Epicenter - Learn about Crypto, Blockchain, Ethereum, Bitcoin and Distributed Technologies - Hyung Lee: Crescent Network – The Hybrid AMM/Orderbook Decentralized Exchange

Episode Date: September 23, 2022

Crescent is an AMM/Orderbook Hybrid Decentralized Exchange that allows trading inter-chain assets in a cost-efficient manner. Crescent is a sovereign Cosmos chain that was created after migrating the Gravity Dex from the Cosmos Hub. Crescent particularly focuses on efficient use of liquidity by giving powerful tools to market makers to use their capital and provide good prices for traders. We were joined by the founder of Crescent Network, Hyung Lee, to chat about his involvement in the Cosmos ecosystem, B Harvest, the Gravity DEX, and combining AMMs and orderbooks. We also hear about Crescent's roadmap and future role within the ecosystem.Topics covered in this episode:Hyung's background and how he got into crypto and the Cosmos ecosystemLaunching Crescent on its own chainThe vision for Crescent's role within the ecosystemLiquidity formation for the projectComparing AMM and orderbooks and the advantages/disadvantages of eachHow AMM and orderbooks have been combinedThe future of AMMsDifference between a market maker and a liquidity providerCross chain dex aggregationMEV and the interchain worldEpisode links: The Crescent EthosCrescent NetworkCrescent on TwitterHyung on TwitterSponsors: Omni: Access all of Web3 in one easy-to-use wallet! Earn and manage assets at once with Omni's built-in staking, yield vaults, bridges, swaps and NFT support.https://omni.app/ -This episode is hosted by Brian Fabian Crain & Felix Lutsch. Show notes and listening options: epicenter.tv/462

This is Episode 462 with guest, Jung Lee. Welcome to Episode on the show, which talks about the technologies, projects, and people driving decentralization and the blockchain revolution. I'm Brian and I'm here today with my co-oes Felix Luch. And today we're going to speak with Jung Lee. He's the founder of Crescent Network. Crescent is a decentralized exchange, AIMM, orderbook exchange on cosmos,

a very innovative project, so we're really excited to speak with him about that. Now, before we get into Crescent, we just wanted to briefly talk about a sponsor this week. So our sponsor is Omni. Omni is your new favorite multi-chine mobile wallet that puts the power of Web3 at your fingertips. You know, in just three tabs, you can stake and manage your assets and over 22 built-in protocols, including all the EVMs, all the major EVMs, L2s, and also non-EBM chains like Cosmo, Solana, NIR, and more. Omni abstracts away the complexity,

but retains, you know, that it's fully self-custodial and makes it really easy for you to get yield on your crypto. They also have multi-chain NFT support, so you can view all your NFTs in one place, and you can show your nice, as an app background. So and in they have an explore section in the app where you can you know see new apps,

new yield opportunities and news. And yeah, so if you want to check it out, you can get it on iOS or Android, just look for Omni Wallets, that's OMNI Wallet or you can go to the website, Omni.5. And yeah, with that, Jung, thanks so much for coming on. Yeah, thank you for inviting me. It is great to be here.

Yeah, so we've known Jung for quite a while. And I guess the thing I remember especially was that in that was, I guess, 2019, there was the Middle Conference in South Korea, and there was the cosmos hackathon there. And, you know, with course one, we were participating there and we were actually working together with Jung and his colleague. And we were working on some cosmos liquid sticking stuff for the hackathon. So I think that was very fun to work together.

So we've known Jung since then, spent a bunch of time there. We were very impressed with his especially understanding of Defi and of sort of liquidity, markets, incentives, financial instruments. So it's great to finally come on an epicenter and talk a bit about some of the work you've done since then. Yes. Yeah. Happy to be finally be in this epicenter talking about things.

So yeah, especially from myself. personally, like meeting these like OG people way back in like 2018 and 19. It feels really good and it feels like family. So I'm very happy to touch with you guys again. And hopefully we can see each other soon in Cosmoverse also. So maybe you can just talk a little bit about your background. How did you get into crypto and how did you get into the Cosmos ecosystem? them in particular?

Yeah. So I majored in mathematics in undergraduate, and then I studied some global finance in master, and I worked as a market maker and arbitrage trader in Investment Bank in Korea and for about 10 years. And then I'm trying to have another chance to participate in this new move of technology, including like AI and blockchain and other stuff. Because from my nature, I was like finding something more innovative and more transparent

and more decentralized technology with a lot of new philosophy, like not very conservative. So from my older job, that was like very conservative environment with very power and money game within. And it was quite boring. And I wanted to have another chance to bring this again in my life to get another energy to move on. And the blockchain just caught my eyes. So, like, at first, like, 2018, I was, like, deep into the Ethereum ecosystem, trying to build some financial application from it. And I was trying that, like, for three months.

And then, finally, I had made a transaction costing, like, millions of dollars. And I thought that it was impossible. So I tried to find another projects that was, like, Cosmos. And I just, like, caught up with this. like big vision, millions of blockchain and connected each other. And I just believed it in one side and then just like keep doing this for more than four years as now. So that was like my history, how to get into cosmos ecosystem.

