Enhancing Liquidity through RFQs and AMMs - DeFi Today Devconnect ARG

Transcript

moderating. So now we have our next panel coming up which is on RFQS and AMMs and enhancing liquidity with these tools. We'll have a great moderator called Danning. She's the research director at Pantera. Uh we have Nikita who is a CEO at Barter, one of the largest onchain solvers. We have Ken the co-founder of Uniswap Foundation. Uh and we have Nicholas who is a senior engineer at Zerox. Okay, please give a big round of applause for our next panel. Hello. Hello.

Hello everyone. Uh, can you guys hear us? Cool. Cool. Um, all right. Thanks for coming to the liquidity panel at DeFi today. Um I'm very happy to be here with our guest panelist to talk about basically the landscape of today's liquidity on chain. Um we've seen a long uh how to say the onchain liquidity come a long way started from the defi um summer from order book now then AMM and then there we have RFQ coming in and many many new innovations. So maybe just to start first I would like every panelist speaker to introduce themsel and also tell us about how does your project is

serving the landscape or ecosystem of the onchain liquidity. Maybe start with Ken please. >> Hi I'm Ken. Um I lead applied research at the unis swap foundation. Um you unis swap is a decentralized exchange uh where anyone can uh provide liquidity um and anyone is able to swap on other users liquidity. That's how we provide liquidity to the ecosystem. >> Hi everyone, my name is Nikita. I've um you know been in DeFi for over five years and we've built Barter. Barter is started as a solver on intentbased dexes

like cow swap and for those of you who don't know solver is like a mix between market maker aggregator and mebbot but we sort of have a honorable goal in mind that we trying to save users money every time people transact on cow swap unis swap x 1in fusion and other venues we try to give them the best rate and save them a few uh and and give them a few extra coins and we are constantly improving liquidity landscape and going to launch a new product called superp position soon. But yeah, that's small TLDDR on me.

>> So, hi there. My name is Nicholas. Um, I work in 0x. Um, 0x and matcha are have been one of the OG uh companies in the ecosystem. Yeah, we know each other >> worked as your exo >> and uh we worked as a permissionless protocol for uh swaps and the aggregation of decentralized exchanges together with um private market makers as other sources sources of liquidity. Um I work in the solver and pricing and RFQ team and uh yeah super super nice being here. Cool. Uh okay. So I want to

also give an overview for people who are not familiar with all the projects updates. So what were the biggest announcement or updates you guys have seen in the space or you know coming from your projects in the past you know like half year a year or so? Um this is a basically question alluding to a cue to Ken to talk about hooks and um okay I'll let you start. >> Yeah so we man what do we do? So early this year we announced uh uni chain uh which is a generalized L2 on built on the OP stack. Um with that we have flash blocks so like 200 millisecond block times uh which is pretty cool for interop and uh there's so much more that

you know uni chain is going to bring into the future as liquidity hub. We also released uh unis swap v4 earlier this year which um is the new version of unis swap where we have uh the singleton contract. So it's hyperefficient and it introduces this idea of hooks which is arbitrary logic that can sit on top of liquidity. So it affects the way that anyone um can affects literally uh kind of all parts of what it means to provide liquidity. It can be something that affects swaps like before and after swap. Um it can affect liquidity provisioning before and after providing liquidity uh while you modify liquidity.

An example of that is um um uh Oiler Swap, right? So Oiler uh has you know these Oiler vaults and um they built a hook called Oiler Swap where you're able to uh extend the liquidity uh that exists in the pool by borrowing from a Oiler swap. And so it'll take out a loan uh at the time of a swap uh in order to facilitate like a much larger swap more than what liquidity sits within that pool. Um and yeah, it's it's pretty awesome because it it allows you to uh create new capabilities for liquidity that haven't been there before. um you know whether it's like MEV capture or MEV redistribution to LPS uh liver mitigation uh something that a lot of

people are talking about uh uh loss versus rebalancing uh and hopefully that in helps LPS become more profitable it helps uh uh equilibrate the execution price across all dexes. >> I actually want to praise our friends from unis swap for the work they do and I agree like hooks were an amazing addition to the defi landscape. If you look at the volume and metrics that were going through all those intense systems in 2024, market makers were starting to get really dominant like uh I I don't want to throw names, we all know them like guys like Wintermute and others. Uh they were gaining and gaining more and more market share and um of course there are like obvious reasons to it. They can

