CROSS-CHAIN LIQUIDITY with Jim Chang of Catalyst AMM

Transcript

foreign co-founder of catalyst it's a fully permissionless protocol to bridge liquidity across all chains and roll apps in today's conversation we'll discuss cross chain liquidity pools Catalyst design and security model we'll talk about trust minimized Bridges and Roll-Ups as a service Mev and impermanent loss in the context of cross chain their Integrations with Neutron eclipse and other chains and so much more I'm also dying to find out why he thinks crypto is just a bunch of theoretical research and won't be taken seriously until real products are built

before I get started make sure to hit the like button hit the notification Bell And subscribe to get notified when I go live every week and remember that none of what we discuss here on the interop is investment advice and if you enjoyed this content please consider sticking with us we're validating on Atmos Quicksilver osmosis and Juno just look for interop invalier set and I hope you've booked your flights to Paris for this summer for nebular Summit it's happening on July 24th and 25th it's two days of technical talks about Cosmos IBC The Interchange and so much more no [ __ ] panels no suits just Tech talks and workshops the second batch of Labor tickets just went up go to nebula.builders for more my guest Jim Chang is coming up next right here on the interop

ER [Music] hey Jim hey thanks for having me on hey good to have you yeah well I'm really excited to be talking about Catalyst today and Diving deep into yeah rollups as a service trusts minimize bridging and cross chain liquidity pools it's a topic that I find really interesting and I think is really sort of apropos currently with you know the ecosystem of Roll-Ups as a service that we see sort of exploding across the entire interchain and on uh on ethereum and Beyond so you know before we get started

I know you've got a pretty extensive background in crypto you worked at Ave prior to working on catalyst so you know tell the listeners a little bit about what you've been up to the last couple years and what has occupied your time uh in and research yeah yeah happy given intro um so uh well again thanks for having me on I'm super excited to share some stuff about Catalyst that we haven't really uh been talking about yet in terms of more details in the protocol uh and and how we kind of see the the vision of catalyst unfolding um my background you kind of touched upon it uh briefly but um I I I've been in crypto for a few years now uh coming up on on six six years now I

believe um and uh yeah my my my passion has always been around um kind of the intersection of uh math and finance and and kind of building distributed systems and so um my background is I I have an undergrad uh undergrad degree in math within Northwestern in Chicago um and I've always kind of been interested in kind of that intersection and so out of college I uh was a part of a big management consulting firm um out in the Bay Area and essentially you know got exposed to crypto um pretty broadly in kind of that sphere right 2017 Bay Area a lot of Ico Madness

a lot of just general excitement about the technology so uh kind of focused my most of my time as a researcher at that firm um kind of working towards those those crypto problems um and then that was kind of short-lived I would say in the sense that I really kind of got the taste of what the magic of crypto was and the beauty behind it and so I decided to to be more of an operator slash Builder and so I joined Ripple as a product manager uh and was there for two years kind of building on the core xrp Ledger work um and uh kind of seeing what it meant to build kind of a monolithic L1 um and then device summer happened I

kind of you know my eyes were open to the the power of uh you know Lego blocks uh for for money and smart contract platforms and so I left Ripple to actually just be kind of a djend yield Farm Summer full time for for bits and then I actually joined an nft startup caught on softball domains whereas the product manager there and also the head of their Dev relations team um and then you know nft land was was crazy I think you know heading the jpeg summer that was definitely a good move uh in terms of seeing where the where the puck was kind of going in the space but I wanted to go back into default and want to go back into my research roots and so I joined Ave as the head of new products there and so I did a lot of

experiments a lot of internal research as it pertains to a lot of infra work like zero knowledge proofs and app Roll-Ups and interoperability and massage and so that was kind of my foundation and my you know kind of solidification of that this is something that I want to do 100 of the time and so I left Ave a little bit over six months ago and and started Catalyst with my co-founder Alexander what got you interested in in this well this emerging field of cross chain Bridges and Roll-Ups as a service yeah I would say um I think what really set me off was uh in this journey really uh but also

emotionally uh was uh was you know this this famous Reddit post uh by vitalik um you know where he said that the future was uh multi-chain but uh not necessarily cross chain um or maybe maybe he said the other way around but basically the the Crux of it was that like you know economic activity would be kind of um you know maintained and retained rather in kind of zones of sovereignty I think it's what he called them right like different ecosystems that had basically a shared base layer uh and so ethereum is Roll-Ups is one of them you know uh Cosmos and and uh the app chains connected by the interchain is kind of another one but there would never really

be you know uh too much economic activity that is moving between them right because you kind of have all these trust assumptions and uh I think when I saw that that post um I was definitely uh kind of like the contrarian in me wanted to prove him wrong which is very hard to do uh but the IBC Maxis got up and they were like what man yeah yeah totally I mean I think for all intents and purposes it was like a pretty uncontroversial post right it was like Hey like you can move things across different chains but it's just going to be not super feasible to do uh through like different zones of sovereignty through different monolithic blockchains so um that kind of like got me really interested in it on top of the fact that you know Ave V3 when I was at Ave came

out with this feature called portals which enables uh you know cross chain lending uh and I spent a lot of time kind of thinking about portals thinking about other pieces of catalyst how you kind of enable at least for a split second like some form of under collateralized lending um and so you know it was a matter of like interest for me personally but also like I I applied that thinking a lot in my day-to-day at Aven so that was kind of like um you know kind of a nice flywheel so to speak for me to get really deep into the rabbit hole for this specific uh segment so I mean I I think you you've spent a lot of time doing research and you certainly do a lot of writing on Twitter and medium and um and you know I'm curious why you

think that all of all of crypto is mostly just theoretical uh and that there's not much I mean how much of crypto do you think is just theoretical and and how much of it is actual real usage going on yeah definitely um I think you're alluding to my my hot take is that right or yeah so uh maybe not so much of a hot take I think it depends on who you ask but I definitely think a lot of crypto is um a lot of thinking uh and and myself you know I'm I'm definitely um someone that is a fault of that but a lot of thinking a lot of research which

