Why Decentralized Trading Has 10Xed in a Few Months

Haseeb Qureshi, managing partner at Dragonfly Capital, and Dan Robinson, research partner at Paradigm, discuss the explosion in decentralized exchange activity on automated market makers from $1 billion earlier this year to more than $10 billion in August. 

• Why trading in automated market makers has 8xed in a few months
• how automated market makers work, and how specific dexes can specialize for the assets being traded on them
• why traders would prefer to trade on a dex vs. a centralized exchange
• how incentivized pools have drawn traders and liquidity 
• why trading on a dex seems can be more appealing than on a centralized exchange
• why trading on dexes now mirrors trading on Binance in 2017
• how “DeFi is the hottest coolest place where people are making a lot of money”
• the difference between the metrics of liquidity, trading volume and total value locked for a dex
• why AMMs are capital inefficient
• what it means that fees paid on Uniswap are about half of those paid on Bitcoin 
• what the purpose of liquidity mining is
• why Haseeb doesn’t think that dexes have network effects and that once professional market makers come in the space, Uniswap will lose market share
• whether users will use DeFi directly through the dexes or through another interface
• why Dan thinks that DeFi is now comparable to financial investment in the 1970s 
• whether or not yield aggregators like yEarn make it harder for AMMs to compete
• what happened with the CRV token launch
• initial dex offerings or initial DeFi offerings, like the UMA token launch
• whether, because of the high fees on Ethereum and the how long its transition to Ethereum 2.0 will take, DeFi activity could move to another blockchain such as Solana
• where DeFi will go over the next few years

Thank you to our sponsor! 

Episode links: 

Haseeb Qureshi: https://twitter.com/hosseeb

Dan Robinson: https://twitter.com/danrobinson

Haseeb’s post on the rise of AMMs: https://medium.com/dragonfly-research/what-explains-the-rise-of-amms-7d008af1c399

Haseeb’s post on how Uniswap could be un-bundled: https://medium.com/dragonfly-research/unbundling-uniswap-the-future-of-on-chain-market-making-1c7d6948d570

Dex volume: https://defiprime.com/dex-volume

Trading volume on centralized exchanges: https://www.theblockcrypto.com/genesis/74539/july-by-the-numbers-a-look-at-crypto-exchange-volumes-open-interest-and-miner-revenue

Incentivized pools: https://pools.fyi/#/?tag=incentivized

Top dexes by trading volume: https://www.coingecko.com/en/dex

Ethereum transaction fees: https://bitinfocharts.com/comparison/ethereum-transactionfees.html 

Uniswap website visits 15x in July: https://messari.io/article/uniswap-total-website-visits-grew-15x-in-july

Website traffic to DeFi platforms explodes in July: https://www.theblockcrypto.com/post/74231/defi-platforms-web-traffic-july

Fees paid to various crypto projects: https://cryptofees.info

Curve reaches $1 billion in TVL: https://www.theblockcrypto.com/linked/75014/curve-becomes-third-defi-protocol-to-hit-1-billion-in-total-value-locked

The Block on dex aggregators: https://www.theblockcrypto.com/genesis/69468/dex-aggregators-user-experience

Delphi Digital on aggregators and why dexes will be like airlines: https://www.delphidigital.io/reports/dex-wars-and-aggregation-theory/ https://www.delphidigital.io/reports/are-amms-the-airlines-of-crypto/ https://twitter.com/ZeMariaMacedo/status/1298589417855168518?s=20

Mooniswap: https://www.theblockcrypto.com/linked/74454/1inch-launches-automated-market-maker-mooniswap

Yield aggregators: https://www.theblockcrypto.com/genesis/75189/defis-yield-aggregator-dao-yearn-finance

How DeFi is affecting OTC desks: https://www.theblockcrypto.com/daily/75870/defi-otc-desks-lunch

Serum on Solana: https://www.theblockcrypto.com/daily/72924/ftx-dex-serum-solana-blockchain

CRV launch: https://www.theblockcrypto.com/genesis/74921/the-story-of-the-botched-crv-launch

UMA token launch: https://www.theblockcrypto.com/genesis/63611/uma-token-sale-retrospective

Serum on Solana: https://medium.com/solana-labs/ftx-chooses-solana-for-serum-a-high-speed-non-custodial-decentralized-derivatives-exchange-c346a27c1f2b https://www.coindesk.com/uniswap-ftx-perpetual-futures

Transcript:

Laura Shin:

Hi, everyone. Just a quick note before we begin. The DeFi space is moving so fast that parts of this episode have become somewhat outdated since the time of recording, which was not even a week before this episode comes out. For one, I say that dex trading volume has 8Xed in August from the beginning of the year, and by now, a few days after we recorded, its 10Xed. I also theoretically talk about a new dex forking the code of Uniswap and launching a token and liquidity mining scheme on it before Uniswap v3 launches. And Sushiswap, which is the definition of that, launched the exact day we recorded, and we were unaware of it, or atleast I was, at the time we spoke. Then, when we discussed the 24-hour trading fees on Uniswap, at the time of recording, they were half that of Bitcoin’s and now a few days later before publication, they have exceeded those on Bitcoin. And finally, we talk about how trading volume on dexes is smaller than that of centralized exchanges and while by the time of publication that statement is still true, Uniswap’s 24-hour trading volume has exceeded that of Coinbase’s which is pretty remarkable. So, if when you listen, it sounds like this episode is old, well…it is in the world of DeFi which moves at the pace of a live streamed video game. With all that said, this is still a fascinating, engaging and fun conversation with Haseeb Qureshi of Dragonfly Capital and Dan Robinson of Paradigm, and I think you will really enjoy it. Now, on to the show!

Laura Shin:

Hi, everyone. Welcome to Unchained, your no-hype resource for all things crypto. I’m your host Laura Shin, a journalist with over two decades of experience. I started covering crypto five years ago and as a senior editor at Forbes was the first mainstream media reporter to cover cryptocurrency full time. Subscribe to Unchained on YouTube, where you can watch the videos of me and my guests. Go to YouTube.com/C/UnchainedPodcast and subscribe today. 

Crypto.com is waiving the 3.5% credit card fee for all crypto purchases until the end of September. Download the Crypto.com app today.

Laura Shin:

Today’s topic is automated market makers. Here to discuss are Haseeb Qureshi, managing partner at Dragonfly Capital, and Dan Robinson, research partner at Paradigm. Welcome, Haseeb and Dan. 

Haseeb Qureshi:

Thanks for having me, Laura. 

Dan Robinson:

It’s good to be here. 

Laura Shin:

This year there has been an explosion in decentralized exchange activity, particularly among the group of DEXes known as automated market makers. Admittedly, the growth is from a tiny base, but the rise is still striking. DEX volume was below a billion dollars in January and hovered at about one billion dollars through May. Then in June it was still below two billion, but in the last few months it has exploded with volume for August having already surpassed eight billion dollars with almost a week remaining in the month at the time of recording. 

Just quickly before we dive into why this is happening, let’s first define the terms DEX or decentralized exchange, and let’s also define or describe the specific type that we’re talking about, these automated market makers, just so the audience understands the distinction. 

Haseeb Qureshi:

Dan, do you want to take that? 

Dan Robinson:

Sure. So, a decentralized exchange is anything that serves the same function as centralized exchanges like Coinbase or Binance but does it on a blockchain or on an off-chain scaling solution for blockchain so that it essentially can be done without a centralized party custodying the funds or managing an order book, and Automated Market Maker is a specific subset of these DEXes that provides liquidity according to an algorithm rather than based on orders submitted by participants, and so the biggest AMM or highest volume AMM is Uniswap, and it provides liquidity according to a very deterministic formula, a very simple one. 

Haseeb Qureshi:

So maybe I can kind of fill in a little bit of a sort of analogy of how you can understand what Uniswap is as an automated market maker. So, I wrote a post about this recently where I describe, you know, imagine that you had a friend that decided that they were going to be a market maker, and for those of you who are not familiar, a market maker is basically somebody who is always willing to buy and sell some pair of assets. So, we often say they’re market making that asset. So, I’m always willing to buy some ETH or I’m always willing to buy some USDC, and I’m always sort of willing to quote the two prices at which I’m willing to buy and sell those assets. 

