A Deep Dive into Batch Auctions | Season 6 Episode 5

Transcript

all right everyone welcome back to another episode of bell curve before we jump in quick disclaimer the views expressed by my co-host 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 this isn't so much a choice that we have to Direct Value and somewhat like play God with the value flows in the industry but rather if they do not receive this value back they will never be profitable they will never be sustainable and we will default back to an extremely centralized set of actors

that offer uh price Discovery through an opaque RFQ system all right everyone we will be back to the program in just a moment but before we do I want to share something that blockor has been cooking up for these last couple of months March of this coming year in London blockworks is hosting D London the largest institutionally focused conference in all of crypto Goldman JP Morgan 72 all in one room so you can know what the big money is doing so click the link at the bottom of this episode it'll take you right over to the homepage and use Bell 20 for 20% off I will see you in Sunny London Town in March this episode is brought to you by Maverick protocol a suite of liquidity tools built around an Innovative amm Maverick helps token projects St treasuries LPS or basically anyone in defi shape their liquidity

with efficiency and flexibility how you might ask stick around and you're going to be hearing about them more later now on with the show hey everyone welcome back to another episode of bell curve very pumped for this episode today Dan and I are going to be interviewing Anna from cow Swap and Ludwig from Cella Labs we're going to be talking about batch auctions as a mechanism which is obviously part of a a build-on from last episode where we described the whole family of orderflow auctions we zero in on batch auctions in this episode and explore how they can mitigate some of that Meb leakage that we've been talking about this entire season really pumped for this this one there were a lot of gems so hope you all enjoy all right everyone welcome back to another episode of bell curve I am joined uh by my

co-host Dan Robinson and today we've got Anna of cwal and lwig from Cella Labs team welcome to the welcome to the program thank you so much for having us thank you for having us yeah uh this is going to be uh a ton of fun um so I I before we sort of get into this episode where we're going to be talking about uh batch auctions and sort of uh relegating me from the Bas layer protocol to applications which I know both of you have a a lot of thoughts about you just kind of give us a quick 30 second uh intro to yourself and the protocols that you're working on hey I'm ludig so working at Ser Labs we're building angstrom it's our kind of response to the problem of um lvr loss versus rebalancing the fact that LPS are not

really sustainable today so we're trying to make them positive in expectation cool and I'm Anna and um I joined the blockchain space in 2017 back then I joined gnosis so as part of gnosis I early on started already looking into Dex designs and specifically into batch auctions and then after working on a few different Dex designs that were only seei successful we finally landed on what is now cow Swap and yeah cow swap is a decentralized exchange that's based on frequent batch auctions um it's an intent based system we launched in May 2021 so I believe we were one of the earliest applications that applied intense in such a wide range and um on

Cops yeah user assigned transactions where um they give price of the or like Des sign the relevant information information such as the buy token the sell token the expiration time and the limit price and um then the orders gets collected in an order book and is then picked up by a competition of decentralized parties that we refer to as the solvers and they compete for the order Flow by maximizing the value that is given back to the users that yeah basically in short is what car is do it that's uh that's super helpful and I think where we want to spend the bulk of our time for for this interview is actually talking about uh solving me at the application layer and uh and I I know one of the themes that Dan and I

Dan and I have been uh poking at this entire season is sort of this win-win situation for both swappers and LPS that we can solve by uh mitigating or stemming some of the me that leaks out from the application so even just talking about designing uh me aware sort of mechanisms and preventing that leakage is I think where we want to spend a lot of time but just for listeners the the last episode that that Dan and I did was on uh orderflow auctions in that whole design space and bat batch auctions are obviously a part of that part of that family so before we get too far into the weeds could one of you uh maybe Anna could I just pick on you could you just describe describe what a an FBA is and maybe some of the uh Primitives that come along with it

for sure and U maybe I can even give a bit of a broader context that might be interesting and if I get too long just interrupt me that's great so freak can B batch auctions is essentially concept that has already been studied within trade five for quite some time and um it's essentially offering a solution that um is providing a um solution yet to a problem that exists in one of the more widely known and adopted financial markets um that are using continuous limit order books and the issue behind limit order books is that they treat time as continuous so in other words there's usually a seral

processing of orders and this allows in a way riskless Arbitrage profits even in an environment where information is symmetric so even if everyone has exactly the same information by ordering them sequentially you have the the incentive to be faster than everyone else in order to extract value and to illustrate this better there's this famous example of how at one point um there was a cable build between the uh Chicago CH Stock Exchange and the New York Stock Exchange just in order to get information milliseconds faster so not even like the blink of an eye significantly faster and why was this build even though it cost millions of

dollar is because it was worthwhile because having access to the information just a little bit faster enabled one party to do front running like basically having this information faster and then take profit from this by front running other Traders and these are basically riskless Arbitrage profits and they are not supposed to exist in a well functioning Market they harm liquidity and also they include um never ending arms race for for Speed and um a market design solution to this problem is essentially if um the the the batch auctions it's if you put time into units or discrete time and um basically process requests to

trade in a batch by using those auctions and um this would have two benefits right one is that you have discre time which reduces the economic relevance of these tiny speed advantages that you have and then secondly it also um by by applying this auction it changes the nature of the of the competition because now instead in of competing against speed you're now competing for price and um so this is essentially like the problem that was existing in trade for is existing in trade for which is known as highfrequency trading hft and I think a very similar problem exists on blockchain today with me and um I would argue that the problem on blockchain is

even magnitudes higher because since you have the public M Pool where all transactions basically linger for up to 50 seconds until the next block is validated um you have advanced players that can now take advantage of like normal retail players and um most people obviously who listen to this are very familiar with MV so um by reordering the transactions and by simply paying higher gas to front run transactions one can essentially um extract significant financial gains and I would say for like there's many cases where single front running or sandwiching attack uh gains profits of a few thousand dollars and um on ethereum and this is where I think the argument

gets very interesting is on etherum you already do not have continuous time right you have block time so every 12 to 15 seconds and your blog is validated so why would you apply this continuous time model in a non-continuous environment and um therefore to me it seems like very straightforward or intuitive that you would basically apply B auctions to blockchain um which mitigates these negative externalities of me so essentially um I'm sure we dive a little bit deeper into this later but um M arises because if you have this sequential execution of transactions right you have like um different prices

for one asset in one block and then they are arbitrarily ordered or like someone else is purposefully ordering them in a specific way to maximally extract slippage and in a batch auction essentially what you do you batch all those different assets like all the assets you ensure that they have that are cleared with one single unique clearing price and by doing this you have not this slippage tolerance anymore that you can extract based on reordering the transactions in a certain way um and yeah if you look into this data specifically I think there's like many blocks where we have seen that a single asset such as something highly liquid as usdc had already a price

