Sovereign MEV with Maghnus Mareneck of Skip Protocol

Transcript

we're live my guest today is Magnus marneck he is founder uh and today we're going to be diving deep into Skip and Mev on Cosmos I'm also dying to find out why he thinks remote work is ngmi before we get started make sure to hit the like button hit the notification Bell And subscribe to get notified When I Grow live every week and remember to know what none of what we discussed here on interop is investment advice and if you enjoyed this content please consider sticking out on osmosis Quicksilver and Atmos and I just want to remind everyone that tickets for nebulous Summit are on sale early bird tickets are on sale and they're almost sold out nebular is The Interchange Builders conference that's happening in

Paris on July 24th and 25th right after ECC if you're interested in hearing about the most Innovative and and forth thinking technical change this is the conference for you so head over to nebula.builders for all the details and get your early bird tickets while you still can my guess Magnus is coming up next right here on the interrupt foreign [Music] how are you doing thanks for joining me

doing well thanks for having me yeah um so today we're going to be talking about Skip and Mev and Cosmos this is a topic that I've covered a little bit on a podcast and honestly it's a topic that I'm you know still like learning about and I feel like it's not a topic that I'm um you know very comfortable with right I know sort of on a high level but at a deeper level um Still Still figuring things out um and what's what's interesting I think what Mev is that ethereum it's a topic that has been you know covered and researched uh for years and I think like people in ethereum have a very good understanding of how many works but in

Cosmos it sort of throws the entire model on its head because we have multiple chains and and also you know some cross chain action that uh introduces other complexities and other opportunities so that I talk about those things specifically and understanding you know how Med is is unique in Cosmos um but first yeah first I would like to get a bit of background on yourself and how you got to be here yeah sure so um I like to say everything in Cosmos has been always up for me because I started skip with Barry at a hacker house that was hosted by Tara in about May 2022 um and so we halfway through that Hacker

House Tara was no longer with us um and so our first experience in Cosmos was oh my God it's so huge it's so awesome Tara is so big and there's all these people here and it's so much excitement and then halfway through that I saw you know a market cap project from 30 billion or whatever it was to zero um but at the end of that Hacker House we had the first version of skip developed and uh you know we we were wondering well like what what else is there right and we were new to Cosmos we didn't really know any other chains and we started talked to Sunny we started off to Jake we started to talk to you know Liam um from evmos and we realized that wow there's there's a lot of interesting stuff going on here um the stack was one of the most

interesting things that we had seen and we decided to just take a bet you said okay let's spend the next couple years um or longer potentially building out more and more piece of infrastructure and one of the big missing pieces we thought was was Mev um we never viewed Mev as something that was purely extractive we viewed it instead as something that was an important piece of infrastructure for safe blockchains um but that's how we got started and so since since you got started like what has been the the progression so did you get started and you know what stage would you say the

company is it now in terms of you know fulfilling its roadmap yeah it's a great question so we're about nine months old um so a little bit less than a year and uh you know the first couple months was really focused on building the team but then also shipping our first product it's very simple it basically allowed for a top of block auction for any Cosmos chain it made a modification to Tender mint which is the consensus layer to basically allow for a external Marketplace to develop around Searchers who wanted to buy the top of a block um that product was was quite successful it was called skip select um it's implemented now ecosystem wide

on Juno it's implemented on injective Atmos Terra II and a couple other chains um running on about 200 plus validators and has mostly done what we wanted it to do which was start to open Cosmos up to this concept of well what if we started to play with the markets in our mempools as validators to allow for healthy Mev to be extracted and what I say by when I say healthy I mean there's a couple kinds of Mev right so usually Mev has a lot to do with swapping um so like you know on osmosis or you know on astroport or Windex or something like that and you can extract the movie by front running users who are swapping and doing

something called a sandwich attack which is generally uh generally quite extractive it's essentially just taking money out of out of the user's pocket that they otherwise wouldn't pay or you can just take advantage of natural Arbitrage that will happen no matter what you do and you can create a competitive market around that and distribute that Revenue back to validators and users and that's what our products do so I think it really did accomplish that first vision so that was the first I would say six months of skip since then we've gotten a lot more ambitious about what we're trying to do with Cosmos in terms of Mev yes but then also in terms of ux and cross chain ux um we've started to build some privacy

Solutions so for example skip secure was our first sort of fun project around this where any user can change their Kepler RPC address where they submit transactions to be one that's fully private so they can basically opt out of the Mev ecosystem because they're fully private no one can ever see their transaction until it ends up on chain um in the future we plan to do a lot more around cross-chain ux and I don't want to get into all of it right now I mean we can go through it piece by piece but that's really our our Focus as we continue to go forward that's really interesting if you just stop on that for a second so these rpcs

that you provide that are that are fully private can you explain how that works and like who's providing that RPC is it is it like a centralized RPC that you guys are running um essentially for the community yeah that's basically right um we right so so skip secure is basically you go to Kepler uh you go to the little settings menu you change your RPC to be the skip secure RPC for that Network um and then what will happen is if skip is available on that chain so basically if it's Juno evmos terra2 or injective we will take your transaction um we will put it into the same sort of like what we call the Sentinel or

basically this like piece of infrastructure that runs the auction um we'll put it at the end of the auction and then send it directly to the validator so normally when you submit a transaction it goes to like let's say Kepler's node and then it gets permissionlessly or just like widely broadcasted out to the network so anyone can see it that's what Searchers will use to conduct any kind of Mev activity they want but via this method because the Sentinel is private and the validator is private no one can ever see your transaction until it ends up on chain you essentially get a direct line to the validator um so it's like if you you know with your validator receive a transaction from a user and you just put it right into your block versus like if it went