So, what a market maker like Uniswap does is basically it’s a market maker that uses an extremely simple algorithm, which we can go into at some point, that decides when and how it’s going to buy and sell, what price it’s going to give for those two assets, according to a simple formula, but the thing that it does that’s so different from any other normal market maker is that one, it has no money of its own, but it sort of raises money from decentralized investors who can give money to the market maker that allow it to buy and sell that asset, and those investors can share in the profits of that market maker, and in addition to that, it’s completely permissionless. So, it all works on chain, it’s all completely autonomous, and that is very fundamentally different than any market maker who has come before it which have all been, you know, normal companies that go on exchanges, that have shingles over their roof and employees. Uniswap does this entirely in an automated way through smart contracts, and that’s a really big difference between what’s come before. 

Laura Shin:

Yeah, and one little disclaimer that I do want to add is that Dan at Paradigm, they’ve invested in Uniswap, so that is something people watching the video might have figured out because he is wearing a Uniswap T-shirt, but we should let people know that because we are going to be talking about Uniswap quite a bit during this episode, and one other thing I wanted to ask about was this idea of the constant function market makers. You know, you touched on that very slightly, but can you just describe a little bit more what that means?

Haseeb Qureshi:

So, a constant function market maker, so we can start with Uniswap. So, Uniswap has a particular function that it uses to decide how it’s going to quote prices, like how much it’s going to charge for ETH versus USDC or whatever two the pairs or the two assets that it’s quoting. Every algorithmic market maker or automated market maker for the most part that exists today uses some kind of very simple, fixed pricing function that can be computed on chain given its reserves or given what it has in inventory. So it’s like, if I have, you know, ten units of ETH and 10,000 units of USDC, what is the rule that decides what prices I’m going to charge for the ETH versus the USDC. 

Uniswap uses the rule called the constant product rule which basically says that the units of ETH that I have, not the dollar value but the units, so literally if it’s 10 ETH, 20 ETH, I don’t care what the price of ETH is, I just care how many units of ETH I have, the units of my first asset and the units of my second asset multiplied together must equal a fixed constant, and so let’s imagine that it’s, you know…let’s say that I’m quoting ETH…or let’s say I’m quoting USDC and Tether, right? Those are two assets that should be worth the same. So, if I have 100 USDC and I have 100 Tether then 100 times 100 is 10,000. Whenever I’m changing the price and I’m quoting somebody some trade against my inventory, I should always make it so that when I’m done the two assets, the two units I have of asset A and asset B in inventory must always equal 10,000 multiplied together. That’s the constant product rule.

That’s the rule that Uniswap uses and it was sort of the first and most influential of this generation of automated market makers that Uniswap really began, but the universe of functions that you could use is actually quite a bit larger than just constant product, and so one of the other large platforms that has kind of really grown lately is called Curve, and Curve uses a different function than the constant product function. It uses this more complicated function that’s a mixture of constant product and constant sum. They call it the Stableswap variant, and basically what that does is Curve is optimized for stablecoins only or other mean reverting assets like, you know, different kinds of Wrapped Bitcoins or different kinds of Wrapped ETH or whatever, and in that universe you basically can assume that probably these two assets actually should trade pretty close to each other. The constant product rule, x times y equals k, it gives you some slippage when you’re making, you know, moderate sized trades, and for something like Curve where you’re trading assets that really should be worth the same almost all of the time, you can actually offer better pricing and you can offer tighter slippage in most ranges of inventory that you can end up with, and so StableSwap, which is what Curve uses, allows Curve to actually be more competitive on pricing because of the fact that it has this different kind of inventory and thus it can offer different kinds of rates. 

So, what we’re starting to see is this renaissance of different functions that are being applied to different AMMs that are better suited for the particular kinds of assets that those AMMs are trading.

Laura Shin:

And so from the trader’s side, what are they using automated market makers for as opposed to, for instance, a centralized exchange?

Dan Robinson:

So, one area where automated market makers and DEXes in general excel is in sort of the long tail of tokens that haven’t necessarily been listed on any centralized exchange. So, Uniswap and a couple of others like Balancer support arbitrary ERC20 tokens, so as soon as something is an ERC20 token, it can be listed on Uniswap and people can start trading it there, and it’s also, in part it’s because it’s permissionless and on chain. Another reason that I think these automated market makers are good for the long tail is because you may actually need some centralized market maker on a traditional exchange to basically go on and actually make a market, both sides of the market to have enough liquidity in a token, and that’s just not going to be the case for most tokens, whereas on Uniswap literally anybody can do that, it doesn’t take any professional expertise. It’s basically just click a button and suddenly you’re providing liquidity in both sides of a market for a particular token, and so there’s proven to be a very long tail of tokens that people would be like to trade. 

Laura Shin:

Now…

Haseeb Qureshi:

One thing…

Laura Shin:

Oh, go ahead.

Haseeb Qureshi:

Go ahead. 

One thing that I want to point out that I think might not be obvious for folks who have been following this space for a while, so you might remember things like EtherDelta, obviously 0x was a very big set of decentralized exchanges or protocol for decentralized exchange, and none of these things really took off in a big way before Uniswap, and I think it’s an important question to like, really try to ponder, like, why? What’s different now? 

And I’m sure we’ll go into that at some point, and I’m sure Dan and I have different theories of why Uniswap has been so successful, but one big, very clear difference actually that I would argue between DeFi now and what decentralized exchange trading looked like two or three years ago, is that two or three years ago it was basically all long tail assets. So, the kinds of assets that you saw getting the most liquidity on EtherDelta or on the early 0x relayers, they were mostly these long tail, really crappy tokens that weren’t getting listed on centralized exchanges, and it’s not that centralized exchanges didn’t list tokens aggressively, it’s that some things were so bad that even centralized exchanges weren’t willing to list them, and now what you’re seeing is actually quite a bit of a different profile in the types of tokens that are trading on DeFi. So, you’re seeing the really high-quality pairs like ETH-USDC or ETH-Tether or USDC-Tether or you know, ETH-Wrapped Bitcoin. You’re seeing those start to gain really meaningful volume, and that’s pretty fundamentally different and it feels a lot more sustainable than what was happening in 2017. 

The other thing that I would say has a big, has a lot of explanatory power for why Uniswap has been so successful, which I think was difficult to anticipate ex-ante, was the power of what are called incentivized pools. So, incentivized pools are basically this idea, so you know, normally if I want to start a new Uniswap pool because I want to market make for some pair, I want there to be some liquidity so people can buy and sell it, you know, I need to provide some liquidity meaning I need to put some of ETH and token X, so let’s say Ampleforth because it’s a token that trades a lot on Uniswap, if I’m, you know, an early holder or I’m a community supporter or I’m the core team of Ampleforth and I want there to be liquidity, you know, on a centralized exchange I could go pay a listing fee, I could go hire a market maker, I could do all the stuff. It’s like, kind of a pain in the a**, it’s very expensive, it’s very difficult. 

What I could do instead is I could go on Uniswap and of course I could myself just provide liquidity, I could just put a bunch of ETH and a bunch of Ampleforth in the pool and let Uniswap do the work, but another thing that I could do is I could actually incentivize other people to go do that for me, and the way that that works is essentially there’s…so Ampleforth has this thing they call the Geyser, and the Geyser, what it does is basically, you can imagine it like every day or every week Ampleforth spits out a bunch of new Ample rewards to anybody who is providing liquidity on Uniswap, and so what that does is it’s almost like Ampleforth can pay a market maker to market make the asset, but it’s paying anybody who is willing to provide liquidity on Uniswap, so it’s sort of this very clever way to incentivize the community to become your market maker, and that’s become a big driver for the success of Uniswap.

Laura Shin:

Yeah. Well, this is actually where I was going to go with my question, is just you know, why is it that we’re seeing this explosion in DEX activity, and I do want to also just put into context when I say explosion, to be fair when I say the August volume is probably going to come in over eight billion, it’s still small compared to the July volume on centralized exchanges which was 109 billion, but it’s not insignificant and it is growing fast. So, you know, what has happened just in the last few months? Is it because of the incentivized pools or does it have to do with the yield farming craze, or why just in these last few months have we seen this explosion? 

