(Bloomberg) -- On this episode of Odd Lots, we speak with Hayden Adams, the creator of the Uniswap Protocol, which powers the world’s largest decentralized exchange. Unlike traditional trading venues, Uniswap allows assets to be exchanged without any custodial requirements, permission, or even accounts.  You can find the episode here. Transcripts have been lightly edited for clarity.

Joe Weisenthal: Hello and welcome to another episode of the odd lots podcast. I'm Joe Weisenthal

Tracy Alloway: And I'm Tracy Alloway.

Joe: So Tracy, you know we recently had the Coinbase IPO. 

Tracy: Yes, we did — a big moment for crypto that was, I think the analogy I most saw used was that it was like crypto’s coming out party.

Joe:  Yeah, I think that’s fair. I saw it compared to the Netflix IPO, maybe some comparisons to the launch of the CME Bitcoin futures in late 2017, but certainly quite a big moment for the crypto world.

Tracy: Yeah. I guess it capped a really big year for crypto in the sense that we've seen this massive institutional adoption or at least a lot more than a lot of people had expected. We saw PayPal saying that it would allow its users to buy and sell crypto. We’ve seen parts of Wall Street start to get more interested. We seem to be very, very close to getting that long-awaited Bitcoin ETF. And so it really feels like all of that momentum sort of built up and then exploded in the Coinbase IPO.

Joe:  Yeah, that’s exactly right. But there is an irony with Coinbase and you mentioned crypto ETF, but there is an irony with Coinbase, which is that, you know, if crypto is supposed to be cutting out legacy financial institutions — like that’s a big part of the sales pitch — but A) Coinbase just went public on legacy exchanges. But beyond that Coinbase itself is kind of in a weird way, a legacy financial institution itself. It handles fiat currency. People send in their dollars or euros or whatever, and from there they can trade various digital currencies, but they’re holding dollars and all kinds of — they’re a gatekeeper and they have to do anti-money laundering regulations and know your customer. And they have all kinds of personal information on their customers. And you have to tell them your name and upload an ID and everything. So although it’s kind of like a crypto exchange, it is a legacy financial institution in some very real way.

Tracy: Well, it’s also a centralized order book, which kind of flies in the face of a lot of crypto ethos about decentralization and, you know, trading between parties without a third party in between them.

Joe: Yeah. Exactly, right. it bears many resemblances to a traditional exchange or traditional prime brokerage for its institutional customer or traditional online brokerage, like maybe a Schwab for retail traders. So it's only like partially new. But that being said, there is a phenomenon that’s growing in the crypto world. And I don't think we’ve talked about it yet, but it’s been growing for the last couple of years. And that is attempts at creating markets that are truly, truly decentralized. So no company handling the trade.

Tracy: Right. So this is a really interesting project from a market structure perspective because normally you would have a third party who’s stepping in to provide liquidity as needed. But if you just have market participants who are trading with each other in a truly decentralized environment, then you have to figure out other ways to encourage liquidity. And I think that’s where this aspect of crypto actually becomes very, very creative. It’s also where weird stuff tends to happen. And I think we’re going to get into that.

Joe: It’s super weird. It’s super different. It's a very different structure. Obviously, if there's no company, there's no one to send your cash to, so that also is its own weird thing. How you solve that, but there's also no gatekeepers. There's no setting up accounts. It's just a very different thing. But there is an argument to be made that a sort of decentralized trading environment is much more true to the crypto ethos.

Tracy: All right. Let’s get into it.

Joe: Yeah. I'm super excited to talk about this, to talk about DeFi, how decentralized trading works. We are going to be speaking with Hayden Adams. He is the founder and CEO of Uniswap Labs. He’s the inventor of the Uniswap protocol. And this is basically a trading system that runs on top of the Ethereum blockchain. And it is a very big deal. As of right now — we are recording this April 27th — over the last 24 hours, one and a half billion dollars of trading volume has been done over Uniswap. That's about half of Coinbase. So here is this decentralized exchange that’s half as big already as the preeminent crypto exchange. It's a very big deal growing extremely fast, but I don't think most people have any understanding about how this all works. And I would include myself in that. And so we're going to learn about Uniswap and decentralized trading and what decentralized finance is with Hayden. So Hayden, thank you very much. Thanks for joining, for coming on Odd Lots.