price better. they have access to unique liquidity. But now like the landscape is different and hooks and other new dexes like eubo, fluid and many many others they sort of provide a viable alternative and a good competition and hooks like oiler swap um you know make our lives uh lives of solvers a lot easier a lot more efficient and we can combine it to beat those big sharky market makers. So I'm definitely excited. Another point that I might add uh it's on a grim side a little bit that recent hacks in DeFi especially like the balancer case made the situation a little bit worse. We talked with a lot of teams who were initially like excited

about hooks and new liquidity primitives and now everyone is slightly scared of um you know deploying and even like hundred thousands of dollars of security audits do not guarantee that you're not going to get exploited. But yeah, I think those two um some of the biggest drivers and updates in terms of volume activity and overall things that are happening uh in the DeFi trading market. >> Yeah, that's quite interesting actually. Uh I think we all like complement each other in a good way because for example um people being scared about his hooks or amm with this programmability uh there's somebody that needs to work

with them and uh integrate them in a way that are um curated right and that's when solvers and aggregators like we do uh integrate both the AMM pools and also the professional market makers not to sort of like beat them but actually to complement with them complement with them uh maybe for getting better liquidity and we also participate in a protocol like parter as one of the one of the solvers so yeah >> thank you >> so what I'm hearing is that everybody is you know working with each other but yet in competing some way um uh I actually wanted to ask a bit more just uh on the V4 hook innovation aspect right like as

a data person what I saw is like the design thinking is to unlock and like uh freedom for a lot of like new design but the standardization of the data decoding was kind of like a nightmare but now has been way better thanks to Jackie from unis swap foundation many helps there um I'm curious also how you guys are integrating all these new liquidity has there been like some frictions or um or in general is like are these new liquidity serving a particular type of needs in terms of the pricing are they better at a certain type of asset class etc. How would you classify these all these new liquidity?

>> Um I guess I'll start because these these guys have done you know the line share work. This is this is really funny because um you're right the the hook data standards have been so like you know and and we spent a lot of time really kind of standardizing um custom liquidity custom curves custom accounting uh because it doesn't follow the XYK curve anymore right anyone can create a hook that actually says I'm going to build a new curve >> and Barter and ZeroX have both been amazing partners in actually being like okay yeah we'll do it you know we'll we'll figure it out and and through this trial and error, literally banging our heads against the wall, like thing making things like not work. These guys have helped us create data standards.

We've gotten a lot of feedback and and actually making sure that it's as easy as possible and also like it it makes like building hooks easier. Um, you know, Open Zeppelin has this uh hooks library and based on the the feedback that you guys have given us or your teams have given us, we've been able to be a lot more stringent on how these custom accounting hooks are built, how hook datas are uh can be standardized. Um, but for liquidity and pricing, it's really tough. I I I'm curious to hear your guys' thoughts on on how you price these things because there's there's like the the quoting side of it in which like, you know, it could be dynamic fee pricing and that's something that we don't really see around other AMMs. So, you have to play with dynamic fees, you

know, in times of high volatility shoots up. Um, and then you also have like things that are like Angstrom for example, which Barter has been, you know, the one of the first solvers to actually pick that up. So insanely hard, dude. They have external indexing like the like all the all of the orders go offchain and you guys index it all and you're able to to rep prioritize and um yeah, I How do you guys think about coding? I'm I'm shocked. >> H I'll be very quick with my response. Integrating hooks is tough. It's requires a lot of research, a lot of tinkering. Uh we have like a dedicated sixperson team who is just focused on

this particular problem. Uh but if you want efficiency, if you want to make money, if you want to be successful, you have to put in the work and combine all those innovations. The way we see it is like we've built our solver router uh as a sports car and then we're winning this Formula 1 race of DeFi. And what does that mean is that um we we we want to give users the best price uh every time when we are you know like quoting on cow swap 1 in fusion and other venues. We try to give the user the best price and um those extra hooks and extra liquidity adds something to our engine that makes it more performant and sometimes give us this small edge that's exactly what we

need to succeed. >> Yeah. >> Yeah. Similarly uh we work with the routing of different sources of liquidity and especially in super fast chains where like um like Solana for example that we have been working for the integration uh this is like every millisecond counts and uh it's actually very hard and it's actually one of our main goals to to reach the best execution uh quality in terms of latency and in terms of best price. >> Yeah. Um so following up on this question but zoom out a little bit let's talk about also RFQ liquidity uh as the name indicates um I want to ask maybe uh a little bit of alpha on the stage as in

like how do you guys see all these comp liquidity source compare as of today? How does like where does RFQ stands out? Where does AMM compete over in terms of like a compet competitive landscape for liquidity? Um I want to yeah I want to have your take on uh what's the best way to route all these liquidity for the best pricing. >> I can take them. Uh so we think about uh the liquidity landscape uh in a nonomogeneous way, right? So you have like the blue chip uh that are very competitive. It's actually commoditized like the swap and then you have the longtail assets where liquidity is not so uh good. Um I think that um sources of liquidity like unis swap for example