spread of usdc East pair had a price spread of 1% in a single block and that's of course not very efficient and so batch auctions removed this this inefficiency thanks Anna um yeah I think that that's a great description of batch auctions I'm I think a lot of us are also a fan of that um of the British paper on batch auctions that I think you're referring to there on continuous versus discreet time um one theme we've talked about on this show is the difference between Mev um and Dex Mev uh so one is emitted by swappers with transactions in the me poool and that can be captured as slippage or or by sandwich attacks by by someone and I can see how a batch auction an offchain batch auction like cow swap can can

prevent that or mitigate that by executing all trades in a batch at the at the same price and so you actually you can't sandwich somebody you you have basically have to trade at the same price as them what about the other kind of Mev that's emitted by by passive liquidity providers on chain and that's that's often referred to as as lever so thinking about I guess for example liquidations um what how does a liquidation occur like maybe somebody sees that um the price orle is changing and wants to like directly back run this transaction um and then take advantage of it and if you look at very of like some of the popular liquidation protocols such as AA compound dydx um they usually um come with a fixed discount um so essentially let's

say that um at a 5% discount um your collateral is being sold and now if you would plug this into a bch auction the order flow um before it gets publicly liquidated you would say Okay immediately once um the price point is hit like basically you are under collateralized immediately um the volume is being sold via a batch auction then you without providing this 5% price discount you could save within the first x amount of minutes maybe you give access to a private order flow for auction that is maximizing for giving users the value rather than just saying oh the cut is

taken pry fall at 5% I was initially more so thinking about the problem of uh stat ARB and specifically sex Dex um so it's the idea that there's some exogenous uh price disc Discovery mechanism that is uh quicker um and that has deeper books where the bulk of the volume actually occurs and here I'm specifically referring to binance um where in effect compared to the MV that originates from users where some user doesn't um communicate a strict slippage tolerance and then there's some direct ordering and extraction through like a sandwich attack what you actually have as a as a worst problem is the

extraction of the stale quotes of LPS and it really occurs and you can think about it fundamentally from the principle of VPS lacking that information and lacking the ability to um change their stale quotes so binance will have a price change and um the LPS don't have time to react and given this they sophisticated actors specifically vertically vertically integrated builders that are going to see that these uh LPS are offering the wrong price and are going to bid up in the uh block auction to be able to be the first to touch the pool the first to effectively extract that value from those LPS and buy at this wrong price and then execute the second leg of the

Arbitrage on the offchain venue so how would you mitigate that so how do you avoid that value being captured by by proposers well I think fundamentally speaking this value is already being recaptured very efficiently today it's just not by LPS we're effectively paying the wrong party and what I mean by this is that if we look at the um proposer bill Builder separation infrastructure today um you have a builder uh that effectively bids in the block auction to the proposer and in a competitive Builder setting U they're bidding most of that value back to the proposer if we're able to shift this and remove that proposers uh monopolistic ability to

decide who gets to execute the Arbitrage you can then have an auction that you run specifically to pay the lp the most to get the right to execute that Arbitrage because at the end of the day they're the ones that are effectively incurring this loss they're the ones that are suffering that that economic loss so this at least what what we're um kind of building with angstrom is this idea of bringing back this auction and effectively directing the value of the auction directly to the party that is losing the money directly to the lp so how you do this um simply put is that you run a very similar ocean to the uh Block Ocean but instead it is purely for the right to to execute the Arbitrage

against the lp and so the lp gets the bulk of that lvr back yeah I I think that's a I I I want to dig into the the construction that you have on that on on angstrom lwig because I think it's really interesting but I do want to I think it's an important point to underscore for the audience because when people tend to talk about Mev they talk about it as almost uh springing from ethereum uh because that's where the most the largest and sort of robust most robust ecosystem for for meev exists today but I think you could make the argument that lot of me actually originates from the applications themselves um and based on the mechanism design of those applications so I would be curious like maybe just an open question uh to the both of you before we get into the weeds

of it you know how do you think about from a high level designing mechanisms that are me aware uh right with the with the ultimate goal being to redirect that me that's getting extracted mostly by proposers at the at the Block the block space auction today and redirecting that back to LPS kind of high level question like how do you think about when you're designing these mechanisms about keeping me in mind generally speaking I'd say you have to be cognizant that at any point in time your application might provide a state where it is purely profitable to execute on it because of some exogenous um kind of movement may that be in price or in some information that your application does not have because smart contracts they're not that

smart really they exist in isolation on the blockchain so without providing them information they will just follow the uh rule set that is effectively written in the smart contract so how I think about it is a being able to provide that information through some mechanism and B removing the proposer control over that application removing the builders and proposers ability to extract value without that application effectively signing off on it because if you have some very smart mechanism but at the end of the day the proposer can sequence a transaction that executes against your state and extracts that value you're going to have to pay them to effectively do it in a way that kind of recaptures that value and that rent is being paid

but that is an unnecessary rent to pay so the most important thing that I'd say is you have to design the application whereby you are cognizant of the fact that some meev opportunity is going to be created but when it is created the application itself has to effectively give the right to someone you have to be in control of who actually gets to extract that value yeah and I think dudwick touched on a very important point which is like as soon as you basically have some form of M um surface usually in the current design space most of the times it's just being totally like at the advantage of a validator right like ultimately you follow the me chain and most of it ends up going for

the to the validator and um so from that perspective I mean I agree that it's really difficult to design in a way that no Meb is exposed in the first place and you can argue that even in a batch auction to some extent that there's value that solvers are like dealing around with but then I think it becomes important that the way you design the competition is in the way that like the the the way you set up the rules of the competition in my view should be set up such that they benefit the user and of course this is a philosophical question and not everybody might have the same opinion on this of like where should the value of M go to but me personally and and working for cup it's very clear that

my opinion is um a I think it is creating inefficiencies the more me you're exposing because ultimately if you expose it you need to somehow redistribute it and that also means that you are giving up block space right at least you need to incorporate an additional transaction of where this m is later on given to so it is making the market less efficient and then secondly from a more like pH philosophical standpoint I would argue that if the user is creating the me opportunity in the first place by have by being basically vulnerable to external people extracting value from them then my view is that that me actually belongs to them

because they exposed it in the first place and it's taken away from them and by creating a market framework where you say okay maybe there is some exposure possible um to create the rules of the competition in a way where you say okay only those people gain the order flow who actually give the value back to where it belongs and I would argue to the user I then in the competition this is a way of how you can design an application in a way that yeah is more user friendly yeah I i' some somewhat add on to the the idea that a philosophical question I think it's a more so an existential question for blockchain in the sense that this value from me is

most of the time extractive someone is losing and we see this with LPS today this isn't so much a choice that we have to Direct Value and somewhat like play God with the value flows in the industry but rather if they do not receive this value back they will never be profitable they will never be sustainable and we will default back to an extremely centralized set of actors that offer uh price Discovery through an opaque RFQ system and I don't think that's a desirable thing we have to make the decentralized applications sustainable by redistributing that underlying value because otherwise the participants that are providing liquidity that are providing this extremely necessary