Yeah, super interesting. And then I think you started out basically as a validator as many in the cosmos ecosystem. Maybe you can talk a little bit about how you made that choice to start there and then how you kind of move to the to the decks space or what you're what you're building today. Yeah. Yeah. So at that time, like POS is even not very universal in the space. Like everything was Ethereum and ERC 20 and a smart contract. So it was very new in the blockchain area.

So I was trying to understand and get some experience and knowledge on this DPS environment and also studying deeply more about Cosmos SDK and Tendermint consensus layer so that we can fully understand and utilize this for a new application layer. So like we started with validator participating in TestNet for about, I think, about like nine months until Main Net launch in 2019. And then started our validator business with participating in different projects. While that, we were like one of the validators who are like deeply reading the blockchain core code basis. including like Kava or Terra or other earlier projects in Cosmos. Even we were participating in several projects as an auditor.

So we wrote some like report on these financial applications. So from this like knowledge and experience, we started to build our own capability to understand and innovate this device. space. So we started to participate in other activities, like working with Tendermint Cooperation and also ICF to build some modules and features for Tendermit and Cosmos SDK. So one of the results was Gravity Decks operated on Cosmos Hub with AMM and AutoBook at the same time. So this is like one of the information who people doesn't know. is that because the UI was Ameris and the Ameris that didn't provide orderbook UI.

But actually in the codebase, order book is already there in 2021. That was like hybrid decks from beginning. Original design was hybrid decks. But we didn't allow, enable it because it is too experimental for Cosmosop. So that was our start to build this one of the Defy application, which is like Dex, with AMM and Audubc at the same time. So that was our first big move into this Defy space. I think there's definitely a lot.

We want to go into there with this like Audubc, AMMs, but maybe just talking a little bit more about. So, you know, gravity was was on the cosmos up. Can you talk about, like, you know, what happened and like how did it evolve or result in, you know, you then launching crescent on its own chain? Yeah. So as a big time believer in Cosmos and Adam, we were trying to persuade the community to have decks in Cosmos up. and incentivized from atom inflation.

So that was our purpose to persuade the community with this. Because for any decks, especially with AMM formation, liquidity incentive is the core part of the functionality. So we believe that there should be some incentive coming from Cosmosop. and that was Adam because the community doesn't allow to mint another token in Cosmosop. So that was only, atom is the only option. So we are trying to persuade the inner community in Cosmosop, trying to persuade to use this atom inflation for this important utility.

But there were some objections. with like hub minimalism and supporting this like new projects in in cosmos ecosystem not by interfering this business with from this cosmos hub like taking all the market share. So we we thought about it and we also discussed about it with a lot of people and we decided not to pursue that because eventually without the decks in Cosmos Up, there will be lots of decks in Cosmos, which is already happening right now. So that would be a more resilient future for the Cosmos ecosystem, having a lot of Dexas with different characteristics and advantages.

So we decided not to have this incentive. for GravetyDex. And finally, we decided to migrate this gravity decks into our own chain so that we can proceed with all the innovation and features we want to build upon. So that was the history for the decision making. Awesome. Yeah, I think before we get into like the features and what you're working on with Crescent. Maybe one thing that would be interesting. We just earlier looked a little bit in the UI, looked at the governance proposals.

And the very first one is about the Crescent ethos, which I found very cool that you make like a voting to kind of approve the ethos you guys have. And maybe that would be interesting if you could tell us a little bit, you know, what's the vision? What are you trying to achieve with Crescent? Yeah. So I think decentralized governance is a hard problem. There's no absolute answer to this because we don't see any future yet.

There will be lots of different approaches to have this robust decentralized community and decision-making process. So we believe that there should be some balanced approach for this governance. There should be some strong leadership from this project builders. And then there will be another decentralization factor from community and token holders. So I think there should be a great balance between this decentralization and leadership. So that is why we. push this at those because we want to have some amount of leadership from our team, but also we wanted to want it for the community to understand our direction and our vision,

which will be very important for all this governance decision-making process. So that is why we write this as an on-chain governance. so that everyone can transparently see how we think about the future of Crescent and can be a reasoning background for all these other detailed governance proposal we are making. So what we have in plan is that we are going to write several ethos in coming months so that we can write in detail about all the different topics and also write about all the changes in environment and changes in our vision and roadmap and directions so that we can keep communicating with the community and write down all of this evidence in on-chain

so that we can make a better trust from this community. But maybe looking at it a little bit differently. or more generally, like, if you think sort of in the long run, you know, if Crescent really succeeds, what do you think the role of Crescent would be, you know, in the crypto ecosystem? Like, what does it look like, you know, a Crescent in like, you know, 10 years when it has, like, worked out amazingly well? So I think there should be several roles we hope to take in Cosmos ecosystem, but like most imminent. role will be liquidity formation for new projects.