I think is really awesome right a lot of the theory behind uh what we're doing from a mechanism design perspective is super important um I think there's a lot of kind of focus on uh research and not so much Focus but a lot of um incentive structures to continue researching more things because it's kind of what's being rewarded right it's like if you have a brand new like Dex Mech design or if you have a brand new tokenomics designer or gas market design or um you know brand new folding scheme for your for your zko processor right like that's the stuff that gets a lot of eyeballs and attention that's the stuff that gets the likes to retweets that's

the stuff that gets the Venture money and so from an incentive perspective you're bringing a lot of people that are have research backgrounds are clearly very smart um but their their incentive is to post papers and say hey like look at this insane New Concept that I'm doing right and everyone's like yes absolutely um I'm not not to say that yeah exactly um not to say that they're not good ideas uh not to say that they're not executing against those ideas but I think you get into the point where like there's a lot of focus on these like novel ideas and maybe not enough focus on kind of the nitty-gritty right which is like okay how do we onboard users how do we streamline this this user

experience how do we you know create a sustainable incentive structures uh to attract liquidity or to attract stake for security purposes right and so um like that is that is something that I think it needs to kind of shift before we kind of reach or position ourselves right for the next Bull cycle or the next wave of of adoption uh from the mainstream audiences because uh that's the stuff people care about right and quite frankly like I applaud people that are you know in the weeds kind of doing those things they're doing you know like account abstraction as a form of like solving the user experience or they're doing um you know like data indexing as a form of like much easier data analytics and

visibility into what's happening on chain right because that that like helps users uh or people who are even billing actual applications like Argus is a an example that I love right they're building very very intense infra um you know they're building you know app chains on Cosmos but um it's for the purpose of building games right and they want to build games that are fun for people on onboard kind of the next wave of folks and so those are those are the things that I applaud right um more so than kind of this this new novel paper that people are buzzing about yeah I mean are you familiar with this uh this post by um uh Danny Grant and Nick Grossman from 2018 the myth of the infrastructure phase

uh I'm not familiar with the post but I can kind of glean what the yeah it is yeah I mean the thesis is that infrastructure and application phases sort of you know follow each other in lockstep and um that they feed into each other and and I think that right now we're definitely in a phase that favors infrastructure building and that that will inspire new applications in in the next cycle and uh you know we've seen this at like three or four times already since Bitcoin was invented right so even even Bitcoin came before there was a Bitcoin Network so the the

um the application came before the the the infrastructure and I think that right now we're sort of in this phase where a lot of infrastructure is getting built and um but another another thing I think that bears thinking about when when thinking about like the usage of crypto is I think we might look back on on this in 10 to 20 years from now and and actually um and actually think that it was useful like when we're in uh when we're sort of in the middle of it uh it might seem like there's not actual much there's not actually much practical things going on but it's all part of the process of moving to the next step and and so I

look back on the early days of the internet and back in those days or like late 90s early 2000s at that time it didn't feel like movie blogging was a very useful thing or like early social networks were a very useful thing to the masses but they inspired a whole new way of communicating spreading information you know connecting with people that was necessary in order to get to like the next step uh which was more broader adoption of applications so I think that yeah with crypto it's sort of similar where you know we might like look at you know all of defy as like a massive casino but it's necessary to get

to the to to The Next Step which is to have like real assets on chain and more user-centric applications wouldn't you agree yeah I mean I I would agree with that um I mean I think I think it's always convenient to look at you know something where you know um when you when you look back at it um you know from like an ex-post retroactive lens you're like wow yeah like that was helpful right um but I think it's almost too convenient for us to say like oh like I think when we look back it'll be worth it right there's kind of like a a sense of optimism with it which I'm okay with right I'm okay with kind of that that

level of optimism I think I think you need to have it if you're if you're building a space and if you've been been building the space for a few years now um I think what I'm trying to say is like that we need to celebrate and we need to incentivize more people that are trying to build towards what we want to be optimistic about right building towards these applications or maybe not even building towards applications but building with like an end goal in mind that is more than just like infra for infrastake um you know like I think I I see some of these projects that are kind of like super serving like the super users of like certain things right

it's like oh this D5 protocol yeah like we're gonna double leverage that with this crazy like staking mechanism so no one can sell and people are like yeah I want that right and you're like hmm like maybe like 20 people want that right or it's like you look at like ZK Roll-Ups and they're like well this is a Zeke this is a prover that is more optimized for this ZK roll-up that and and everyone's like yes like we want that you're like well maybe like two teams want that and they don't have users right so like things again I go back to Argus which I love where it's like they're they're kind of like on both sides I think people get really bored when you think about like building

applications or like yeah like wallets right like you could see an nft on a wallet or something um but Argus is really great because like we're building games people love games it's true I love I love games personally but underneath it they're still innovating at the infra layer right like they're building these like sharded like game like game chains that like all feed into some evm bass chain and using Polaris and they're on Cosmos and they have IBC like those are things that people get excited about but like I want to applaud that because they actually have like an end goal in mind right yeah yeah that makes sense um so you you wrote this piece a couple of months ago about Roll-Ups as a service and that whole ecosystem

