On The Brink with Castle Island - Matteo Leibowitz and Eric Stone on Uniswap V3 (EP.218)

Episode Date: May 31, 2021

Matteo Leibowitz of Uniswap and Eric Stone of Flipside Crypto join the show. In this episode we discuss: Uniswap v3 and how it differs from v2 and other AMMs Broader implications of Uniswap v3 for DeFi Uniswap v3 metrics and how the system is being used today Uniswap treasury and grant program How Flipside is engaging with Uniswap and other DeFi communities Follow Teo on Twitter @teo_leibowitz and learn more about Uniswap at Uniswap.org. Follow Eric on Twitter @theericstone and learn more about Flipside Crypto at flipsidecrypto.com.   Sponsor notes: Copper is transforming how institutional investors engage with digital assets by developing award-winning custody and next-gen trading infrastructure. Headquartered in London, the firm is scaling rapidly across Asia and North America to bring its suite of products to a wider pool of institutional investors. To learn more visit copper.co or reach out on Twitter, @CopperHQ Aave is a decentralized, open source, and non-custodial protocol where users can deposits and borrow digital assets, and earn interest on those assets. Head over to aave.com to experience and learn more about DeFi.  

Hi everyone. Today's episode is brought to you by Copper. Copper is the global provider of blockchain infrastructure solutions for institutional investors who are actively trading their digital assets. They have an award-winning custody application that's connected to 25 of the world's largest exchanges, and they ensure safe storage and movement of assets for some of the world's largest crypto hedge funds, market makers, family offices, and high net worth individuals. Copper's clear loop is the first off-exchange settlement network for digital assets, and it's the safest way to trade balances in custody across multiple exchanges without requiring unchain movement of assets. head over to copper.co forward slash clear loop to learn more. That's copper.com forward slash clear

loop. And today's show is also brought to you by AVE. AVE is a decentralized open source and non-custodial protocol where users can deposit and borrow digital assets. You can earn interest on those assets and head over to AVE.com, that's AAVEE.com, to experience and learn more about DFI. Today we have the pleasure of sitting down with two key builders in the industry. Teo Leibowitz, strategy lead at Uniswap, the most popular decentralized exchange by volume, and Eric Stone, data scientist at Flipside Crypto, which is a crypto data and analytics firm and a CIV portfolio company. The focus of our discussion was on Uniswop v3,

which is the latest upgrade to the Uniswap decentralized exchange protocol, as well as Flipside Crypto's involvement with Uniswap through its community grants program. We touched on a lot related to the latest upgrade, including the core change in the form of concentrated and customized liquidity provision, the early success of V3 in facilitating capital efficiency, and the implications of liquidity provider positions as NFTs and their impact on yield farming. Among other topics and hot takes sprinkled throughout. So with that, let's dive into the conversation.

Brought down by bad mortgage investments, Lehman, which has 25,000 employees, will be liquidated. The federal government loans, American International Group, AIG, $85 billion. This is a different kind of market and the Fed is asleep. The federal government is stepping it to stabilize Fannie Mae and Freddie Mac, the two mortgage giants that have been threatened by the housing crisis. The Bank of England has pumped 75 billion pounds more into Britain's ailing economy with a new round of quantitative easing. You print a couple trillion dollars and all of a sudden people start to worry.

So out of this worry, we have something called the Bitcoin. Bitcoin. Hi, everyone. I'm Ria Boutoria and today we're here with Teo from Uniswap and Eric from Flipside Crypto. Hey, guys, thank you so much for joining us. Hi. Hello. Hey, so we're super excited to dive into Uniswap V3 and Flipside Crypto

and get a better understanding of what you guys are doing with data dashboards and analytics and a little bit more into V3. But before we get into that, I'd love to hear about your respective backgrounds and what led you to Flipside and Uniswap respectively. And maybe we can start it off with you, Eric. Sure. I'm Eric Stone. I'm a co-founder and head of data science at Flipside Crypto. We started in late 2017, actually, as an investment fund in the crypto space and have evolved into a data analytics provider. So we have an open platform that makes it super easy for anyone who knows SQL to create dashboards and do analytics on crypto projects, primarily on Ethereum, but more chains are coming.

We just announced our collaboration with Terra, for instance, so that'll be rolling out very soon, very excited for that. We work with crypto projects directly to build bounty initiatives so that crypto products can source data insights that they need and grow the base of analysts who know their project. So analysts get paid to do the hard work, and the projects get analytics and dashboards that they need for their users to make better decisions about how to use the protocols. And my background is in statistics primarily. I am a statistician, and then I've sort of branched into the field of data science from there,

working anywhere from government and intelligence all the way to tech, online education, and now crypto. So that's me. Awesome. What about you, Taya? Hi. So, yeah, thanks for having me today. I'm Taya. I lead strategy at Uniswop Labs.