service will like cease to provide liquidity because they just won't have any Capital left because they're going to be continuously arbitraged and continuously at a loss L you mentioned some of the professional participants in the in the system and I think in um and and as well as the centralized parties that might actually take it over if other if other uh if we don't design better mechanisms um there are a lot of these kind of professionalized participants in in ethereum today um in I know in cow swap has solvers and angst some of angst's design there are there's a role for professional market makers there all right everyone we will be back to the program in just a moment but before we do I want to share something that blockers has been cooking up for these last couple of months March of

this coming year in London blockworks is hosting Das London the largest institutionally focused conference in all of crypto we are gathering 1200 of the world's largest asset manager so think trafy macro funds Crypton native funds big allocators and financial institutions so Banks payment processors Etc all in one spot it's very rare to get the likes of Goldman JP Morgan 72 whatever all in one room so you can know what the big money is doing we're diving into the themes that they care about so we're talking about the intersection of macro and crypto where we are in the cycle real world assets so everything from stable coins to unchain treasuries to tokenized assets it's going to be a blast but the other reason you really want to go it's London baby center of the world at one point you got Pub culture you got fish and chips great

beer it's going to be a blast so because you're such great listeners to bell curve there's a Code Bell 20 that's going to get you 20% off so click the link at the bottom of this episode it'll take you right over to the homepage you'll see all of our speakers and use Bell 20 for 20% off ticket prices are going up soon make sure you go use that code I will see you in sunny town in March I'm curious I guess Anna um how do you see the role of the solvers in uh in cow swap um and who who are those those solvers and what what powers do they have who are the solvers um I think it's actually in a variety um of different actors um we do have a few individuals who initially I think just um nerd

sniped to getting into the competition then um after time I think our competition got a bit more sophisticated by more market makers entering the competition um who of course um by adding their own private liquidity have a structural Advantage um and then we also have protocols um such as urine who essentially just want to facilitate specific use cases that are maybe not in the interest of the broad variety of servers because those trades happen not frequently enough and they're not providing enough profit opportunities but are still um for protocols specifically interested to interesting to be um executed on it's interesting

because of course um we I think all protocols that have these external whether they're called solvers fillers or resolvers are dependent on them because you ultimately want to ensure that you provide a good experience to your users meaning that there is competition I the less competition there is the more risky it is that actually at one point um there is no profitable execution or no no good execution for the user the risk is somewhat limited because of course the user asss their order with the limit price so the limit price is guaranteed per smart contract guarantee um but the slippage tolerance is something that could theoretically be

at risk if let's say there is suddenly no server competition only one server executes there has they have no incentive to provide a better price than anyone else because they're only the only one competing on this and then theoretically they would be able to extract this slipage so from that perspective I think all protocols that are working on this essentialized competition um yeah really require a good set of actors that are competing with each other but I think in a way you can even compare that to to blog Builders and Searcher environment where essentially you in a similar way are relying on a strong decentralized competition so maybe just to double down

on this question a little bit because I feel like one of the the common components of these more Mev aware designs you're relying on this sort of third party network of everyone's got a different name right if solvers or fillers or Searchers or whatever and I feel like maybe intentionally or not the sort of even when I went into doing our me season and this season and the sort of ideas that Searchers are these little like atome oneperson shops you know writing code and stuff like that and I feel like more and more that is not the case and it's a smaller group of entities that you might think that are now what people refer to as like integrated Searcher Builders which are basically hedge funds and my sort of understanding if I had to really guess at it would be that Mo it's mostly the

same entities that are either fillers on Unis swap backs or or solvers on cow swap or or whatever like can you can you like any any ularity that you can give for listeners about like how similar are these entities across all of these different platforms is it mostly these sort of hedge fund entities and then how do you see the future of that market shaping out because if this is the direction that we're going in terms of app design I feel like it's important to have a a competitive market you know I don't know if it needs to be thousands of independent Searchers but like how do you see the F future of this sort of uh Market of sophisticated entities in terms of the people that I've seen on the solver side on the Builder side and on the Searcher side um solvers are like

as Anna mentioned quite a diverse group of people at least they were in the beginning you have these small teams from these very random backgrounds that were working in the beginning and a lot of them are now no longer able to operate and so now you only really just have the best of that initial diverse group and then the hedge funds that kind of opted in um but realistically it all converges towards statar it all converges towards uh market makers that are just more efficient and so I I think I'm somewhat frustrated at at this idea that we have all of these different roles like as you kind of mentioned they're all converging towards the same person because a builder um is a better solver and a builder is also a better provider of liquidity on any um

underlying liquidity source to Arbitrage and amm so they're also a better Searcher um and if we don't design applications where there's also this sense of we can operate kind of collaboratively there is um some force in us all optimizing for a problem and not having a single winner that that takes all you're always going to have that and so without um kind of applications that design that um with that in mind the competition will die down to like a select few um hedge funds and entities that are running like very sophisticated Builder infrastructure and are able to build on like block end on latest State because they're aware of the block template so they're cognizant of all of the state and so even for like

a simple routing uh problem for a solver for example they're better equipped so should we try to prevent this from happening or or should we just lean into it or embrace it we should aim to maintain extremely permissionless competition so anybody that wants to opt into the market can operate um in terms of limiting that like somewhat force of centralization it's extremely difficult because these um these entities have economics of scale they have greater efficiency um so it's very hard to design applications with with that in mind especially when you have privacy that enters because then there's an advantage for private order flow so I think we should push against it um but we should do so not with not by like

forcing it and introducing unnecessary efficiency but just simply coming up with better uh application design where that privacy and that like inherent hyper competitiveness where you have a win or takes all isn't isn't like an existential property of the system if you you can have some winner of a very sophisticated actor but still have others that contribute to the solution and that are effectively pushing the solution to a more efficient end result I see that as like a positive outcome L I want to get back to U what we were talking about before about the two different kinds of Mev um in decentralized exchange with the dominant kinds uh one that's suffered by Traders uh by having their their trade sandwiched or suffering slippage and one

that's suffered by passive liquidity providers so um my understanding Ang uh tries to solve this problem by separating the auction into into two pieces um do you want to talk about how that works and get more into the details of how angstrom solves this yeah absolutely so angstrom at its core is looking at the current kind of supply chain and saying well we've already found like a really efficient way of recapturing this value so let's reimplement that at the lp level and how you do this is you have to a remove um kind of random execution on state where anyone one can execute you have to gate it and we do so by gating it behind uh effectively verification of consensus so you have um this network that is

dedicated to our application which we call the guard Network which is run um and can be run by anyone that effectively provides stake and you're effectively peering around all of these transaction and orders and then structuring them within a bundle within this single transaction that the proposer or the Builder can't effectively break up so you're removing their power to sequence and Order transactions and you can break up the bundle into two components the top of bundle uh where the underlying Arbitrage is executed and like the rest of the bundle where user orders are effectively batched at a uniform um clearing price and how that works and why this is actually useful is that the top of bundle is the arbitrager that is