could you describe what that ecosystem looks right now like what that ecosystem looks like right now because I think there are a lot of people and certainly to me before I saw like this really nicely laid out graph that like has all the categories it was also a little bit hazy yeah I mean if I had one takeaway that I want people to kind of take away from this like crazy like mental map that's like categorizing these like 30 projects into five different buckets it's that um well I think there's two takeaways one takeaway is that like this modular slash roll up Centric future is is here to stay

and what are the implications of that right what are the implications of people literally clicking a button and having a blockchain just like spun up for them uh trivially and maybe even from a no code fashion um so I think that's like the big takeaway is like this is happening and you can see the Project's doing it you can see the excitement you can see the Venture dollars backing of this right the second takeaway is that it's a very confusing and crowded field right like 30 projects all kind of laser focused to do the same thing and there's alliances based on ecosystem or based on values or what have you but the fact of the matter is like

um you know there's there needs to be some specialization or some I'll say it maybe like consolidation that needs to happen or else you know if you're if you're this like you know application developer that everyone you know loves to talk about I if you're an application developer please ping me I don't see too many too many of those in the in the bear um but if they're going and they're saying yeah I wanna I wanna build a blockchain right I want to build blockchain for for nft mints or something like that right for my collection uh having 30 options is like untenable right um so that's basically like the two takeaways that that I'm seeing but you know if you want kind of a more um like a more direct answer to to you

to your question of me walking through the ecosystem I think broadly it comes down to like ecosystem alignment right and I think like that's kind of how I grouped it um I think there's a couple ways you can slice it one is like okay like do you believe in you know like full modularity right so it's like you could build an execution layer you can you can build a DA layer you can point to different DNA layers you can have a settlement layer you cannot have a settlement layer blah blah or or are you more like of the monolithic sense right and so like Dimensions like no we're just building app chains and they're all kind of generally connected to one another using the dimension Hub through IBC right um so it's like one piece another piece is like no code versus code so it's like

okay are you an SDK kind of like the cosmos SDK we can like add in a bunch of modules and that's how you build it or you like a no code right you're like uh oh like here's a GUI and I just like you know like check all the boxes of the things that I want and then I make it from there um but then again like like I said I think it always goes back to ecosystem I think you can find similarities of thoughts and values on the cosmos ecosystem around sovereignty versus an ethereum ecosystem which is around you know quasi-modularity uh and then if you go to the you know Celestia Celestia is like full full modular like pick and choose like any anything ever right like more more and more modular layers are differentiating themselves in the stack

every day but um I think like but broadly that's kind of like a somewhat helpful map of a very confusing picture yeah one one thing that comes to mind here and you you sort of touched on this when you were mentioning you know the the choice that Developers uh the the choice set that developers have when picking a platform you know it also feels like I mean I hope that there's going to be sufficient demand for all these different um security layers such that they will remain secure so you know for things like ethereum I think

that there will be sufficient demand there but there's there's so many other um uh data availability layers and and uh and sort of security layers that are underpinning all these Roll-Ups as a service will all of those survive and then you know the next question beyond that is those that don't or those that don't have sufficient security all the applications that are built on top of those will have to migrate or move somewhere else you know if if if that doesn't play out um because it feels like the industry is moving towards decentralizing even State machines right so we have like a broad choice of State Machines of of uh of consensus and datability and settlement

but what is the cost of security I think will is something that's going to play over the next couple of you know years yeah I I think there's like an existential um like problem that's kind of like looming over the space where it's like okay like number doesn't go up forever at least especially as it pertains to like security you know like you know also for liquidity but let's let's look at security um for for now and so if our mechanism for security is literally proof of stake so it's like you know how much numbered you have that you know can incentivize people to verify and finalize blocks um you know without kind of uh without

violating the protocol or being malicious um there's going to come a point where it's like okay like we're kind of at this phase where finite amount of money that's staked in securing these systems and a growing number of applications and value being moved on top of it and so how do you mitigate that right and so I think that's where like shared security becomes a really like uh prescience thing right uh and and people are kind of tackling in different ways right we of course have replicated security with with ICS in the hub uh we have mesh security uh that I know sunny and the team think a lot about uh kind of more this this uh Democratic nato-like approach uh we have

eigenlayer right where they're kind of restaking the ethereum um security we have Babylon chain that thinks a lot about leveraging checkpointing on on bitcoin as a form of security and so we essentially like their their forms of like Leverage essentially which I actually don't think is a bad thing right I think with leverage comes Capital efficiency and that's how literally traditional markets work right now um and and I I find that a very fascinating topic uh and I think it's going to be a big lever we need to pull in order to overcome what you just discussed absolutely all right let's let's talk about Catalyst uh because we could go on for this sort of high level stuff for for hours what what is catalyst and what is the vision for the product

so catalyst is uh a universal cross-chain liquidity layer I think the motive for Catalyst was born from um kind of the acknowledgment of of what I was talking about before right it was like that we're entering this world of a proliferation of chains and in in that kind of end state of millions of chains um everything that we know breaks right uh everything like composability of of dapps um you know all the infrastructure that we have as it pertains to Wallace as it

pertains to RPC endpoints as it pertains to data indexing so on so forth that breaks um you know managing multi-chain deployments breaks governance breaks so on so forth right I don't need to belabor the point um and you know recognizing that um we decided that the core issue with a lot of these kind of infrastructure or these kind of dapps is that they've taken kind of like a copy and paste approach right or they're like okay like we did some work uh for ethereum or or for a Cosmos app chain oh there's more of them now I I guess