So currently, there are lots of new projects coming into Cosmos. Within a year, I expect there will be like 50 to 100 new projects will come to Cosmos ecosystem, and they all need liquidity for their tokens and other tokens they are creating, such as stable coins and liquid-staking tokens and other stuff. So we believe that it is a very hard job for us to have all of this liquidity in the ecosystem. And there should be a better way, smart way, an efficient way to provide this liquidity formation with sustainable methodology. So we believe that Crescent is one of the teams who can do this role at best. we do our best to take this role for customer ecosystem.

When you say liquidity formation, do you have in mind, you know, like crowdfunding? So basically a project wants to raise money and, you know, sell some token to the market or like just that there's like, you know, liquid markets for these tokens so that, you know, people can sell them and buy them. or you mean, I guess, a mix of both? What do you think is the most important aspect to liquidity formation? So I think fundraising and liquidity formation is a little bit different topic, which can be a little bit related. But I think our role will be more on this liquid deformation. But still, from this process of liquidity formation,

there can be like additional token sale approach, which is called like IDO. So I think that can be another role crescent can take, but in more like a connection to the liquidity formation process. So liquidity formation is more like an ongoing liquidity so that any token holders can buy or sell this token at any time. This is completely different from fundraising because fundraising only allows you to buy this token within specific time period. So liquidity will be more like a continuous liquidity for users to trade anytime they want. All right. Cool. I think that leads us into the meat of the discussion where I think you already mentioned a bit that Gravity Dex and I think also Crescent basically has. this hybrid AMM order book approach.

Maybe for the listeners or also because it would be great to hear you talk about it, can you compare kind of what is the difference maybe also between AMM and the order book and what are the advantages or the disadvantages of either of them and how you chose to combine them in the end to create a better product? Yes. So we saw all this innovation coming in in 2020 with Uniswap and other Dexies to come, like Coff Finance and Balancer. And we believe that there is some area where AMM was popular in Ethereum ecosystem with very certain reasons, including listing was like a power game within. centralized exchange. And also gas fee is too expensive for decks to provide

orderbook utility. So that was like more centralized orderbook decks was there before that but it was not very successful because it was too centralized and gas fee doesn't allow Ethereum ecosystem to build order book decks. So there was some reason for AMM and there was some reason why orderbook decks was not successful in Itherium and I saw all of this like story of evolving this tax feature and utility and we we saw that actually in Cosmos there's no reason to ignore this great utility which is order book decks but still AMM provides a great way to form initial liquidity because like market makers will have a lot of hard time making this market

because of this huge risk of inventory management for this unhatchable tokens. So I think there will be like some equilibrium point where in some circumstances AMM can do better and in some circumstances, Audubu can do better. So why not take all of these advantage in one marketplace? So that was like very obvious conclusion from our team to build this from start. So when we build this gravity decks, most easiest solution will be just building this AMM, like copying UNICEF version 2. That was like easy conclusion.

But we didn't do that because we saw obviously. order book will be the better approach in Cosmos. So we did, instead of building basic AMM, we built hybrid decks from start. So now we are doing our best to take these advantages from these two different features at the same time. So I think utility is not the all for our project to success. because we need like a company business effort, partnership, and market maker incentive,

and LP incentive plans and strategy, all these things come together to make the project successful. So we are trying to do that right now, doing more business and doing more partnership, so that we can bring more users and partners and token. within the crescent decks. So can you explain, like, how do you combine, how do you make a hybrid text? Like, how do those two sort of interact with each other? So this is more like how you see AMM. So if you see AMM as a stand-alone innovation

from nowhere, then it is very difficult to understand AMM in a concept of orderbook. But if you see everything within orderbook perspective, and AMM, you can see that AMM is restricted version of orderbook with very restricted freedom for market makers and traders. So traders cannot do limit orders, only market orders. And market makers can provide liquidity only following this constant product model formula. So this is like a very small subset of orderbook. This is, so when you see this space of freedom for this design for the decks,

then it is very clear to understand that AMM is a subset of orderbook. So when you see this way, then combining these two is much easier, because you can see AMM as orderbook and order book feature. And then we built everything based on order book, and then we place this AMM on order book base ground. So what we did is to describe this AMM liquidity into multiple limit orders on orderbook so that we can calculate each amount in each tick of the orders, which is completely perfectly following the constant product model. Then we can derive exact

number of amount for each tick. So we just place all of them within the order book, and every order is coming from users and liquidity pools, they are matching each other without any prioritization or special treatment, they just act as another limit order and they just meet each other with the orderbook matching system. Another special thing about Crescentex is that we use batch execution, which prevent MEV from start. So this batch execution is matching this order each other, without any time prioritization within a block. So if there's multiple orders in one block with different sorting, but still there's no prioritization of order matching within a block.

So it doesn't matter at all. So like Skip Protocol is doing M-AV-Solution, like providing ways for traders to lobbying the value, validator to make their transaction at top of D block, but still either your transaction at the top or bottom, the results will be exactly same. So there's no need for MEPA at all, because there's no difference. So this is another characteristic we have in our decks. So I guess, yeah, we definitely want to go a bit to that since that seems to be a very important feature obviously of Crescent.