Dan Robinson:

I think the DeFi boom in general and the yield farming craze particularly have been big drivers of Ethereum usage, and DEXes I think serve those uses particularly well because if you’re doing something on chain with tokens, you’re borrowing them on compound, or if you’re minting DAI on Maker, or you know, participating in yield farming in any number of these protocols, you typically don’t want to go to the trouble of depositing to an exchange, waiting for the deposit time to expire, making the trade, withdrawing it. You’re actually going to spend more potentially in transaction fees from the deposit and withdraw than you would from an on-chain trade, and so it’s much more convenient for you just to go to Uniswap, and this is true even if you’re trading a fat tail token or fat tail pair like USDT, USDC, or USDT-ETH, and I think that, having actually something to do on chain with tokens has really increased the usefulness of DEXes, which in my opinion provide right now a better, a more convenient user experience than centralized exchanges even with extremely high fees and slow transaction confirmation times. 

Haseeb Qureshi:

I agree with that wholeheartedly. I think there’s a really amazing story about how DEXes and DeFi has started to look actually more convenient than centralized exchanges. In the post I shared a story of a friend of mine who was telling me, you know, there was some hot token that he was interested in trading, and it got listed in a few places, and he was telling me look, you know, I could go, look up on CoinGecko and try to see which exchanges it’s listed on, and how many of them are legit, and you know, where has the most liquidity, and then set up an account and send my funds there, but it’s just too much work and so instead, like, I’m going to go, I’ll just click a couple of buttons and buy it on an aggregator or on Uniswap or on whatever, and you know, that to me was a little bit of a revelation that, like, oh, sh*t, people are going to use DeFi because they’re lazy, not because they’re ideological, not because they care about noncustodial trading.

But the other element that I think is also very important, especially in this latest bull run that we’ve seen kind of centered around DeFi is that you know, so much of what made Binance successful in the ICO craze was that Binance was the first to list a lot of these assets, right? And so if you wanted to get in early, and of course so much of crypto speculation is about getting in early, you had to go to Binance because that’s just where things got listed first, and now you’re seeing that actually happen on DeFi. DeFi is where it’s listed first, right? COMP first traded on Uniswap before it traded on any centralized exchange. UMA was trading first on Uniswap. So many of these assets are first available in DeFi before they’re available on any centralized exchange, and so if you want to get in early, if you want to get in with the cool kids, if you want to get in with what all the influencers are doing, they’re all on DeFi doing it all direct, and I think that’s another thing that’s driving a lot of the excitement for people to get onto DeFi. Of course it’s driving up gas fees like crazy, but in a way it’s become part of the game that DeFi is now, you know, the hottest, coolest place where all the people are making all the money, and that of course incentivizes people to say hey, sh*t, I want a piece of that. 

Laura Shin:

One other thing I was curious about is so Uniswap, but like, when I asked this question I was asking about yield farming, but Uniswap does not offer any kind of liquidity mining so yield farming is not possible there, but a number of the other AMMs are doing that, and so I wondered, do you think that the uptick in volume that we’re seeing on these DEXes that are not Uniswap but ones that do have a token, do you think that’s something that will be sustainable or do you think that the volume that we’re seeing on those DEXes is just, you know, kind of a flash in the pan just because people are being incentivized to put their liquidity there? 

Haseeb Qureshi:

Well, two things I want to say to that. So, one is that I think incentivized pools are basically isomorphic to yield farming. They’re kind of the same thing, just in a different term. 

Laura Shin:

Wait, what does isomorphic mean? 

Haseeb Qureshi:

Basically the same thing. It’s basically the same thing, right? 

Laura Shin:

Oh. Okay.

Haseeb Qureshi:

It’s just sort of a different name, because yield farming is this idea that like, okay, I put up my assets and I normally get some nominal yield, but that yield gets juiced by some protocol rewards, right? Uniswap has the same thing, right? I’m supposed to put my money in the Uniswap pool, I get some yield by the trading fees and the impermanent loss, but that gets juiced by the incentive, you know, there’s incentives in the pool.

Laura Shin:

Oh. Oh.

Dan Robinson:

On a per pool basis. 

Haseeb Qureshi:

It’s the same thing but it’s on a per pool basis. 

Dan Robinson:

You’re talking about stuff like the Ampleforth Geyser here. 

Laura Shin:

Right. Right.                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                               

Haseeb Qureshi:

Exactly. Like Ampleforth or Synthetix or you know, any of these incentivized pools that are getting a lot of action.

But it’s important to recognize that what that does is not drive trading volume, it drives liquidity, right? Trading volume is a totally separate deal. So, because incentivizing a lot of liquidity up front gets these big, fat, juicy pools, right, and says, right, look, there’s so much capital that you can trade against, but if nobody wants to trade, it doesn’t matter, it’s just going to sit there and do nothing, right? The actual, the real interesting story is not the liquidity, the real interesting story is the trading demand, right? Where is all this trading volume coming from, because that’s not incentivized. Nobody is getting rewarded for trading against Uniswap. They’re trading against Uniswap because they like the prices. That’s what’s really interesting.

And I think, you know, the fact that the pools are really big, I mean, so most of the highest volume Uniswap pools are at this point not even incentivized. They were in the earlier days, but at this point they’re not, and the same thing is true for Curve, right? Curve has incentivized liquidity because the people who are depositing funds into Curve, they are getting, some Curve tokens, but people who are trading against Curve, they just want to swap stablecoins, and Curve has a tone of volume for people who are getting nothing except actually getting trades executed.

So the liquidity, what it’s showing us is that there is actually latent demand that’s there. If there was no latent demand what you’d see is a ton of liquidity with almost no trading volume, and that’s the opposite of what we’re seeing which is what’s so interesting about what’s happening now. 

Laura Shin:

And let’s just parse this out a little bit more, because I actually was going to ask a question about this, but we can cover this now. You know, when I was doing research for this, I did see things like, Uniswap had obviously the highest trading volume out of all the AMMs, but then something like Curve, which was smaller, had more total value locked in it. So, can you…I mean, I think you kind of already started to answer this, but can you just walk us through a little bit more, like what does it mean for an AMM or what does total value locked mean for an AMM and then what does trading volume mean, and why…so, if we see a lot of total value locked in one place but then more trading volume elsewhere, like what does that mean? 

Dan Robinson:

So, you can think about the liquidity that’s locked in a pool, and locked is really not the right word, and we all use it and feel a little bit guilty about it because it’s available to be traded against, right? You could think about that as kind of the maker side of the market, the maker orders where somebody is pooling liquidity and it’s available at any price in some quantity. The volume is the taker demand, and typically the makers are, they’re sort of neutral lot between these two assets, and they’re just trying to make money off of the fees, and the takers are trying to actually trade there.

And what we’ve found is that to some extent liquidity drives volume, the quantity of maker orders, because it determines how much you can trade on this without pushing the price too far, and so it’s not the full story, and in fact Uniswap right now has lower liquidity than Curve or Balancer total in its pools but much higher volume, and I think that’s in part, it shows that there’s some stickiness to volume aside from liquidity, and that’s true even on the same tokens, although the other thing that matters is whether this is liquidity in tokens that people actually care about trading.

And so, I agree with Haseeb generally that volume is maybe a better indicator, although it can be confounded a bit if its liquidity would be unprofitable otherwise and is there, people will trade against it, but potentially that could dry up when incentives dry up. 

Haseeb Qureshi:

So, I think the right way to think about TVL, which is total value locked in a protocol. I mean, it’s a fun headline metric, we all talk about it and it’s really stupid, it’s kind of like trying to measure how good a blockchain is by looking at its hashrate, which is like, it’s like one thing and it kind of matters if you contextualize it correctly, but if you don’t, like, it’s just not the main thing to think about. 

I mean, the most cogent way to describe what is TVL for automated market makers is like, again, if you imagine that this is a market maker, you know, this is just literally a market maker that’s buying and selling things, that’s the size of its balance sheet, you know? This thing, if it’s got 500 million dollars locked, then its balance sheet is 500 million dollars, and that allows it to do a lot of trading because if you have 500 million dollars, you can absorb a lot of volume. 

