Cross-chain Arbitrages in the Realm of Blockchain Interoperability - Burak Öz

Transcript

[Music] welcome and today I want to talk about our latest paper uh cross-chain Arbitrage in the realm of blockchain interoperability this is a work that we have just finished and uh actually submitted to Archive and it's on today uh here is the list of my uh valuable co-authors and we hope this study will bring some into this Dark Forest of cross chain me because we have studied so far meev in atomic uh ecosystems or instances uh in singular blockchains uh but cross domain M especially between

blockchains themselves are usually on theoretical studies so far so we will bring some empirical evidence into this stuff that's actually happening and I will try to leave some 10 minutes at the end for questions we have a long session so it should be fine okay okay let's start with the motivation so why do we even care about this stuff so we know that the blockchain ecosystem is not a monolithic structure where we have a single blockchain and all the activity happens there but we know that we have different blockchains with different properties like block times consensus mechanism uh sequencing mechanisms and so on which then creates this multi-chain ecosystem that we have today and especially in the last years because

of these inefficiencies in layer one networks like ethereum regarding uh scaling so the number of transactions that they uh execute per second or in general the speed of the network we also have these Layer Two Solutions which emerge to scale the underlying layer ones and we know that defi applications decentralized Finance applications remain at the Forefront of this growth with uh like daily trading volumes in billions and total value locked again in billions and we know that what happens is that when we have such uh Finance activity happening we know that uh the blockchains can result in states which could create profitable opportunities to act upon and the canonical example of

that are for example Arbitrage so like if you see two markets in the same blockchain offering the same assets but that they have price differences you can buy on the cheap one sell on the uh expensive one and then make profit from the difference and the nice thing about um blockchains is that they can provide Atomic EX execution meaning that you don't have to take any risk in the uh Arbitrage so it will happen if it's successful and making you money and otherwise it will uh fail causing you no risk in uh attempting the Arbitrage and in general we call all these activities of um acting upon profitable States maximal extractable value on blockchains that's like all the value that you can

extract by inserting transactions sequencing transactions censoring transactions and so on and this concept although we always discuss it in blockchain context it's really anything in life like you can think it like buying a Taylor Swift ticket before everyone else on Ticket Master and then selling it one week before and probably you're going to make a lot of profit now because this is a crypto conference we are interested on me in blockchains and as I said me has been heavily discussed on single chain environments and then more recently we also started discussing in Cross domain environments uh but mostly As I said these studies were either on theoretical level or most recently focused on me

that you can extract between centralized exchange like binance and blockchains but not necessarily too much activity has not activity but too much research has been going on between the blockchains so me that you can extract blockchains and we have recent studies on the potential value that you can extract so like by just looking at the States you can actually come up with a a estimation for that but we have no onchain evidence of this stuff actually happening so we had this nice paper back in 2021 I think uh called Unity strength that attempted to formalize cross domain meev so to give a super uh simple um explanation the idea is that if blockchains are these individual domains

then meev on a single chain is simply the maximum balance increase that you can achieve by doing all the um actions that you can do given a certain state so like given that we are in Block 1 million what can you do to maximize your balance and then cross domain meev is simply this but now assume that we have multiple blockchains and blockchains and it's they are all on a given State and then what is the cumulative set of actions that you can do to reach this maximiz balance across all blockchains and assume that it's they're all denominated in the same numer year so we have a common like USD kind of value that we can then try to optimize so this is cross domain me as a

definition and to put it into an example if we look into single domain me so a classic uh single chain Arbitrage we know that like on ethereum we have various markets which offer the same assets so like let's assume we have vet usdc and usdc usdt and usdt vet markets which you can sell vet for th000 uh 2,000 USD and then swap to usdt in the same rate and then sell the usdt for two vet making one e uh revenue or one vet Revenue uh minus the transaction fees for profit but in general pretty profitable and the cool thing about this is that you can do this automically as I said so in a single transaction you can