Hayden Adams: Thank you for having me on. I'm really excited to be here.

Joe: How did we do in that introduction? 

Hayden: Fantastic. Actually, I'm extremely impressed.

Joe: Okay — it's all downhill from here, I promise. But if you're impressed with the introduction, then I'm happy. But let's just start really big picture. It's like ‘DeFi.’ I see that all over the place. It's a super popular buzz word. I'm sure. If I looked at Google Trends it would be a straight line up. How would you describe what is DeFi? 

Hayden: I think a good place for starting talking about defy is Bitcoin. You know, Bitcoin is this decentralized system for storing and transferring value over the internet. You can think about it as magic internet money. And it has these properties that people care about and people like about it. It’s provably fair, it's secure, there's no single centralized party that controls it and it's this global system, that can be accessed from anywhere in the world. But you know, it's also still limited in being money, right? It's limited to storing and transferring value. And early on in the blockchain days people started to think about what are other applications of blockchain. And so Ethereum was born and Ethereum makes it easier to build various other applications that have some of these properties people care about, about Bitcoin, that it — it's provably fair, that anyone can audit its entire history that no one controls it — and apply that to other types of applications.

And it does this with what are called smart contracts and smart contracts is essentially code that is run on a blockchain. And so the same way that every Bitcoin node verifies every transaction on Bitcoin, every Ethereum node verifies the execution of every program running on Ethereum. And so DeFi is essentially the idea that finance goes far beyond just money and storing and transferring value. There's entire other world of lending, borrowing exchange, insurance, synthetics and options and other types of derivatives. There's an entire financial system. And we can take some of what we've learned from Bitcoin and apply and build systems that have some of the properties that we care about in Bitcoin to this broader class of financial use cases. That’s essentially DeFi. Uniswap is a DeFi application living on Ethereum. It's the most popular application and it applies basically this to decentralized exchange.

Tracy: So I'm going to jump right into something that you guys do that's different to some other exchanges, but you have basically every coin in existence able to trade through your platform, including like ones that I think are probably jokes or scams. I think that's fair to say. Why did you decide to do that and how does that differ from some other competing platforms?

Hayden: I think that what's really interesting about Uniswap is it makes it incredibly easy to create new markets, and trade on markets. And so it ends up being extremely good for the long tail of assets. And that gets into this kind of unique infrastructure that UniSwap… Uniswap is not a traditional order book exchange. Uniswap is what's called an automated market maker, or it uses automated market-making. And what that means is it essentially allows anyone to spin up their own market. So in traditional exchange, you essentially have two main types of participants. You have professional market makers who are constantly putting up buy and sell orders. And then you have, you know, retail traders or takers who are executing against those orders. And it can be extremely hard to create liquidity in the long tail of assets because the professional market makers that are necessary to the function of order books, they don't necessarily -- it's not worth it for them to maintain inventory or to bootstrap the longer tail markets. Because one way to think about it, right. You know, it's only really worth it for them to kind of market make on the largest, most popular assets.

And so with Uniswap it adds what's called automated market-making, which allows basically anyone who wants to create these markets and deposit assets into a smart contract. And that smart contract will automate the process of market-making for them, such that for them it's basically a passive experience. And so in two minutes, someone can spin up a new market, create liquidity in it, and they don't need to be extremely sophisticated. They don't need to have this market making background. They don't need to work with other professional market makers. And so we kind of remove this gatekeeper in the creation of liquidity.

Joe: So let's talk about how automated market-makers work. As you pointed out, in a traditional market there's someone in the middle, they always are posting a bid and an ask, and there's a little gap between them. And then anyone can come and take either side of it. It works very different in the sort of liquidity pools of an automated market maker. Explain the basic functioning of it.

Hayden: Essentially each market on Uniswap is a smart contract on Ethereum. What that means is that it's a little program, it runs on Ethereum and something that's really interesting about smart contracts is it's code that can hold funds. And it's code so you can basically create arbitrary logic. And then that logic dictates how the funds stored in that smart contract behave. And so an automated market maker is a way of coordinating market participants within Uniswap. So, you know, there's two classes of users, right? We call them traders and then liquidity providers, or takers and liquidity providers. And creating a new market on Uniswap is incredibly easy, the same way that anyone can kind of create a new account on Bitcoin or on Ethereum. And anyone can deploy new code to Ethereum and anyone can transfer value on Ethereum. Anyone can create a new market on Uniswap.