through like 50 other different nodes before it ended there okay that's interesting I guess that has also implications on I mean so um what what's the Sentinel The Sentinel is our was our first product which was basically a uh you could call it like a I guess a Matchmaker or an auction house where a bunch of Searchers on one side would submit bids to be at the top of the block and then validators would receive transactions from The Sentinel and put them at the top of their block so it basically created a two-sided marketplace with validators on one side who were mining the the box and then Searchers on the other side who wanted

to be at the top and then the Searchers would compete and say okay you know I found this back run and I'm willing to pay fifty dollars for it and then somebody else would pay 75 and someone else to pay a hundred then the top winner would eventually get you know selected by the Sentinel and sent to the validator and we've seen those bundles go as high as multiple thousands of dollars in one bundle so I think rhino uh Rhino validator the other month they had a three thousand dollar or three thousand Juno bundle um that you know was mined and was a you know exciting thing for for them because I think they gave a lot of it to saving rhinos uh but but those those auctions can be can be very competitive and and

can be very lucrative as well okay well we'll come back to that um in in a couple of minutes but um yeah so I mean we talked a little bit about like good Mev and bad Mev and yeah I've heard other people characterize you know these different types of enemies uh Sunny Agarwal gave a talk at nebular Summit last year about uh about this very topic and so osmosis the strategy uh with regards to me the so can you um can you sort of you know explain what are the different types of me so you already mentioned sandwich attacks as as a quote-unquote bad type of Meb um what are the other types of Mev yeah there's actually very few types

that are very common and then there's what we call the long tail which is like a variety of one-off opportunities that Searchers will find and you know look for um the primary form of Mev so let's take the ethereum ecosystem where there is a lot of Mev right about a billion dollars a year um it is split basically between sandwiching and Arbitrage and then there's an additional bucket for I would say liquidations so sandwiching is the concept of basically forcing a user to pay their maximum slippage right so you say I'm willing to trade X for y and then

somebody comes in above you and buys a bunch of x to make it or sorry buys a bunch of Y to make it more expensive and then sells you back or sells the Y afterwards once you sort of did your trade so essentially what it has the effect of doing is artificially raising the price for you um so to the point that you max out your slippage and then selling at that higher price so we view that as generally bad or just purely extractive it would be fundamentally illegal if this were like a real Capital Market like in the US like that that's like a definitional market manipulation um the other half of the ethereum but but it makes a lot of money in ethereum

the other half is arbitrage which is basically okay you bought a lot of whatever and now there's like a price and balance between the pool that you bought it on and like a bunch of other pools on osmosis so there's like an Arbitrage that can take place this was happening before skip ever came into existence it was very prominent on osmosis it was about six million dollars last year um you can see satellite.skip.money if you want you know to see the full amount that was captured and so that's really what skip focuses on is we focus on creating more competitive markets around Arbitrage we don't create markets that allow for any kind of

sandwiching um we viewed that as like chains can enable it if they want to if they want that kind of Market but so far no chains have have decided to do so and so we try to you know stick to what app chains want from us so so what you're saying is that there's like essentially two types of maybe there's a sandwich attacks which are generally considered to be extractive this is the act of getting getting a user to pay is maximum amount of slippage uh or forcing a user to do so by intervening in that transaction and then on the other side uh you you have Arbitrage which has existed since

markets have been a thing feet and it was you know recognizing the opportunities and discrepancies and pricing between different markets and capitalizing on those opportunities and and here generally you know we would see that or at least it is characterized as non-extractive Mev uh but but more like you know on this good Mev um are there other types or this is primarily to so you must liquidations too um liquidations would fall in which bucket and um you know why why are those more I mean those are obviously more common on ethereum because there are more lending

markets on ethereum uh than than in osmosis currently or in Cosmos generally yeah so right so so I would I would categorize liquidations and then there's this other kind called Just in Time liquidity attack which it I I would both view as relatively benign as well um so like I would put them probably in the good Mev camp they're not as extractive as sandwich attacks but they also have some of their own problems so liquidations is the concept of um and and lending markets by the way like rely upon this kind of Mev to stay healthy which is basically somebody's loan went bad and there's some kind of uh you know reward for for liquidating

that loan to basically pay back the collateral and so what will happen is Mev Searchers will find those opportunities and then what we call atomically or all at once liquidate or borrow liquidate and then sort of like claim the reward as like a riskless profit um it it makes a lot of sense but it's also relatively small actually relative to the other first two types and then we might get into this more in Cosmos but we don't really have it today because we don't have like uniswap V3 clones yet except forges out and I know that you know osmosis is doing some stuff but just in time liquidity uh attacks are are basically

someone wants to execute a big trade and a concentrated liquidity pool and then a Searcher will see that intent and then provide a bunch of liquidity right at the band that they'll be executing over to basically grab all the swap fees from that and then prevent anyone else from getting like their proportional share so it kind of screws the other LPS but on the other hand it gives the trader a better execution price um so it's very different from sandwiching which totally destroys your execution price so I would call it like benign but also like it does screw a different party in the ecosystem which are LPS which I think are also important um okay that's interesting I mean I see that as just you know

markets being markets right I mean you have like Market participants that are able to provide a better price because they have some Advantage like having more liquidity and and so you know if that that exists then I just like yeah I would also see that it's like quite benign um yeah interesting okay uh so where does Skip fall in all this and like which types of Med does Skip and like typically you know companies like skip uh that are focused on providing Mev Solutions where does where where to skip fall and what types of me does the skip uh address is it mostly the sandwich attacks uh like this bad Mev or are you providing Ultra solutions that address the other types of maybe these like Arbitrage and you know these

just-in-time liquidity uh yeah so the good news is Cosmos is actually pretty resistant already to sandwich attacks um this was something that we noticed from the beginning right so one of the one of the most I think misunderstood aspects of Cosmos versus ethereum is that we don't have gas auctions or we don't have yeah we don't have gas ordering so when you go to Kepler right and then you select the high gas option or you add an even more gas that doesn't affect where in the block you're executed right so every the the the the transactions are ordered by the time that they're