incentivized to effectively Arbitrage back to True price but they're competing against all of these other Searchers to effectively bid and and provide the biggest uh kind of Kickback to the lp so only the um Searcher that effectively contributes the most value back to the underlying LP gets to execute that Arbitrage after this for the rest of the bundle you have uh a frequent batch auction that provides this uniform uh point of clearing price and how you arrive to that is you look at all of the user orders and conceive of them as limit orders and you can construct and find the point of intersection between the supply and the demand so the supply is in u the asks so sell orders and uh the demand is in the buys and as you

find that point you can effectively arrive to that single fair price but you add in the constraint whereby LPS are effectively only providing liquidity at that final price and what that means is that they're no longer offering that stale quote because all of these market makers and user orders are effectively providing that information whereby the lp is now cogn of the fair price where users are users and market makers are willing to trade at and they only offer liquidity at that point now uh so so both angstrom and the Des for angstrom and cow swap have batch auctions in them and batch auctions limit in some ways what you can do with an individual trade I'm I'm curious Anna How cow swap thinks

about the composability and whether um forcing pushing users into this into this bash auction model limits what they're able to do with their with their um assets on trade on on on on chain or whether you have whether you have new solutions to to get around that do you have an example in what way it limits that because for example cowo also um launched pre- and post order interaction right that we refer to as cow hook which also makes it now very composable in the sense that you're now able to come up with any pre-order interaction pre Swap and any post swap that you want to do whether it's like um disolve my um my whatever maker CDP now I want to swap it

and tab into Aro for example like all those things or I want to swap and then trade cross chain like all those are possible I don't know if that's what you were hinting at yeah that's that's I'm asking about I think about programmatic orders or or is that is that that's what you refer to so how how does that work um so it's two separate things um one is essentially just allowing that you add any type of um on interaction before and after normal Swap this is what we refer to as cow Hooks and then with programmatic orders this is actually something that we literally just went live with this week and it's actually pretty cool it's um take it's leveraging ESC 1271 so it's now allowing um any smart contract based system to

essentially place any type of programable order in the future Um this can be used to generate multiple discrete orders um basically endless even if you want to create a 100 different t- orders for example you can do it with a single signature um if you want to update those by for example saying I want to change the date I don't want this to be executed every day but instead every week or I want to change the slippage you can adjust this for all 100 orders within a single signature and um yeah you can assess a proposed uh order against a set of different conditions so you can really come up with anything you can come up with stop loss with DCA orders or to come back to the to the topic of today

of essentially like the issue of liquidations think about um you have um your assets uh wherever and then you want to protect it because you you're um getting close to being liquidated um and you're thinking okay I actually don't want to continuously monitor monitor my position but instead I want to automatize an order that says um once I'm closely to be liquidated I actually want to buy more of my backed Assets in order to secure my position you can automatize this entirely with cop today or you can even think the opposite if for example um you have your backing in Isa Isa is uh increasing in price as it is right now so essentially you would be

able to loan more against the East that you have already deposited you can automatize this well and automatically via house SW um yeah take take more loans out automatically without individually having to monitor your portfolio I've got a question how long does it take about just TimeWise to run a like a batch auction for either cow swap or angstrom um for us we so we do it every block but there are optimizations that can be done in terms of gas efficiency because for us like we're able effectively to compress down all like user interactions within unisoft to a single actual Unis swap trade which makes it far more efficient so if there are very few orders you can

effectively prolong the the duration of the auction to collect more and then gain that gas efficiency but in times of effectively like high uh swap count like high volatility you can have one every block for us it's also 12 seconds but with the caveat that we are not yet synchronized to block time um instead we start an auction whenever an order comes in but this is something that we probably want to work on beginning of next year that we have sync time with block time so we have one bch per auction that's that's super I guess what I'm what I'm trying to get at is here um I'm trying to understand if there's a different swapper like the profile of a swapper for a for an FBA like a cow

swaper an angstrom is different than Unis swap per se um like for instance I would I would guess especially like during the depths of the bare Market when there's not a lot of organic onchain trading activity there's a good amount of like sexex arbs uh that are making up the majority of like certain uni pools um I is it as effective uh to do that through an FBA type mechanism like I would guess it's more organic order flow that's moving through there right yeah I also for us we basically never compare Unis swap we one two three um volume with with our volume but we only look um specifically at Dex aggregator volume in comparison because all this Arbitrage you would never have

on a batch auction right not even on a DEX aggregator and um so I think yeah it's and I mean there's the numbers vary but I think people there's like estimates that um Arbitrage volume on Unis swap is between 50 to even 80% or so yeah it's from the last estimates that I saw was is about 80% for us uh so we still have the Arbitrage volume for the sex de because it it definitely provides this gain and efficiency especially in a competitive setting right you have the best of the best that are competing to provide what is effectively the best Oracle on price because their p&l is dependent on it um so we still have that uh portion of volume that is going to occur once we launch um but we don't have the sandwich

volume because by construction you can't sandwich um a batch where you have a uniform clearing price hey everyone wanted to take a quick second to shout out this season's partner Maverick protocol now many of you probably know Maverick is an Innovative amm which they are but reality there are a lot more than that as well Maverick is a suite of tools for defy users and builders that allows them to put liquidity where it will get the most work done since Maverick launched in March they've been gobbling up market share and at the time of this recording which is the end of September on a trailing 7-Day volume basis Maverick is now a top three decks by volume and they support over 50% of the volume on the L2 ZK sync a chain Maverick enables LPS and token pairs to

process higher volume with limited tvl which allows them to support some of the highest levels of capital efficiency for lsts like rap Ste another very cool feature is something called Maverick boosted positions so that allows protocols looking to bootstrap their token liquidity to Target the shape of liquidity of any token pair with surgical Precision Maverick is backed by some of the leading institutions in crypto Founders fund panta coinbase Ventures Finance lab they are all backing Maverick in their Vision to revolutionize the next generation of defi daps and helping to build their liquidity in all market conditions click the link at the bottom of this episode let them know that I sent you thanks guys one thing we talked about a lot of the show is whether there's a role in

the long term for Passive onchain liquidity providers like have been provided by amms um Anna do you do you think in the long run we will have a lot of passive onchain liquidity um I think it might change in the sense that um I think for the very liquid token pairs usdc um e rep bit bit coin um essentially you have a lot of very efficient market maker liquidity um then and then okay maybe you have like some very efficient models on curve for stable coins but I would assume that for longtail tokens you will always require onchain liquidity like I can like market makers typically have a

portfolio of maybe the top um 20 assets and then even in an fragment like even if you look at a batch auction model that is able to do ring trading such as C was able to do it in order to like really leverage um uh multi-dimensional order book liquidity you still will always have like certain assets that will be hanging around in an order book if there's no on liquidity for quite some time and I think what will probably disappear is maybe amm that we know from today because we know that LP provisioning is not uh Capital efficient and it's not profitable in most cases but I do think