complete this Arbitrage without really taking any risk and because of that it's super competitive because pretty much you can go on YouTube right now and find a video explaining how you can do it copy paste the code and good luck finding anything because anyone can do this it's highly competitive and because of that the returns are super low in unless you're finding a really like exotic Arbitrage maybe involving like six Hops and weird tokens which the people can't attempt now this is single chain Arbitrage a cross-chain Arbitrage would look like this where you would have some of the markets on a different blockchain and still like if you have the capability of uh making a swap on

ethereum but then also continuing that swap or continuing the Arbitrage say on arbitrum well then you can actually make profits just like you were making profits with a single chain Arbitrage the main difference here is that you have to let go the atomicity property the I and that's because these blockchains they are not acting together they are not synchronized in that sense so yes you can have your still ethereum swap transaction going automically but it doesn't guarant you that the arbitrum transaction is also going to be successful and this introduces a lot of execution risks because maybe one leg of your transaction could fail and making the whole thing uh go sideways or because you need to like make one of the trades

first and then send that tra um output to another blockchain you're losing valuable time how do you send that there are bridging Solutions but they have cost associated and if you're are not going to use a bridge then you have to have inventory on multiple networks which is then fragmenting the inventory that you have so this is again something more difficult than a cross-chain Arbitrage where you will have inventory only on ethereum and that's it and because of all these challenge associated with uh cross-chain me or cross chain Arbitrage execution in general you can also think it as well this then has to be less competitive than our Atomic Arbitrage because well it's much harder to get into this like

executing this stuff is much more difficult and this could then mean that the returns are potentially higher than Atomic carbage R because you don't have to let go that much of the transaction fees to the block Builders to the validators because maybe you don't have to prioritize your transactions that much so another question we might ask is that okay yeah we can do these cross chain arbitr but we also know that we can sexex Arbitrage these individual blockchain so what I mean is that yes we can find a price discrepancy between ethereum and arbitrum on the same yeah say with usdc pool but it's potentially more profitable to Arbitrage them against binance than against each other

because with binance you have lower fees and potentially deeper liquidity meaning that you can make even more Revenue so then one might ask yeah in that case why would I attempt even doing cross chain Arbitrage and in fact this is the common skepticism we have in the community about about does cross chain Arbitrage actually happen and one way to counter this would be that blockchain interoperability Advanced we have faster bridging Solutions more secure we have nature Bridge of the pro blockchain protocols themselves but also we have these multi-chain Omni Chain Bridge which enable uh bridging between various blockchains so then this might potentially enable cross-chain Arbitrage and this could be like an incentive to

do them instead of sex STX arbit ing them but still the question is does anyone actually do cross chain Arbitrage and in this paper we show that they're actually getting done so the methodology although it's much deeper um described in the paper I will just give a brief overview so the idea is that we try to find two hop arbitri between these nine networks that you have you can see uh some l1s and then mostly l2s we try to like get the uh blockchains with most defi activity happening on them and then you can see that all these chains have their different ordering mechanisms block types mol types and if

you go deeper uh if you try to think about it this would mean that like as an arbitrager now instead of being a pro at ethereum only you have to optimize for I don't know optimism as well because it has a different mechanism or binance smart chain and use maybe the private me pools that binance has to offer now and this again complicates things and uh more broadly we introduced two Arbitrage strategies that is a sequence independent Arbitrage and a sequence dependent Arbitrage which I will talk about briefly but in general we look we try to identify these two strategies across these nine blockchains over a one-ear period from September 2023 to August 2024 so okay what are these two

strategies super simple this sequence independent cross-chain Arbitrage is the Arbitrage where we assume that the arbitrager has inventory on both networks and they don't have to bridge the assets from the Hedge leg to the other leg uh to the other blockchain to complete the Arbitrage and this is potentially more advantageous because you don't have to wait for the bridging to be completed and you don't pay the bridge costs but of course this is again potentially more complex because you need to have inventory on multiple blockchains the other one is a sequence dependent cost chain Arbitrage because in this one we assume that the once this swap is done on the starting leg it's bridged to the other blockchain and you have to wait for the bridged assets to