submitted roughly so we call this a first come first serve mempool where you know if you submit after somebody else you would expect to be in the block after them so adding more gas does not put you higher up in the block it doesn't change the ordering within the block so what that means is it's very hard to front run someone right because if you sub submit and then I want to be like get in front of you I can't really do that because because of the the way that the blocks are structured because it's first come first serve it would be very hard to like restructure and get on top of you right um which basically means it's impossible to execute a sandwich attack and we've seen that and we verified that

repeatedly going back through the history of Osmosis where there would have been a lot of Sandwich attacks if it were possible and we saw like wow no one's really doing this right so we didn't want to introduce that as like a new concept or we didn't want to like be the party that introduced sandwich attacks we wanted to give instead a more competitive market around what was happening which was this latency based Arbitrage back running right where you know seps emits a big Swap and then I'm a Searcher and I want to like background it and do what we call cyclic Arbitrage afterwards and so that's what we really created markets around and all of our products basically do this like each one of our products is

is geared in some way of cap helping capture the revenue from those backgrounds and distribute them to a variety of users we've done it on Juno and evmos and Tara in the form of an off-chain auction we've done it on osmosis in protocol via proto-ref where the protocol itself will do the background and give it back to users and governance and then we've also done it now um by interfacing with gaps directly that create the swaps so like Astra Port um maybe in the future Windex and other places uh that will actually do it and then redistribute it back to the user so that's really what we focus on right now so yeah part of front running uh how

does this back running uh work and you know what what are say arbitragers um trying to capture with a with a back running Mev strategy yeah so like if you just take the example of of Osmosis um you know let's say there's five different pools that um you know all have banana token right and they are roughly priced they should be priced in dollar terms the same right so you imagine if in one pool banana token is priced at a dollar and the other one is priced at ten dollars what would you do you would buy a bunch of it at one dollar and then sell it at ten dollars

right you just make what we call Arbitrage or riskless profit right so at a baseline all the pools should be balanced right the price of banana should be the same across both but what happens is every time somebody comes in and trades on one of those pools then the price of banana there will change right if they buy a lot of banana the price will go up they sell a lot of banana the price will go down so either way there's going to be now some difference in that pool uh of banana token versus all the other pools that were not touched right so there's like basically money on the table it's like there's a twenty dollar bill on the table and before skip what would happen is all of these I don't know it's called

Searchers call them whatever you want Bots basically would search for these opportunities after people's trades and say oh that guy did a big trade let me now buy from that pool and sell into a pool that has a higher price right and basically the first person to grab the 20 bill off the table would own all of it right so what we do is we create competitive markets around getting the 20 bill so essentially we ask people to form a line and say okay how much are you willing to pay to capture that Arbitrage and then someone says well it's worth twenty dollars so I'll pay five dollars right then someone else will say okay well I'll pay 15 right it's worth twenty dollars so it's

still a profit and then eventually someone will pay well I'll pay 19.99 right to bot to to purchase this opportunity and that's what we've really seen happen is that now that we've created these competitive auctions people are paying 99 of the value of the Mev or the Arbitrage to capture it and that money goes back to the protocol right so now the protocol is recapturing the value of its own Mev and that's really the point of skip right is we help you create the markets neutral markets to make these people compete for these opportunities that you're creating and then you get the money back and can decide what to do

with it and that could go to validators could go to users could go to governance you know community pool um you go to rhinos whatever it is right like it doesn't matter to us it's it's mostly about giving change the ability and then the decision about how to capture and then where it goes right so arbitrages are paying up to 99 of the opportunity cost uh is this something you see consistently like like the that it you know it it pushes the boundaries of like sort of you know reaching close to 100 or or does it vary in what would you know what are some of the criteria that you've observed that would cause that to vary is it like you

know less people uh auctioning on that opportunity or are there like other uh criteria here yeah so I think it's a little bit of give and take so there's some things that are very obvious that create Arbitrage and the ones that are obvious and easy to understand and easy to um capture become very competitive very quickly so for example what we do on our end is we actually introduce these open source Bots called Skipper that allow for anybody to participate in Med by just with with very minimal programming knowledge they just go they download Skipper it's an open source repo

everything we do is open source pretty much and they just run it on their computer laptop doesn't require a lot of infrastructure and they'll automatically be doing these basic arbitrages between the pools and so the the opportunities that Skipper uh captures have become extremely competitive basically 100 right so you're it's it like it would be very hard now to just run Skipper as it is and make like a lot of money immediately in the beginning people were actually but now it's it's a lot more competitive what we've seen people do instead is you know the people who are I would say more I guess sophisticated are adding

additional pools and finding additional crossed protocol arbitrages and then starting to do that and then usually the way Mev works is you find some opportunity you capture money from it and then over time other people realize what's going on and then they start to get into the game and then it becomes more competitive right and that's the natural way and you know I think that's how you incentivize Innovation is and initially you encourage people because it's profitable because no one else is doing it and then other people come in and eventually just gets good for the ecosystem right um because you need Searchers right Searchers are important to find these opportunities you know they're dedicated on this stuff they have teams around it and you know they're the ones who are

who are being creative right um but eventually I think I think the competition gets very high and then the protocol is the one that benefits so so who are these arbitrages like who are the big players in Mev that you know most people have never heard about who's extracting the most value and making the most money here so different parties I think on Cosmos versus uh ethereum and ethereum there's there's some well-known mov shops so like Snoopy you may have seen on Twitter is one uh they do a lot of uh Arbitrage or like stable Point Arbitrage um there's also uh SCP is kind of an I I wouldn't call