I think when you say it like that, basically there is one within a blog, there's one clearing price then for the different orders and you would have maybe you know that you have three orders and they would all get the average price in the end. Is that how it works? Am I understanding correctly? Oh, actually, we have two ways to match each other with batch execution. Still, in batch execution, there can be multiple surprise within the batching. And in our first design in Gravity Dex, that was we used our universal surprise model,

which is what you described. we have only one sub-price for every matched pair of orders. But now we upgraded our decks to have more matching within a block. We need to allow multiple shop-price matching price for each block. So now we have multiple batching price within a block. but still we don't have any time prioritization, only price prioritization. So we pursue more optimization, but still we keep our fairness of trading for this like sorting of orders within a block.

So with the AMMs, right, I guess also there there's been a lot of, you know, movement, how this works, right? Because the traditional model, probably, you know, the significant number of listeners are familiar with is, you know, you have some pool. Maybe it's, I don't know, EF to die. And then as a liquidity provided, I can just put in EF and die, you know, in equal proportions into this pool. And then, you know, I get some trading fees based on, you know, the people. trading and then the price changes, right, then kind of sort of it will sell, you know, maybe even buy, die or the other way around.

So you have this impermanent loss to deal with. But, of course, the nice thing is as a user, as a liquidity provider, I don't have to think about it. I mean, the decision's very easy, right? I just put money in there. Everybody puts money in there. It doesn't really make a difference. you know, there's no

maybe where the sophistication does come in is understanding the risk you're taking, understanding the impermanent loss and how to deal with that because that's not so simple and that I think is very confusing for people as well. But of course we already have with Uniswap V3

basically you know a big change where I think I can be like, okay I'm going to put money into this pool but now I'm going to specify that, you know, I'm concentrating this liquidity on a certain price range. And of course, that already goes a little bit in this hybrid direction, no, because it's, it's now all of a sudden that decision is an important decision. You know, where do you concentrate this liquidity? If you know, you concentrate it in the wrong place, maybe you liquidity doesn't get used and you just don't

run any fees or maybe otherwise you take more risk. So already is something that I think as a normal user, you don't really know how to deal with this stuff, right? That's a normal equity provider. So what, how do you see AMMs in the future? Like, how do you think they will evolve? Oh yeah. I just want to like share some situation in the future. In the future, like, how do you think they will evolve? Yeah, I just want to share some situation in Ethereum first, because, like, obviously, Defi's in Ethereum is, like, heading to the future, so we can expect the future of cosmos by researching this Ethereum ecosystem. So, like, one year ago, UNICEF version 3 launched, and now,

If you can see this liquidity, the version 2 liquidity is almost like done right now. There's no much liquidity left and all the liquidity is coming from version 3. And if you see this liquidity formation within all these pairs, you can see that this liquidity is like very dynamically moving. So this is more like a very sophisticated strategy for liquidity. provider to provide this liquidity with very frequent rebalancing of the position and like more proactive liquidity providing methodology. So I think that even within this situation, we also heard that SISO service launching also same thing, which is called Trident, which will be coming very soon.

And I think in Sushi sub also this new feature will dominate all the liquidity because obviously it is more capital efficiency. But difficulties is not the reason for avoidance. Even all the difficult solution can be utilized by smart people and this can be a better solution if it is more capital efficient. But exactly, directly, the retail users cannot use it, or if you use it, then it will be very difficult to manage. So that is also another fact. So what will be the solution? And just briefly, you say, so when you say capital efficient, what you mean is basically that, let's say somebody wants to go and they want to trade, then, you know, one way, to measure it I get is sort of the slippage of like, you know, how much does your trade,

does your price move away from the current market price, you know, a certain amount of money. Let's say you're trading $10,000 worth. And then would you define like the capital efficiency that, you know, the uniswap v3, you will need a much smaller amount of capital in this pool to basically provide the same kind of, you price for somebody trading $10,000 than in like Uniswobb V3, V2, something like that. So I think we should define the word liquidity. So the liquidity is not TVL. That's for sure. So that was like one of the biggest misunderstanding in defy space and Dex area. Everyone is watching for TVL in Defi Lama or Defi Pulse and other dashboard having like TVL 1 billion grade.

But it doesn't like prove all this utility because the liquidity and TVL is completely different within this like modern Defi world. So like this is liquidity should be defined as like a, when you want to trade a token without moving the price more than 1% or 2%, and then how much amount you can trade. So that is like in coin market cap, you can see liquidity plus minus 2%. So this is like very basic approach to measure this liquidity because it can, it, it, calculates all the amount of liquidity you can consume immediately within this price range. So this should be more like a measurement for the liquidity. So from this point, having same liquidity from smaller TVL means capital efficiency.