arrive to complete the Arbitrage so you cannot do the individual steps independently but you have to comply with the sequence of the things whereas in the first one you can even do this step two so like the profit leg before the Hedge leg because you have the assets already Okay so let's jump into the analysis results so from September 2023 to August 2024 we found 26,8 cross chain Arbitrage executed across these nine blockchains with a daily activity increasing like 3.2 times from where we start to where we end right now and then there are some individual spikes but in general you see the upwards Trends in general uh overall these arbitres traded 465 million in uh

volume but of course take this with a grain of salt this is potentially a lower found because um estimating uh profits or in general USD volumes of these uh individual transactions is quite challenging because they could happen on like any kind of tokens and multiply that with nine different networks so actually spotting like the dollar price of these uh individual tokens are quite difficult and when we can't do that then we can't really account for their volume in this total volume so this is most probably a lower bound and then what's interesting here is that we found that these arbitr generated close to 11 million in revenue and 9.5 million was kept as a profit and

on average they had like a 67% profit margin now to put this into comparison the previous Studies have shown that Atomic Arbitrage so Arbitrage that happen on a single chain usually have a profitability rate less than 10% and that is because if you don't let go like 90% % of your profit Revenue then someone else will and they will do that Arbitrage and you can't keep profit simply but with the cross chain Arbitrage we see that you can keep a lot of re of your revenue and that is because again not many people are attempting to do this so you don't have to let go that much of your Prof of your earnings the similar thing also applies to sex Dex Arbitrage so the same study also claims that sex deex arbitr can

keep up to 65% of their profits and we think this is because of the similarity of the in the nature of these Arbitrage both of them are much harder to do than uh than a single riskless Atomic Arbitrage hence the competition is potentially less than these Atomic arbitres when we look into the individual blockchains which these arbitres ured on we see that the most volume passes through ethereum with 36% and the countwise like ethereum arbitrum Bay are all similar you can see like Bay gaining some traction over time but in general this is an expected thing because we assume that for these profitable token pairs ethereum is

usually having the deepest liquidity and hence if you want to do like a profitable Arbitrage usually you do it with ethereum against something but of course a lot of activity is also happening on other chair chains and we will uh try to look into like why don't they have that much of a volume this is another way to to show uh how these chains interact with each other not interact but how the arbitres are distributed again on the left hand side you can see that countwise for example yeah binance smart chain and arbitrum Arbitrage are also prevalent but if you look into volume and profit anything with ethereum is like much more prominent than the other uh chain Pairs and if you look into the

individual pairs the top four uh pirer with most volume and profits include ethereum in it what's interesting here is that um binance smart chain and arbitrum and binance smart chain and polygon have like the second and the third most activity and yet uh profitability wise or not profitability but profit-wise in total and volume wise they are much lower than uh ethereum involving block uh blockchain paays and this is because of the tokens that are arbitraged in between them so we will see in shortly that the token payers that are involved in this cross chain Arbitrage are usually these meme coins or like utility

coins which most of the times don't have that much um liquidity behind them so they are usually arbitri for really short periods and then never used again and because of these smaller Arbitrage that happen like constantly between for example binance and polygon they don't add up to like a lot of volume but and yet the count is too high what's really interesting is this dur uh distribution of Arbitrage durations so what we know from traditional Finance is that like if you read these famous books um Flash Boys uh you will see that they comp these high frequency trading firms compete for micros seconds milliseconds and yet we see that cross chain Arbitrage yes like

a 34% of them take less than 10 seconds but you see the long tail of the Arbitrage duration so this is like the difference between the last leg and the first leg of the Arbitrage and this is like super unintuitive because like we would expect this stuff to be happening like blazing fast and yet you see that for example ethereum polygon Arbitrage the most frequent Arbitrage pair they have like a median duration of what 20 minutes this is crazy and we actually find that this is because these arbitres use the ethereum polygon native Bridge which has a really long bridging time and yet you see still a lot of them are happening so how does this happen this is probably