Now, of course, like, compared to normal market makers, one thing you can’t help but notice is that all of these automated market makers are incredibly capital inefficient. Like, the amount of capital that’s incurred is like absolutely insane relative to the amount of trading volume that it’s doing, right? Just by the absolute numbers it doesn’t really need that much liquidity, but you know, what’s happening in a sense is that the liquidity, the balance sheet of Curve is getting subsidized by this airdropped CRV distribution, and the same thing is happening in incentivized pools, right? Like, these things are getting subsidized by whatever teams are subsidizing these pools, and so that may mean that normally, you know, at equilibrium, if there were no subsidy, really there should be less liquidity, and you know, the returns compared to the risk and so on, it’s obviously hard to measure all that stuff because nobody really knows what, you know, how much risk should you be taking, how much should you be rewarded for taking a certain amount of risk? Everybody is sort of, you know, just finger in the air, kind of take a rough guess, but it’s very clear that one effect of liquidity mining is that it is somewhat distortive of the sort of natural rate of…sort of what you should expect to be the risk-free rate and the price of risk in these markets. 

Laura Shin:

Yeah, and one thing I do also want to point out for listeners is this is another way to contextualize the volumes that we are seeing on Uniswap. At the time of recording, the fees for the last 24 hours on Uniswap were roughly half that of the fees on Bitcoin. It was about a half a million dollars on Uniswap and then about a million dollars in fees on Bitcoin, and you know, that’s quite remarkable because Uniswap is nowhere near as old as Bitcoin, but actually I want to maybe throw that to you guys, is it fair to even make that comparison between the two because they’re obviously completely different things?

Dan Robinson:

I love the sort of fees paid as a metric in general, and I think maybe more than I like total value locked, and I think it’s fun especially to compare the fees earned by these applications on Ethereum to some of the smaller chains that are still around from 2015 and before. 

Haseeb Qureshi:

What, like Bitcoin? 

Dan Robinson:

Yeah. I do think obviously there’s ways in which it’s incomparable, and one way actually is that those Uniswap fees don’t even count the Ethereum gas fees that people are paying to use Uniswap, that’s just the transaction fees that are going to liquidity providers, and then on the flipside, the Bitcoin one does not include block rewards, which are for miners the primary source of their revenue, and that’s really what drives a lot of this miner demand, and then obviously again these are very different kinds of services, but I do think it’s useful to show just what are people actually willing to pay for. 

Laura Shin:

All right. So, you know, one thing that’s interesting here is, I mean, it’s not like Uniswap is completely dominant but it is quite dominant, and I just wondered, what does that mean for decentralization?

Dan Robinson:

So, Uniswap the protocol is completely decentralized in the sense that the contracts aren’t changing chain and can’t be stopped by the team, and if the entire Uniswap development team disappeared tomorrow, Uniswap V2 would continue to operate, the interfaces are open source. That said, I think that’s one of the benefits of decentralization, is that you can have this kind of network effect building up in one system without having the kind of platform risk that that would cause in sort of more traditional internet businesses or in any other kind of businesses that have existed before, and so that’s I think one of the benefits, is that it could be that Uniswap sucks up all the liquidity in the market, but that’s fine. 

That said, I think that’s not necessarily going to continue to be the case, and I think there’s going to have to be more innovation in this field. I don’t think Uniswap V2 is going to be the final word on liquidity and just have sort of an infinite lock on the AMM market. 

Laura Shin:

Well, there’s going to be V3 coming out soon anyway, so…

Dan Robinson:

Right. That’s what I was referring to. Uniswap V3 will be there. 

Laura Shin:

Dan trying to be oblique. 

Haseeb Qureshi:

Yeah. So, I disagree with one of the things that Dan said which is that Uniswap has a network…

Dan Robinson:

Finally. I came on this expecting Haseeb and I to disagree on everything and so far we’ve been completely in sync.

Laura Shin:

Take out your boxing gloves.

Haseeb Qureshi:

He’s been eyeing me this entire time waiting to start a fight, and I’ve been telling him, look, I love you, I love Uniswap, but he doesn’t believe me. 

First, we can get to that maybe in a second, but the other thing that I want to say in response to Laura’s question is that I don’t think that…so, I agree with Dan obviously that Uniswap is decentralized in the sense that nobody controls it, but I also disagree with sort of the spirit behind the question, that the idea that having something that’s sort of outsizedly successful on chain is a sign that something is not decentralized, right, because I think you should sort of expect it to be natural that there’s some power law to like, what is the most valuable thing and what ends up becoming most used by people or most beloved or most effective, and if there’s some kind of scale effect to that thing then naturally it will end up becoming really big. So you know, it’s sort of like asking is crypto not decentralized because Bitcoin is the biggest one and it’s got, everyone loves that one? Is that kind of not decentralized? And so in the same way I think Uniswap being sort of the right now, it’s the number one largest gas guzzler on Ethereum today which I think speaks to how much people love it and how much value people get out of it, but I don’t think that’s any mark against being decentralized.

Dan Robinson:

I totally agree with that. That was very disappointing as a point to disagree with me on.

Haseeb Qureshi:

No. That was not the thing I disagree with you on. No. 

Dan Robinson:

Oh. Okay.

Haseeb Qureshi:

The thing I disagree with you on is the network effect. 

Dan Robinson:

Oh.

Haseeb Qureshi:

I don’t think Uniswap has a network effect, but…

Laura Shin:

Yeah. Well, that actually leads to my next question because I was kind of curious, I’ve seen people chatter about how another team could fork Uniswap and for instance, like I guess part of the fee now goes to the Uniswap team instead of all of it going to liquidity providers, and so someone suggested, oh, someone could fork Uniswap and then have all the fees go to the liquidity providers, or that someone could fork Uniswap and then launch a token with it before Uniswap does. So, how durable do you think Uniswap’s current lead is?

Dan Robinson:

So, I’ll let Haseeb answer that question because he is invested in a project that is doing something like that. 

One quick correction there is that there is a protocol fee switch in Uniswap V2 but it isn’t currently turned on, so right now all fees go to liquidity providers.

Laura Shin:

Oh, right. Right. 

Haseeb Qureshi:

So, this is actually…I was just chatting with somebody about this yesterday. I think it’s a very interesting question. 

So, there actually is a protocol. We’re investors in 1inch which is a decentralized exchange aggregator, basically meaning it lets you kind of route an order across, like, every single decentralized liquidity source. So, you could route it to Uniswap, Curve, Balancer, whatever, it just finds you the best price across everything, and so they launched their own version of Uniswap, a fork of it called Mooniswap which has a slightly different design, basically uses the same front end, kind of the same underlying mechanics, and Mooniswap, I think right now they’re at like 100 million in total liquidity. Their trading volume is still quite a bit lower than Uniswap’s, but they’ve managed to garner a lot of liquidity in a very short amount of time, and in a way I think this is not a complete validation, but it’s something of a sign in favor of hey, you know, Uniswap does not have a traditional network effect, right? 

And part of the reason why I say this, and I think this is actually an interesting question, right, Mooniswap is doing liquidity mining, so there’s going to be a small amount of the 1inch token which gets liquidity mined to early LPs who, you know, put capital into Mooniswap to help capitalize it and give it liquidity. So, naturally what that does is it juices liquidity, right, it brings a lot of capital into the pools and makes it so now trading can happen with lower slippage and so on, and that’s great, and it makes a lot of sense. 

The question is, could Uniswap counter by having their own version of liquidity mining and sort of pull it back in their favor, right? And I argue that actually they couldn’t, and this is kind of a weird argument, and this might totally off-kilter for this podcast, but I kind of want to talk about it with Dan so I’m going to go ahead and nerd out on it. Here’s the idea, right, when you do liquidity mining, basically what you’re doing is you’re rewarding early adopters, right? You’re rewarding people who come in early before it really makes economic sense for them to provide a lot of liquidity, you’re sort of subsidizing that early liquidity in exchange for them having some ownership over a future revenue stream that’s going to come when this thing comes at scale, right? When the thing is really big and has tons of liquidity then, you know, the people who got in early are going to get rewarded by getting some of the profits that would otherwise go to those latecomers, right? So, in a way it’s sort of a redistributive transfer from latecomers to early-comers, right? Okay. 

So, it makes sense that Mooniswap can do this. Mooniswap can do this because Mooniswap has no latecomers, right? It’s a tiny protocol and they’re rewarding the people who come on early, but Uniswap is already at scale. So, right now 100 percent of Uniswap users, all of the liquidity providers, they are getting rewarded with Uniswap profits, right? If you turned on liquidity mining in Uniswap then everybody who gets those profits would get, like, they would all be getting Uniswap tokens which would give them just a portion of the profits that they were already getting 100 percent of, so it’s almost like a stock split or something, it basically doesn’t do anything to the amount of revenue that they get because they’re already getting 100 percent of the revenue of Uniswap. 