There’s basically a smart contract that dictates the creation of new markets. Anyone can basically create new assets, right? You can also create new assets on Ethereum very easily. So anyone can create a new asset on Ethereum, and then they can add that asset to Uniswap. And in the same way that Ethereum is decentralized and permissionless and can be accessed anywhere in the world, the same exists for Uniswap markets. And so anyone can create a new market by calling what's called a factory smart contract, which deploys a new market for two tokens. And then they can create liquidity in those two tokens by basically depositing a sum of two assets into that smart contract. So you might deposit some USDC, which is a stable coin, and some ETH, and then that creates a marketplace between ETH and USDC.

And then what’s very different is that rather than having just market makers posting bid and asks, essentially the smart contract manages the market making for you. So you just deposit capital into the contract. The contract automates the pricing, the price updating, the rebalancing, and then people can immediately start trading against it. And so you're not matching up buyers and sellers, you have this smart contract, which will always buy and sell in either direction in either asset and then people who trade against that. So you're not coordinating between people, but you're actually coordinating people to a smart contract.Tracy:So this is where the liquidity provision becomes quite important, right? And different to other types of traditional exchanges. So you have to incentivize people to contribute to the liquidity pool in order to be able to provide liquidity for, as you put it, the sort of long tail of crypto assets — things that people might not naturally make markets in.

Hayden: Yeah. So what's really interesting is in a traditional order book structure, essentially all market-makers are competing against each other. Basically the first person to put up an order is executed. In an AMM, or an automated market maker, it essentially pools liquidity across thousands of different liquidity providers. And together they function as a single market maker, sort of sharing the same strategy in this automated market maker. And so you might put in a hundred dollars, someone else might put in a thousand dollars, someone else might put it in a hundred thousand dollars — and that's all pulled together. And market makes us a single unit. Uniswap basically has built-in fees. And so every time someone makes a trade, there's a fee taken on a trade and that's paid out to liquidity providers, as are called. So liquidity providers are taking on some price risk, because the automated market maker is managing their liquidity for them and buying and selling tokens. And there is some price risk being taken on, but that's compensated for in fees being paid by people who want to trade against it.

Joe: Right. So explain that. So you mentioned, for example, a pair trade between ETH and USBC. And if I look at the website CoinGecko right now, that is actually the ... it looks like that's the highest volume-traded pair. So if I had some ETH and I had some USDC, I could put both into this pool and then what? I'm locking it up for a defined period of time, or for period of time, and then how much am I getting paid? Explain to me exactly the mechanics of what my incentive is to lock them up there.

Hayden: You’d lock it up for as long as you want, essentially. So you can lock it up for second a day. Well, I mean, the time’s s block, so 15 seconds is the minimum, or a day or a year -- however long you want. Let's say you put in a thousand dollars and the entire pool is a hundred thousand dollars. You are now 1% of that liquidity pool. And so you are earning 1% of the trading fees on that pair. And so in terms of the profits right now, there's a 0.3% fee taken on every trade. So today's volume was $1.2 billion. And so, $1.2 billion times -- I can just do the math right now -- but so it's a 0.3% fee. You're getting about $3.6 million in fees today on Uniswap distributed to liquidity providers. One way to think about it is, if a pair is earning a thousand dollars in fees per day and you're 1% of that pair, and then you're earnings about $10 in fees per day. If you're 10% of it, you're running a hundred dollars per day. And so the returns are very different across different pairs. There's tens of thousands of different trading pairs and they're all doing different volumes. And so returns have been all over the place.

Joe: Just to be clear, the trading fees for end users. So not everyone has to lock up — some people might just want to go from one coin to another. Those are constant, but what determines how much the liquidity providers get ultimately is based on their share of the pool. So in theory, is that how the remuneration structure works for the liquidity provider? Just like, how much of the whole pool of their stake is?