So we talked a lot about liquidity provisioning now. We also mentioned market makers. can you maybe kind of detail what is the difference between a market maker and a liquidity provider and maybe what's the role of the market makers in Crescent? Yeah, so because we are hard-billed decks with AMM and AutoBook at the same time, it is very important for us to make a synergy with these two different approaches. So AMM provides most robust liquidity. Also, in a micro sense,

a liquidity pool provides better liquidity in a very narrow price range, which is like narrower than 0.2% or something. So if you are swapping very small amount, then it will be very efficient to use this liquidity from AMM because like AMM is providing liquidity exactly from current price to very near near the current price. But this order book market maker will provide liquidity not very near the current price.

It will be distanced from current price for about like 0.1% to 0.5%. For the more volatile token, it will be like 1%. So there are some different aspect of liquidity provided within this micro-liquidity area, different distribution of liquidity, and different characteristics also. So when you trade in AMM, if you trade one time and then the price will move, and then it will not come back until arbitrator trade back. So if you trade again, and the price will be worse, because you already move the price.

So this is like inefficiency of AMM. But for the orderbook market maker, the market maker will provide the price, and then if you consume it, and they will replace the order again at this exactly same price again. You can consume it again. So I call this rechargeable liquidity.

This is not the case in LP, but the market maker can do this because they can hatch in other market. So these are like very different characteristic for these two different liquidity, and there should be a good way for our community and governance to decide and calculate the contribution of liquidity coming from different methodology and measure this support of liquidity from different methodology and we need to distribute our incentive wisely so that we can create this very healthy competition between LP and order book market makers so that we can optimize this capital efficient liquidity providing incentives.

So probably a good amount or listeners have actually provided liquidity before, you know, whether that was on uniswop, curve, sushi swap, or, you know, like, osmosis, very popular, also on Crescent, right? So there's like many different places where people, you know, have kind of your normal people or like, okay, I can provide liquidity. What's, you know, what's going to happen with that? Is there a future where, you know, normal token holder can still, you know, provide liquidity on these exchanges?

or do you think this will all be replaced by, you know, professional, sophisticated market makers that, you know, can just play this new game that's arising? So I think great thing about blockchain ecosystem is that everything, every different protocols can be connected each other without any trust assumption or verification assumption. So everything, everything can be connected very well with this code base only.

So if we can trust this algorithm, then everything can be connected very well. So I believe that even though the management of capital will be much more difficult in range liquidity, concentrated liquidity world, still there will be like professional fund manager participating in this capital management and providing capital from retail investor for this fund. So which is called, we are calling this Managed LP. So managed LP is a fund operated by professional fund manager, but invested by all these like retail investors. So within this minimal trust within this fund manager. We can trust this manager to manage this capital with very restricted functionality within this decentralized tokenized fund. So I think the future will be

more like a ranged LP fund operated by professionals, but invested by retail. So one of the example is called Arachis in Ethereum, which is providing this platform for users. So this fund manager has very restricted freedom about how to manage this fund. So it is quite safe for users to invest into because they don't allow any send out all these assets within the fund. Only they can do is like a minimal amount of swap or like changing the range of the LP, that is all the things manager can do. So I think more of this functionality will be available in the ecosystem.

So I think retail investors can participate, but the actual operation can be more delegated to professionals. All right, that's cool. So we can remain being LPs in the future, too. that's good um so we wanted to i guess maybe switch topics a bit now we talked a lot about kind of liquidity in the context of this decks of this single network now obviously the cosmos ecosystem as you mentioned already potentially will have like thousands of of networks and already has a lot of different dexes and yeah maybe it would be interesting to to hear your

thoughts, how like you can aggregate liquidity across these different ecosystems and maybe at the role that Crescent plays in this story? Yes, I think Dex aggregator is one of the interesting topic to discuss also. So I think in Ethereum, it is a transitional phase right now, like one-inch users are trading maybe more than UNISOP front-hand users right now. So this is like a golden cross of Web 2, evolving to Web 3. So like Web 2, Web 3, the definition is that when the traders are locked into one front-hand, which is Uniswap, then they cannot utilize other liquidity coming from balancer or Sushi-Sub

or other products, so you are jailed into one website so that your backend is also jailed too. So this is web 2 in our definition. But when you use 1 inch, then you can reach out to all these decks even like very early one because you don't need any trust on this. If your shop is successful, then any decks you can use because you believe in results, not the service provider.

So 1 inch is more like a Web 3 service provider who are connecting all this existing back-end service by one front-end. So this is like the change of service within Web3 environment. And unfortunately, in Cosmos, we are in Web 2. Most of the users are locked in certain decks, and they are jailed there until now. So we want to free them. So when we want to free them, there should be aggregator for Cosmos, right?

And there are several projects coming in, such as Rango Exchange. But I'm looking deeply into a solution called squid from Axela. So the problem within this Cosmos ecosystem is that all this environment is asynchronous compared to Ethereum. Ethereum is everything is synchronous, but in Cosmos because all this application is in different blockchain, this asynchronous create problem of delay. and latency of like trade execution. So like especially like IBC transfer is really slow and unstable. So we need to wait like one minute or more for this one execution of trade with IBC transfer in it.