and liquidity pools like AMMs are like always on. They're always going to be available. Sometimes they can offer uh a better a better experience for the user but at in the same at the same time then you have like depth and liquidity that professional market makers can bring like RFQ that uh add up on top of this layer. It's like one layer on top of the other and it's kind of like a fabric that it's the execution layer for the user to be seen as seamless. Right? This all this complexity is abstracted for the user. So uh I don't see one beating the other but actually in in the future

all these converging also I I heard you say about the uh before swap and these hooks sometimes uh the offchain uh prices and the offchain uh participants can bring signals to the onchain market uh that could make the prices to to be more liquid and the spreads to be tighter. So it's kind of like a uh collaboration. >> Uh let me uh jump here with um our newest product called superp position and we were inspired by RFQ guys and uh what we observed on the market with like intents cow swap solvers that uh market making liquidity is becoming more and

more dominant and what do market makers do? they route to the hot wallet and um we as an onchain primitive we always felt like we are fighting for the little guy improving life of the end users and we don't want to become or lose to like those big sharky market makers and we thought why don't we do the same we can also start routing to user wallets and unite user wallets as a new liquidity layer distributed liquidity layer and it allows like new things to be built for example self-custodial yield. I've uh just came up with this analogy today while explaining the concept of superposition. But uh imagine like you have different stables or if you're like

a normal person. Imagine you have like uh I don't know euros, dollars, British pounds in your wallet and you don't care like imagine you have a,000 pounds and you know approximate number of dollars how much that would net you. But you are asset agnostic. You don't care which money you hold. You just want them to generate yield increasing numbers and uh you can allow us to tap into your liquidity under condition that we will give you more than what you currently have and we will start routing and maybe at one point in time you will have a Japanese yen in your wallet which you don't use or not going to be spending soon but uh you start earning yield.

It's self-custodial. It's growing and no one has done it before in DeFi. Combining RFQ like pricing mechanism with onchain settlement is really really powerful. We obviously cannot rely on just one or another. We definitely need more AMMs. We need order books um centralized order books. We need uh RFQ systems. Hopefully there is a room in the world for superp position. the one I just described and um yeah this is one of the like RFQ kind of innovation that we are extremely excited about.

>> Yeah. Um so we have super position is so cool. I I just learned about it today. Um we have a unis swap x um where anyone can be a filler as permissionless. Anyone can sort of provide quotes to to trades that come in um on the unis swap x API. Um but I I I guess like stepping back a little bit uh the thing that we think about when we think about swapping and and lping is we we approach it from like an asset um asset class standpoint. So we actually work backwards from like stable stables uh ETH volatile pairs or you know like BTC volatile pairs um and then like longtail assets uh RWAS etc etc. But we we like to be very specific

on like how people trade on the assets and how people LP on the assets, right? someone who LPS similar to you know what Nikita just said um when we someone who's uh LPS and stable stables are really looking at sort of a range right you're looking around a peg uh sometimes obviously it goes you know you know past the range of um you know 10 bits of uh USDC USDT around one um but generally you stay around this like tight range uh volatile pairs there's some lagging prices uh across sex indexes Um and then longtail assets are sort of you know it's wherever liquidity lies which normally sits in AMMs. Um so when we think about RFQs and we think about unis

swap x and how people fill orders in unis swap x it is a lot of times in sort of like larger volatile pairs um stable stables where there is a lot of liquidity already. So anyone who wants to fill orders like have a stable or have a have a lot of stable assets just sitting there ready to go. Um and volatile pairs obviously you don't have the lag in in um uh uh block times. Um and then for longtail assets you know you want to be on AMM because that's where it's most liquid. That's where it started and that's where most LPS are going to stay. Um so we we don't really just only think about how we fill prices but we also think about you know what is the liquidity provider like you know what is the person who's actually um you

know quotequote like market making or filling orders? What are they doing? how are they what is their hedging strategy and we work backwards from there to think about how hooks uh play a role. Um and so when we think about the future of RFQS and we think about the future of AMMs um it really comes from user behavior and the assets in which uh you know the users are using >> yeah that makes sense. Um I want to follow up on in terms of how work work with market maker. Um I know there's like the the evolutions of RFQ design back when I was at working at Xerx you know there's this term RFQM RFQT based on different you know like workflow how user sign their uh wallet approval and then the market maker confirm the quotes

and later on they will confirm the actual firm quotes. um there's some kind of like out of a serial selection where if you're a sophisticated trader, you know how to hold their quotes until very last millisecond. And so everybody in the space like working with market maker starting to trying to figure out the problem of how to protect market makers quotes um to avoid you know like toxic flow basically and but then it is a really hard question to figure out. So there's I believe there's still I I'm curious how today working with market maker look like and if there's still like friction like that to solve. Um maybe start with Nico.