Hayden: What we're calling fees is actually much closer to what a spread would be on a order book. And there's no exchange fee, right? The whole point is it's decentralized. And so you have participants who are creating these liquidity pools. And then there’s this spread essentially, which is what the 0.3% is, which is collected on every trade. And those are paid out to liquidity providers, proportional, pro-rata, proportional to their portion of the liquidity pool.

Tracy: What's the downside of providing liquidity in this way? Because I imagine without someone in between the trades or a traditional market maker, you could for instance, quite a big spread, I guess. Or at least, prices may move before a trade is actually executed on?

Hayden: Well, so there's risks to being a liquidity provider. Being a liquidity provider is similar to being a market maker, which is that you take on price risk. Basically you take on risk in the divergence between the two assets, right? Because you are putting up two tokens and you're getting out two tokens, but you're not getting them out at the same ratio if there's been a price change. And so,  Uniswap is constantly being arbitraged against other markets. If there's a very large price movement, you might have sold some of your token — Uniswap essentially sells on the way up and buys on the way down -- And so if there's a very large price movement, you might've sold some tokens at a sub-optimal price and lose some money there. But you're earning fees along the way and very frequently or very often that makes up for it.

Joe: So someone could put in 10 ETH and 10 USDC, and okay, I'm going to forget about it for a year and collect some yield. But I do run the risk of if ETH crashes in the meantime.

Hayden: Yeah basically. But it could crash and it could still be profitable in a world where enough fees have been collected. One thing to think about here is if you're thinking about maybe the biggest market pairs in the world, right, Uniswap still can compete on these. But as I kind of highlighted before, let's say some new asset gets created. It just got created today, let's say, and they can't immediately just reach out to professional market-making firms and say ‘Hey, we created this new asset. We're one of 5,000 assets that got created today on Ethereum. How do we...’ And I'm barely exaggerating there in the numbers. Essentially all the creators of that asset need to do is they basically can deposit some of that token, one token and some of another token into Uniswap, and there's immediately a trading pair that people can trade against immediately. One analogy is it's almost like the user-generated content -- like Netflix versus YouTube, but for liquidity. Netflix, you kind of have this limit in the amount of content you can create. Whereas this is like user-generated content but for liquidity.

Tracy: So just on that note, I guess one of the problems with decentralized, user-generated content is that it could migrate somewhere else. And I have to admit, I just heard a little bit about this (I had some people actually asking if we could do an Odd Lots episode entirely on this) but could you please explain what happened with Sushiswap and this idea that they siphoned off your liquidity?

Joe: Oh, this is the good stuff here.

Tracy: Yeah, so people call this a vampire attack, which immediately means we must ask you about it.

Hayden: Yeah. So funny enough, you know, the siphoned-off thing is really funny because before Sushiswap launched, Uniswap had $300 million in its liquidity pools and after Sushiswap it had about $1.6 billion. And today it has about $9 billion. It definitely hasn't siphoned off liquidity. But essentially Uniswap is a decentralized protocol, right. And something that's sort of core to smart contracts on Ethereum is this idea of open source software. Uniswap was basically built entirely open source. So all the code is publicly available so that people can verify the kind of workings of it. And basically the whole idea is users don't need to trust anyone with their money, they're just trusting the code that's being run. And so all the code with Uniswap is publicly available. And so essentially what we saw with Sushiswap was someone created a fork of Uniswap the same way, we've seen forks at Bitcoin and forks at other crypto platforms, and they released a new token. They basically said anyone who uses our version of the protocol will be earning Sushi tokens.

And so they kind of incentivized people to basically put their liquidity in Sushiswap instead of Uniswap to kind of try to compete with Uniswap and grow their own liquidity. And for a awhile that had an effect where, there's a little of bit added context here, which is that when you create liquidity on Unitswap, you get a token that represents your portion of that liquidity pool. I had mentioned that if you're earning, if you're 1% of the pool, you're earning 1% of the fees, there’s actually a token that says ‘I own 1% of the DAI and USDC in this contract.’ So what Sushiswap did is they basically incentivized people to deposit their Uniswap liquidity tokens. While they were deposited, they were earning fees, basically they were earning Sushi tokens. And then at the very end of this, I think it was like a 10-day period, basically they migrated all the liquidity over to their own system. The actual result though was, funny enough, a huge increase in Uniswap’s liquidity during that period of time and post-migration actually a huge amount of that liquidity stayed and the overall system has grown. So it's been quite beneficial in the long run for Uniswap. But it was a very interesting time for sure.