So if you use like Van Gogh exchange, then sometimes it takes like more than a minute to complete a trade. So this is not the good UX for users to experience. So Axeluxkid is more like skipping all this IBC transfer layer by utilizing liquidity relayer. So liquidity relay will do all this job for the behalf of the users and then get this reward back from the users after the swap is coming. So in this model, they can erase all this latency caused by interchain communication, but only like if there's a block confirmation coming in one second in crescent and like another block confirmation in two seconds in osmosis, then you can do this like interchange swap or within two seconds. like routing crescent and osmosis at the same time. This is practically not possible right now because of this transfer between blockchain,

IBC transfer will take at least several tens of seconds, but if you remove this, and this is like a matter of second. So I think this solution can bring more practical sense of interchain dex aggregator, because it removes all the unnecessary delay from this. So I think, like, IBC transfer is not a dependent variable for liquidity relayer and interchain swap functionality. So, like, this, we want to, like, work with them. We are communicating with them so that we can provide this option for, the users. Users does not know the benefit of this aggregator, so we want to promote this

and marketing this so that users can use more of this aggregator for themselves, because your swap price will be much better than use only one dex front-hand. So within this environment, ecosystem can be more healthier because the competition will be not about marketing, or business, or front-end lock-in, but more efficient liquidity, more cheaper trading costs and stuff. So it is always good for users to have this better, healthier competition, rather than spending a lot of money for marketing and the community locked in to one platform. So we believe this Web Street world will open all this new environment for all these new decks like Cusera, say all these deckses are new, but it is very difficult for them to attract new users.

But this aggregator will allow them to attract users without marketing effort. So this will be better for all this ecosystem. So one question on this, and you maybe kind of touched on it, but the, you know, one of the things that if you, in the cosmos ecosystem or in general, right, if you have things that are like cross chain is that, you know, you can't make these atomic transactions that in Ethereum, right, like you have some, some transaction that, you know, like automatically does different things on different chain, different, you know, using different smart contracts and you know everything works at once. So do you think this is like an issue when it comes to this cross-chain dex aggregation? Or like how is this, what are the implications of this? Yes, there should be some issues. So relative sense, Ethereum ecosystem is better at aggregating everything.

everything because it is synchronous and asynchronous environment will prevent some optimization for users. But still, it is better than aggregating nothing. So it is more like relative sense. If you can do 90%, and then still 90% can be quite good. So everything is like very statistical. So if you trade like 100 times, and your trade is most optimal for 90 times,

and that is good utility already. So I think asynchronous environment can result in non-optimal shopping route, but still when you achieve this for most of the time, then the decks aggregate gator makes a lot of sense. Cool. Does it change anything? you mentioned that Crescent basically gets rid of front running or MEP in the sense of like on the decks with the batch patches.

If we're talking about this aggregator interchain world, does that change the equation somehow? Do you or like I guess how do you generally see MV in the interchained world on a high level maybe? So, like, there was some data coming from Skip Protocol about the MEVE situation within Osmosis. So they had about like 6 million of profit from MEPE trading. So I think it is quite large. And when our ecosystem grow, the MEVE problem will be increased. over time with more amount of money extracted by all these smart traders.

But still, I think most of the profit coming from Ossumazis right now is with this USTC, which is like an older version of UST from Runa. So it is very, like, restricted area of MEV execution, because, like, most other pairs does not have significant MEV problem right now. That is what data is telling us right now. So maybe MEV problem is, like, overstated, especially in cosmos ecosystem, where, like, gas price is much cheaper. and like the defy is much smaller than either of right now.

So I think it is a little bit overstated. And in interchained world, still, MEVE can be a tool for extracting value. But still, I don't think it is like preventing any core functionality or creating significant economic problem yet. But I think it is very sensitive within very narrow range of applications, such as like auction and trading venues like Texas. So in those applications, I think it is better for our architect to build this application. in a way that we can prevent this from start

so that we don't create a solution problem which does not need to exist. So I think that is one of my philosophy. Okay, yeah, definitely thanks for expanding that. It makes sense to me. I think maybe I guess you kind of touched upon it where you talked about a little bit about UST just now and that a lot of the MV came from this.

Now, obviously that was a big event in the cosmos wider ecosystem, the Luna collapse. And as far as I understand, also Crescent had like Luna assets listed. You know, maybe can you talk a little bit about how that impacted the Crescent and maybe also the lessons you learned from that? for Crescent as Dex, but also, yeah, maybe the Cosmos SDK or in general the ecosystem. Yeah, so we had a great lesson from this Luna crash. One of the biggest lesson is that it is very dangerous for us to decide any canonical solution for any tax-related liquidity or tokens or bridges or other solutions.

Because I think in cosmos especially, it is about variety and robustness or resilience of the ecosystem. It means that a minor project should be respected enough to provide all this connection for the users. So our philosophy has been strengthened from this experience because even though the lunar was, Terra was one of the biggest ecosystem in cosmos, it just vaporized within one week. That is like how blockchain evolve. And even for current status, a very small project can rule the cosmos within six months. that is also true. So we are more humble right now.