them Infamous but they are a massive shop and they've kind of subsumed a lot of the other Mev shops and they do a lot of what we call sex to Dex Arbitrage which means they'll be arbitraging between finance and ethereum um in Cosmos we know some of them because they're on the Discord and we like them a lot I think they're a little more hobbyists right because there's not that much Mev that you know to them it's usually one one person who's just very exploratory and interested um I would say one of the best teams in this is white whale uh they have an incredible Bossa Nova is an incredible Searcher who can create these Arbitrage Bots and has open sourced them for the community I would I would highly recommend checking out their open source

bot um that uses skip but they've done a really good job and then what we've noticed is a lot of these Searchers are are people who made a lot of money doing this in Terra in the in the good old days and have now migrated to the rest of the ecosystem but we don't know who they are um most of the time these people are anonymous okay um so so you would it be fair to characterize the Meb the ecosystem in Cosmos as you know still fairly immature and and sort of in a growth phase or perhaps you know like the early days compared to other ecosystems like certainly ethereum

yeah I think the reason for that is none other than the fact that you know um Mev is highly tied to volume and volatility and we just don't have as much as ethereum right I hate to break it to you but like you know we're getting there uh but it's gonna take time and you know the volume on Juno or osmosis or you know any anything else in in in uh Cosmos is just does not rival ethereum Methods at this time I I really do believe that this is about to change yeah what what is the Catalyst that that causes that change in your opinion um I think it is a combination of

I think a lot of things have to change um so one thing is we need better custody Solutions because you know volume and liquidity is highly tied to institutions and we don't have institutional custody Solutions in Cosmos um fire blocks doesn't support most of these chains Anchorage doesn't support most of these chains and we need that right like institutions legally need that or or from a security standpoint require that to come in and so I'm excited to see the development of that I also think for a long time we haven't had native usdc um Native uscc is also important for institutions to trade and you know gives a level of credibility to the ecosystem

that I think is is is turns heads right in terms of where people can deploy their Capital so I think Noble is going to be a big thing and then lastly I think the the I just saw something we flipped in terms of Cosmos developers versus ethereum Developers that's not to be like looked over right like the the arrival of dydx the arrival of say the arrival of bear chain Archway Duality a bunch of these protocols they think will have a lot of will attract a lot of interest right and turn a lot of heads and I think especially if dydx does well that's an example of a eth based protocol right from starkware coming over and being successful in Cosmos I

think that's going to be a like a canary in the coal mine for a lot of these other protocols being like what have we deployed our own app chain right it worked well for dydx we should do it and so we have faith in in those things developing the right way so you said something interesting here you said that you saw somewhere that Cosmos just flipped ethereum in terms of its the number of its developers yeah like what a number of actors to help us they're they're you know they track these things as as one of the you know ecosystem metrics and right now we're the top ecosystem so you mean to tell me there are more developers building on Cosmos than than

salinity developers I find that really hard to believe I mean like I know the cosmos is growing and there are lots of developers but but yeah yeah dig up the metric um but you know and also you know it's possible that metric was biased I don't know but that's it seems to be that we were neck and neck for a while but less and you know I guess after a couple of recent events things have changed but I think it's it's in some ways it kind of makes sense to me because in Cosmos we're developing l once right all these app chains are full Sovereign l1s and you know the amount of capital that that can attract is enormous as we've seen right say raising an 800

million valuation there are changes announced they raised at a 420 million evaluation those are going to be fully fleshed out you know 10 to 40 person teams um that I think you know that there's a lot of excitement about you know what you can do as an L1 and I always go back to the example of you know the third largest blockchain ever was built on Cosmos right it was Tara I think we're gonna have that again like we have the stack and we've already proven that our technology can do that um it's just a question of finding an application that and an ecosystem that that can do that again and so you know we we've I guess since we were born into that time we you know

we see that in the future as well um this is also because mag counts as a thousand devs um maybe yeah they could be that maybe our devs are you know count as more adapts in the ethereum test um so let's let's talk about the product a little bit uh you know you mentioned earlier Sentinel so tell tell me about the different parts of the skip stack if you will and you know how that integrates with validators and who interacts with the different components of of the of the sac and

um yeah like you know I I as as as a validator I'm actually like Curious and interested in this uh but also maybe other people out there like interested in implementing a skip on their validator or maybe leveraging some of the stuff that skip has built to you know try to make some money um with uh with Mev yeah great question so I'll talk through our first product skip select which is available on Juno Atmos tarot 2 and objective so as I said before what we do is we try to create a neutral marketplace right where we don't assert any influence or control between Searchers on the one hand and

then validators on the other hand so you may not realize this as a validator but you have a lot of power because the way that tenderment works right or the consensus algorithm that underpins all Cosmos chains and more uh is it's a round robin system right so who creates the block well first it's Seb then it's imperator then it's Cosmo station then it Loops back around right and it has this round robin system when you are selected as a block builder in that system you have unilateral control over what transactions go into your block meaning if you wanted to put in whatever you could right if you wanted to reorder whatever you also could so

basically the validator has a huge amount of power in terms of how blocks are built right but on the other hand validators usually don't have and shouldn't have the level of sophistication to do the Mev capturing on their own blocks themselves right because they don't know where all the opportunities are they don't want to build all the different Bots required and all the different you know um I guess like observation tooling and liquidity management required to do that kind of Mev capturing so the the idea here is that validators have control over their blocks and that Searchers are the best ones to capture the Mev

so what we do is we create a competitive market for the Searchers that's the Sentinel so the Sentinel interfaces with the Searchers there's RPC calls that they do into the Sentinel and they say hey I want this transaction at the top and then a different Searcher says Hey I want this transaction at the top they battle it out and then basically 400 milliseconds before it's time to create the block we choose the winner and then we send it to the validator and say validator here is the winner of the highest paying Mev transaction and this is the amount that they want to pay please include it at the top of your block the validator will say okay great I'll include at the top of the block and thank you for the