we'll like redo that work and just deploy there right or or stand up um kind of infrastructure there um yeah in a way well I'm talking about more of like like a sushi approach where it's like right yeah like let's let's just kind of settle into these different places yeah um and maybe there's some learnings along the way right maybe they're like oh like we can probably streamline our config and make it much quicker or or we can probably reuse some components but it's still manual work right um I think it's same to be said for like Alchemy when they expand to different chains right um and so our thought process was okay what how do we make something that's like automated

right as there's new chains it is trivial basically to be on those chains right it's not redeployed it's not uh you know managing config is on a governance Vote or anything it's just it's just there right and in that design space we thought the most important thing was liquidity right um and so that's kind of like our Flagship focus is very much like liquidity that meets you where you are as easily as as kind of we see the future unfolding right and so so differently you know we're all the millions of chains

catalyst is able to be in all those chains automatically and they're built in a way in which you can facilitate swaps with any chain that catalyst is on right and so it's kind of like this network effect of like okay there's more Roll-Ups being kind of propagated Catalyst will be there automatically and we can talk more about how that's the case uh and the way that catalyst is is kind of constructed is that you know you have assets that live on roll up a roll-up a wants to talk to app chain B um there's a catalyst contract there and you don't really think too much more about it it you can just trade between them you can swap you can you can move value and so we very much believe in the

concept of like orderless blockchain you know I think the analogy of like countries and borders for for blockchains is pretty opt and so we kind of want to dissolve those borders and make it incredibly easy to connect them together um and kind of like this motive came from you know my time at Ave where I was thinking about what what doesn't an app roll up look like for uh for the products and uh under the Ave kind of umbrella and uh one of the big things that was kind of holding me back was like okay great like we have a new roll-up um or we have a new app chain and uh you know it's hyper optimized to do all the things that we need to do right uh and capture Mev it has you know optimized

gas prices for certain op codes uh has like our or no you know a native gas token that is beneficial for for the protocol um well how do we connect to uh our deployment on polygon or our deployment ethereum our appointment at Avalanche and every team that we spoke to was like oh I don't know you know maybe you can do a multi-sig and we're like like we've seen how that goes out um and so yeah like we built it in a way that that can help those teams now right now you can now you can spin a roll up and you're like oh like my roll-up spun up now what oh yeah now I can actually trade from anywhere because of catalyst and so that's kind of uh the motivation and kind of the tldr of the product

so so you you you mentioned that Catalyst would kind of be there automatically and it would be fairly trivial to uh to onboard catalysts onto new roll-ups how trivial are we talking about it is that something that Roll-Ups themselves uh decide to do or or or is there some sort of you know Central governance or decentralized governance mechanism that uh decides where Catalyst gets deployed yeah the net of it is that it's up to the the roll-up right um so the way that Catalyst works so I think there's two two kind of like

layers to it so it's like what can Catalyst do like what is it able to do and then what should it do right and so on the base layer of what can it do it can support permissionless um deployments and asset swaps because effectively how it works is like we can talk more about that in a sec but uh you know there's pools of assets that live in each chain and there's a kind of a intermediary um kind of a county measure that's being passed as a message in order to allow for the uh the equal kind of transfer of value between those pools right and you know it's a it's kind of like a one-to-end

relationship where one pool and one chain or one roll up can connect to the end number of deployments that we are uh alive on on catalyst um so that's what it can do what should it do I think we give that decision making power to the roll-up Builders um and so if it's a matter of hey like we just want this we want it out of the box we want to connect to anyone perfectly fine if they want to be more mindful of of the connections uh to which chains they want to interact with and mindful of the expansion of that it could be governance for them right for them

um and I say all of this within the caveat that like this is kind of our somewhat of an end state so to speak where we can enable this but of course we want to be mindful as the product or as a protocol is deployed on Tesla and then Mainland and expansive chains of course we want to be mindful of of where we expand and and if it's like worthwhile time for a maintenance perspective so all that to be said like in the end we'll have permissionless deployments and we defer like the expansion power to the Royal teams but as we work towards that in our terms of our different click stops um we will have to have some level of like short like diligence of chains want to

deploy on before we enter the permissionless setting yeah but yeah that's just kind of the Playbook of sound uh software development no that makes sense and so I I wanna I wanna just clarify first here for for those who maybe haven't haven't got this yet but there is no there is no Catalyst chain Catalyst doesn't have a central a coordination mechanism uh that sits on like a you know a cosmos chain or or some other sort of you know Sovereign zone or or state it is fully decentralized from the perspective that every Catalyst deployment uh sits on a chain or roll-up so that's one thing and and then the

other thing that you mentioned which is really interesting and sort of figuring you know building a mental model of how this works is that every Catalyst deployment and every pool on Ever every Catalyst deployment so let's say we have Catalyst deployment on evmos and there is a usdc pool there and we have a catalyst deployment on polygon and there's a Matic pool there um and then there's a catalyst deployment on you know arbitrum and there's whatever their native token is um all of those assets are effectively in pools with each other so this is this one to n relationship so I can swap my usdc formatic and Matic for the

arbitrary token and and they're all kind of part of this one big um pool of assets that sit on different chains is that correct that's correct okay so that that's that's really interesting because it sort of breaks the the idea of liquidity pools and amms as we've conceived it so far if you look at Ave right you you most most of the time you have this pool of two assets and then and then and there there is multi-hop you know for example on osmosis pools you're able to sort of go through pools to to swap assets but here everything exists at the same level

they're all in the same like semantic level and effectively at the same um level of being able to swap with each other so I think it it's like a novel design that I I don't think exists yet especially across chains yeah definitely um a couple points that that I want to maybe um like shine a little like a little bit more light into um so I just want to kind of um reiterate that we are not our own chain um and for and I think the beauty of that is is that we have um some level of of credit neutrality with that right like we're not being

