme in crypto is regulation finally colliding with our sector in a big way especially within the context of the United States right the SEC has taken a very sort of opinionated stance and approach and there's probably it's only a matter of time bless you before that comes to Mev as well and you know it has struck me that sometimes we talk about Mev as this like hey we are as a community need to make a decision about the way that this should

work but the United States and our regulatory apparatus has its own opinion right about how value you know should be trans uh you know transmitted in financial systems and a very specific definition of what best execution means for instance and you know to put my own opinion forth here I think a situation where users are front run or sandwiched but some amount of Mev is passed back to them would not fit that you know their specific definition of best execution today so let me just pose that to you it's first of all do you agree with that statement and second of all what how do you think regulation kind of bumps into Mev

I think that's a great question and we have to caveat this I mean it's always speculation right yes um I think I think if you zoom out far enough then U.S financial regulation and crypto or at least the consumer protection part of it is actually very very aligned because um what crypto at the end of the day is all about is creating more fair and Equitable markets for users where there are fewer Central parties who can exploit them in some shape or form and you sometimes have to make compromises in the short to medium term saw our systems today they fall short of that and not just because the system is

poorly designed it's more so that the applications on top of it are poorly designed and this in turn may be kind of this may be a result of like lack of privacy lack of throughput Etc right um but uh I think the Silver Lining is that you know the execution that users can get on public blockchains is getting better and better every year so if you if you just compare um kind of the execution quality of users I would love to see that charted on a graph and I think it's it's going up you know parabolically and I think it'll continue to go up as we learn to build better exchanges and as we learn to make the the base layer more imv resistant Etc

nonetheless so I would also agree that yes sort of the crypto does not currently fit into any Frameworks of uh what best execution means in traditional Finance I think that's okay I think that um definitions can change and I also think Regulators are pragmatic and use a kind of I think they apply their Force where it's the most necessary except if they get some kind of of course orders from above that are more politically or of kind of National Security interest um and the the third point would be that what we need to build our systems in a way that

they are just not subject to the regulation of a single party because of a single regulatory regime can make the rules in crypto then crypto just as a concept has just failed and that's why I'm so anti-latency minimization and so on as well because this just creates the conditions for crypto to centralize in a single geographical regime and this is something that we have to avoid we need to make it so that that crypto has power sent us in the US in Europe in Africa in Asia in China and so on and I think that's when crypto research reaches its

its maximum resilience and the and not single regulator can come along and kind of enforce its own preferences or utility function on the system as a whole they maybe they can say well as a U.S validator here's what you can and can do right but this will only have a small effect on the system as a whole because the other value that I might still do what they want right and so I think in order to build a system that is truly robust then you you just need participants from many jurisdictions and you need a way for their own utility functions to kind of make it into this hot pot of utility functions that that

creates the the total guarantees of the system and I think you know I think the fair caveat to say a sum of the subject matter that we cover in this season is going to be discussing what's currently happen happening but some of it is necessarily going to be speculating a little bit about how things are going to play out in the future um and so we'll do our best to caveat when that's the case but um I think we covered all the major topics that we're going to be getting into I think this is going to be especially if you're a nerd around Mev or how blockchain systems work this is going to be a wildly fun season so yeah I'm really looking forward to to dig into it and we've got a great first episode lined up with Justin and and tarun so

we'll we'll dig into it more then yeah so it's great I look forward to that episode as well all right my friend I'll talk to you soon foreign [Music]