>> Sure. Uh yeah I live that day today. So yeah um there's always some exposition from the point of view of the market makers to the market when they are accepting these uh quotes or um posting these quotes for the aggregators to integrate. And uh one of the things that helps this to uh overcome this issue is first latency. So you don't want participants in the market to be uh not trusting you right and be defensive.

And uh one of the things that helps with this is more information. So um we have changed or improved our the way we negotiate in this back and forth with the market makers uh using uh permit 2 and using our new version of OTC before um to um make the the negotiation with the market makers the the seamless. So um these days I think this this is there's also the RFQ and the AMMs with programmability prop AMMs that are like new primitives that are coming to to the space that help um transmit this information in a in a more in a quicker

way to prevent this kind of like toxic flow, these bad actors that are going to always be there in this adversarial uh environment, right? And um yeah, I think that's something that it's ongoing and needs to be improved and needs we need to work on that and that's why we wake up every day thinking about the solutions. >> So let me pick up the question of propm and start from this angle. Um it's funny that propm like began on Salana and there are multiple reasons for it like technical they didn't have like good RFQ systems and just blockchain features uh sort of pointed this direction but also Ethereum orderflow is extremely

fragmented you know we have like a lot of orderflow coming from unis swap from wallets from cow swap and uh there is like no single point of entry while on salana there is jupiter and jupiter you know like manag managed to capture most of the market share and u propm emerged as they like provide like better quotes and uh they are almost like 40 or 50% of total order flow happening on salana right now but propm are getting like commoditized I would say they are not in touch with like end users and they are also like fighting for like small basis points of improvements which is good but

not um not you know like a major innovation difference. However, we feel like this trend is coming to Ethereum as well. Like for example, Kyers swap, they um have built their own hook, I believe on top of UniV4 and they the only aggregator that has access to it and you know they're like mixing their own liquidity with unis swap amm programmability with their own pricing methods which is quite fascinating. And you know it seems like they've market share increased so they're clearly doing something right but uh I would say and like a small problem with RFQS on Athereum mainet well props are coming uh

liquidity is getting more fragmented maybe also partly due to hooks and just due to other um like new designs but I would say RFQ market is heavily concentrated. We pretty much have maybe two, three major platforms. Hash flow, BBOP native. That's about it, I guess. Maybe I'm missing someone. Apologies uh in case. But um yeah, we have like we need more RFQ systems. We want more competition. Competition leads to better execution and better prices.

And I uh I wish we had more partners to innovate in this particular niche. >> Thoughts on propm? you want to jump in or >> I do have thoughts. I I mean it's it's interesting, right? Because prop like prop liquidity means that like you can generally almost always provide like better quotes, you know, like you're you're holding your own assets and you're sort of saying I know what I can I can sell on like the central exchange. I know, you know, other buyers. I can always play with whatever liquidity. Um but it is it is interesting, right?

Because uh if you do that, then you sort of knock out like onchain LPS, right? you're always going to give you're you're almost always going to be able to give better pricing than onchain LPS are just based on you know the blockchain. Um I'm going to shill for a little bit. We we recently announced the you know this big unification proposal for unis swap. Um um one of the things that that's a part of it is the PFDA so protocol fee discount auction and in that um anyone generally a serger or you know anyone can actually uh auction to take for a couple blocks um uh zero fees for a certain pool. Um, and that creates a new I I think for the first time in a very

long time, we're able to like play with the the fee dynamic for uh LPing, right? You if you're able to trade no fees on a ETH USDC pool, that gives you a competitive advantage over everyone else's trading in that block, but also an advantage over all the other dexes. Um, you know, with with propm and even with like just stable pools in general, right? Like you're seeing fee compression um lower and lower, right? because you're able to do that. You're you're playing on tight margins and you're only trying to uh get that little delta and like that's enough because you're over a long volume. Uh you're much more profitable. And so with PFDA, we're letting people sort of at least I like to think about it as like an