we might come up with designs with amm designs that become more um promising and more profitable for liquidity providers and um for example cop this is just like a side project that we did a little bit of research on so it's not really our main focus but um we looked into cow amm and this is essentially um looking at the opportunity um of amms where essentially you would only solve um trades and badges and at a uniform clearing price so essentially you would collect orders a multiple of orders and see if rather than sequentially being traded if they

all traded what would be the final price date of the amm and that state is the one that is being given to all orders in this batch and the looked at it mathematically we calculated it and this is why um this is actually more profitable for IP providers even if you remove then the entire protocol fees because obviously you also need to set some sort of incentive why you would be trading against such an amm and for liquidity providers this is yeah this would be a more profitable way and of course if if they could choose against like providing liquidity on a normal amm versus on a cow amm it uh yeah if if they go purely by if if you

forget about other other factors such as popularity knowing uni SW more um being first mover advantage and so on if you ignore all those factors and just look at it from Pure economic profitability um it would be a laric move for LP providers to to to shift their liquidity over I'd say that they're they're going to be able to do that very soon um and on uniso because we're building on Unis V4 but effectively like the design that you've kind of uh described there is very similar to what we're building um but I think we found that there is an improvement to be made on that design which we implemented um which is to say that that design by definition is leaving money on the table um and why I

say this is because in the effectively function maximizing amm in the amm where we have these batches and everyone is trading at the same price what occurs is you at least theoretically are converging towards two price because as someone makes a trade they are pushing the price naturally and then that is the new um effective price of the amm and some sophisticated actor might realize oh well there's still money to be taken from that and they'll again push and then they'll progressively converge towards that true price but where we kind of flipped that design is that process is inefficient in terms of arriving to that price because at the very end there is some Delta that will always occur due to transaction

uh fees so gas and then just swap fees and you're necessarily going to lose that um so where we kind of change that model is instead have an initial party that is responsible for arbitraging back to that like effective true price and if that effective true price is incorrect let's have the rest of the bundle um through Market maker limit orders and user orders contribute to bringing that back to the effective price without having that additional cost um but in terms of uh kind of liquidity provision and passive liquidity going back to your question um I think that passive liquidity will still exist but not as we know it and understand it today all of

the solutions that have kind of been designed for providing efficient liquidity onchain in a decentralized sense thinking of Iraqis and various other um LP management uh Protocols are plagued by this problem of lvr they're plagued by this uh the the fact of life that they're continuously getting Arbitrage like this is the problem that we initially worked on and why we kind of shifted towards angstrom is because we realized that regardless of our knowledge and cognizant of toxic flow regardless of how much information we have on price if we know that an Arbitrage is going to occur we cannot bid the value in the block auction to the Builder for them to actually shift and re execute our transaction that rebalances our position

because to do that we have to compete the arbitrager and the arbitrager is paying the Builder from their profits and their profits equate to the LP's loss so in effect what you actually result to is as an LP you'd have to bid just as much as you would have lost and doing so is is strictly unprofitable so or you can like effectively do a lot of work and do your best to be cognizant of toxic flow and move right before but still have to pay that loss because otherwise the Builder doesn't pick you or you just don't do anything and you effectively get Arbitrage so what we kind of realize is that if you remove that loss there is quite a bit of incentive to provide liquidity in a very efficient automated way and so passive

LPS as we know them will probably go through these third party services in like hopefully a decentralized applications and be able to provide their liquidity and have it managed for them and then it'll be very efficient so long as you remove that inherent loss that occurs every time there's some toxic flow I've got a I've got a sort of follow-up question for you here lwig and actually just wanted to maybe change Tac for a second so I think we might not have made clear here that angstrom actually is going to be a hook on Unis swap V4 so I would actually love to get your your thoughts here on what was the what was behind the decision to launch as a hook on univ V4 as opposed to launching your own Standalone decks yeah um so for us uh building an

amm and implementing that smart contract not very easy right it's already significantly difficult um it takes a lot of work and you have to audit it and which is extremely expensive and these amm are holding a lot of capital um and then there's more so the network effect that Unis swof has the sheer like dominance in terms of people's habits of of trading when I want to trade on chain and I think oh where am I going to trade often I'll just intuitively think okay let's go to unisoft they've like effect they've spent a lot of resources in the marketing um end of the game to push that image and I think that they've somewhat won on that on that end so for us to say okay let's build this entire

new design let's have a more efficient system and on top of that let's um compete at the marketing level just seems like an inefficient uh way of going to Market and I think that that is why so many hooks are very happy to tap into the Unis swap Network effects because LPS clearly trust unisoft they trust the smart contracts they trust the quality the code so if you're able to innovate in finding like a more efficient way of clearing trades and a more efficient way of providing value to LPS and you base yourself on like the base uh Unis swap smart contract you can still be just as profitable you don't have to leak value to Unis swap itself but at the same time you don't have to spend so much time working on all of those other things that seem quite difficult and very expensive okay one

one uh qu or follow-up question to you on that um so basically to to summarize here the advantage is sort of there's the Unis swap brand and you already you have a bunch of flow right so trading flow that's coming maybe through the Unis swap front end or or wherever um so is that am I basically summarizing that correctly that that's sort of the the advantage and then I guess my I sort of a technical question actually to both you and Dan maybe is how does the how does the routing sort of work there like what's the what's the decision that the Unis swap front end or whatever makes in order to route to let's say Ludwig's hook as opposed to you know other hook XYZ then lwig do you think about your competition as other hooks sorry I've got a lot of questions for you on this

obviously I'm just curious how it all works in terms of routing uh that's kind of an unsolved problem at the moment and Dan correct me if I'm wrong but hooks had a fair bit of complexity in routing and so I think that's where the shift in Unis swap X compared to an on-chain router makes a lot of sense um so you kind of depend on that filler role um to find the most uh efficient path to a given trade um so most of that I think is going to be pushed off and this idea of onchain routing being efficient I think is somewhat of an idea of the past because you're operating on state n minus one you're operating on the the previous uh State and as the block is then created there's a ton of orders that might hit that state and then you

kind of do not have the route that you would have hoped for um in terms of competition generally I see like competitions other hooks um other people that are at the base layer trying to find more efficient design and better ways of making LP sustainable Anna why why not build km as a unisoft before hook can we convince you to do that um it's just not really our Focus right we were just thinking about how can we make amm more efficient um could be an idea to play around with but yeah definitely I I think we will stand away from building km as a standalone project because just competing with UNIS on