So we are a development studio that contribute to the Uniswap protocol and its surrounding ecosystem of products. So, yeah, a bit of background on myself. I sort of started my career in crypto as still fairly young in the grand scheme of things, and started out as a research analyst initially independently, and then I joined this outfit called The Block. that was sort of just as this sort of suite of products that we now know as as defy were beginning to come to market and so I spent quite a lot of time focused on that area also sort of digging into some data related stuff and yeah you know in my capacity at uniswop

labs I sort of get to apply a lot of the insights from my research phase in this more sort of practical fashion. And yeah, very accepting stuff. Awesome. So Dexas have really risen to prominence as like a killer app within Defi among a number of applications. And Uniswap has really led the charge with, you know, volumes across Uniswap and other Dex's rivaling centralized spot exchanges. Would love to get your take on what factors you think are driving the significant adoption and how Uniswap has maintained its dominant position despite pretty intense competition from generalized or specialized dexes. I think there's a couple reasons. One is that I think people are increasingly realizing that decentralized exchange protocols can offer user experience on par or actually superior

to that found on centralized exchanges, you know, by using sort of unhosted wallet like Metamast. it really serves as this sort of universal portal to a suite of financial products. And so it's this very seamless process to actually get started training on these platforms. There's the sort of interoperable nature of decentralized exchanges with other financial services as well. And so, yeah, by using decentralized exchange, you no longer have your capital siloed within a particular service. I think yield farming that kicked off in a big way in May of 2020, really accelerated adoption rates by bringing more liquidity into the ecosystem. And so suddenly, decentralized exchanges could start to offer execution quality on par with centralized exchanges. And then I think a really big part of it as well is the permissionless nature of decentralized exchanges on the listing side of things.

And so Unswap V2 offers access to over 30,000 unique tokens. By contrast, Coinbase lists about 60 different assets on its platform. And so, you know, for users that are seeking to trade assets sort of along the long tail, their only option, really, is to start playing around with decentralizing strangers. I'm sure Eric has a couple others. I mean, I think that's the fact that you can. can get liquidity on a new token or a token that Coinbase hasn't decided. They're going to cover that finance isn't listing yet.

That's a big factor. I think the yield farming is also spot on, clearly a ton of bootstrap liquidity on both Uniswap V2 and Sushi swap. That's sort of what you're started in the D5 summer with pool two's, meaning stake liquidity tokens being themselves staked on yield farms in other protocols is now sort of the becoming the industry norm for new launches in many respects. So I think that's driven a ton of the liquidity side. And I think the swap side is related to that, you know, getting your hands on governance

tokens that you believe will whatever give you governance rights or potentially become valuable. That's a fun speculative side of the market too, that decentralized exchanges. a lot. Yeah. I think another thing that really came into focus last week was the resiliency of decentralized exchanges in turbulent market environments relative to centralized platforms. And the fact that you can rely on that despite the crazy fees that you, you know, you still have to pay last week, I think is definitely a factor, at least for me, you know, that has driven me to using Dex's versus centralized exchanges at certain points.

time. Did Gway make it over 2000? I think so. But yeah, it sounds like this sort of perverse argument, but I think does ring true that, you know, these decentralized financial services are actually somewhat more conservative than their centralized counterparts, both in the sense that they offer, you know, 24-7, 365 guaranteed uptime, but also, you know, by sort of operating in this non-custodial fashion, it also eliminates counterparty risk entirely. And, you know, anyone who's been involved in the crypto industry for a couple years now knows sort of the risk that counterparty risk can introduce. So, yeah, I'd like to think about those properties too.

Yeah. I'd love to talk about Uniswap v3. And before we really get into the weeds, you know, I'd love it if you could lay out the salient changes. in V3 versus V2 and relative to other AMMs. And then would love to know what each of you are personally most excited about in V3. Sure. So I can give a little background on V3 relative to V2. I think the sort of crux of V3 is this concept of concentrated liquidity.

And to understand the advantages of concentrated liquidity sort of need to start with the default state that previous AMM designs, you know, popularized by Uniswad V1 and V2, brought to the table. So previous AMM designs, you start with this sort of equal dollar value of two assets. You have this sort of two asset inventory and you deposit these assets into these pairwise markets and these pairwise markets that exist as Ethereum-based smart contracts. and the smart contract contains a pricing algorithm, and the formula there is X-Dens why called K. So practically what's happening is that you're reserving small amounts of inventory

across the entire relative price curve between these two assets, between zero and infinity. So, you know, no matter where the market is trading at any given time, this smart contract, this, you know, univiswap pool will always be able to offer a taker, a trader, some amount of liquidity. So that type of design has some nice properties, namely that you sort of have guaranteed liquidity at any price point. And then it's also fairly straightforward to implement.