because or the tokens involved they don't have much activity going on and they are super risky so people are not even like uh attempting to hold them hence you can like do a swap on ethereum bridge them and still 30 minutes later you find a profitable state which you can act act upon and complete the Arbitrage and again this is something conflicting with what we know from sexx arbitri because sex Dex Arbitrage say on ethereum versus binance on binance the prices are getting updated on the order book like every millisecond and on ethereum you wait until like the very last second or millisecond to like submit your ethereum leg of this trade

so that you have like a profitable Arbitrage and act like as close to the order book uh price as possible but here we see that like that's not the case at all and again this is because of the tokens that are involved in sex Dex Arbitrage versus the cross-chain Arbitrage so these are the 10 most uh profitable token payers and good luck if you recognize any of them um for example this Omni token is if you look it up on coin market cap um it's the first layer zero inter reparable uh supporting token payer a token that you can uh Bridge across blockchains hence the activity on

it and we found like 1,944 distinct token payers which is like quite High compared to sexx Arbitrage which usually happen on like 60 pairs or whatever and again this is all supporting the fact that like these cross-chain arbitres they are not like a profitable ongoing business that you can always Arbitrage n mty usdc payer but you Arbitrage it for some time and then you move on and good luck with with getting rid of your liquid uh with of your inventory you have on that token this is showing a quite interesting thing this is the profits that they made from these top 10 token payers and you can see that for example

that Omni vet pair it's like it has the spike and then it's pretty much never arbitraged again and really similar behavior for all these profitable token pairs again supporting the fact that there is no wet usdc or wet wrapped Bitcoin pair that's like constantly arbitraged uh in Cross chain Arbitrage realm right now if we look into the bridge usage in these arbitr so the difference between the sequence independent versus sequence dependent Arbitrage we see that like only 32% of them actually use a bridge and 67% happen without a bridge maybe this is a limitation of our methodology but that we didn't detect some Bridge operations in between but still this

gives us a hint that the bridges are right now not uh efficient enough for supporting cross chain arbitraging you see on the right side A A Time series of like how many Arbitrage happen per month per Arbitrage type and you see that the sequence independent Arbitrage they are increasing actually but the bridge using ones they are almost stale so again um and you can discuss this further from the point of view of Okay then if Bridge are this slow and like they are not really helping a lot to do these cross chain ARB arbit then probably the entities that can do these arbitr are again going to be these like potentially highfrequency trading companies or those

which have like inventory across multiple networks and know how to hedge and so on so the again the barrier to ENT enter this Market is then increasing and across among the ones which used brids we saw that uh at mostly layer zero which is like an infrastructure that you can also build your Bridge upon is used followed by native bridge what's interesting here is that although native Bridge had uh relatively higher costs and longer durations overall they had a much higher average profit than the multi-chain bridge and this could be like uh pointing towards a trade-off between

Bridge security like you would trust your high volume Arbitrage with the NATO bridge and then if you want to do really like maybe sub dollar arbitres then you would use these um multi-chain Bridge which are potentially faster and yet they always have this risk that like something may happen like you can think like an attacker model where the bridge would like simply yeah see your transaction and then they would complete the Arbitrage instead of you on the other network we didn't look into this but just saying okay so then what could be some security implications of this stuff so we found that so in this um seminal paper Unity is strength they say

that like the thing that we fear about cross domain me is super traders that is these trading shops which can successfully execute like these me strategies across multiple blockchains and because of economies of scale they simply become more powerful as time passes on and no one can simply get close to them and in fact we found one evidence of that in cross J in Arbitrage this c74 guy he is right now like doing 43% of the whole Arbitrage volume overall like 32% but you can see he's increasing his um prominence and what this can lead to in

the future is that like if this guy is the only one capable of doing these arbitres well then he can simply say okay like if we have block Builder like mechanisms on different blockchains eventually he can simply say hey I have this really valuable order flow like do you want to access it this is the price you have to pay me for this and hence simply have this upper hand which can then lead to centralization in different networks which we have seen happening on ethereum and for example like sex Dex Arbitrage have been shown to be a prominent resource of this centralization these are all the um arbit resur that we have identified which conducted more than one person % of the total cross chain Arbitrage uh