preferential to our own execution environment uh we just Deploy on on any other execution layer and you know facilitate the transfer value uh between them and so we're definitely a new breed I would say we are a true natively cross-chain protocol or dap whatever you want to call it app protocol uh in order to facilitate that right and we feel very strongly about about maintaining that um we may eventually have um you know a home but that would be more along the lines of um of of like coordinating essentially like

coordinating all the different deployments uh as opposed to having its own kind of like State layer so it'd be more like a more like a data layer so to speak governing the protocol upgrades and things of matching and then that yeah that's what one thing that having a centralized chain you know helps with is that coordination mechanism yeah definitely especially as kind of the the multi-chain governance infrastructure uh ossifies in many ways and allows for kind of that that trivial um you know upgrade and I know axelar thinks about that a decent amount as well um so that that's one piece uh and the second piece is like we um really believe in the concept of

um um like well there's there's a number of things but I think the system that we're outlining is is capital efficient in the sense that you know every chain is just leveraging liquidity they have for the full swap there's no kind of additional liquidity that's needed for some sort of you know intermediary bridging token uh so to speak um and so that's that's capital efficient uh and and pretty kind of simple by Design uh from from that part from the protocol perspective um we also really feel strongly about the concept of leveraging native assets for swaps as opposed to some form of like uh a lock-in mint mechanism that we see with other token Bridges uh I think

by eliminating that element we minimize the attack vector or you know minimize the the kind of ability for an exploit that we've seen historically uh by a significant amount right and so um the way that you know we architect it by having native asset swaps is you know pools are isolated um so you can only ever really attack a specific pool and set up like an entire contract that holds you know 180 plus million dollars of different assets um and uh and there's a number of other ways that that we're kind of ensuring security knowing that this is a very highly scrutinized field uh but if you get it right you know it's it's it's obviously going to be well worth the

well worth the trials and tribulations so explain how pools actually talk to each other and and there's like different parts of that question obviously there's the the messaging protocol or the trustless bridging protocols that allows pools to send messages to each other across different chains and ecosystems that are perhaps not you know leveraging a unified messaging protocol like IBC for example so that's one thing but but I think there's also like how does um how do pools get information about other pools because um from what I've read catalyst is able to

initiate a initiate a swap by um without knowing any information about liquidity or of another state and so I'm curious how that works and you know if you could explain also this this concept of unit of liquidity sure so one of the design principles that we had when um constructing the Catalyst protocol is um how do we create the most extensible lightweight deployment that we can in order to service this future that I talked about right millions of chains and roll-ups

um and how do we make it the most performant that it could be right uh and so with those design principles in mind um you know there's trade-offs that need to be made right I don't necessarily see it as a trade-off but um I think you know from a from a formal kind of theory perspective it is a trade-off and so Catalyst as a system whereas a protocol is the composite view of it uh does not have knowledge of of the global state right um or or said differently um there's imperfect information um at what that's known for each deployment of catalysts right

um and how that effectively works is by removing the need for State synchronization between each singular deployment of catalyst it allows for a lot of the things that we talked about right that we're that we're solving for it's more extensible is a very lightweight protocol um it is um you know trivial to kind of expand the different things it's very elegant by Design very simple in its flow um so so how does that actually work in practice right and so I always like to say that there's two layers to catalyst so the first layer is uh the actual um smart contract deployment um which for all intents and purposes is

is just the math it's just the math library right and then underneath that is kind of a messaging interface what you're alluding to right like leveraging IVC in order to pass a message because at the end of the day you need relayers to relay information and yeah you can't do cross train without it um and so let's skip the second part but we can talk more about that if if you'd like to I spent a lot of time thinking about first arbitrary messaging protocols um and and I write a lot about it it's definitely one of the things I'm really really fascinated on the space um but on the first piece right the math the actual you know the actual uh contracts that hold the the deposits

um how it works is essentially and like I said there's no State synchronization right so how it works essentially is you could think of each contract of catalyst as its own kind of box right it has box there's assets inside of it um there is generally a concept for each asset like its own price curve right and so you know x times y equals K if you look at a pretty simple uniswot V2 model is something that people are familiar with from an amm perspective um and Catalyst we we take that concept and we essentially apply it to each asset but um

maybe we won't go there maybe maybe that's a bit too complicated but um essentially you can think about it's like each asset within this box has its own price curve right that is and the x-axis is you know the the actual asset and then the y-axis is is less important in this case um but that that's all there is right so there's no kind of like knowledge of outside state of other chains and so I think what's good about that is then you don't need to kind of keep track of different metadata that that com that comes in from different chains and you're like oh I understand the balance of that the Avalanche deployment or the polygon deployment or what have you right because then that becomes very complicated to keep track of all that state from different deployments and so

we kind of just get rid of that right and so how does an actual transfer work is that this box wait wait I just wanna I just want to just stop there for a second because this this price curve um I mean typically in a traditional the price curve it has both assets and the price of both Assets in the curve here if you're talking about a single asset what is it pricing against um yeah is it pricing against US dollars or is there like some other kind of generic uh price uh that it's yeah how does that work you could basically think of the y-axis is like

held equal for all assets within catalysts essentially um I don't know how much more I could share but uh like the way that I essentially like if you were to look at a swap you would see that there's two assets on different chains right let's say it's a cross chain swap which actually you know Catalyst doesn't even need to do cross chain it could do local swaps within its dollars yeah um but you know let's say it's cross chain you'll see that they they each of these assets have their own local price curve and when you overlay them like when you