We don't believe in decision-making, all this centralized into one solution and stuff. That is why we are providing incentive for both bridges, axler and gravity bridge, because we are not sure anymore. Everything is humble, but still small projects should be respected within this variety of ecosystem, especially in cosmos, that is the philosophy.

So we believe that there should be lots of different stable coins, lots of different bridges, lots of different liquid staking in one deck. So like variety of choice will be our philosophy for Crescent, and we are going to pursue this philosophy to provide better, more option to users. But I guess the downside of that being that, I mean, what's the argument? The other argument would be that now you're fragmenting the liquidity and the users maybe are confused and don't really know what to use. Do you see that problem too? And if yes, I guess how do you address that?

Yeah, so that is very interesting problem to solve. So, like, for example, if you have, like, different kinds of stable coins, then obviously the liquidity will be separated within one decks, but still, the liquidity is connected each other. If you have, like, IST versus USDA from gravity and USDA from axler and then die from gravity, those like all of these like stable coins pairs are connected each other. So if we have this like multi-hop swap within feature, so which is coming very soon in Crescent, the liquidity is connected each other.

So there will be like multiple routes you can take to have this best swap price. So every liquidity is connected with this smart features. So if we can utilize these smart features within the decks, then we can connect all this liquidity so that users can utilize this liquidity. Let me have one example. So if you want to trade A through to D token, and you have like pairs with A with B and A with C, and then B with D and C, then B with D and C with B. D. Then you have two routes, A to B to D and A to C to D. And then you can utilize routing through B or C to be result in D. So actually you are utilizing both liquidity. So if you're swapping large amount of liquidity, then you will separate this amount of swap into two

route, A, B, D, and A, C, D, then it is just like swapping from A to B plus C combined to D. That is like exactly same result. So if the liquidity is connected each other very well, then the separation is not the correlation. One other thing we wanted to touch on, I guess this ties into what we talked about, you know, in the beginning, you know, the hackathon we did in Korea together, right, which was around liquid staking. I think back then there was like, you know, different liquid staking ideas. You know, we had been working on this kind of delegation voucher idea that, you know,

we sort of also built on in Korea. But you had another sort of other design for liquid staking. And of course, liquid staking is actually also one of the things that, you know, built natively into Crescent, I'm actually not sure if anyone else's built liquid staking, you know, into the kind of base protocol, like the way you guys have done. So can you talk a bit about like, you know, how does liquid staking work in Crescent? Yeah. So first of all, way back in like 2019, I really enjoyed brainstorming with you guys talking about different ways to build this liquid staking problem. And it was great motivation for us to study further

about this liquid staking area. So we were researching different kinds of solutions. But in our Crescent, the solution is not very universal. This is a restricted solution with some assumption, which is called liquid staking validator. So we assume that there is a wide list validators who are participating in liquid staking activity, and they will separate the delegation power evenly each other. So in this assumption, liquid staking is much easier. We just distribute all this power, and then when you underlegate,

then you will just undelegate all this same power from all these validators and give it back to you. So this is like very basic stuff. And also when you have enough staking rewards accumulated, then automatically in the end block, the module will just rebut. withdraw this reward and restake. So this is quite basic stuff.

But from this innovation, we have more interesting to come, which is called liquid farming. So liquid farming is not very universal word right now. There's no much application using this word or using this concept for the defense. But it is increasingly important within this current cosmos ecosystem right now. So there are lots of capital locked inside LPs. So users are trying to reutilize this capital, but it is impossible for them to reutilize this capital for other purposes. So, for example, if there's a pool with atom versus Q-atom, which is coming from Quixilver,

and there will be like farming rewards coming from Quicksilver and Crescent. So the APR will be quite fruity, like maybe 30% or 40%. And then the reserve tokens are just atom and Q-Adam, right? And these are all atom, actually. So if you want to use it for like collateralized lending, then you can receive all this farming and use it as a collateral to lend some stable coins. But if you use a collateral for Q atom, then the stacking rewards is all you got. Like farming rewards will be much higher. So this will be a good solution for users to utilize.

this pool token as a liquid farming solution. So this will be like a minted token coming from LP position, but auto-compounding this farming rewards. So liquid-staking, liquid farming, a similar concept with different purpose or target to liquid-requify. So liquid-farming is liquefying the LP position. So this is another concept we are building. I think like there's like next month we are going to launch this.

Okay, cool. So just so I understand, the liquidity position, because you just have, there's not the range liquidity in Crescent. So basically you, every liquidity provider is the same. There is no, as long as you provide liquidity, they are fungible between, each other and then you were able to use that. And is that actually a stable pool or how do you like I guess is it using the constant market the function is, do you have different functions actually?