payment right a tip you could call it um additionally validators since they have to register a connection over P2P gossip right we're just a note the central is just a note so you just all you have to do is peer with us but there's nothing else required um when they register they also tell us whenever you give us the tip send this much to delegators in the network as well right so there's a percentage split there the validator can say I want to keep 100 and don't give the network anything that's fine no problem or they can say I don't want anything I'm very uh altruistic give 100 back to the network and the delegators right they could also do that or they could do any

combination in between um and so that's basically that's basically like the structure of it right a validator would go to our site register for an API key that allows them to peer to The Sentinel then all they need to do is uh basically patch their version of tender mint with Mev tendermint or Mev comment uh that allows them to peer to us and then also to put whatever they receive from the Sentinel at the top of the block so it's like a small reordering of their blocks and then that's all they need to do it takes about five minutes is what we've seen um and then the Searchers are I mean I could go through that separately but

there's a whole different set of things that they need to do they need to set up um oftentimes they'll run Skipper and then they'll make some modifications sometimes they'll run their own nodes to find mempool transactions by the end of the day what they're doing is they're submitting bids and groups of transactions to our auction uh to be considered sort of in in that option um yeah so so the so the validators are the validators just like run software let me guess they they patch their tournament software um and and that allows them just to interface with with Sentinel and to essentially receive um bids from Searchers on the other side

of that market that are uh you know bidding for that block space that are hoping to capture some of um some of that Med um so it's fairly straightforward I feel about like validators like they like they just have to turn on skip to make this additional Revenue yep feels like because there's no other kind of assembly required uh so on on the Sentinels sorry on the Searcher side though so you mentioned this other piece of software uh Skippy um what what what's the utility of Skippy and and how our Searchers using it is it sort of like a framework or some sort of some sort of an SDK for identifying Mev transaction Meb and sort

of like submitting that to to the to the marketplace yeah that's exactly right um it it's it's as I said before it's it's a plug-and-play Searcher bot so you just download it right you grab it from the open source GitHub repo and then you press run um well you don't press run you type you know whatever the command is Right python main.pi whatever it is um and then what it will do depending on the network you're on is find Mev opportunities and then submit bundles to uh to the option to try to get included right um and it you know early days especially just made people money and what it had the effect of doing was opening the uh

was democratizing uh sort of the entry point into being an Mev Searcher so sometimes we saw valid editors being Meb Searchers we saw people who very little technical experience doing it and a lot of people like are excited about this stuff and they want to see how it works and get involved right it's like a little bit of a hobby project but what we've seen is some people really enjoyed it and then actually got much more serious about it and started to really add to it Upstream changes um and it just generally made we generally helped build out I think that ecosystem of Searchers to the point that now there's over 300 active Searchers every day um who are submitting bundles we get

over a hundred thousand bundles on some networks on some days um and then we manage that so that was really the purpose was to you know on all sides of the network make it more accessible and to not have it be anywhere sort of like behind closed doors right validator side all you need to do is run software um you don't maybe you don't understand how and maybe works all you know is you run the software and then money gets deposited into your account magically magically right and you know that you're not screwing users in the process that's like a big thing for validators right then on the Searcher side all you do is run this software and then you just make money magically right and you also know that you're not screwing I think those

people care a little less about this but you also know that you're not screwing users and that you can add to it to make more money if you want to get more into it okay so so like vanilla Skippy like anybody can run out it will search for me the opportunities but then you can you could sort of tweak it or or augment that software in order to look for other more complex Mev opportunities but I mean if everyone's running the same software and like that software where's the competition I mean they sort of have to uh be um you know like be looking for Unique I mean the

opportunities because otherwise like everybody would be getting the same ones right and then there would be no competition right yeah I mean the competition first comes from price so if everyone's running Skippy and everyone says there's a parameter it says I'm willing to pay 10 of the value of the Arbitrage in the form of a tip then the only per the person who wins is the person who puts 15 right and then the person who wins is the person but it's 20 right so very quickly goes to 100 um but if you wanted to find an opportunity where you can only bid where you can win at only bid 10 you have to go elsewhere right or you have to add to Skippy or it's actually called Skipper um and then basically you know find other pools yeah exactly

I mean I I confuse it with the peanut butter for a second um okay and so yeah talk about so I'd like to get a little bit into like this you know when when I when I heard Sunny talk about I mean the you know he and and other people as well is like capturing you know the the the the good Meb and delivering it back to the community Through pool or you know some other means like channeling those those revenues to a dow um at what point does that happen and

like how are protocols like today thinking about this and what are the most common ways in which Protocols are you know utilizing that that Mev uh yeah great question so we talked about the first product a lot that was the first six months of skip we've been focused on a completely different set of things since about four months ago so when we talked to Sunny and and Dev who had a big you know uh architectural influence on this stuff we the the original design of skip right this this off-chain auction with the Sentinel didn't work for osmosis

um it's not that it didn't work from a from an infrastructural standpoint or technological standpoint it didn't work because reality is you are placing a lot of trust in skip right we're running this software like in the Sentinel you don't really know what's going on right I can tell you that it's honest but you know like we're trying to build trustless systems and so we thought about well what does it look like to do this fully in protocol right where you don't need to trust the auctioneer right you don't need to trust that they're running a fair auction and that we're not taking taking ones out and putting our own in or something like that right you know that everything's being done right

because it's being done by all the validators in an open source way right and they can see the logic and so that's when we built this thing called protarev uh protarav is something that's built on osmosis it's specific to osmosis right now um where it's built as an additional module in the cosmos SDK stack that interfaces with the gam module or the amm module on osmosis like their native swapper um and it essentially will auto back run every transaction without using an external third-party or marketplace right um there was a lot of open questions about if this would work right we're like like if we put all this logic into