liquidity would probably be a nightmare it would be very difficult I I guess just to get the the other side of the coin here is is there any downside to to building on a hook obviously the the upside is is huge there but uh not to put you on the spot there a little B but do you see any any negatives for the negative trade-offs with the hook there's the obvious concern that you're less gas efficient because at a technical level you're operating within the design of unisa before and you could be operating with a design that is applied to you right um we aren't bothered At All by that because as I mentioned like we're going to arrive to swaps that are like substantially cheaper than Unis swap V4 itself because of our batching because of our ability

to compute everything off chain and then compress everything down um so if you're operating just a standalone hook and doing all that logic on chain it can get a bit more expensive but I don't see it as that like that big of a problem honestly there are also some um just to be fantic there are also some gas efficiency benefits of V4 right when you have multihop trades on on unisoft V4 because of the Singleton design it's actually cheaper some some of the costs end up actually getting amortised across multi the multi uh the multihop swap so there's some benefits particularly if you're anticipating it being arbed against other other Unis V4 pools i' I'd say so like yes there's some benefits but I think that I think architecturally it doesn't impose limits

that could that could increase the gas absolutely yeah I think those benefits are somewhat marginal compared to the cost of operating Purity in their framework but if you're operating within like the ecosystem of hooks you're definitely better off just having a hook yourself all right so maybe closing question for for the two of you so um one of the one of the the questions that we've been asking ourselves this this whole season is do you ever see a future where um you know we can move price discovery on chain outside of like the longtail of sort of crypto crypto donative assets for maybe like the majors like a Bitcoin or E like L B I just heard you say something that was like H that's just not ever really going to happen so I I'd be curious like do you ever see a world

where we can move price Discovery off of the the binan of the world and onto the dexes or is that kind of a pipe dream or and and if it is possible like what are the things that we need to improve to to get there there's a there's a big zero to one that has to happen I don't want to say impossible because being bearish on on technology in the long run make makes you look like an absolute fool um but in the near future like I don't see it happening because at the end of the day like looking at salana for example and a lot of people that operate in defi on salana will tell you that they have these super efficient order books and transactions are so cheap and everything's great and you can effectively uh have like similar efficiency to centralized um exchanges I

don't think that's true and I think it goes back to how transactions actually end up on chain um as as soon as you have a builder let let's apply this theoretical framework where transactions don't cost anything as soon as you have a builder that has a limited capacity of transactions that they have to kind of post on chain um and they get to decide what gets included you can you immediately arrive to the problem where if you have let's say a limit order book which would be necessary for on chain price Discovery they can censor those transactions they can choose which transac actions get included and which don't and so if you have a stale quote uh and a market maker is trying to

cancel their limit order well the Builder naturally is going to say well no that that transaction looks great to me I'd love to execute and be your counterparty there and so if we don't solve that problem regardless of how cheap the transactions actually are you're still facing that loss because you don't have the ability to cancel your orders you don't have the ability to shift your limit orders as quickly and as as efficiently because there's that rent that has to be paid because another party will be able to extract value and you'd have to pay them more than the value that they'd be able to extract that's an interesting concern to think about but in general I am more bullish I think ultimately that's the end game right

that is the goal for us why we work on this is that we do believe that at some point and I'm not saying in the next year or the next two years but that ultimately um price Discovery is taking place um on blockchain um and that the majority of trading volume is taking place on chain is the goal right this is why we essentially but this is the future that we are all building towards and um you see more and more assets being created on chain and you see more central banks looking into building on top of uh different blockchains so I I do think that um again not in not like in the next two three years but that in

the future I I I do believe that price Discovery will take place and I'm I'm sure um if you think about those problems long enough we will also come up with solutions for those I think those those problems are very hard those problems are extremely difficult um because you'd have to find a collaborative way to arrive to cancellations you'd have to find a way to limit the ability of the end proposal ER and reduce their financial incentive to zero to censor anything which is an extremely difficult problem that like we're currently very much played with we first a have to fix scalability and ensure the throughput necessary which is

also very difficult and then you also enter the other problem which is decentralization like creates um latency between nodes you have an idea of a mle or you have private order flow both create their their own complexities like I'm I'm not one to say like this is never going to happen I just think that for a significant amount of time this will not occur on chain and we are going to need zero to ones in the fundamental technology that we're building upon to be able to do this because there is no smart mechanism designed today that will make that possible I mean you could probably even think I I mean this is this is going of go in a very deep discussion but like um think about ways of if you could apply penalties for if a cancellation has been

expressed or if the wish for cancellation has been expressed and the um the rule then basically says if the transaction is still being executed despite that expression that the party the executing party is being punished slashed in whatever way I don't know like I'm I'm sure if you play around enough with Game Theory there like some ways of yeah like if you do that you'd have to have full data availability so those transactions have to be posted somewhere and everyone has to hear them everyone has to see them to be able to slash so to slash a censorship you need full data availability because if you don't nobody knows that that that transaction was seen by the ultimate proposer it doesn't it it can just be like it doesn't necessarily um have to

be full data availability as long as you have some proofs to verify the transaction right you could still have privacy and at the same time have incorporate some sort of proof system that would validate whether a order has been cancellate canceled or not I don't know it's getting complex and I'm not saying that there's a solution right now but I think like if we think hard enough and long enough about the problem there might be some ways of circumventing it no we'll get there we we'll definitely get there but it's gonna take a lot of work on that note of optimism I think we can leave it uh and and lwig this has been a ton of fun and a really great episode if folks want to find out more about you uh follow you or the projects that you're working on what's the best

way to do that um for us on cow swap basically at cow swap is our Twitter account and personally um Anna Ms George also on Twitter is the best way for us uh at Labs on Twitter we're going to be releasing some really cool me data very soon and our docs so I'm excited about that um and personally at Van B.E on Twitter excellent all right team this was a really fun one thanks so much for joining thank you having us all right what a great episode that was uh that was a real thinker a lot of um lot of eye opening moments for me in that one that good yeah yeah me too and they got they got pretty deep into it I think

also recognize a lot of the caveats maybe to some of the concerns about about the kind of these kind of solutions I think so too yeah I think so too so let me ask you this question that I sort of pose to both lwig and and I'd love to get your response here so you you could look at me as either originating from the sort of the base layer of ethereum and the design there and a lot of the me mitigation solutions to date have been sort of focused on the eth layer they're moving sort of the priority gas auctions offchain in the form of you know me boost or or what have you the other lens that you could look at it is through with Meb actually originating at the application layer through the mechanism design of protocols like Unis swap how do you

ultimately view that yeah so I think I think there's certainly some kinds of Mev all right so as as for where does Mev originate I think ultimately everyone designing protocols on top of ethereum is is trying to build it to build their applications or their protocols in a Mev aware way or they they have to just because of what the base layer um provides and then but I think on the flip side when people are thinking of the base layer they really should be thinking maybe a lot more and I think they they have started to think a lot more about what kinds of applications it makes possible and what um you know what affordances you provide to people who are building applications on top so I