set the y-axis equal to each other it forms x times y equals K or or another very popular price invariant right um okay so it's like a conditional almost so that's why I'm saying like the y-axis doesn't really matter in this case uh because we hold it equal and only care about these two x-axis uh between the two different assets are being swapped um sorry that might be way too theoretical but uh that's hopefully helpful uh if not no worries we'll we'll try to we can we can double click more into that um but essentially going back to my original piece we have this box right these assets and so when they receive a

message when they receive kind of uh you know the unit of liquidity uh message that that we kind of pass um they just they just take it right they're like okay here's like this Universal accounting measure um and I know now that I need to withdraw x amount of units of liquidity right for for this specific asset and so you have this box it receives this message it goes cool three units of liquidity whatever that means I'm gonna convert it into the amount of Juno that I need to deposit to this address right and that's it that's it's it's it's a it's a very simple

box and so where does that message come from well it comes from the other side right and so in this case let's say a user is trying to sell usdc on Noble and they're trying to get Juno right and so they sell usdc again another box on Noble this one then calculates okay like you are depositing a thousand usdc uh that will be you know three units of liquidity and then you're sending that message on IBC to be received on the Juno side and again that whole process happens again it goes three units of liquidity okay I'm going to convert that to Juno and you send it to the user right and so that's effectively how the math portion of catalyst works and you can kind of see where the messaging portion happens

to with the IBC message passing okay so assets have this unit of liquidity which is a value that all of the pools can understand and convert back to the to their asset and and that allows the receiving pool uh to to figure out what the what the what the corresponding amount of tokens are for the swap is there some sort of communication happening between all of the Catalyst deployments to agree on this unit of liquidity is that

required by the math or or is not not required and the agreement is implied with the deployment essentially okay because Catalyst at the end of the day is is math right um and so that that's where like the agreement is everyone calculates this this the same way um and so I kind of laid out the instance of like noble to Juno but that that's a one-to-one use case and again like we're I think the power of catalyst comes in the one to many where you deposit again in this case a thousand usdc and Noble and you go

you know what I actually don't want Juno I want notes on Kanto or I want you know Matic on some polygon ZK evm and uh you can do that and you can see now where like the extensible design of the protocol allows for that right because again these are all disparate boxes that don't require any sort of you know synchronization of any kind they just receive this message um and are able to do so um so that's that's where the magic of catalyst comes in so when we're talking about IBC you know to people who listen to this podcast it'll it'll be obvious how that

works and how those messages are passed but when uh when dealing with chains that don't have IBC so for example like going from Juno to polygon um or you know like Solana or whatever I mean like each of these chains are going to have some protocol or bridging protocol at some point presumably that allows messages to pass so we have things like polymer and um like uh you know LaGrange for instance they just announced their funding today um are is catalyst just optimistically using you know what's available or are you building your own messaging protocol at some point is that is that the goal

or you just use what's there or are there any are there any requirements in terms of the types of messaging protocols that you could use so there's no requirements again we've built catalysts to be modular in that in that respect and so said differently you know um if if folks are familiar with kind of how I how I view modular interoperability protocols we Outsource our verification layer um and so we leverage the verification of other protocols uh in order to ensure the validity of the message that's being passed upon receipt right um and so we built it in the way in which we can have basically any message passing protocol and any verification

method in order to do so um with that said as you could probably see from this flow passing a valid message is the most important thing for catalyst and so we have a insanely High bar insanely insanely High bar for the message passing protocols that we're working with and so that's why we're proud to be in Cosmos we're proud to use IBC if IBC expands or when IBC expands into other environments whether through polymer or electron or or kind of another team focus on that we would like to use IBC but if there Comes A Time In which there is something that is equally trust

minimized and and kind of battle tested and scalable as like IBC zkibc we are happy to use it as well uh because again we we have built catalysts in the way in which we can leverage any message passing protocol so it's not a matter of can it's a matter of who kind of meets the standard that we're holding for for these for these protocol Integrations okay interesting um so I I guess from that from that answer we should assume that your Catalyst will favor highly trust minimized um message passing protocols that don't you know things like like ZK things like

IBC where essentially chains are verifying the state of each other and the but but less so the bridging of the of the sort of you know multi-cigs and so on yeah I I think that's the most important thing for us um I will say maybe as a side note I think the the ZK or the snorkification of of verifying um makes sense for every interoperability protocol I actually don't really see a interoperability protocol that's not at least exploring that route uh because I see the writing on the wall they know that that's where kind of the

technology is moving um and so that doesn't proclude us from kind of working with any any of the popular arbitrary message protocols right now but um I think over time you know we're we're in ourselves building security mechanisms within catalyst that that may allow us eventually to relax those trust minimized conditions um but as it stands right now like we're we're very very mindful of of maintaining that high level of security with the partners that we pass messages through very cool I want to talk a little bit about

um impermanent loss and you know people are generally familiar about how impermanent loss works on on an amm that that has by two asset pools um how should we can think about our impermanent loss in the context of of catalyst does it introduce new Dynamics for improvement loss that are not obvious [Music] no I wouldn't say so um the impermanent loss kind of um calculation that you can kind of do is is similar to to what users would probably do for a balancer or you have

like multi-asset pools um so with that said like we don't introduce any new types of permanent loss outside of maybe you know whatever kind of um Divergence that might happen from a from a cross chain perspective my introduce um but you know I think I think the apartment loss that we introduced in the system is is is similar to something people are already familiar with with balancer and even then you know that's not enough for us we can't just say yeah it's like the eye all that everyone faces like we we think very very deeply about how we actually overcome some of those IL barriers uh because we think it's