I guess we didn't ask this in the thing. But do you use, is it possible to kind of utilize different curves in Crescent for let's say stable coins versus. you know, just a constant product market, the one? Yeah. So the ranged LP feature is just like Unisop version 3, utilizing constant product model. So this constant product formula will be used for shopping those two tokens within predefined price range. So our calculation is that when you provide liquidity within 10% price range, then the leverage of liquidity will be 40 times about. So if you provide like 1 million liquidity for Q Adam versus Adam, then it is equivalent

to 40 million liquidity in osmosis. So this is 40 times capital efficient. So yeah, that is how we formed this arranged LP, which is completely different from the approach coming from curve finance. Curf finance is using a different curve, but ranged LP is providing

LP with the same formula used in UNICEF version 2. Okay, so you predefined this range for this specific pool that's done by the team essentially and

And that's how you achieve the fungibility, I guess. But I guess maybe that goes a bit into detail. But I'm wondering, because I guess the Q atom might accrue rewards as far as I understand, do you have to change the mid price there, the liquidity? And how would you do that? Yeah. So like it should be manual right now. So like we are going to provide like this is.

10% is like very practical range of price because Qatom is accrued emulating about max 20% of APR right now. So in quarter it will be like 5%. So 10% is like a wider enough range for this pair. So we create this pair and then after one quarter we create another pair and then we distribute farming to this new pair. So people will migrate this liquidity into this pool, and we are going to provide front-hand

so that with just one click, they can just migrate this liquidity to new pool. So this is like a manual process right now, but there are new approaches from this like car finance that the price can be automatically, the price range can be like automatically rearranged from this Oracle or like surprise, past shop prices. So we are currently researching ways how to mitigate this so that we don't need to like constantly like migrating this liquidity to new pools.

So that is like one of our roadmap. Yeah, I mean, maybe talking a bit about roadmap in general. Yeah, tell us like what are the main milestones that are coming up for, you know, Crescent in the next, you know, two years. So imminently, like next month we are going to launch farming version two with this Riquid Farming Fehrase. So in farming version 2, we allow farming distribution within token pair, not token pools. So because we have ranged pool with one token pair, like Adam versus Bitcoin, we will have multiple pools.

Already we have multiple pools. But right now, we are distributing the incentive, targeted. to each pool right now. But after farming version 2, we are going to distribute this incentive into this token pair, Adam versus Bitcoin. And then from this liquidity contribution

from each pool is calculated automatically so that we can distribute these rewards in a very fair computation so that we don't need to disarmes a farming rewards for each pool, but only for each pair. Then each pool will compete each other for their liquidity providing. So that is how Unisab is doing right now. So Unisab, if you see Unisub, there's like two kind of pools,

Ethereum versus Dai with like 0.3% commission fee and then 0.05% commission fee. And based on circumstances, the trade is executed within these two pools, and these two pools are competing each other to provide optimized liquidity for the users. So this is like same thing in farming version 2. So now farming version 2 will allow pools to compete each other for the best liquidity they can provide. So this is like imminent feature. and we are going to launch also after-swap, after-batch swap, which allows multi-hop swap swap. So like if you want to swap A, routing to B, C, and then to D, then this will be allowed in our next upgrade. So this is also in plan.

Yeah, we have several more. But that is, I think, most imminent upgrade right now. We are going to also launch managed LP within, I think, like, five months. So managed LP will allow users to invest into LP fund so that they don't need to manage their position, but delegate this to a professional fund manager. So this will be also launched. only next year. Cool.

Very nice. Well, thanks so much, Jung, for coming on. It was really a pleasure to speak a few about, you know, Crescent and about AMMs and markets in crypto in general. I mean, this is obviously an area of, you know, enormous innovation, enormous speed of development. So I'm really excited to speak a few about it. I think it's also becoming an area of great complexity, right?

that's like getting hard. I guess like so many things in crypto, right? That's getting harder to like wrap your head around it for sort of the, you know, the normal crypto person. So hopefully this was helpful also for listeners to like understand a little bit. Like where is this all going? And, you know, how can people still participate? You know, so I'm excited that there's going to be this managed liquidity coming so that,

you know, normal people can still participate in all of this. And so yeah, thanks so much. I'm excited to see where Crescent goes and where it goes next. And hopefully we can have it on again at some point in the future to get an update about how things are going. Yeah. Thank you for inviting me and looking forward to see you guys in Cosmoverse. Thank you, Hugh. See you soon.

And thanks so much for our listeners for tuning in. If you want to support the show, make sure to leave us an action's review. Let us know on Twitter what you think of the episode. it. And thanks so much. And we look forward to being back next week. Thank you for joining us on this week's episode. We release new episodes every week. You can find and subscribe to the show on iTunes, Spotify, YouTube, SoundCloud, or wherever you listen to podcasts. And if you have a Google Home or Alexa device, you can tell it to listen to the latest episode of the Epicenter podcast. Go to epicenter.tv slash subscribe for a full list of places where you can watch and listen.

And while you're there, be sure to sign up for the newsletter. So you get new episodes in your inbox as they're If you want to interact with us, guests or other podcast listeners, you can follow us on Twitter. And please leave us a review on iTunes. It helps people find the show, and we're always happy to read them. So thanks so much,

and we look forward to being back next week.