I think the Uniswop v1 contracts were a couple hundred lines of code, if that. The downside of this zero to infinity price curve model is that you have a lot of liquid. a lot of capital reserved at price levels where you don't really expect to see any trading activity and you don't really expect LPs to liquidity providers to be generating fees, right? So let me give you a practical example. Take a stable coin to stable coin market like USDC versus Tether. Over the long term, you sort of, you know, assuming Tether doesn't blow up in some spectacularly,

fashion, you expect these two assets to trade at this relatively constant price of one, right, one to one. But if you take the X times Y equals K model and you were to deposit some assets into this USDC, USD market, you're actually only reserving 0.5% of your capital for price activity between the prices of 0.99 and 1.01, which is really why you expect traders to trade in this market and to generate fees as a liquidity provider. And so that's incredibly capital and efficient, right? You're just reserving a huge amount of capital for these price levels where you will never really see trains.

So it sort of sucks as an LP because you're not sort of using your capital efficiently. And then it also sucks as a trader because you're not being offered significant depth around the mid price, right? So you're getting bad execution relative to other vendors. And so V3 introduces this concept concentrated liquidity, which allows liquidity providers to specify a lower bound and an upper bound in which they would like to allocate their capital. So rather than depositing across the entire price curve, you can sort of selectively deposit in price ranges at which you would expect to see most volume and therefore generate the most trading fees. And so if you think about it from the LP's perspective, they can now sort of afford to allocate less capital relative to a V2 liquidity rider and roughly similar fees at absolute level. And then for traders are just being offered significantly more liquidity around the mid-price.

So getting part of an execution. Yeah. That's definitely the most compelling shift, right, is from the perspective of the large LP. to be able to concentrate their position and write that the trader ends up getting a more stable price closer to the middle of where the tick is. So one of the cool things about the way that we surface our data is that we have a combination of events and contract states that we can look at. So actually, every time that someone on the LP side or on the liquidity provider side

adds or changes or removes liquidity to their positions, we record that in a nice, easy-to-read format so you can see how that's shifting over time, how often people are updating their positions and how much fees they're accruing and realizing along the way. And of course, over the last few days, you've seen a lot of sort of scrambling among LPs to move down on the lower ranges of the Rap Death pairs. We've seen a ton of volume on the stablecoin pairs. I think USDC was over a billion across its pairs yesterday, which is quite something. And yeah, actually, like someone is literally asking right now on our Discord about how to look at the bounded liquidity positions in our data.

So really, really fun to see people exploring that and actually using it and frankly making a ton of money off of it from the LP side. Like it's, it's impressive how quickly the bigger LPs have been able to rack up quite impressive sums. So I think personally, I'm actually very excited. We're building a tool right now that lets you sort of forecast. how much you might get on a daily basis if prices remain somewhat stable, just to give you a sense of how wide or narrow a band gives you how much more, how much less fees to help you make that sort of risk assessment when you're placing a position. That's been what's sort of kept me from doing it because I'm not in that upper echelon of the

multi-million dollar LP. I aspire to be someday, certainly, as should we all. But for now, I'm excited to be able to analyze the data and to look at what the intelligent LPs are doing, sort of follow their lead, and let myself assess that risk, you know, with the help of the data. I will just add to what Eric said, the byproduct of concentrated liquidity as a feature and the flexibility that it affords LPs is that LPs can suddenly, you know, set these. or execute these strategies according to their risk preferences as well, right? So you might have these very, very sophisticated, well-capitalized LPs who are willing to take on a ton of risk and have, you know, additional tooling that they use to set these strategies,

but then you also, you know, might continue to have relatively unsophisticated, less capitalized LPs. And you will see the sort of difference in these strategies reflected in the different ranges that each party sets. The other really important thing about V3 that I wasn't necessarily explicit about is that it now means that the UNSWProticle can cater to every single type of market.

So previously you just had this sort of default pricing formula X times Y equals K, which turns out does actually serve certain markets fairly well, things like stable coins, a continuously see quite a lot of volume on V2 and so on. But it certainly can't serve like-kind asset markets particularly well, because, again, you sort of expect most training activity to take place in this fairly narrow price range.

But then, you know, X times Y quits K also can't serve exotic assets particularly well either, right? You know, assets that might have something like Theta decay or some type of expiry and so on. And so now, you know, basically you have complete flexified, as an LP to tailor your strategies according to the market that you're providing liquidity. And that's really exciting. You know, Uniswap as this universal, really living up to the name now, is this universal exchange protocol.