equalizer. like an average LP is able to see this and be like, "Okay, I know that I'm going to get slipped because the like you know the Binance price is moving this much and you know the I have 12 seconds to at least in this block I have 12 seconds to be able to cover up my losses. So I can actually hedge as an LP and trade across other DEXes to arbitrage across like what I'm seeing on Binance. Um I see it as like this this really cool new primitive for for people to compete with uh proprietary liquidity. Um, but I don't know, it's yet to be seen. I I I want more people to kind of like build on top of it and and you know, do more creative things with it. But I think more than anything, what I'd like to see is um creative

solutions for onchain LP so that like proprietary liquidity doesn't just like win out. It is like such a huge capital fight risk for all AMMs and and Dexes and liquidity in general for the space. um you know it it has its benefits but I think every liquidity provider big and small plays a role in you know keeping the health of Ethereum DeFi safe. >> Yeah I kind of agree with you. So it's kind of like uh we don't go want to go to the extreme world where concentration beats uh in a way that it gets better execution or uh a lot of fragmentation which is in in a sense a curse but also a blessing for us because it's our work.

So the idea would be let's keep it balanced in a way that there's concentration but also these possibilities of n uh smart ways to build liquidity pools with programmability and let us like solvers uh make it like so the end user doesn't suffer this fragmentation right >> I I would generally agree with the panel about the opinions about prop amm um I from first glance I thought uh I thought it's basically RFQ in a more elegant way like a format uh implementation of Salana because of the nature of the gas is so cheap. You're able to basically stream or your levers onchain which already RFQ been doing offchain in a

way. Um I would say though maybe there's a contrarian take I'm cur curious also if anyone else have a discussion later on but um I think RFQ is better than the propm in two aspects. One is uh we've been hearing now that all these Salana propm is suffering from the toxic flow issue because you can basically have a bot a MV bot programmably basically checking the price of propm and figure out if there is a profitable arbitrage and take the flow when the pricing is off. So uh they are I believe fighting against that. Of course you can do the sim um uh thing that that's like whitelisting just the Jupyter aggregator front end. Um but I mean it's a similar thing that when you are doing offchain

RFQ you probably is um like the API only allow pings from you know partner known partners. Um but there's another aspect I think RFQ can do better for the market maker compared to probate MM which is um when you are a aggregator you are sourcing the price um between AMM and the RFQ liquidity right and your goal is to provide a price that is better than you know just available out there for the AMMs so you can basically design dynamic fee where the market maker doesn't need to provide the optimum optimal price but already maybe a slightly better price than AMM so to potentially even help to keep a basically like margin for both the aggregator and also market maker. Um you

can figure out how the incentive work later on but I think this is actually some um intelligence the proper AMM are not considering right now. Um if that makes sense. Yeah. Um cool. I think we don't have much time. Uh any last thoughts from everybody about what to look forward to for the ne in this space of liquidity in the next maybe 18 months 12 months >> real quick. So I think that uh based on the new technologies that are decreasing latency and think making things super cheap. I think we can uh improve our protocols and also make the information more readily available for users and for our partners to decrease spread and

increase uh the trust. For this we need to build and yeah we're hiring. >> Uh I would make a bet that in 12 months the market would look different than it looks now. I think initially like Dexes and AMMs had a lot of flow and then aggregator started winning and then intentbased systems emerged. Uh hopefully in 12 months uh not hopefully I mean definitely uh the orderflow landscape will be more efficient and different you know we'll see new names new projects both from liquidity sites from execution side from where like wallets and where people like are coming from. So yeah, it's something to watch and to study. I think it will be

exciting. >> Um, I want uh more hooks. I want uh LPS to be profitable. Uh like like just objectively on markouts more profitable. I want um yeah better execution quality that's like uh I want I want actually in the next 18 months I would love if DEX like onchain pricing is actually the true price and we don't actually refer to Binance price as true price anymore. Let's just refer to like onchain prices as true price, right? I mean, we have to get there, but like you know, we got a power team here. Um, what else? I think uh uh oh, can I chill?

>> I can chill. Okay, shill. Um, so part of the unification proposal that I was alluding to before also has um we have the uni burn, but it's not a regular burn. uh we actually have it split out so that there's a token jar which collects fees for from unis swap um and then it has a separate fire pit which burns the uni token and someone a searcher takes the fees or t yeah takes the the token fees. So this comes from like all V2 pools uh some V3 pools. I would love for people to take that token jar and do other things with it. Take the token jar and then give it back to LPS have more yield generating strategies. make LPS more profitable or take it make pools deeper and make the pricing better. Uh we can have like

jitting into pools with the token jar so that we're actively always making prices better so that all onchain liquidity is true price. Boom. Done. >> Really happy to leave that on a bullish note. Um with that, thank you all the panelists and thank you everybody for coming. >> Thank you. Thanks. >> Thank you guys.