So, my argument is that what liquidity mining really is is it’s a way for you to juice your early growth and to get early liquidity on board, but once you’re at scale, liquidity mining doesn’t really do that much for you. That’s my argument.

Dan Robinson:

So, right. I think you said a lot of interesting things there, but the piece that I’ll pick out is that you said that you think Uniswap is at scale, is mature, and I think there’s quite a long way to go. I think, in fact, right now the amount of liquidity in Uniswap is a small fraction of what I think ultimately it will be, and same for volume. 

Haseeb Qureshi:

That’s fair. 

Dan Robinson:

So, I would say it’s still early, as with Bitcoin and as with the rest of the space.

Laura Shin:

Yeah. That was pretty much exactly my thought. I was like, Haseeb, this is nothing yet. All right, but we’re going to talk in a moment a little bit more about the future of Uniswap and also talk about just the future of trading in cryptocurrency, but first a quick word from the sponsors who make this show possible. 

Laura Shin:

Back to my conversation with Haseeb Qureshi and Dan Robinson. So, Haseeb, I think you wanted to expound a little bit more on the network effects in automated market makers. 

Haseeb Qureshi:

Yes. So, I don’t believe that…so Dan earlier made the claim that Uniswap has a network effect, and I don’t think this is correct. I think you can clearly argue that AMMs have a scale effect, meaning that the more liquidity they have the better pricing they offer, and that’s great. However, let’s say I have a big pool of ETH-USDC and I have another pool of Ample-ETH or whatever, the ETH-USDC pool has no effect, it doesn’t help me in any way in getting people into the next pool that I start, so each marginal pool doesn’t really benefit from other pools that exist within the same protocol. So, when I think of a network effect I think the idea that nobody can compete with Uniswap once it gets to scale because of the fact that it just doesn’t ever make sense to list your pool anywhere else other than Uniswap, and that’s not really true. 

If you’re bringing your own…I mean, for something like Ample, right? They’re sort of bringing their own community, they’re bringing their own volume, and if at some point when Uniswap turns on fees they’re paying some of that volume to another set of users, at some point they might just decide, like hey, we’re just going to fork Uniswap or we’re going to use some other protocol that doesn’t charge fees, and we’re going to take that liquidity in-house. And so I’m not certain that you’re not going to see the same sort of thing that happened with 0x where they created this…you know, when 0x was a public good and everybody loved it and found it really fantastic, everyone was using it, and then once it turned on fees and once it started becoming, you know, you sort of ended up having this real question of like, hey, do I really want to be paying a third party for getting what is a nicely provisioned public good but actually not something I can’t just fork and build in house, that you won’t see the same sort of thing happening once Uniswap tries to do that. 

Laura Shin:

Okay. Yeah. So, we’ll have to see what happens, but one other thing that I wanted to ask about was Haseeb, you had written a couple of really great blog posts about automated market makers, and in one of them you said that over time you felt DeFi would probably attract serious market makers, and at that time we’ll see more complex markets, and the follow on effect will be that Uniswap will decrease in market share. Why do you think that?

Haseeb Qureshi:

There are a few reasons why I think that, and I want to caveat all this with saying that Uniswap V3 may really fundamentally change my analysis of Uniswap because from what I’ve heard it’s going to be very different, but let’s just take Uniswap V2 so we can actually talk on something.

Laura Shin:

Dan is just smiling. He’s not going to talk about it, but…

Haseeb Qureshi:

I know. Yeah. That’s right. That’s right. 

Laura Shin:

But we are looking at him.

Haseeb Qureshi:

That’s right. That’s right. 

Let’s take Uniswap V2 as the sort of core example. So, Uniswap V2, like I mentioned earlier, is a very inefficient market maker. Like, most market makers don’t need 100 million dollars on their balance sheet in order to make a market with pretty tight spreads, especially if they have pretty strong priors, or if they’re able to hedge, or they’re able to do other things that allows them to lower their risk relative to something like Uniswap. 

So, the problem with Uniswap was that Uniswap doesn’t know anything that’s going on in the world, right? It basically has no signals, no inputs into its pricing function besides just how much of asset A do I have and how much of asset B do I have. That’s literally the only thing it knows when it decides what prices to quote, right? Of course if you look at real market makers, if you just have arbitrary levels of intelligence, right, what would you do? Well, look at what real market makers do, they like use machine learning and they look at liquidity and they model volatility, and they do all this fancy stuff and they also, like, try to hedge so that they’re not taking on momentary risk, and they don’t really need as much balance sheet to absorb all of that trading. If things are going one way or another they can draw credit lines, they can do all sorts of stuff, right? They have unlimited freedom to do whatever they want.

And so now it’s obviously true that up till now DeFi has been so small that professional market makers just really had no interest, or like, the kinds of professional market makers who had interest in DeFi were very, very small and not very well capitalized, and they pretty much all traded on order book exchanges which also didn’t have a lot of volume, and so you sort of had this, like, negative, this kind of death spiral that was happening in DeFi market making where the order books were thin, they didn’t bring attractive market makers which led to order books being thin, and it sort of was very difficult to break out of this trap. 

Now it’s very clear that that has changed, right? Retail is here. Retail is trading on DeFi, and the first thing that every market maker I’ve talked to has seen is like, holy crap, if DeFi is going to be the next frontier, we want to be there because we can see, look at all these profits that Uniswap was making, right? All these profits that Uniswap was making is not because Uniswap has some brilliant pricing advantage, it’s not because it’s like way more gas efficient than any other way of providing liquidity. The reason why is it’s the only thing there. It’s the only place where you can actually get meaningful liquidity today. 

What I argue is that once professional market makers set up shop, which is not trivial, it’s going to take them a bit of time because it’s complicated and there are regulatory hurdles and all this stuff, but once they do, they are going to outcompete almost all AMMs that are pricing difficult to price assets, right? So, by difficult to price assets I mean things like ETH-USDC, which is a complicated pair to trade, right? You know, you couldn’t trade that from your bedroom very effectively. If you’re trading things like stablecoins or you know, Bitcoin, different wrapped Bitcoin things, then that’s pretty easy and you can use I think algorithmic market makers fairly effectively to you know, capitalize on fairly…you know, the hard part of that is inventory management, it’s not really pricing, right? But for something like ETH-USDC, I think you’re going to end up getting much more competitive on chain pricing because it’s a very liquid market, it’s a thing that a lot of people want to buy, and Uniswap is just going to get competed with by smarter, leaner, more intelligent actors, and I don’t think that algorithmic market makers are very well equipped right now to compete with those kinds of actors.

And when that happens, right, it’s not just that Uniswap sits there and is like, oh, man, I didn’t get as much volume this month. What happens is that liquidity providers are like, wait, crap, this is a crappy deal now, I’m not getting a whole lot of volume, I’m not really getting any trades, and instead what I’m getting is the only people who are trading with me is only when my pool is mispriced, and that’s when the market makers, the professional market maker is like, no, I’m not going to do this trade, that’s crazy, but Uniswap is like, oh, yeah, sure, x times y equals k, let’s do it, and what you’re going to get is this sort of inverse scale effect where now suddenly you’re getting this adverse selection where the only people trading with you are the people who are basically getting good prices, better prices than they’re getting anywhere else, and the professional market makers are actually getting the sort of majority of the retail volume as opposed to the arbitrageur volume. 

So, that’s my argument in a nutshell of why I think AMMs are going to recede from being the kind of core liquidity sources in DeFi.

Laura Shin:

Well, and then what do you think will emerge? 