It's definitely seemed from where liquidity is showing up that the liquidity is getting allocated to the quote unquote right fee bans. We saw sort of in the first few days like there would be, I think there were two or maybe even three different versions of the tether wrapped Eath pool, and then eventually everyone figured out to move it into the smaller interval, lower fee version so that you could concentrate your liquidity to an even more precise interval. So it's been interesting to see that too, that like there's this freedom to let the market figure out what the right of the currently three available fee bans is. Yeah, definitely a fascinating aspect of it. I am curious on your take, Teo,

on, so you get all these nice sort of LP and swap features. There have been certainly many hot takes on our social media networks about the implications from the traditional pool to approach of like staking liquidity tokens and yield farms. That personally, I don't think the takes are very well informed. It's actually another thing that I was going to say I'm excited about is to see the innovation in the defy space and how it, how creative. of TFI developers and builders use this new NFT and the liquidity provision side of things to be creative and thoughtful and innovative in sort of how that gets integrated into the

yield farming side of the space. But I'm curious about your take on that. Yeah, maybe just a level set for the audience, TAYO, if you could just explain this new concept of LP tokens as NFTs. I think it makes sense because You know, like you said about this concentrated liquidity change because they established these different liquidity provision strategies and price ranges within the pools that their positions are technically not fungible anymore. So they're representative NFTs. So I think what Eric was asking was how does that impact the ability of LPs to use their positions to farm yield? Right. Awesome. So, yeah, Rhea, I think you sort of gave good.

background there, but yeah, just to sort of summarize that, you know, in Uniswop v1 and V2 and and sort of previous AMM designs, every LP is sharing the same strategy, the sex and X-10-Y equals game strategy, and so you deposit some assets into a pool and you are issued what's known as an LP share, which represents your sort of pro rata share of the underlying liquidity and that can be redeemed at any time. And it has this nice property of being fungible. And so you can use these sort of depository receipt tokens and use them across other services like lending protocols and borrow against them or across liquidity mining programs by staking them into some kind of shared contract. And then this liquidity mining contract will

issue continuously at some rewards, again, on a pro-rata basis, according to how many equity-fider shares are sitting in that contract. So sadly, you know, V3 does complicate this to some extent, and that's sort of the trade-off of introducing this, you know, infinite flexibility for LPs is that you do break the sort of fungible nature of LP positions. And so as an LP in V3, you are still issued a sort of depository receipt token, but it comes in the form of a non-functionable token rather than a fungible token using the ERC20 standard. Does that mean that liquidity mining is off the table?

As we know, no is the fortunate answer. So we don't have anything in sort of a production state already, but there are numerous parties working on the implementation of a liquidity mining contract that actually, you know, functionally works very similarly to previous liquidity mining contracts. Basically, LPs will stake their NFTs into this contract and rewards will be distributed pro rata according to virtual liquidity rather than sort of liquidity itself. So let me just give me a quick example.

Imagine again, you know, we look at a stable coin stable coin pool and I decide to deposit a thousand dollars. between the prices of 99 cents and $1 and $1. So, and imagine, you know, at the time, you know, these stable coins are trading at a relative price of one. So I will be providing an amount of liquidity within this $99 to $1.1 range. Now, Eric comes along and also deposits $1,000 of liquidity, but instead he does it between 90 cents and $0.10, so a far wider range than my range.

Very diverse. Very risk of us. Eric has concerns about TABETHA, clearly. So while we've both deposited a thousand dollars worth of capital, I'm actually providing significantly more depth around the midprice itself. And so my virtual liquidity is higher than Eric's virtual liquidity. And so if we then, you know, both take our NFD tokens and deposit them into this liquidity mining contract

and someone is sort of disrupting rewards through that contract, I will receive a higher portion of those rewards relative to act. The fancy thing about this new version of liquidity mining contracts, and all credit to Dan Robinson, a research fund right paradigm, for producing the spec that people are working on, is that you can actually participate in multiple liquidity mining programs simultaneously just through this one contract, which is nice. So previous, the liquidity mining programs that I guess we all know,

in love, you sort of had to pick and choose and decide which ones you want to participate in, you know, according to which has the best risk-addressed yield. Now you can participate in, you know, any number of liquidity money program simultaneously just through this one single contract. I think that's a nice feature. Got it. That does sound alluring. It is.

Yeah, you should check out the stack. And I think we're expecting that to go live in a couple of weeks. so not too far down the road. Yeah, that's something that UGP actually funded, right? They are funding the audits for that. But, yeah, actually, Paradigms, Entrepreneur and Residents, Omar, is working on implementation alongside some folks at UNSWLAMS as well. Got it.