you know a tough kind of problem at that I think especially as a as of late has kind of gotten a lot of conversation around you know how do you protect against IL how do you you know what are mechanism designs in order to overcome it or mitigate it excuse me uh we're very much a part of those conversations uh and and we care about deeply because I think in the long run I all is something that shouldn't be something that LPS need to be exposed to necessarily or should be compensated for uh Julie so that's that's kind of like our perspective on IL and what about Mev how is it captured and redistributed so we we have plans to capture Mev

within the system uh I think from a philosophical lens um I'm definitely of the camp of what we call believing in Sovereign Mev so Mev that is being generated and extracted and or leaked into a system um should be captured as much as possible and harnessed uh for the benefit of the protocol generating it right um so I think this is somewhat of a different uh take than some other Mev shops but um this is something that I think we we feel very strongly about and so how we apply that um you know philosophy to catalyst is essentially Catalyst deals with cross chain

um it deals with a variety of different domains and so it is a very new exciting and um kind of uh um confusing and complicated design space of like not only doing cross-domain mov um you know calculation and and and simulation and and therefore mitigation but also doing it in an environment in which you're you're dealing with more than two domains right we're dealing with three four five and eventually we're going to be dealing with with millions of domains that we're connecting um and so with that said like the answer that I can give you is that um yeah this is definitely an active area of research that we think very deeply about and we collaborate with a lot of organizations in the space in

order to calculate and eventually capture that but we we do see that as a very uh high priority for us and something that can offer a sustainable yield for for LPS within the system uh ultimately like what we're trying to do with the IL mitigation and with the Mev capture amongst other kind of research topics is very much like the you know ability to compensate LPS for kind of the the price Discovery um risk or duties that they take on uh and and making sure that there's sustainable yield in the protocol in the long run so one thing I was thinking about it just came to mind here is you know we

discussed earlier how for for instance optimosis does uh routing of Trades through pools in order to be able to access liquidity that isn't immediately accessible between um between two assets um or at least facilitate swaps between assets that don't have a pools that requires a coordination mechanism that I guess osmosis probably has within the protocol with with Catalyst will this be possible will it be possible to do multi-hop swaps across different domains and how will that work would there be some sort of a mechanism that sort of has an overview of where the liquidity is and

facilitate those routes uh the sure answer is yes um that we can do multi-hop and that we can do some sort of uh some sort of Route optimization um the reality of it is that uh it's not something that we are focusing on um you know ahead of mainnet but we do we do see the the value of doing multi-hop the value of um kind of offering the the best price execution um for for users um and you know I I actually think that we can do it quite trivially I don't I don't actually think we need maybe you're alluding to this but I don't

think we need like a central coordination uh Hub in order to do so okay really interesting yeah yeah um and I mean obviously like it wouldn't you know I I think the trade-off there by not having a a blockchain to do the coordination is that some of it has to be off-chain computation right um that's just that's just me being intellectually honest um said differently like uh instead of maybe saying like you know like implementations like multi-hop and price routing and and off-chain computation versus on-chain um I think we're this is something that we at Catalyst care a lot about which is kind of the evolution of

um like how we view a transaction where it becomes more of like intense and execution right so I'm sure you've heard a lot of work on intense as it pertains to like anoma and Suave and the such and so that's something that we we do a lot of active research on as well and so the intent for the user in this case is hey I want the best price for for you know this this asset and so we can facilitate that through uh sophisticated execution uh whether it be single hot multi-hop whether it be gas sponsorships whether it be you know something uh that might be cheaper when you introduce latency through some sort of like limit Buys so we're uh we're definitely very very mindful of the research on intents and how we can kind of uh Implement that

within catalyst cool um yeah I'm learning so much here about how how catalyst is constructed and it's sort of starting to come together uh more uh more clearly yeah in the future sorry that's uh maybe maybe a little complicated to start but you know it takes it takes cycles and it it's learnings for me to explain it more clearly yeah no definitely um so in the future can can we so like one thing that I see here is that we may have I mean also also the case for you know amms that exist cross-domain now is that we have assets that that live on different chains or different domains that are

representation of the same asset right so you might have like usdc on one and wrap uscc on the other you might have like crap ease and E's um is it is the is the Future Vision for for Catalyst to in some way create meta pools of all of these assets to have even more um uh optimization of liquidity and and and capture the best prices for these assets or you know is this something that will remain sort of siled on on different ecosystems so that you know if you're if you're swapping usdc you're always going to be swapping usdc for like some asset on the Chain the destination chain that you need to go to or will it

you know would it be possible for Catalyst to say oh there's some liquidity over here I'm gonna go grab some of that to uh to fulfill this trade as well um what what do you mean by metapool well um I guess like so here's an example let's say you were going from from Juno to um to osmosis and on the destination chain there there isn't enough like you want to go from Juno to osmo or maybe this is not a good idea but like let's say you want you want to go from Juno to osmo the other way right and there's not a whole lot of osmo in Juno um and and you're asking to swap more osmo

on the destination chain then there is on that chains pool would Catalyst be able to go and get liquidity from another chain in order to fulfill that swap go go on another chain and find liquidity that is just sitting there on another Pool of the same asset gotcha okay um sorry that's slightly different than what I what I interpreted uh metapool is the B um but well I guess metapool would be like a all of the Juno or all of the osmo or all of the usdc across all of the domains it just knows that that there's liquidity on all these different chains and if there's not