trade volume what is interesting here is that on the second column you see the type of the uh entity that we have identified so by type we mean is it an externally owned account so like issuing transactions always from the same address or is it a smart contract that they use to conduct these Arbitrage so in atomic Arbitrage even in sexex Arbitrage which like the blockchain activity happens on a single chain we usually see smart contracts getting used so I as a me Searcher develop my own logic deploy a contract and then spam that contract and if it finds an opportunity it executes otherwise does

nothing well now I ask you like how do you do that for cross chain Arbitrage like how do you ensure that when I call a smart contract on ethereum it knows that the polygon price is also profitable so I can execute this and because of this lack of um synchronization between the blockchains that you cannot really say to a Smart contract right now like only execute if the polygon price is also uh profitable we don't see a lot of smart contracts getting used for this but actually EAS and you can ask okay how does that happen we think the idea is that you have these offchain scripts which simply uh track all the prices on all these

markets and then you can apply your same arbit Discovery Logic on them like find Cycles whatever and then simply issue them either simultaneously or issue them via a bridge like adopt a bridge logic and this is how you then synchronize uh the signals coming from different blockchains and what we find interestingly is that the four smart contract arbit Treasures that we have identified the two of them have nine failing transactions for every one successful Arbitrage they have and like we show the top two uh networks that they have these failing

transactions on both of these two guys have them on mainly on polygon and this brings up the question how can we analyze Network congestion or fail transactions in the cross Trin realm so the question is if we study failing transaction on a given blockchain like we can I don't know look at a smart contract and say yeah this guy is interacting with this contract and this contract is producing a lot of failed transactions but can we really conclude that that contract the failed transactions is because that guy is like solely doing an activity on that chain or do we actually have to consider this

more in a cross domain um environment because the fail transactions that I have on polyg for example right now could be because I'm trying to do an like ethereum polygon cross chain Arbitrage and maybe I have the ethereum leg secured but I'm still spamming polygon to get my transaction in so I think this requires like when we model this congestion we need to consider that the causes of these failing transactions could also be these cross-chain signals that these guys are trying to process and then finally we look into the private Mol usage and for this we focused on ethereum we also looked into uh bundles at binance smart chain but

didn't find anything there but on ethereum we see that uh I think the um daily number of private transactions increased like five times in this uh time period and then in general like you can see like why they are um attracted by private transactions because like you see that they have a higher average block position and why does this matter because like the successful execution of almost all of his strategies depend on acting upon on a state before the others or like not leaking your strategy in a public men pool and that's why like using private men pools could be quite advantageous but now the question could be okay like still yeah I can have my

ethereum transaction relate privately but if my optimism leg is public then is this still a viable Strat Y and so on and this is a good question like right now we don't have a service which would guarantee you privacy on both ethereum and uh say arbitrum simultaneously but like with these share sequencing Solutions coming up maybe they will provide better atomicity and better privacy guarantees which then can again increase the prominence of these cross-chain Arbitrage that we see today so then finally discussion we had two items here the sequencer in incentives is one of the big ones this is also the one of the

biggest things discussed in this uh Unity strength paper that is um the threatening thing with this cross chain me cross domain me is that the sequencers of these individual networks the validators of them whatever they are in the best position because they simply get to choose what goes into the block and order them censor them whatever now if you want to do a cross chain arbitr it's not enough that you get the ethereum leg going on you need also the polygon leg the arbitrum leg and hence I might have an incentive if I'm a sequencer on say arbitrum to do a deal with like an ethereum block Builder or an ethereum validator so that I can somehow avoid this nonatomic execution that I have to fight against right now