You know, another challenge with AMMs generally is this idea of impermanent loss, which is essentially like unrealized losses relative to a buy-and-hold strategy when the price of one asset in a pool moves relative to the other before accounting for the trading fees incurred. Does the design of V3 lead to reduced impermanence loss, or is it more nuanced than that? Definitely more nuance. I think Eric might have some opinions,

but I have fairly strong opinions on this topic. So I'll share those first, I guess. I do a place to start us by answering the question, I guess. And the general rule for V3 around impermanent loss is the narrower the range you select, the more risk of impermanent loss. And this should be relatively intuitive, right? So, again, if we go back to, you know, the stable coin example, selling a price range between 99 cents and a dollar one cent, you know,

as soon as the price of one of these assets falls below 99 cent, you will basically have 100% of your inventory concentrated in a single asset, and you will have effectively purchased that inventory at this price of 99 cents, right? And so if we imagine that the price of that asset is continuing to fall, and you had instead set a wider range, you'd sort of be averaging down your purchase costs and you'd have a better purchase cost relative to someone with this more narrow range. And so the flip side, no pun intended, also applies, right, which is that, you know, the broader range that you set,

the lower your risk of impermanent loss. But, you know, you're also earning less fees relatively, right? So going back to this virtual liquidity concept, you're providing less liquidity, and so you're earning a smaller share of the fees. And so that's sort of the consideration that you have to take into account as a prospective LP, you know, what range do I sort of expect these two assets to trade within over the period of time in which I want to provide liquidity? And yeah, is there some advantage to be had by setting a wider range versus a small range and so.

Yeah, I mean, I think you're spot on, like the, it's all about the, the risk assessment. I like your point that like basically you can think of outside the tick migration as the like indirect limit order in a sense. You end up acquiring one side or the other if you're out of the money, which is, you know, you can see that as a feature as well. I think it's actually called out in the in the SWATB3 docs that you can, you can effectively trade that way. It isn't. Yeah. Yeah. So, you know, where my strong opinions come in, well, and I guess, guess the first part isn't really even an opinion. This notion of impermanent loss exists across any type of market structure, right? Across AMMs and Central Limit order books, you know, if I'm making

markets on Coinbase Pro today and I set a bunch of bids and asks and those bits and asks will get executed, I'm also going to end up with my inventory concentrated in a single asset, right? And so I'm going to miss out on either additional upside or I'm going to realize this additional downside if the price keeps falling, right? So it's not an AMM-specific problem. It's really just the realities of market making. This notion of impermanent loss is actually correct in the AMM context because returns are path independent, which means that if the price goes down a lot and then up a lot and and down again and up a lot and whatever and returns to your entry price, then you're going to be a sort of flat plus plus fees. So that is a nice property. The part that I really strongly

disagree with is that LP return should be benchmarked to a 50-50 hold portfolio. And I think this is one of these weird, sort of very crypto-specific. Yeah, it's simply a crypto-e evolution. It's a very pure crypto-specific. It's a very pure crypto-specific evolution. And I think the reason why the industry is sort of coordinated around this benchmark is because everybody is just structurally long crypto perpetually. And so people approach these LP opportunities with these pre-existing inventories of crypto assets, right? And everybody wants to maintain some kind of long exposure. And so naturally they sort of benchmark LP returns to, you know, the counterfactual, which is, you know, just holding and continuing to maintain pure long exposure,

right? But really, you know, being an LP is just another passive investment vehicle, you know, very similar to, you know, buying an ETF or investing in a mutual fund or investing in some kind of venture vehicle or head front and so on, right? And really, you know, you start with some dollars or some common meaning of exchange, whatever, and then you either make money or you lose money. You know, if you think about this sort of impermanent loss concept benchmark to 50-50 returns, you know, according to impermanent loss versus 50-50, you're actually losing money as the price of these assets goes up, right? And you still maintain some inventory as the price goes up, although eventually, you know, you sort of get stopped out. So the idea of losing money as prices go

up is just entirely perverse. Whereas, you know, if you're benchmarking versus 50-50, then you are still sort of net down and so on. So I'm in the camp and I think it's a pretty lonely camp right now. It's a quieter camp, but I think it makes a bigger than we think. But yeah, I think, you know, over time, more and more people are going to approach Alping starting with dollar denominator capital and they will just naturally start benchmarking versus the risk-free rate.

But that's a crypto risk-free rate or just treasuries or so on. So yeah, sorry, that's my impalmendent loss spiel. I like it. I think in general, it may be an unpopular opinion because they're certainly very vocal, smart people who say a lot of correct loud things about imperman loss. And if your benchmark is, as you say, a 50-50 hold

and crypto is going up at a rate of however many hundreds of percents every year like you wanted to, then, yeah, being in a stable coin, ETH pool will not make you as much money if ETH is going up on, you know, that sort of factor. But, you know, in a bear market, it being in a pool that's half a stable coin means you didn't lose as much on here on the ETH downside, too. So, you know, I think in permanent loss, like many of the factors that we've gone over in deciding where to put your liquidity, how you hold your assets, what you actually have in your portfolio, how risk averse or how strong an appetite you have for risk. It's all part of that equation. It shouldn't be the only thing you consider. It certainly shouldn't be a reason that you like don't engage with Uniswap B3. I think the relationship between impermanent loss and the returns you're getting on your fees as you concentrate your liquidity. scales very similarly. So you don't have to like apply a new mental model or anything to how impermanent loss works in B3.