remember a few years ago nobody who was working on ethereum um you know before before 20 20 before 2020 basically almost nobody who was working on the ethereum base uh layer even thought about M as a topic um it just wasn't you know it wasn't it wasn't really a and that when the flash uh flash Bo 2.0 Paper came out by Phil Dian um uh and others I think that kind of made it oh this is an interesting academic you know concept that you have this and then you know we we sort of saw the entire all of ethereum get flipped on its head when when Flash spots launched launched meev gu and you realize oh man this is actually much more important at the Bas layer than we realized I think ultimately what we're seeing right now

is a consequence of Mev unaware protocol design and we were all just a lot all fairly naive I think before excuse me and not really thinking about about um how Mev would would practically arise um we've seen the consequences of that at the protocol layer and that's requiring me applications to be a lot smarter in how they uh they design mware applications my hope would be that that's that fixes that the protocol layer could actually change um change some of the some of at least the decisions that applications now have to make yeah I I tend to agree with that I I I do I do Wonder like I'm I'm interested now that I feel like me has really dominated the the popular discourse to to such a degree that basically everyone's been nerd sniped by

it at this point and I do think this next generation of Builders when they're building their applications they do have Mev is not just like a small side consideration you could clearly you know hear both from from an and Ludwick like that's at the Forefront of their mind so actually like yeah lwi is is a Mev native Builder basically I he didn't really get into it but you know before they had done some um some work on the quy fiding on on Unis Swap and so they they've been research on that and so he's very aware I think of what it's like to be um to be a liquid figher and actually be on the other side of this and I think generally like these Builders who grew up in a meev aware environment just have a different a different perspective on it than for

people who who started with ethereum sort of thinking it was like it was just like oh Bitcoin plus um Plus Code that's super interesting not that you can probably comment directly on you know sort of Paradigm strategy here but do you feel like there's a bit there there is a really big difference between some of the younger more Mev native sort of Builders and like if so what is what is the biggest difference before between like this gen sort of the the love drgs or the the ones that came before well I think the Boomers like me try to try to be as up to dat and stay aware as I say that it makes me sound even even older and more out of touch um but yeah I mean generally working with people like like Frankie who I think you know um yeah kind of really like lives and breathes

this stuff you start to and you know I think I think Phil Phil and Charlie were ahead of the game on it but sort of have always have always believed this in their bones um you sort of start to it's about really the attitude is about thinking okay what's the worst case that could possibly happen you could hear that talking to both Anna and lwig here is when they talk about any solution they'll go and here's the here's what it could lead to here's the here's the ultimate possible equilibrium for it um I think that's I think you know it's important that people actually still actually can do that um and I think it's you know yeah that's a way in which the world has really changed is that I think we we start to recognize okay what's the end state of this mechanism what's the

actual equilibrium rather than this very fragile um non-equilibrium I think often that that uh things start off in when they when they initially launch yeah so what how would you define like maybe the not that the toolkit is the wrong word but the different set of strategies or how some of these more Mev aware applications look like today as opposed to their counterparts from even two or three years ago you know know one one commonality seems to be I feel like the cow swap intense model that I feel like they did a pretty good job of leading where they're relying on this offchain network of uh Builder solvers but like what are some of the Common Core Design decisions that you see being replicated across this this nextg of sort of amm or Dex designers yes so a few things I

think intense versus transactions is one where often not not every me solution I think flash Bots has done an impressive job building Mev around transa the transaction model But ultimately I think in tense gives you a lot more flexibility potentially to work with Mev mitigations and so I think that's uh that's a tool you see used a lot um I think things that move that are slower than a single block are often used um for for certain kinds of Med mitigation you see that with with cow swaps um uh longer batch auctions and that's partly beneficial because you end up you know individual blocks have have this a single proposer right now in ethereum and that's that's can be uh a big constraint for meev mitigations because you have basically this Unlimited Power by this individual potentially profit

motiv motivated entity to censor so uh protocols that take more than one block to resolve um is one solution but on the flip side another kind is protocols that take that find finality in less than one block so if you look at what most uh L2 sequencers are doing with faster sequencing slower faster block times you know two seconds or less um that can actually mitigate a lot of Mev um because slower block times means you have uh less less loss versus rebalancing for example um and less less opportunities to reorder transactions because there's fewer transactions to reorder so I think those are those are you know some of the things people can play with in this in these toolkits and one final one that I think you're starting to see but is is um the tech isn't uh is still kind of getting more

developed is a little more privacy in the M poool um and and protocols that allow you know the encrypted transactions or um or you know threshold encryption for being able to finalize the set of transactions before you before uh Builders or proposers are able to see them um and generally being able trying to play with privacy so so uh uh trusted execution environments as well and Flash buts has has done some research work on this things that mean that you don't actually have the Builder doesn't necessarily or the proposer doesn't have access to the to all the information about the about the block before they built it I think that that's another tool that I think is is just important for for avoiding yeah some of those informational uh disadvantages that L was talking about on the on the Pod yeah I think um you know it was

interesting to hear also L we we didn't get uh super into the Weeds about about the the guard work in in angstroms design but I thought that was a really a very cool thing which is basically sort of this side car that execution clients can run to basically run their own auction where it separates top of Block M have opportunities from rest of blocks that all just happens Upstream right Upstream of where the the the block auction actually happens like do you do you think that's like you know credits to the team at Cella labs for I think coming up with a really Innovative design do you expect to see sort of more of that where you can almost imagine this like stream of of me flowing right in One Direction maybe some of it originates up here or whatever most of

it got extracted at the at the Block space layer I feel like you're starting to see there's just sort of more vanilla orderflow auctions and then some of these really Innovative designs like angstrom has you know where whereby you're redirecting that flow back to LPS or swappers or whoever it is like do you expect to see more more more of that yeah I love their design um and this this General category it's it's sort of an interesting middle ground between what would you consider like a like a a whole full l two right which has you know with a sequencer which has which is like reaching the sort of pre-c consensus on an entire complex blockchain State versus something you know very relatively simpler like like an orderflow auction that's an order by order auction um and this is something that's you know this and I cow swwa I

think as well it's something somewhere in between it's this protocol that has rules it operates on more than one transaction and it and it has more complex um validation logic but you know it's not necessarily a full L2 it's really just for coming to consensus on these particular these the say order transactions and one nice thing there is that unlike an L2 you know you may not need to prove this really complex state in order to like you know uh interoperate with L1 state or or to exit from this uh from this quickly and because it's so much simpler you could um you know like these basically do all their execution on chain but you could imagine one that uses that uses EK potentially to um to actually have that be more efficient but it's you know still a lot easier to prove than a full