the validators is it going to really slow down block times it's going to be really computationally intensive is it going to require a lot of capital to do these arbitrages which are generally Capital intensive and we what we figured out now that's now that it's launched and been running for a month and made forty thousand dollars approximately for osmosis is no and it doesn't actually do any of those things that um there's ways of doing these things extremely computationally easily um and still very effectively you can actually since it's in a protocol you can mint the liquidity that it needs to do the Arbitrage in real time you just mint it out of thin air and then you burn it by the end and just keep the profit you call this Flash minting

um and so we're like there's so many cool properties of what you can do when you're actually inside the protocol that you can't do when you're outside in the Sentinel and so really the focus has been over the last couple months of being like okay how can we like generalize this thing that we've done for osmosis which I would actually call relatively groundbreaking to do Mev inside the protocol and have it be recaptured by the protocol how can we generalize that outside of Osmosis right and so that's that's really this product that we're building called protocol owned Builders or pob which I would General generalize as saying taking the Sentinel and putting it inside the protocol where validators can fully

verify the logic it's a fully open source and then it's it's really com it's really controlled by the protocol right um that's really the focus of what we've been doing which is basically putting these auctions inside the protocol allowing Searchers to still compete Searchers are great as I talked about before Searchers are basically the product of skip right like they're the ones who are doing the work um but to do so in a fully verifiable trustless way because it is something that is built into the logic of the protocol yeah that's interesting I mean I think this is one of the things that that really set apart you know osmosis is

Vision osmosis is the vision of osmosis and that of say atom 2.0 and the allocator [Applause] [Music] um and I remember when that whole debate was going on um sunny and other people in the osmosis Community saying you know block space auctions are are essentially are great but if you can just build Mev uh uh auctions right into the protocol is sort of eliminated for like this whole you know module at the um at the validator level and and also probably more more effective at battling

Med because because it's right in the right there in the protocol I you know at the time I felt that that was like a really interesting uh point of view and and uh you know did fully understand it then but I started to come around to to you know to understanding and and how valuable it is so talk a little bit about you guys have put out a research article on on your medium and you've also released this um Med satellite which I'll link to in the show notes um which uh which is sort of a data dashboard um that that shows the the Mev has been captured on osmosis can you talk a bit about the the you know the experience of having

launched this product on osmosis and what you've learned from that and and also you know the amount of Mev that's been extracted and what's been sort of yeah interesting from that from that perspective yeah uh so I generally I mean there's an account that the Mev goes to on osmosis I'll send it the private chat I don't know if you want to share it afterwards but this is the cap this is the module account of protarev so this is you know we're not touching this this is just running on all the validators on osmosis so if you're an osmosis validator you're running this code right now um and it is just built into the processing of transactions in the osmosis State machine to back run and then capture Mev

um it's captured about I want to say about 50 000 it says forty thousand here but there was a profit split with skip to build this out which was 20 decreasing by fifty percent every year um and uh yeah I mean we've been pretty excited about the performance of it it's it's capturing 80 of the 80 to 90 percent of this Mev on osmosis by and introducing basically no overhead to the validators like I don't think any validator should have noticed any increase in sort of their requirements since the sodium upgrade um there it is right and it doesn't only just capture in osmosis it captures an atom that captures an axle usdc and we

could add additional tokens you know with government governance wants to um to to capture you know other other forms of of capital um I'd say The Learning Experience here was just like we sort of went out on a whim to build this protarev module we didn't know if it would work honestly um we thought it was very all the existing knowledge right on ethereum has put this in an off-chain system right it's too computationally intense it's going to overload the validators right and this was really kind of bucking that and saying well it like let's just build it inside the protocol and see how it does and so far we've been very excited about how it's done um and I think the osmosis team has also been

very excited um so the learning there is that in protocol makes a lot of sense we want to move that direction over time we don't want people to trust skip ever except in the form of Open Source Code that they can verify and that that's where we that's where we want to go and in some ways it's kind of a little bit scary to make that jump because you know our original business model was like well if every If people really like this off chain system maybe we can charge for it as an infrastructure business you know and you know people would pay for it and it made a lot of sense because we had our own infrastructure now we're just developing open source code that anyone can fork and copy and whatever right um and that's a little bit scary to us

but you know what we've learned and what we've seen in crypto is if you just continue building the right thing that works for protocols you'll find a way I think I hope I pray that you know protocols and foundations will reward you for that um and you know right now we're trying to negotiate with foundations to continue to be a partner in that and recapture and I think the nice thing about working with skip is it is a value creative thing right like we're we're recapturing for the protocol value that was otherwise going away right and so is the other model is the previous model dead then is is like you know Mev auctions as they existed previously and as I guess they still exist in ethereum uh are are they going

are they just going to become obsolete or sort of phase out of existence as protocols uh Implement Mev at the protocol level because of all the advantages that it has um it's a great question actually so we don't plan on deprecating um the sensible anytime soon because a lot of validators still use it and we want to maintain that and by the way it's all been free as well we've never charged for our products um so I think we we plan on maintaining the Sentinel on chains as long as they don't have POV so I think once we once pob gets to the point where chains really

like it and it's been proven and they really want it um we will start to transition validate or transition chains away from the Sentinel and then into pob which is this in protocol basically equivalent right um but we're still a couple months away from that and if a chain for some reason does not want pob I think it would just be a discussion right saying okay we can continue to maintain this for you no problem um but you know would want to understand why like I think it would be an interesting question why they want the off chain system more um but uh I think if we had it our way or if we if right now our thesis is that the end protocol will win out

um because it has these properties as I was talking about in terms of capital efficiency and and also in terms of you just don't have to rely on skip for like off-chain infrastructure right it runs on stuff that has to run anyway for validators so I think over time that will be that will be the winning solution so when you were saying earlier like that you know essentially in protocol the protocol can mint this liquidity and then uh and then the protocol is doing its own flashlight they're really meant to me right like flash loans and other ecosystems it's just liquidity that's sitting in like let's say Ave or yeah compound but here they're they're literally creating it