It doesn't magically go away, but it should be effectively the sort of same relationship in terms of the increased fees you're getting for your concentrated liquidity. So, you know, if you really care about improving loss, it should sort of fit the existing paradigms for Benjamin. Yeah. And correct me if I'm wrong, but even taking this 50, 50-50 benchmark as the status quo, the fees that you earn should more than compensate you for potential IL.

Like, is that something, Eric, that you've looked at the data for? Yeah, I mean, that all depends on how much swap volume there is and how much liquidity you provide and how narrow the bands are now in P3, right? Certainly, if you benchmark against a 50-50 hold, there are many instances where you would have been better off holding the two assets than providing liquidity, even net with fees. So this is why the debate rages on. In these powerful bull markets, it is frequently true, particularly for uncorrelated assets that impermanent loss can be worse for you as an asset holder than just holding.

But that's not necessarily the only reason you're providing liquidity, right? There are lots of different reasons to do it. It's a more reliable source of income. You're making money on swap volume. You're not making money on appreciation of an asset. It's a very different way to think about your assets. And apples to apples comparison. Yeah.

Yeah. I mean, it's certainly part of the equation. And you can certainly point to instances, especially in the first three months of this year, where liquidity providers were not making as much money on their fees as they would have just holding. Yeah, as Eric said, it's just, very, very case-by-case-case dependent.

Another sort of factor in play is the size of liquidity pool itself. So there is this, I haven't really dug into this data in a couple months now, but I remember not too long ago looking at LB returns over the past, I think 12 months in the S&X East pool. And over this period of time, S&X hit, I think, you know, 30 or 40X, to be on some absolutely insane run-up. But what was really fascinating is that, you know, relative to a 50-50 portfolio, whole portfolio,

LPs and that pool had actually outperformed because there just wasn't a lot of liquidity relative to volume in that market. And so they were just raking. So they were more than compensated for sort of giving out some of this SNX upside exposure. Right. That's a good point.

It's all about finding your pools, right? It's risk, how you decide where you're putting your money and, you know. Yeah, but it's also just a difficult, you know, it's a difficult thing to measure in the aggregate as well, right? Because your returns are basically a function of, you know, your entry price versus your exit price and then fees generated in that sort of interim period. And every LP has a different entry and exit. thing, right? And so you do sort of have to measure it at this very granular level, which can be difficult. Another thing, you know, Eric, I think you touched upon earlier is this idea of liquidity providers having to manually adjust their ranges as the price of assets changes.

My understanding is that these adjustments are not automated yet. So, you know, it's up to the LP to update their range. How do you envision that playing out? Do you think that there will be tools that will help liquidity providers kind of manage their position. Is that something that Flipside is working on? Or is that UNISWOP could potentially fund just further research into and development of tooling? Yeah, I would say for now it's about analyzing that and seeing what the patterns are and, you know, looking at how expensive it is, how it varies by how why. particular positions are and how much liquidity they have in them, how much time elapses

between a tick departing a range and that range moving toward the tick. We've seen so far, even with some pretty dramatic swings in some of the bigger pools, that the larger LPs are more than willing to move their positions very quickly after they stop running money. And I think that reflects the substantial upside that the new fee equation is giving them. I would say Flipside would love to be a part of helping anyone out there build such a tool. We are probably not the ones to build it, but the ones to help understand how you would build it. Eric, you also built this pretty awesome dashboard, just looking at some of the key metrics on Uniswap V3. I would love for you to just expand on what some of those metrics are,

especially around capital efficiency, participation, liquidity. and based on how you're tracking them, how do they compare versus V2 and other exchanges, centralized and decentralized? Yeah. I guess the easiest way to get access to this dashboard is following Flipside Crypto on Twitter or me at the Eric Stone on Twitter. And I think it's my PIN tweet right now. So that's pretty easy.

The idea with this dashboard is just to walk people through the key metrics that characterize liquidity, volume, fees, and efficiency on, you know, swap B3. I think I compare V2 and V3 on the capital efficiency front. But for the most part, it's focused on V3 as liquidity increases overall. I think it's, yeah, it's up over a billion right now in terms of gross or real liquidity across the pools that have migrated to V3. you can see a breakdown of deposits and withdrawals or liquidity removals and liquidity additions below that. We've definitely seen an increase on the deposit and withdrawal side over the last week in the market.

People are clearly attracted to the increased volume that's happening there. And you're seeing that in terms of more liquidity being added to particularly the stable coin tethered pools. I shouldn't use the word tether as a verb. I apologize. We also get into a breakdown of looking at real liquidity versus the virtual liquidity representations, which the startling thing, if you look at that one, is particularly for those narrow band stable coin pools where you can be very precise with your liquidity, the virtual liquidity numbers are immense.