L2 so I think I think playing in that design space where you have these kind of like side chain like um you know limit some somewhat limited trusted uh uh separate consensus processes for for things like ordering um or censorship resistance I think it's really powerful yeah yeah I tend to agree with you um I I would be curious because I'm sure you just thought more about the the Dex model than I have that but one thing that you know I wish we could have maybe poked at a little bit more with on it even is just sort of the average uh the profile of the average swapper of something like cow swap as opposed to uh Unis swap or something like that my my my high level understanding of this kind of comes from payment for order flow where the more the higher percentage of

your uh you know sort of the transactions that are passing through your platform or organic kind of more prog derogatory kind of dumb dumb transactions B the more valuable that that ultimately ends up ends up being you know I I'd be curious like how how do you think about um the the trade-offs maybe that even from like a Unis swap perspective of like taking on some of that more uh toxic order flow but like having higher overall volumes versus someone like cow swap who's has very toxic order flow and then actually some of their the the liquid like I was going to ask her about this uh this deal with balancer where actually balancer gives them better execution prices because you know they know they're not going to get run over by the toxic orderflow like how do you kind of think about those two

different business models so to speak yeah so I think ultimately the value index is it's it's true that it comes from the from non-toxic orderflow you would expect you would expect it to I think it's a little maybe a little bit subtler than that but um so in general I think you want as much of that as possible one thing I didn't push back maybe maybe I could have on what Anna was saying was for pairs like eth usdc are we going to get you know off offchain fillers have have all this inventory so we you know what we might as well fill offchain still like really large trades on cow swap they're filling an onchain inventory generally um I think you still need that onchain inventory part there's just so much more of it and I think it's

possible that just being exposed to to toxic order flow um if if you've got short enough block times if you've got you know the right sort of the right fee level actually isn't as much of a disadvantage um in in if you're still again gonna ultimately get this like large flow that ultimately like offchain failures just aren't willing to take so I'm not sure that if that's if that's as directly answering your question but I I actually still think I think it's really important to have this this uh base level passive liquidity on chain you know still still really very much believe this that like this massive pool of e usdc I'm not sure it has to be as deep liquidity pool as it is right now but having it on chain is just really valuable um and it makes dexes a lot more usable than they would be if you

had to depend on um on offchain liquidity for for um for large order flow so I do think it's you know I think ultimately my point is things like cow swap you know you they yes they they can try to filter out some of the of the toxic flow But ultimately they're still going to the same place there still ultimately a lot of a lot of order flow there is getting filled on this onchain from onchain liquidity and so you're still it's not like we're never going to actually get rid of passive liquidity I think yeah I tend to agree with you on that now maybe like last closing question for you because I realized actually first of all if if you want you know an advertisement for for hooks in the future should get lwig back on because even talking I caught up with

him for a little bit before this episode and you know hearing from him describe the benefits of launching a hook as opposed to his own standal on Deck was you know definitely made me may think pretty heavily but like can can you give us some sort some sort of insight into I don't know almost like the ranges of hooks that are being explored today or like what the general interest has been because I'm realizing I'm kind of like wow Unis swap like V4 is is coming um it was interesting to this is the first time I've really spent time talking to someone who's building on it it got me honestly feeling pretty excited about it and I was like I should probably be paying more attention to this so I mean can you give us any sort of sense of what that ecosystem is looking like

today I have to admit you know I was hoping and you know we were we were trying to design unisoft V4 to be something that that a lot of projects could build on but in the past um you know I think some great projects have built on top of Unis but in general in the past whenever um we've like put out something you Unis has put out something and said like oh let's let's hope other people build something on top of this it takes a long time and you know you get very limited amount I think of of stuff actually being built on top of it the Oracle is one example where it took a while you know the V2 launched an oracle a year later the V3 Oracle came out I think during that time like very few projects even made use of the V2 Oracle

because it was kind of hard to integrate and you just would have had to go to some trouble to do it and same with V3 I think it turns out it takes a lot of effort to get people to build on top of your on top of of something and so I was I wasn't too you know I thought okay maybe you know some people will build this but we'll see like a couple projects and and it'll be um you know it's at least it's be an option for people to do but maybe like you know we'll have to like go out and push and find people to build on top of it but actually just the response has been tremendous I think just so many people actually want to build on it there's been tons of inbound interest um I think I think people a lot of people who would

have built dexus but felt intimidated as like say and then was saying about by Unis swaps um uh perceived first mover Advantage but have they had a cool idea that for something they wanted to build is a decks um a lot of people I think a big part of it is just the message of feeling okay like now I'm actually I've been invited to basically build this uh buil to try out my deck design on this and like rather than feeling like I'm I have to go compete with this thing I'm actually going to be doing it as a on top of this platform I think it's it's just had a you know profoundly different effect I think on on in terms of getting new people to build to it from anything else that unisoft has done before and I think it's really cool like it's I it made me

really excited about building it as a platform um and I think the uh yeah I think we're gonna see some really exciting things I think lever reduction is probably the main um interesting the most interesting thing we see built on top of it but people have come up with new ideas for for oracles um built on top of hols people have come up with um new designs for for sandwich prevention and these other kinds of other kinds of me prevention that think make make use of really clever um clever algorithms uh in there and and you know people are doing it for using it for other things I think some of the more out there ideas you know you could build like a lending protocol or something on top of it right you can try toch it even Beyond um uh uh just you know what you would think of as

a traditional decks and you know I think we'll see which ones really which ones really work but I'm I've just been trying to be encour as encouraging as possible for people who are who are building on top of it because yeah I think say that same thing L was saying um that you know it's great to to sort of be able to build this uh and take advantage of some of those liity Network effects and brand Advantage yeah all right last question for you here uh and this was actually something that you asked Anon lwig um when we were talking about the Builder Searcher market and solver filler you know Searcher entities being a relatively small number of pretty sophisticated actors you're like do you think that this is something that we should fight against or try to prevent so that's my question to you do

you think that this is something that we should try to prevent and strive against or is there just some natural economic forces and we can just kind of let this happen I think it's important to keep it competitive I think it's important to reduce barriers to entry I think it's important to reduce um you know gain some scale um and and you know sort of the extent to which there's natural monopolies in the space I don't necessarily think it's important to if you've designed a competitive game um some people are going to be better at competing at that game some entities are going to be better at competing of the game than others and they will tend to win um I think if that's if it's still a competitive mark Market I think that's that's fine and ultimately like that kind of thing you'd expect if if they

were to disappear someone else comes in and takes it place if the only way that they're out competing everyone is just by being better at it by being more competitive um so I come down somewhere in the middle of it um I do think there there are ways in which that can be that can be very dangerous um and I think I you know I think we it's good that there are forces pushing back against that um and in crypto I think we always want to have forces pushing it back against anything that that that seems to risk a natural monopoly that could threaten the um the censorship or decentralization res or or anything censorship resistance or decentralization of the of the space But ultimately yeah to me it's most important to just design the game to be as competitive as possible rather than trying to to to hobble for example the

ability of a very competitive party to be able to compete yeah I tend to agree with you all right Dan this was a really fun episode next week we're going to be talking about uh cross chain uh with Hart lamber which is going to be a really fun one so I will see you uh see you next week see you next week