sufficient liquidity on a destination chain it can go in pool liquidity from somewhere else that's yeah like in my mind when I think about a metapool so the short answer is yes um the medium length answer is that the way that we're thinking about like the five-year road map for catalyst is essentially focusing on what we think is really really Pro like pressing And Timely which is this kind of proliferation of chains and ensuring the extensible connections and automatic connections of all of them together permissionlessly um the second kind of step for catalyst

excuse me uh comes in the form of capital efficiency and so there's a number of ways that we're thinking about Capital efficiency as it pertains to you know I'll share why not concentrated liquidity uh rehypothecation um virtualized liquidity is a form of in this concept you called metapools right so so Universal liquidity um uh or aggregate of liquidity through different domains and so that is a better performance to describe what I was talking about oh no worries virtual pools yeah yeah so so that's kind of what is on the horizon for us absolutely uh we're making very very significant inroads on

that kind of horizon for us right of capital efficiency amongst other things um just quickly I when I heard the term metapool in this kind of scenario you outline um I I came to the mind of like there are chains with canonical assets or maybe not even Medical but like metapool plus you said like the proliferation of different representations of assets um yeah and uh and I know you know there's some really good work being done with like uh ICS unwinding for like ics20 transfers um yeah and such uh but um my my take is that like there will be canonical assets that live

on each chain as like that's where the asset lives and all the coordination between those assets will happen through like you know upside differently through like ICA like account abstraction or kind of like remote controlling and Catalyst will be kind of the infrastructure to facilitate that um got it yeah that's very cool my tank on it at least but we're not necessarily building for that world we're building for uh good when you build for that world you assume like homogeneity and we're not doing that right we're building for like a heterogeneous world and so um we're we're more and more building codes what I was talking about where it's like Capital efficiency through the form of virtualization very cool

um so I've never delivered it before so sorry I've never shared that before I'm never shared what we're doing in V2 but you heard it here first um so you you guys have uh recently announced that you're working with Neutron and eclipse and uh just so happens that uh someone here in the chat was asking about how you plan to acquire liquidity on your Neutron deployment uh can you talk a little bit about those Partnerships and uh yeah how you plan to attract liquidity knowing that well I think Neutron will probably have its own decks and probably other D5 protocols that are built there yeah definitely

um and so we're super excited to to partner with the neutron um Team I mean you know they they have not only incredible backgrounds but they have a demonstrable track record of building great products uh in crypto and so it was kind of uh a privilege for us to to be able to to work closely with them on on uh you know Catalyst and and having Neutron be one of the the first homes for catalyst um for for our deployment um and you know I think I don't need a belabor kind of the the story of neutron necessarily but I think

it's it's a welcome addition to the space that will unlock a lot of uh kind of value to um to The Interchange in the cosmos ecosystem and for us like I think what was exciting about Neutron is very much you know what what they're bringing to the table and what is very exciting about or should I pose or no no worries no I'm just taking a picture but um I think what Neutron gets excited about is uh is kind of again the the possibility of uh retaining liquidity within within their chain but still allowing the facilitation of swaps with other with other Sovereign

chains and eventually Beyond to ethereum uh and and allow for this kind of like bridging ability as well uh permissionlessly right and so that's kind of like what they're excited about and I think from ethos perspective we're both excited about this element of permissionlessness uh enabling kind of a borderless blockchain kind of vision um to answer uh the question about uh liquidity on the neutron deployment um I kind of alluded to this but we think a lot about sustainable yield you know as it pertains to um you know as it pertains to Mev capture as it pertains to Il mitigation very active

research that that we spend a material amount of time diving into and we'll we'll deploy with with Catalyst when it heads in the main net um so that's kind of like our our foundation right of of what we want to do and of course uh you know on top of that uh would be more of like an open point on on what uh what incentives looks like but I think foundationally we think that those levers will provide the sustainable yield that I think D5 broadly has been has been looking for so what's the what's the roadmap uh what can people expect from catalysts and how can people get involved today if

they want to start working with you guys sure so uh roadmap wise [Music] um very very uh busy month I would say oh you're welcome I appreciate it um I like I like I like lcs's pfp I don't I think it's like rock climbing or something um that's cool uh but yeah we have a busy month ahead of us like I said we have our white paper um that's gang peer reviewed and we'll release that and and with that will come a lot more of the details of what I just discussed very poorly uh around around the price curves and and the conditional matching of them um

we also have our test Nets that will come out uh very soon uh um sooner than I think people will will realize uh so that's that's that's going to be no it's cute uh scores are cool um oh sorry there's a private chat and so I saw it um but uh yeah you are talking about the uh the squirrel that is in the profile picture of the person who is commenting for those listening and not watching uh yeah sorry I I forgot not always a video format um but uh yeah so so our test net will come out uh in the coming months uh and then uh that would be kind of like our preliminary

release candidate we'll have more release candidates over time we'll have quests we'll have you know incentives uh for our test net and then we do anticipate a uh late Q3 midnight launch uh later this year that's uh yeah great and uh I mean I I it's it's sort of strange to talk about a main net right when there's no chain it's it's I feel almost like we need other nomenclature to describe this this mesh of deployments on different you know maybe it's a main mesh or something like that yeah you guys could be the first to coin it yeah yeah we we gotta we gotta be better about that I would say uh but uh yeah that's kind of uh

our timeline so busy busy nine or so months ahead of us great well I'm really looking forward to seeing this uh play out and looking forward to the test net or the test mesh I'm gonna start playing around with uh with Catalyst Jim thanks so much for joining uh on the podcast today I hope those who were watching the live stream learned a ton and certainly those who will listen to this after the fact will be really excited to learn more about about Catalyst going uh in the future so we'll link uh to all the documentation everything in the show notes so people can follow up there thanks again cool thank you foreign