like minimize the risk that I have and as we have shown that the prominence of these cross chain arbitres are improving over time we can think that this incentive to collude or like have collusion between these sequencers of different blockchains could improve and we propos these shared sequencing Solutions again to like level the playing field because if I as a regular Searcher can simply guarantee that I could have like an atomic execution of my cross chain Arbitrage between arbitrum and polygon or whatever then this would like enable to me to compete with these potential um sequencers that are

colluding and then getting their execution guarantees so it's a bold statement to say that these sequencer will will collude and so on but still we need these shared sequencing solutions to at least like come over the uh problems we have it like long bridging times or uh the non atomicity right now we have in Cross chain domain and then finally we uh proposer risk optimized strategy for those arbitrageur who cannot really afford doing a um sequence independent Arbitrage so the ones who cannot have like inventory of these uh meme coins on multiple blockchains and the idea is that uh say you have

inventory on ethereum but not on arbitrum then you do the ethereum leg of the trade and then take a loan on arbitrum so of course you have to collateralize something but that could be in in a like more token and then get a loan of this risky asset complete the Arbitrage simultaneously so just like a sequence independent Arbitrage but then on the first leg breach the assets to the other uh blockchain and when they arrive close the loan position now you hedge the risk uh of holding that shitty coin because you take a loan and immediately uh finish the Arbitrage and yet still you have the assets to close the loan position because they are

coming from the other leg and you have no rush when they arrive because the Arbitrage is already over and yet still you can do a Arbitrage like you are doing a sequence independent Arbitrage and this is an actual example of that where the guy uh arbitri wet wrapped Bitcoin pair and earns 0.5 wet and then closes the uh loan like I think six seven minutes later now this is an interesting example because it's a wet wrapped Bitcoin Arbitrage which is a pairer available on sex and we think that like to to be able to do Arbitrage like this on these more

popular uh payers you need like yeah liquidity because like the guy is taking a loan for 2.62 wrapped Bitcoin which is like not nothing and uh you need either fast bridging Solutions or strategies like this to like hedge the risk of like holding uh holding inventory on multiple chains but simply use loans and so on like another way to think about this would be like shortening the asset on on a sex but then I guess it will be a sexex Arbitrage again so this is an example of how you can do uh cross chain Arbitrage while being uh less less exposed to risk

so conclusion we show that cross-chain Arbitrage yet there uh while there is a lot of skepticism are actually a reality H but the majority of them don't depend on bridge because of the delays and the costs associated with them we see that already there's a significant centralization am among who does them and we see that cross chain Arbitrage could actually lead to a different way of network congestion that we should think about and we finally found that private muls as happening in pretty much everywhere in um me extraction especially on uh ethereum are getting adopted faster and faster and finally we discuss uh how uh cross J andme with it

with with its growing uh prominence can incentivize transaction sequencing across blockchains to minimize the risk of the execution and finally we share uh we propose a strategy which would uh be less risky than holding assets on different blockchains and and that will be it and this is the archive link Super Fresh just two hours ago uh uploaded and that's going to be all for now I think minutes for questions okay perfect thank [Applause] you first

hi first of all thanks uh for this great talk and all the research you've done it's uh amazing I have a question um about traditional Finance markets like the for Forex Market where you have all those traders in arbit Treasures who create so many um like millions of uh price indications and they kind of help the price um and um the the whole Market what is your opinion on the crypto Market with those are which Treasures do we need this in the crypto Market do we need to increase interoperability between chains to make this flourish and doesn't this only benefit a few super Traders like you like you mentioned yes um I mean fair point

so yes we need this because if we are not going to have a single Dex operating across all blockchains then we're going to have price discrepancies and to have efficient market we definitely need them to be arbitr so that we have we don't have price discrepancies and yet we have this concentration of who does them and I'm not sure if this is something EV inevitable uh or like if it's something that we want to avoid because you see similar Tendencies pretty much in like every Market in blockchains right now like if you look into ethereum you see that there are only two entities building 95%