And I have struggled to articulate the way to think about virtual liquidity properly. It's not actually how much can be swapped on the pool, but it represents how you relate to the underlying assets to each other. It's used to determine how much liquidity, a given participant or a liquidity provider maintains in the pool, and it's used to provide fees. And I guess also it'll be used for the sort of fungible version of liquidity tokens. So it's definitely a huge part of it. Yeah, I think, sorry, just like a strict definition of virtual liquidity, is the relative liquidity required in a V2 pool to support the equivalent deaths being provisioned in V3.

So I don't have Eric's dashboard in front of me, but I think some recent numbers, tens of billions of dollars of virtual liquidity in the stable coin to stable coin pools. Is that correct, Eric? Yeah, exactly. And so, you know, there isn't actually tens of billions of dollars worth of capital in those pools right now. I think that's somewhere on the order of a couple hundred million. But what that number is saying is, okay, in order to provide the equivalent depth in V2, there would have to be tens of billions of dollars allocated across this entire price

curve from zero to infinity. And then some small portion of that would be provisioned around the mid price. So, yeah, it does really sort of speak to the scale of capital efficiency that V3 provides relative to V2. The other thing I really like in the dashboard, is being able to look at V3 versus V2's capital efficiency, which it's summarized. I think it's like the last two weeks by day. It's basically just looking at how much swap volume each dollar of actual liquidity supports in each system.

It's a little bit unfair as maybe the wrong word, but given which pools have the most volume, there's a bit of a self-fulfilling reality happening there. But it does illustrate that with, you know, there's still much less liquidity. there's a bit over a billion on on v3 and I think something close to six billion still in v2 yet the amount of swap volume that those two you know tbLs are supporting is essentially equivalent they you know right around a billion or more on each b2 and v3 lately so you get up close to a sort of 50-50 breakdown on swap volume with considerably less liquidity on the v3 side which is impressive. Yeah, because you have like a, as an LP, your return on investment and significantly

higher. And then you account for the opportunity cost of being able to deploy that additional capital elsewhere and not keeping it locked up in the pool, which is, it's been really impressive to see that for sure. And this dashboard is a part, if it correct me, if I'm wrong, of a broader initiative by the community for data and analytics tools for UNISWAT participants. specifically through the provision of a grant, right, from Uniswap grant program. What was the process of, you know, creating this specific initiative, soliciting support, and receiving the grant? And then what exactly is this grant funding and who can participate?

Yeah. So there are actually, interestingly, two initiatives. The grant, I think you're alluding to is, well, yeah, there is just one grant. We are running a separate initiative, which I can talk about. The grant you're talking about is from Uniswap to entice the community to make compelling dashboards that highlight the data in V3. And right now, there's equal grants for our platform and for doing analytics and the top entries submitted from both platforms will receive a reward.

And I don't actually know that much about the other part of it, but there's some potential like, you know, analytics committee that could be formed from the folks who make the best judge dashboards. My role in the, in the actual grant side of things was more like getting the data ready and, you know, chatting with the Uniswop team on the dev side to make sure we're characterizing data properly and coordinating with our biz dev people. So I have less really specific insight into how the grant process actually works. I know that Uniswap governance is is excited to deploy parts of the Treasury, to entice the analytics community, to engage with data, and to make really great analytics tools, dashboards, forecasting tools.

We've seen a lot of really cool stuff come out already as this data hits the, whatever you want to call it, the data warehouses. So, yeah, so our role from our perspective is to give anyone who wants to use the flip-side data platform to participate in this contest the best possible cleanest access to uniswap v3 data that makes it really easy to start actually doing analytics once you understand the fundamental concepts of the platform. We have ideally provided tables that reflect each of the core components. So you can look at position changes, any action that an LP is taking, any swap that occurs. Anytime a pool is updated, whether because of a swap or an LP action, we update the core

stats for the pool to show you how much TVL there is, how much virtual liquidity there is, what the underlying tokens are, all that fun stuff is done for you so that you're not stitching together sums of complex events and trying to do your own decimal adjusting and trying to figure out what virtual liquidity means and how to convert out these complex sort of exponential numbers that represent the Oracle prices, all that fun stuff we take care of for you. So you can just dig into the data and answer questions about things like, what is the shape of the liquidity in each pool and how often are people changing their positions? Are the biggest liquidity providers becoming less aggressive with their positions as the market gets more volatile?