Haseeb Qureshi:

I suspect what will emerge is a larger ecosystem of actors that looks more similar to what you see in CeFi. So, now what does that mean? There are many different ways you can imagine this playing out. So, in my blog post I sort of paint this picture of, you know, you can imagine a world where there’s a really simple software stack that any market maker can deploy that basically pushes out a smart contract and an API endpoint that anybody can ping that API endpoint, get a price quote, and then automatically submit it to an on chain contract which prints the trade, and so the contract kind of looks and feels like Uniswap, the assets are all on chain, it sort of has the same affordance to the end user, right? It’s like, whether I’m using Uniswap or using this or using that, like, I don’t really care. All I want is a good price, and if that’s true, you can imagine, like, a network of market makers from all over the world, none of whom you have to trust, you don’t have to sign an agreement with them, you don’t have to get to know them. All you have to do is just make sure they have a standardized contract and take their API quote, plug it into the contract, and it prints the trade at the price that they give you, and you can do that. You can do that today, it’s actually not that hard, but you know, most of these market makers are not DeFi natives, so they haven’t ramped up on any of this stuff yet, and of course up until now there hasn’t been any reason to. 

So, that’s one vision of how this could play out. I think there are obviously many others. It could be that order book exchanges start becoming dominant. It’s hard to say in advance, but what I can say is that I’m fairly confident that the landscape is going to change and that centralized market makers are going to play a larger and larger role in providing liquidity on chain.

Laura Shin:

And so, will that then…so basically that, you know, as you said, will be more like an order book decentralized exchange, but then in terms of, like, the interface, would that be something like 1inch exchange, or what would be the interface for that, or is just something like 0x?

Haseeb Qureshi:

So, I imagine there are going to be a lot of sort of front ends for DeFi, right? Like, right now, if you think about what DeFi is like to use today, right, there is, like, 20 different protocols, you go on DeFi Pulse to see what the newest thing is, and you Google it, you look it up on Twitter to see if it’s real, you do some trades with it, you try to look up, like, what orders, what books do they have, what actual trading pairs they do. It’s a pain in the ass. It’s super complicated. Like, this is not the way that DeFi is going to work at scale, right? Not every user is going to have to underwrite every protocol and figure out which ones are real and which ones are not. 

At some point there is going to be some, like, sort of almost like a brokerage experience which is sort of what aggregators are today. They’re sort of the closest thing to that brokerage experience, but there are also things like, you know, Instadapp or Zerion or things like that that show you, you know, look, you go to one place and you say, hey, I want to trade USDC for ETH, or I want to take out a loan, or I want to do this thing, right, and I don’t care about the brand, I don’t know what Uniswap is, I don’t know what Balancer is, I don’t know what any of these things are, I don’t care what they are, all I want to do is do the actual thing I want which is I want to buy something or I want to trade something or I want to invest something, right? That’s going to be what DeFi looks like at scale, to my mind, and when that happens…

Right now, you know, most of the users today are going directly to the protocols, and maybe something like 20 to 30 percent are going to aggregators today, and I think a lot of that is that people don’t really know what these aggregators are, it’s pretty new, and it’s pretty complicated to even express what’s going on on a lot of these platforms, right? Like, it takes a podcast like this to even explain what even is Uniswap to a lot of people who have kind of heard the name and just know things are happening, but eventually, you know, a year from now, two years from now, three years from now, it won’t be quite as exotic to be like, oh, yeah, I’m trading on DeFi, you know? It’s not that weird, and at that point I think there are going to be much simpler affordances that make it so that liquidity pools like Balancer or Uniswap are not the front ends. They’re not the things that users are directly interacting with. They’re sort of like wholesale markets. They sort of work on the backend. They’re somewhere in the supply chain of what users are doing, but they’re not the place where users are directly going because they don’t need to. There’s no reason why they would be committed to a single front-end if what they want to do is just to do the thing, not the protocol.

Laura Shin:

Yeah. So I know Dan is going to want to rebut that, or I think, but before he does that, I actually just want to kind of help listeners along. 

So, when we’re talking about aggregators, think sort of like Kayak or Expedia for travel, and that’s just what places like 1inch are doing for automated market makers there, just showing what is the best price across all the different DEXes, and then people can use that to trade rather than just going to each individual one and comparing.

And I will also say that I think these ideas are kind of percolating because I did see Jose Maria Macedo of Delphi Digital, he wrote something about comparing basically DEXes to what he called airlines and then saying, you know, the aggregators are where this is going to go, and so in his view, he was saying that DEXes, they need to focus on their supply side, like on the liquidity providers because the retail, the traders, that’s not going to be how they’re going to differentiate themselves, which you know, I have that in the show notes, his tweet storm and his articles on that if people want to look because it’s pretty interesting.

But anyway, so Dan, I know you disagree with Haseeb’s idea that order book DEXes will dominate in the future, especially like once we get to more high throughput chains, unlike Ethereum 1.0. So Dan, what do you think is going to happen?

Dan Robinson:

Yeah. So I think, as I understand Haseeb’s argument, it’s that professional market makers who have access, he’s right, to much more detailed information and can do much more sophisticated strategies than an on chain market maker will essentially outcompete Uniswap and offer a tighter spread and therefore draw away the high quality volume, and I think my response to that primarily is an analogy, and this is to look at in 1974 what you might have said to somebody who thought that they could compete with active managers of money by simply buying all the stocks in a list and just holding them, and this seemed like a crazy idea at the time, and it was compounded with another crazy idea of mutualization which was of running an actual mutual fund that was owned by its investors rather than being publically traded or having a sort of an independent management company, and Vanguard obviously founded around that time, now the second largest asset manager in the world, the largest mutual fund.

And I think that shows some of the benefits of simplicity and of low costs. So, market makers, professional market makers that Haseeb is referring to here, often have very high margins, they pay their employees a lot, they can often have a relatively higher cost of capital, and generally they don’t think it’s worth it to get out of bed for basically the kind of profits that right now people on on chain liquidity providers are willing to do, and so I think the advantage that automated liquidity providers have is that you can get much more capital deployed during the strategy by sharing all the equity benefits essentially of investing in the strategy without the liquidity providers and keeping sort of the overhead minimal, and by just making it as simple as one click just to provide this, and you don’t have to diligence some particular market maker on their strategy, and I think that’s part of my…and this is very speculative, but that’s part of my vision for how Uniswap could not only compete in DEXes and for crypto trading, but could potentially change the face of I think market making across the financial industry. 

Haseeb Qureshi:

So, three counters that I want to make to that. So, the first that sort of going in reverse order from some of the claims that you made, so one is that I agree, I don’t think market makers are going to summon as much capital on chain but they don’t need to, right? The whole point is that Uniswap and its like are very, very capital inefficient relative to normal market makers because they have to be, right? Like, in Uniswap we call inventory liquidity, which is a really weird way of thinking about inventory, right? Like, normally for a market maker liquidity and inventory are two totally different things, and you can provide liquidity with your entire inventory or not, and you know, Uniswap sort of uses a very narrow band of its inventory to do most of its trading. 

The second thing that I would say is that, you know, if you believe that Uniswap is not at scale, it sounds like you’re also implying that DeFi is going to grow really dramatically, and so if the bet is that market makers aren’t going to get out of bed for a market that’s going to grow dramatically, then it’s like, either DeFi is going to grow really big and it’s going to be an interesting pie for market makers to attack or it’s not, it’s going to stay small, and I think I agree with you, or at least to your prior claim, that DeFi is going to grow quite a bit, in which case I think the pie for market makers is going to be attractive enough for them to get out of bed and to attack this new market, and of course the thing is you don’t need that many market makers to do this. You know, you don’t need a Jump or a Tower or somebody really, really huge in order to get somebody on chain who can outcompete Uniswap. Like, you really only need one or two competitive market makers to get a really strong ecosystem going.

And then the third point that I would make is that, you know, the point you’re making about passive investing sounds like a great bull case for a set protocol, but I don’t think it really describes what Uniswap is, because Uniswap is not a passive strategy, it’s actually an active strategy, it’s just a very simplistic active strategy. I will always quote x times y equals k, and like, of course if that strategy is tremendously successful, right, like, you could get a centralized market maker to do the exact same thing, right? They do have the balance sheet. If the APRs are as sustainable as they are, somebody will absolutely deploy 200 or 300 million dollars to just put on on chain capital pools and just start earning those trading fees, right? Like, if that’s the thesis, then there’s no reason in principle why Uniswap and decentralized LPs would capture the lion’s share of that. 

So, all that being said, I think I agree with you that Uniswap has changed things, and there’s no going back from I think the new ideas that Uniswap has brought to the table, which I think were very difficult to anticipate in the beginning, right? I think incentivized pools and bootstrapping liquidity, these are really, really new phenomena that were theoretical before this and now it’s like very, very clear how it works and why it works and that you can’t take the genie back in the bottle. 