Tracy: So what were you actually thinking when this was happening? It's a 10-day period when people have basically, you know  they're leeching off the protocol that you created in order to steal your liquidity. How did that feel? I'm just curious what you were doing during that timeframe?

Hayden: First some context, right. I had been working on this project for over three years at that point. So I know the first two years I was the only person working full-time on Uniswap. And then I've kind of built out a company over the past year and a half, two years, and someone came along, and then probably in five days they created a new token and they forked the entire protocol and said, ‘Hey, you know, use this one instead.’ A huge portion of those tokens were reserved for the people who created it. There is this open question in DeFi generally, which is how do creators capture value and how are creators rewarded for their efforts? And there was this element of maybe a little bit personal to it, but at the same time, that's also part of, you know, the fact that it was open source also fair game, right. It's part of the ethos is the fact that you can kind of have competitors spin up. Uh, and then in the long run, it's, it's worked out

Joe: So many questions. So many things ... I'm trying to think which way to go next, but I will say, so Uniswap on May 5th — so I don't know exactly what day people are going to be listening to this -- but on May 5th, according to something, you are launching a version three, and we can talk about some of the upgrades and why you launched new versions, but you are releasing that code under a different license. Am I correct?

Hayden: Yeah,  correct. We're using something called the business source license, which essentially, all the code is public, all verifiable, you know, people can build on top of it and integrate it. And for the first two years or two years at a maximum -- it could be less — but two years I started this hard cap, essentially. You can't, you can't work it in a production center.

Joe: So in theory, you can’t be Sushiswapped. You can’t be vampire attacked for at least two years?

Hayden: In theory.

Tracy: Does that sort of fly in the face of the decentralized ethos? I get that it's always a balancing act, but it feels like one of the big selling points of Uniswap was that it was truly a decentralized exchange. And if you move away from open source ideals then maybe it isn't.

Hayden: I think that it's truly decentralized in that it's all completely verifiable. No one controls it, right? There's still this no control, even the kind of license -- there's this sort of community governinge system that I haven't even mentioned, or we haven't gotten into, but the community governing system actually does have … so the Uniswap community does have the ability to kind of grant exemptions to that license, or even reduce the kind of two year period where it can't be forked. One of the kind of funny things is that there is this sort of community of users and developers around Uniswap. And when we were releasing the Uniswap V3 code, there was kind of this feeling of what would the Uniswap community want? Does it want to be Sushiswapped day one, does it want to be Sushiswapped day two? Or does it want to have time. The kind of long-term moats almost, if you think about it, of decentralized protocols, the best example would always be Ethereum for these types of things, but the long-term value is in the network effect, in the community that builds up around it. And so the idea here is essentially to kind of protect the Uniswap community in the short-term and in the long-term, there'll be this sort of massive network effect built up around it.

Joe: Explain the role of the Uniswap token in the network, because there is in addition to the exchange, the protocol, like everything, there is a token associated with it. My understanding right now is that the token doesn't really confer anything. It's not equity. I don't think from what I understand, the token holder actually collects any of the trading fees or the spread or anything. And yet the token has gone absolutely nuts. And if you bought it a year ago, or when it came out, you'd be sitting on a fortune right now. What is the role of the UNI, the coin?

Hayden: Yeah. So I think that it might even be helpful here to back up and give some added context. The value of DeFi and the value of Uniswap is in its decentralization, right? And decentralization, there's kind of a lot you can pack in that word. Right. Does it mean that no one controls it? Does it mean that there's no kind of central points of failure, and there's different ways of achieving decentralization, right. One way of achieving decentralization — and we've seen a lot of people talk about it — is through automation. Right. So what that means is basically relying entirely on code essentially, right? So smart contracts can have logic that's built in, and that logic can basically lead to a very decentralized world, right?