These are all questions you can answer pretty quickly because we're combining. I think I alluded to this earlier. If you know Ethereum data, our ability to combine both the content, contract state side of the equation with the event side of the equation means that you can answer more questions, especially about a complex project like SWATV3, where knowing what the state of the contract is on these complex positions tells you a lot more than just what you get from the emitted events. So that's our hope is that the analytics community is enticed by that, sees the value of that upfront work, and can make some pretty cool visualizations and tools to help

LPs and showcase the benefits that swappers are getting as well. That's awesome. I think this whole concept of treasuries funding the community to build tools is really awesome. And just being able to see how quickly some of these projects are deployed relative to, you know, a centralized entity trying to build the same thing in-house with a limited set of, you know, participants working on these respective problems. That's been really awesome to see.

maybe as a last question for you guys more broadly what do you think the impact of v3 will be on the entire defy ecosystem you know like what other protocols does it enable or affect would love both your thoughts on that yeah i'm i'm happy to give this a try so so i think the sort of very immediate implications of v3 is more efficient capital allocation for liquidity providers and better execution for traders. As we discussed earlier in this conversation, I think DFI and DECS were already on this very competitive trajectory pre-V-3, at least over the past 12 months,

partly due to UIX, partly due to superior asset coverage. And I think V3 is only going to accelerate that adoption rate, because now DECS, or V3 in particular, can provide execution on part with, centralized services. I think as far as like some second order effects and some cool ones, you know, V3 can be used for things like stable coin peg the fence. So if you guys or listeners are familiar with MakerDAO's peg stability module, V3 can basically step in and provide a similar service there. I think another really interesting as well is if we think about these

lending protocols like Maker or compound RV and so on, all of the which are over-clateralized and have some kind of liquidation threshold, part of the reason why liquidation thresholds have been so conservative. Historically, I think averaging around 150%, obviously it does change on a per cent basis, but roughly 150% is because there needs to be sufficient margin for liquidators to come in and actually liquidate this collateral and come out with some net positive proceeds. And all these liquidations are happening on chain. So, you know, liquidator will step in, return the borrowed assets, take the collateral,

and then try sell that collateral on a decentralized exchange. So if a exchange protocol like V3 can offer better execution, then suddenly this means that these lending protocols can also probably afford to lower their collateralization thresholds. And so V3 makes other protocols more capital efficient as well. I think that's just going to be a net positive. the entire industry. I like that one.

I live in constant fear of liquidation. So I look forward to that brighter future. I think I'm also interested in the potential has been teased for V3 to be released on a layer two. I don't want to name network names. I can't actually remember if you guys have said a network that you're planning on. Is it optimism? Am I remembering right?

I am extremely interested to see how it works there where fees are essentially a non- consideration, there's so much a part of the equation right now in terms of having to figure out, particularly how often you're going to have to adjust your positions as prices change. So when fees are no longer a part of that equation, becoming extremely targeted and aggressive with a narrow price range will be the norm, presumably. And there you might see the development of more of the sort of automated position management smart contracts crop up. And I would expect that. And I'm super interested to see. It's hard to sort of forecast what the eventuality that is, but hyper-efficiently managed positions that have super deep liquidity

certainly benefits the swapper and has ramifications that I can't quite fully fathom on the liquidity side. Yeah, I think just the sort of optimal LP strategy is very dependent on the execution environment itself. One more slightly left field, broader implement, application is the business source license that Uniswap V3 uses. So previous versions of Uniswop were licensed under GPL, which means that everything is source available, but also that the code can be modified for commercial purposes. And we saw this happen a lot. In fact, you know, the number three decentralized exchange protocol on the market today

is a Unswap V2 clone. So Uniswop v3, the team decided that a business source license, which basically means that everything is source available, but it can't be modified for commercial purposes until this two-year period has elapsed and then it does immediately transition to GPR. The team decided that that was in the best interest of the Uniswop community. And I sadly agree there. And I think the implication there is that I think we probably have set a bit of a precedent for the industry.

and I do expect to see adoption of that type of licensing moving forward. I think, you know, as fantastic as sort of the DFI summer was for the DFI world, it also raised a lot of these sort of complex questions as to how can we sort of develop these protocols and attract innovation in a sustainable fashion in a way that sort of continues to incentivize parties and protects IP and so on. And I think this is a really nice way of navigating that challenge. That's awesome. Well, this has been an extremely enlightening discussion.

Thank you both so much for joining us. Before we say bye, what is the best way for listeners to connect with you guys? Yeah, for me, it's following at Flipside Crypto and me at the Eric Stone on Twitter. Right now, you can go to FlipsideCrypto.com and participate in the bounty challenge from Uniswap as well as a couple of other ongoing values that we're running for compound and another separate one for Uniswap as well. So yeah, that would be the best for me. Yeah, on my end, I'm also on Twitter, Teo underscore Leavowitz, if you manage to spell that.

I don't blame you if you can't. If you're not already following Teo. We'll add it. Or otherwise, yeah, go follow the Uniswold Labs team at Uniswap. Awesome. Thank you guys so much for your time. Thanks for listening to another episode of On the Brink with Castle Island. To find out more about Castle Island, visit castle island. Visit castle island.vc. To listen to all of our podcast episodes, please go to On thebrink dashpodcast.com or just click on the tab in our website.

Thanks for listening.