That said…

Laura Shin:

Well, when…

Haseeb Qureshi:

Go ahead. Go ahead. Sorry.

Laura Shin:

Oh, no. Go ahead. Finish.

Haseeb Qureshi:

I was just wrapping up this. 

That said, I don’t think when you look at the market share two years from now, I don’t think it’s going to look the same as it does today.

Laura Shin:

Yeah. Well, one thing I also wanted to ask about was this introduction of yield aggregators, like yEarn Finance which basically seeks out the highest yield from any kind of DeFi product, whether it’s a lending protocol or an AMM or a yield farming scheme, and I just wondered, do yield aggregators make it harder for AMMs to compete because it just sort of feels like right now people can sort of, you know, all flock to one or another very easily with just using yEarn, and I just wondered, like, how does that even affect the ability for some of these AMMs to survive? 

Haseeb Qureshi:

I don’t think it really affects them in a direct way. In an indirect way it does potentially because it allows these big, gigantic pools of capital to just, like, kind of fly around at a moment’s notice, and that’s a little bit scary, but…

Laura Shin:

Yeah.

Haseeb Qureshi:

…it’s sort of like they were sort of doing that anyway even before yEarn came around because lots and lots of whales were just sort of following the best liquidity mining rewards, and right now it’s pretty clear that liquidity mining is here to stay, and I don’t think you can blame the giant pools of capital flinging around left and right on yEarn so much as the underlying thing that yEarn is responding to, which is huge amounts of subsidies being paid out that primarily whales are capturing right now.

Dan Robinson:

Yeah. I think it is an important development. I think it’s something that we all thought was basically inevitable eventually, which is that on chain liquidity is going to be quite flighty, it’s going to seek the highest yield, and I think the protocols that want to be sustainable sort of have to design around that fact, and it’s a good fact in many ways. I mean, the means that you can, with some incentives, for example, just sort of summon a huge amount of liquidity that you wouldn’t necessarily normally be able to, I do think it poses extra risks, and like essentially almost any of these strategies are going to have some discretion by governance about what kind of pools are safe to put money into, and so I worry about it from that perspective, but I think it would be a great outcome if essentially people can just pull their liquidity as passively as possible and have it safely allocated to whatever on chain is providing the best benefit. 

Laura Shin:

Okay. So, one other thing I wanted to ask about which is just sort of a one-off, it’s like a little bit separate from everything we’ve been discussing, is the CRV token launch, which basically what happened there is that an anonymous developer supposedly not affiliated with Curve…so CRV token is for the Curve AMM which is the one we were talking about focuses on stablecoins, and this anonymous person or supposedly anonymous person decided to launch the open source CRV token contract and paid about 8,000 dollars in gas fees to do so, but a lot of people were surmising that this was actually somebody who either was affiliated with the team or that it was somebody who the team had orchestrated this with, and I wondered, you know, what do you guys think happened there and what do you guys think this incident says about how AMMs might launch tokens going forward? 

Dan Robinson:

So, I only know about it what I’ve read on Twitter, and it didn’t seem to me that anybody had an extremely strong theory about who this person was or what they were doing. I saw sort of a lot of speculation flying around about it.

It is kind of just a remarkable fact that this can be done and sort of a weird fact about DeFi and about token launches that yeah, basically whatever happened, if somebody random actually had just deployed the contract and like, got a sufficient community momentum around it, I could see how the team wouldn’t actually have a choice, and that’s a very weird thing where protocols essentially are not under the control of their creators, they’re under control of whatever the community wants to say is the CRV token, that’s the CRV token. 

Haseeb Qureshi:

Yeah. I think crypto has enough speculation so I try not to add to it personally, but I do think philosophically it’s fascinating, the idea that, like, the community itself just sort of decided which CRV token is the CRV token, and I think this is one of those things that will eventually show up in, like, a philosophy textbook someday of what makes the token the token of that community other than just people on Twitter deciding that it is and then starting to trade it and give it a high price, and the team…you know, if in fact the story as told is correct, that’s kind of fascinating that, you know, they sort of were like, it wasn’t up to them which CRV token ended up being the one that they had to sort of canonicalize as the core part of their protocol. 

Laura Shin:

All right, and one other thing we touched on briefly was some of these initial DeFi offerings or initial DEX offerings like the UMA token launch where they just put a liquidity pool, was it UMA versus ETH or what?

Dan Robinson:

That’s right. 

Haseeb Qureshi:

Yes.

Laura Shin:

Yeah. On Uniswap, and that garnered a fair amount of criticism at the time. Can you explain why that was and whether or not you think that means we will see, like, a lot more of those or very few of those? 

Dan Robinson:

Yeah. So, I would say, and I think most people would agree, Uniswap is not a fantastic mechanism for initial price discovery, certainly not if you put a huge amount of liquidity out there, and so one metaphor here is to initial public offerings.

Haseeb Qureshi:

You mean if you don’t put a huge amount of liquidity out there. 

Dan Robinson:

I’m saying, no, the more liquidity you put the worse because essentially the initial market maker ends up selling a bunch of…whoever is the initial liquidity provider ends up selling a lot of tokens at a price that isn’t close to where it ends up. 

Haseeb Qureshi:

Right.

Dan Robinson:

And so I think there have been some other experiments around this, and I think, you know, people have done sort of batch auctions, so yeah, one metaphor here is to initial public offerings where there is a lot of criticism of banks for mispricing initial offerings so that there’s a pop on the first day, and that’s to some extent an opportunity cost of the company. I think just initial price discovery is very hard, and you don’t want to necessarily sell too much at whatever initial price you set, no matter how well you valued it, and yeah, that said, I think within, you know, a matter of minutes I think in the UMA token space it was trading very close to where it was trading the rest of the day, and so you will get price discovery, you just may get it with some very weird trading prices at the beginning of the process. 

Haseeb Qureshi:

Yeah. I think there’s still some science to be done on exactly how initial price discovery should happen on DeFi. Obviously UMA I think was the first to really do this, and I should note, UMA is a portfolio company, they were the first to do this and I think they got a lot of criticism for it, but they really did kind of show, they’re like hey, here’s probably what you should not do and here’s what potentially goes wrong, and then there were other teams that have sort of neglected that lesson and sort of did the same thing like bZx and a couple of others. I think people are still playing around with, like, okay, can we do some kind of batch auction on Mantra or Gnosis or whatever? There are other folks who are kind of coming up with other ideas, but it’s still sort of unchartered territory.

The other I think question that nobody has a really good answer to yet is how much float do you need out there on day one for there to be good price discovery just in terms of there being enough supply? So if you have, like…you know, look at CRV, right? CRV was sort of linearly releasing CRV tokens out for trading which led to this really ridiculous first day fully diluted price of like 45 billion dollars which made it worth more than Ethereum, and then it’s slowly been ticking down as more and more supply, more float has been coming onto market, and you know, you saw the same things…we’ve seen this lesson many, many times in crypto with Zcash or with Grin that having this sort of effect where a tiny amount of float comes out on day one and it slowly, linearly increases until at some point it stabilizes, it probably behooves tokens that are thinking about getting some kind of price stability on day one to try to make sure there’s enough float out there initially for that price discovery to be useful. 

But I don’t think anybody has done a full, you know, a full post mortem on exactly what are the right parameters for a new token. 

Laura Shin:

Fees on Ethereum are extremely high right now. In the last two days they were as high as almost seven dollars, six dollars and sixty cents, and at the time of recording they were still at two dollars and 40 cents, and obviously Ethereum is still in the midst of this lengthy, multiphase transition to Ethereum 2.0. So, Ethereum 2.0 is not going to be able to accommodate this kind of DeFi activity for a while, and meanwhile we are seeing that FTX is going to launch a DEX called Serum on the Solana blockchain and the FTX founder, Sam Bankman-Fried, calls Solana 10,000 times faster and one million times cheaper than Ethereum, and Serum apparently will also be interoperable with Ethereum. So, I wondered if you think…and it’s not even just about Serum, but I just wondered in general, do you think that there’s a real risk that the DeFi community or ecosystem could move to another blockchain? 

Dan Robinson:

I think my views on this are pretty close to Haseeb’s, and he explains it very well, so I’ll let him answer.