Where you're not relying on any people or any company to follow through. You're just relying on the code to execute as defined. And so at its core, Uniswap is decentralized in that the logic for Uniswap — the protocol -- is entirely in these on-chain smart contracts. One of the benefits is that it's not custodial. So people who put their tokens in, they're not trusting any third party. If you deposit liquidity into Uniswap that liquidity, the sort of right to claim that liquidity — that liquidty sits in a smart contract and that smart contract is not controlled by anyone or anything other than Ethereum's consensus. Your private key — if you have an Ethereum private key — that gives you the right to claim your portion of the liquidity. No one else can claim it.

And you're not, you don't need to trust anyone else. This type of system, that's the most important version of decentralization. But there are things that can't be decentralized in that same way. Another way of achieving decentralization is relying on almost market dynamics. And so a really good example would be like proof of stake in Ethereum is basically achieving decentralization and proof of work, I guess, and Bitcoin is,  achieving decentralization through economic incentives. Essentially people have an incentive to act in a good way, or they're losing money. That's another version of de-centralization and an incredibly important one. But when you talk about, a decentralized ecosystem, there are still things, there are still sometimes decisions that can't be perfectly made and don't have perfect answers from a market incentive standpoint, right?

Those types of decisions are still important towards building out this decentralized finance ecosystem. The way that I think about Uniswap and the Uniswap token, which is a governance token, essentially its role is taking on decision-making in the best way possible, in the most decentralized way possible. Making decisions that can't be immediately automated and can't be … It’s essentially coordinating the human element of the decentralized finance movement and of Uniswap, which can't just be perfectly automated or it can't perfectly rely on market incentives. That's essentially the role of the Uni token. And so something that's really cool is Uniswap token holders can't for example just withdraw all the funds from the swap protocol, right? That's a decentralized, it makes it more decentralized, right. But there are kind of these other things, right? Like long-term community ecosystem, growth and development.

There's a lot of value that can be kind of created by the Uniswap protocol. And there's a lot of sort of human work and effort that can be coordinated towards building value and in growing the protocol. That’s sort of the role of the Uni token. So the only kind of, or the main kind of governance action that it's taken so far is basically spitting out this, this grants committee. So you know, what the unified governance is, is it's an on-chain governance system so Uniswap token holders can vote on actions from a smart contract. And that smart contract has a treasury.

Joe: To be clear, could at some point, the Uni token holders vote to pay a de facto dividend to Uni token holders down the road?

Hayden: Well, Uniswap governance is a smart contract and it's very broad in what it can do. And so definitely, Uniswap token holders or Uniswap governance has this thing that we call the fee switch, which is a sort of, built-in, kind of, fee that can be taken off the, you know, basically it can come out of liquidity provider fees. So right now liquidity providers earn .3% but these Uni token holders could vote to add up to a .05% fee, which is a pretty low fee kind of coming out of that. That could theoretically be kind of given, you know, it's sort of a governance system, so it can do whatever it wants with the fees that it got. Or whatever the Uni token holders want and whatever the community wants. So the Uniswap governance system has it's like built-in treasury. 

And so far, it's basically the main things that it has done is basically fund other teams and projects and developers building on top of the Uniswap ecosystem. So there's other kinds of teams, building analytics platforms and other interfaces, and those kinds of gifts have been funded out of Uni token holders who voted to fund that. So there's this community treasury. It technically has $14 billion worth of Uni tokens in it. And so it doesn't necessarily need trading fees either because it can sort of use the Uni tokens for a disbursement, but in the future it could theoretically collect fees.

Tracy: So, I mean, it does seem like you guys are working on quite a lot at this moment in time. We spoke a little bit about the new protocol, but what do you think is next for you in terms of your own company's development, the technology and the wider crypto market? So we, we've just seen this big moment for Coinbase, which we talked about a little bit in the intro, but where do you see Uniswap actually going over the next couple of years?

Hayden: Yeah, I think that something that's really important to think about when you're thinking about DeFi is how early on we’re in DeFi. DeFi is essentially what we're trying to do is modernize financial infrastructure. And so there's all these benefits, right, that are kind of inherent, you might've seen stuff like very centralized exchanges, whether it's crypto exchanges like Coinbase and Binance, or you know, non-crypto exchanges, like Robinhood or brokers like Robinhood or Ameritrade. They all have these sort of downtime and there's these inherent benefits to DeFi but it's also early and so there's also sort of these early downsides. One way to think about it is we're in the early days of the internet. Right? And so transaction fees can be high at times, it can't process that many attraction transactions in the early days. Right? So it's kind of like AOL days, where things are still a little bit slow, but they're getting built out very rapidly.