and then burning it it's like a transient aberration of liquidity used for an Arbitrage and then it's got and it doesn't affect the network or token amounts at all hmm interesting what's crazy yeah at this point is osmosis can mint not only osmo they can mint Adam they can mint usdc they can mint anything right because these are IBC assets and as a protocol you can amend any IVC asset so they you know commonly right like in proterab it will be minting atom and then burning atom or minting usdc and burning usdc so that you can actually do Arbitrage in any token not just your native one

all right okay that's really cool yeah walk me through this this uh this platform so this is med satellite uh and it's sort of data dashboard that uh displays the amount of energy extracted on ethereum or sorry on osmosis so what what are we looking at here yeah so what you're looking at here is what we call cyclical Arbitrage so these are transactions on osmosis that start in one token let's call it X and then do a bunch of swaps through a bunch of different other pools to end up with more than they started with in X right so let's say you started with 100 osmo you do a bunch of these swaps and by the end of that one transaction because all the swaps are in one transaction you end

up with more osmo magically right almost so you're looking here I mean if you if you look at the big spike that is can you guess what that is uh me yeah that was Tara yeah that was Terror exactly so what was happening there was there was a huge amount of volatility on some of the pools mainly the ones that had UST and Luna and so there was a lot of crazy price differences that were going on and a lot of Arbitrage required to make the prices equivalent which meant a lot of Mev right um so uh that's what you're seeing here are basically the amount of profit being captured in riskless Arbitrage uh via

these cyclical cyclical Arbitrage transactions and so over time we expect this to basically Flatline right because there should be no more of this because proterab is now doing it for the protocol and recapturing all this Revenue back for osmosis and its community and the distribution of those funds is going to be decided there's already a governance proposal that was passed I think um in terms of how that will be distributed between validators and users and community pool and whatever else um yeah what was is the was the void that usdt sorry usdt created it it uh as a sort of lack of a stable coin after this is that what what was

that contribute was that a contributing fact to this sort of flatlining compared to like previously how things were sort of you know Med extracted were sort of you know continuously growing like it was the absence of a stable coin to uh the to sort of to blame for for a consequence of of this uh you know flatlining of the Mev uh was that something else I honestly think the flatlining of the Mev is a consequence of the decreased liquidity in Cosmos and on osmosis as well post narrow right I think I mean yeah what what Tara did for Cosmos was just bring a bunch of liquidity and a lot of excitement to protocols and I think

since then it has decreased right and we're in a bear market so what you're seeing here excuse me is basically a consequence of lowered volume and volatility um there have been spikes there have been days for example the usdc collapse where there was a lot of like there was a big spike I think there was ten thousand dollars of Mev captured on one day which in today's market is a lot on osmosis um and I think we'll see that again one thing I'm really excited about for example is the the liquidity that moves over from ethereum to be in dydx which has on some days more volume than ethereum itself um is going to probably flow through

osmosis right and so I think uh I think a lot of I think this is going to grow rapidly is is my sense yeah yeah I think that uh dydx launching as a cosmos chain and usdc launching as well is just going to bring like a ton of liquidity into the space I think like we're about to we're about to have some some some nice um some happy times in the cosmos ecosystem I think I have to say to say it blindly uh yeah so what you you were you were talking earlier or we mentioned this a little bit before um that you guys have been working also on IBC unwinding and in this packet

forwarding middleware so just as a recap so IBC on whining is the um the act of unwinding IBC transactions back through the original chain in order to uh to get to do an IBC um sort of transfer uh rather than going straight from like one chain to another and the reason why you want to do that is to eliminate uh you know uh you want to keep fungibility and aspects and so this is particularly useful in the case of USD C and cable will be that that's I mean business model will be like that these

transactions will fall back through the chain uh and so you're working with uh strange live on implementing this how does that play into any V but I I hadn't considered that uh that uh unwinding transactions was uh for a part of that yeah I mean this is part of a broader um so like one thing to understand about Skip I suppose is we don't have a chain right we don't have a token um there's no there's no skip token right we are I think laser focused in General on growing Cosmos because we integrate with all of these Cosmos chains as our customers and as our partners and we are highly incentivized

to see them bro right on on all parts of our roadmap um it's it's fully dedicated to trying to grow Cosmos um and we have I think a team now with Sam Hart as well who has identified a lot of the big problems between Cosmos today and beating the ethereum ecosystem and our Ambitions are to try to move in that direction as fast as possible so to answer your question what is IBC and winding up doing Mev it has nothing to do with that there's no there's no relationship right like we we don't plan to charge for that service uh and we don't plan to somehow have it be part of the Mev extraction story or I mean

prevention story what we realized was wow this is a pretty ux right when you send your tokens from chain a to chain B and now they're unusable and lost because you you forgot to send it first back to a different chain then to a different chain and then to chain B that's just like I think that's a problem that I faced before is one thing um and I think it's a problem that a lot of new users will face and so our thinking there was well like this is a big impediment in our minds between Cosmos now and the cosmos that we want to see in terms of growth we should just contribute to this and make it a public good

um so that's really what we've been doing is we've been you know really uh really pushing um in terms of pack and forward middleware adoption and really uh and we've already made a spec and implemented a design that would allow for this to happen as an API and talking to wallets about integrating it um but really we really just believe that it's something that needed to happen yesterday um and we didn't really see it as much movement as we wanted to see and so we were trying to push it Forward um I think it's actually every one thing that we talk about a lot internally is a big impediment to Cosmos growing is that there's not enough ownership over the cosmos ecosystem from any one party