Haseeb Qureshi:

What are my views? Hold on.

Dan Robinson:

The Wall Street analogy.

Haseeb Qureshi:

Okay. Okay. Okay. Got it. Got it. Yeah.

So, my take here is that I think at this point it’s basically over, in the sense that Ethereum has already solidified itself as being the Wall Street of crypto, right? Now, there is a very real question that, like, look, Manhattan is now crazy crowded, the rents are insane, like, traffic congestion is just unthinkable, and so if you’re not a gazillionaire you probably don’t want to live in Manhattan anymore, right? But that doesn’t mean that all the finance in the world is only going to live in Manhattan, right? There are going to be…like, even with gas prices being as high as they are, obviously they are this high because people are willing to pay these prices, and those people are mostly whales and people who are trading at very high volume and so it’s worth it to them to trade on these venues, but eventually there are going to be other places that open up, other venues that sort of become, like, the analogy that I’ve given before is like sort of becoming the Chicago to Ethereum’s New York, right? 

There will be some second place with some kind of bridge, whether it be a Layer 2, whether it be an interoperability solution like Polkadot or Cosmos, or whether it be another Layer 1, there is going to be some second place where people can go and probably initially that place is going to service a lot of the lower expense transactions. So you know, folks like you and me who aren’t really looking to place 10,000 dollar trades on any given day, we’re going to be trading, you know, for closer to one cent or five cent fees on these chains that don’t have as much liquidity, they aren’t going to be as juicy, they’re not going to have as crazy yield farming, but they’re going to place where at least in the early days we’re going to start doing some of the normal stuff that we used to do on Ethereum.

And so I suspect that’s the way you’re going to see some specialization among chains for DeFi, is there will be some chains where the really high value transactions take place and those chains will probably be more congested, but then there will be places where lower value but higher throughput stuff can happen which will likely be offloaded to the newer places where DeFi is going to migrate.

Laura Shin:

And do you think that might be what happens with Serum?

Haseeb Qureshi:

I think Serum is much more of a verticalized bet than I think what I’m describing with some of these other chains. It’s sort of like…I think Serum is kind of trying to king make Solana right now, where Serum is like this big, fat, super impressive project that I think both Solana and Serum are betting that like, okay, this is going to be the big thing that makes Solana work because Serum is going to be so big and awesome and attract so many users that it’s going to cause this ecosystem around it.

And this was actually a big thesis that a lot of investors were entertaining back in 2017, was this idea that like, okay, yeah, Ethereum is kind of working for some stuff, right, but when you really get the killer app, that killer app is going to decide who wins because they’re going to decide, hey, Ethereum is too crowded, it kind of sucks, EVM is painful, we’re going to go to this chain and then they’re going to bring all their users, right? In a way that’s kind of like the story of the Libra and the story of some of these other big things is that they’re going to bring so many users that they’re going to cause this ecosystem to either migrate or have a new parallel ecosystem develop around them. 

Whether that will happen, we’ll see. I think it’s tough. I think that Serum has a really, really big lift ahead of them, but if it works it could obviously be really, really incredible. 

Dan Robinson:

I’m looking forward to the Serum launch in part because it’s going to be an order book-based DEX, and so this will sort of test the question of whether order books for DEX work better if you eliminate the scaling constraints.

Haseeb Qureshi:

I don’t think it’s sufficient, but it is necessary. 

Laura Shin:

Yeah. Well, so we’ll see what happens. All right. So, let’s now switch to predictions for just maybe, I don’t know, the next few years, and when you give your predictions for what will happen with AMMs I’d also be curious to know what you think will happen with centralized cryptocurrency exchanges.

Haseeb Qureshi:

Dan?

Dan Robinson:

And this is on what time scale predictions? 

Laura Shin:

I just think maybe, I don’t know, like three years. Three to five, I guess.

Dan Robinson:

So, I will lay down a marker which is that I think Uniswap V3 will be viewed as a really significant development and will change the AMM landscape, and that’s easy for me to say, but if I ever come back on this podcast you can make fun of me if I was wrong. 

Laura Shin:

Well, you know, I probably will have to…

Haseeb Qureshi:

Can I make fun of you anyway? I’m joking. 

Do you think Uniswap will still be the number one venue for trading in three years?

Dan Robinson:

I think that’s the way to bet, and I will make that bet with you, Haseeb, maybe offline. 

Haseeb Qureshi:

You’ll make a bet with me. All right. All right. We can do it on air too, I think Laura likes these things. 

Dan Robinson:

I think mostly likely, in part because I think they have a head start and also because I’m just generally extremely impressed with the innovation that’s been going on with the team and I think the community at large. So, that’s one.

I do think we’re going to see some highly specialized AMMs come out, and I’m thinking especially of options and maybe of stuff like prediction markets where you could see something that’s very highly optimized for a particular use case, and so I think we’ll see some of those, but in my view I think Uniswap, and especially with what it’s got coming down the pipe, will bring up the lion’s share of volume. 

Laura Shin:

Haseeb.

Haseeb Qureshi:

So, my prediction, unfortunately I don’t have any really exciting picture to paint, but I think trading volume is going to continue to grow. I think that DEXes will eat more and more of centralized exchange volume, and you will continue to see this phenomenon that a lot of tokens, especially if they’re on Ethereum, are going to start trading on Ethereum before they start trading on centralized exchanges, which is going to continue to drive people into the front doors. 

I think that you’re eventually going to get to a place where the experience…so one, I think we’re already starting to see this but you’ll see it more and more, centralized changes starting to integrate DeFi and to basically offer some of these services directly in house, because of course they want to be at the front end of crypto too. They see what’s going on here and they realize it’s a threat, and you know, these guys aren’t stupid and they’ve got big businesses to protect, so I think that’s going to be one thing that we see, and I can see in that sense, like you know, like we mentioned before, sort of DeFi becoming almost like this back end to the services that people actually directly interface with, and it might be that people still end up trusting the exchanges and spending a lot of their time there, but as exchanges get more regulated or get more onerous or just get more annoying or are sort of too overprotective of their customers, those customers might choose to go direct, and if they do, they’ll go natively on DeFi and they’ll do all the crazy stuff that Dan and I are talking about.

I do think that a few years from now it won’t be that weird for people to be trading on DeFi. I think it’ll still not be like your mom, but it’ll be like your smart tech friends who will all be…it just won’t be that hard. You’ll tell them, like hey, just use this app or use this website. It’s like, pretty good. Make sure you don’t touch any of the weird stuff, and it’ll route your orders to the right place and you’re not going to shoot yourself in the foot, right? And that’s most of what people need in order to navigate DeFi. 

So, I suspect that things are really only going to become easier and easier to use, and that’s the thing about technology is that centralized exchanges, we kind of already know what they are, you know? There’s not that much more room for innovation other than just adding more products or making the thing slightly smoother, DeFi has so much to go, and the thing about technology is that it only ever gets better, and that’s the scary thing, is that it’s difficult to put a cap on like, how good DeFi is going to get. It’s hard to know, and right now DeFi is only…the only thing happening in DeFi pretty much is spot trading, and of course we already know spot trading is a tiny portion of all the trading in crypto, and so I think that’s going to change as well, but that’s going to also take some time.

Laura Shin:

All right. Well, we will see what happens. All right. Where can people learn more about each of you and your work? 

Haseeb Qureshi:

You can find me on Twitter @haseeb and you can find some of our writing at dcp.capital where we publish research. 

Dan Robinson:

And I am at @danrobinson on Twitter. 

Laura Shin:

Perfect. All right. Well, thank you both so much for coming on Unchained.

Haseeb Qureshi:

Thanks for having me, Laura. Thanks for doing this, Dan. It was fun.

Dan Robinson:

Thank you. It was. 

Laura Shin:

Yeah. It was really fun. Very excited to see where all this goes. 

Dan Robinson:

Totally. 

Laura Shin:

Thanks so much for joining us today. To learn more about Haseeb and Dan, check out the show notes for this episode. Don’t forget, you can now watch video recordings of the shows on the Unchained YouTube channel. Go to YouTube.com/C/UnchainedPodcast and subscribe today. Unchained is produced by me, Laura Shin, with help from Anthony Yoon, Daniel Nuss, and the team at CLK Transcription. Thanks for listening.