of ethereum blocks if you look into the validator set like you see Lio coinbase so centralization is going to happen I think the critical question is what is the danger of that yes centralization in a decentralized world sounds like uh bad but besides the meme is it actually a threat for example on ethereum block building site the threat is censorship because two entities deciding what go go into every block can simply say we're not accepting any transactions from the area of Zurich um with arbitr again if only like one guy is doing all of them you can ask the question okay if this

guy is for some reason out of the game can someone fill his shoes that's a valid question um right now I would say in Cross chain Arbitrage there isn't that much competition so if that guy is out of the game maybe no one will fill his shoes but in many of the markets we have like the block Builder Market I would bet that those shoes can be filled so if there is enough competition and if we have Solutions against the things we don't like about centralization I think we are fine with having like yeah a centralized Market but yeah cross chain arbitr right now is maybe not one of those markets which we have enough

competition hi thanks um in terms of actually identifying the arbitrages per se what was well two two kind of questions what was the methodology that you went through to actually identify the Arbitrage was it based on the trades based on the wallets or a combination and secondly once you've identified those arbitrages is there actually a strategy which could be a copy trade strategy which could follow those arbitrages available yeah um so the methodology like you said it's based on the trade action so metadata about the trades like the um volume the tokens involved and the time stamps and then uh just by

applying cisic on this this can produce some false positives because like randomly two guys can do like opposite side trades on polygon arbitrum and that would be a match so to minimize that we also looked for some kind of an um association between the entities this could either be from directly the signer addresses or it could also be from the smart contracts that they use but of course the risk here is that if you just look into the contracts that they interacted with if both of them interact with Unis Swap this doesn't indicate anything so there we apply a logic that we check that we uh don't know what the contract is doing that is it's not a

labeled smart contract we have this like list of 1.2 million contracts that we know what they're doing and this is like a list curated of Def apps nonme contracts whatever and if it's not in this and of course think it like this this is not only on ethereum but we have to combine together these nine different blockchains that we studied so it's a long list probably still a lower bound but um we then check if it's a contract that's not in this list which can potentially indicate that it's a Arbitrage contract and both guys are using the same Arbitrage contract the second question uh what was the second question all right once you've

identified that this guy this this wallet is actually a successful arbitr is there actually an ability yeah uhhuh U so the thing is that because most of them are not actually using smart contracts it's hard to like you can't really understand their offchain logic but what we see is that this is not actually something we mention in the paper but pretty much like that ca74 guy all of his arbitres are yeah yeah so you can so okay two things first of all all these EAS they always use aggregators the idea is

process the signals outside see that prices have a discrepancy and then leave delegated to aggregators which will find you the best price to do what ca74 does assuming you have liquidity enough it's still risky because like let's say the guy does like one leg or there are like four minutes apart the two legs now you see that ca74 is doing something on polygon and let's say you front ran him like copied his strategy he actually uses public mol usually so you can do this but now you don't know where this is going like you don't know if he's going to head like profit on arbitrum on binance maybe he's not even doing a cross-chain Arbitrage there's always that that risk as well

because these guys actually uh are doing a lot of rebalancing as well because if you don't use a bridge then your assets like if you always do the same direction trades with this not using bridging then you are draining your inventory over time so you have to rebalance and we actually found that like they most of the times rebalance like either after the Arbitrage directly or like every X minutes um so it's risky to just copy because you don't know when or where the second leg is going to happen hey thanks for your nice presentation um did you only looked into

this nine blockchains or did you also looked into smaller ones I mean speciic when you speak about low liquidity it's in small chains which are Maybe dagging because of heck it's a very huge possibility of Arbitrage did you also check this yeah we didn't check uh any other network because the problem is that the way you um identify these trade transactions are by looking at events like swap events transfer events and transfer events are Universal but swap events depend on the actual Decks that emit them and yes on ethereum you have these common Unis swap like balancer

events but if you study nine different networks like they have different dexas that have different signatures we in total identifi like 108 different protocols used and now if you go into even smaller chains you have to be aware of like what is emitting a swap and like what's an active Dex there and we didn't go into that direction yet but yes that's a fair point that this stuff could also be happening on those chains or maybe even Solana so yeah [Music]