that we need to see right there's a lot of chains that focus on their own development many exceptions to this I must say but there's a lot of chance of focus on their own development and their own tokens and their own develop you know their own developer ecosystem but they don't focus on Cosmos right we need more people to take advantage to take ownership over the cross chain story right and the cross-chained ux experience and so we figured you know if we want to encourage people to do this we should start right and so this is our start is trying to figure out sort of how to do that that solve this ux problem um ultimately I guess you could say it is going to Advantage us if it works out because we think it will lead to just a better ux and more volume and more

activity which means you know more Mev and and more you know value of skip overall um but you know I think that's that's the long game yeah very cool yeah I'm really looking forward to to this module being uh sort of implemented across um all the cosmos change so that that this this problem right so there's this impediment as you put it um is finally solved um so let's let's talk a little bit as we wrap up here the the future of Med and Cosmos I mean we've already discussed a little bit you know the the prospects of Mev um sort of being internalized in

protocol and like Osmos is sort of leading the way on that what are some other things that you think are are coming in the future that people should look out for and and I guess that are sort of specific to Cosmos because you know there are there are many things here that I think especially in the case of uh you know cross-chain transactions and as we Implement IBC or um mechanisms that allow IBC uh tokens to move across ecosystems I think a lot of these things are going to play into you know the the sort of Med story what are your most uh interested in and what gets you most excited yeah

um so I think I I think there's a there there's a uh like a a combination of in in Mev there's there's two sides of of the market that we talk about one is the block space side and one is the order Flow side right um meaning the people who have access or control over the nbb marketers the block space is the validators right they control their own block space and what goes in it and then the the users are the ones who actually create the orders and the transactions that create Mev and you know create activity um and so we focus a lot on the Block space side I think what you can expect to see from

skip is a lot more on the order Flow side uh so a lot more tools and Integrations with some of your favorite protocols um that are already in the works and some already pass through governance to integrate at the user side so that users can actually Define if they want some Med or if they want a part of the MVD they're creating right so ultimately what people don't realize I think is users have all the power here right validators yes have power but users also have power because they're the ones who are creating the orders that create the mbv so I think over time what you'll expect to see is a as a market or like an agreement between the two parties of how nav is distributed and we want to

make sure we're building solutions that enable for so that that communication and trade-off um and you know have Integrations on all sides of the market I think long term what you can expect to see from skip is some more experiments calling them experiments because this stuff is very tricky of uh what it looks like to start to explore the cross-chain mov space so what does that mean like in a basic form that means what does it look like to do a cross chain Arbitrage right where I'm doing a swap on osmosis and a swap on Juno and I'm capturing some value between them you know those chains are connected over IDC but otherwise they're fully asynchronous Sovereign l1s there's

there's some stuff that we're going to be putting out that's an exploration of how you could do that right so maybe some Alpha here is uh you know Sam Hart was one of the original designers of the scheduler which is an atom 2.0 which was an idea to sort of Leverage the concept of block space Futures to schedule blocks together so that you could do cross chain Arbitrage he is now at Skip and we are talking about this idea a lot and you know may have um we had some problems with the initial design but we may have some improvements that we can share um and I think I just noticed that earlier I I referenced the scheduler when I I said allocator but I was actually meeting the scheduler but yeah

go on oh yeah yeah all good um and then I think also what we think very deeply about are how do I so if what you care about is sort of like this cross chain flow over IBC relayers play an important part in that story and what can you do to have relayers be incentivized and also capture a part of the med right and we all know that relayers need incentivization right now um and so we're we're dedicating some time and and we have a dedicated team to exploring this so I think basically overall what you can expect to see are more players in the Mev ecosystem that skip can create Integrations for where sort

of this can be shared in a more broad way between different pieces of essential infrastructure like relayers but then also with users um for a lot of that to be open sourced and discussed openly as we have done in the past and uh you know eventually some idea of how to tie these things together and how it's sort of overall benefits cross-chain ux um I think that's that's skip end game right is is some sort of unifying mechanism of all these things and what about crosstain me the across ecosystem so crossing the boundaries of Cosmos and ethereum and you know as things like polymer come online and other cross ecosystem IVC

implementations do you think that there will be a big opportunity there and uh you know is is the market uh sort of prepared for that and yeah I do think there will be an opportunity there um is the market prepared for that I think uh yeah I think I think in terms of like what it will look like to have those IVC connections I think you'll see a lot of liquidity flow I think you'll see a lot of price discrepancies of assets on different ecosystems um and I think over time you'll see tooling developed to make those price discrepancies lower right and to give like execution capabilities for people doing Arbitrage there um and I think you'll also see

incentivization of the relayers who are doing the information passing between them um but I think that is that is going to take some time um and it's also going to probably going to require a little bit of a restructuring about how we think about relayers in in the IVC World um there's actually some really good research here if anybody's interested in the uh ethereum ecosystem about how they they incentivize relayers for cross-roll-up communication um I think some of those ideas seem applicable to Cosmos okay yeah I'm really looking forward to seeing how this plays out I mean I think the the cross chain EV and specifically like you know the crust ecosystem of EV

um space is going to be really interesting with a a whole lot of sort of innovation I think about you know how like how to capture that Med but also protocols you know uh leveraging that Mev to the best of their abilities to you know flow value back to users which I think is like the best way that we can um leverage Mev um yeah Mike thanks for coming on the podcast and uh you know giving us a a high level uh overview of honeyde but also a deep dive on on uh on a lot of these topics uh certainly help you know broaden my understanding and um hopefully we can get you back on again at some point to getting a little bit deeper on on some of these specific issues and also as you

know as the space evolves I'm happy to get you back on at some point uh to get an update yeah thank you for having me and to our listeners thanks again for tuning in I remember that we go live every week on Thursdays so if you're interested uh in this content if you like the content we're using uh please consider subscribing and uh also delegate to our valuers we're on Atmos Quicksilver at osmosis and you know what probably we'll be integrating uh skip at some point it seems really interesting thanks again