I think that over the coming years, we should expect DeFi and Ethereum to scale up from being able to do what it can do today, which is 10, 15, 20 transactions per second, to being able to do hundreds of thousands or millions of transactions per second. We've seen a kind of explosion in the usage of DeFi. So Uniswap didn't exist two years ago. Today it’s doing $10 billion in volume, per week. Six months ago it was doing about $2 billion and a year ago it was doing about $50 million. So we've kind of seen this explosion. I think it's attracting a lot of new users. We've seen a kind of similar explosion in users, but it's still in the like maybe 500,000 users range. But I think over the coming years, these sort of inherent benefits of DeFi are going to become more apparent and the downsides are going to be slowly worked out.

Joe: Well, let me ask you a question, Hayden. You talk about users and you talk about the benefits of  DeFi and I have to admit the technology and the sort of rethinking of what a market maker can be is really cool. And the liquidity -- I'm super impressed by it. However, when you think about a traditional stock exchange, one of the nice things about a stock exchange is you can raise money there and then go do something. So for example, you could be a company like GameStop and raise money on the stock market and then go build physical video game stores, which may not be the best business these days, but whatever it is, or you could build something really exciting.

When I look at -- when you refer to like using DeFi it just seems like, okay, you have a bunch of people who are trading, that’s cool, and you have people who are sort of lending or staking and earning yield. That's cool. But like, who's doing the borrowing? In other words, where are you competing with TradFi I guess for like actual people who need liquidity or people who want to raise money. It seems like a bunch of trading and lending, but where is sort of end user who's getting something out of the system that isn't just more trading. Because also when I look at all the top coins, everyone just seems to be, it's like this like Russian doll of, well, this is the thing that … it's just more coins.

Hayden: That's a really good question. And I think that that gets to what I was saying about it being in the early days. And so where DeFi is now won't necessarily be where it is in a couple of years. I think that, in you know, five, 10 years, it won't be called DeFi, it will just be called finance. So part of what we're trying to do is build better financial infrastructure that has inherent benefits to it. That will sort of benefit the entire class of existing financial use cases. There is a lot, right now, it is sort of living, like there is kind of this almost closed ecosystem or closed system of crypto. And within crypto Uniswap already kind of works better than most centralized exchanges, right? And these decentralized lending systems are far more used than other ones, but outside of it, it hasn't quite penetrated. 

In the early days of anything it’s more kind of diehard users and early adopters. And that's definitely the phase where we are. And over time though, basically the, the kind of UX barriers will be reduced and the kind of security will increase and the kind of throughput will increase. More people will be exposed to the very real benefits that underlie it.  Because there are, there's sort of like fundamental benefits and value to DeFi. It's true that those are not all being accessed by a wide enough audience. And what we're expecting over the coming years is for that to change. And we're expecting for more people to get brought in, but that's an ongoing process and, and we're still in the early days of that.

Joe: But can you give me an example of how someone who needs to raise money for something that's not just more tokens, or someone who needs to borrow for something that's not just a token development thing, gets value from financing in the DeFi system, as opposed to the traditional ways.

Hayden: One kind of really easy example is ... we have this nice advantage of living in the U.S. and having, at least mostly trusting our banks and having a lot of access, we still have pretty good access to financial tools, but there's definitely a huge portion of the world that doesn't have as good access. And probably would love to even be able to open a U.S. bank account. A very, probably early adopt use case that I could see growing is first off, having you a U.S. bank account would be great, just being able to store U.S. dollars. But then beyond that, being able to earn a return on them. And so these decentralized money markets that we're seeing, along with these kinds of decentralized stable coins that we're seeing, could give people in third world countries and developing economies, exposure to less risky assets and even yields on those assets, which I think will be a pretty big early use case.

Tracy: Do you worry at all that as DeFi becomes more accepted by traditional finance or more incorporated into traditional finance, that it loses, I guess, some of the ethos or culture that began it? I'm thinking this idea of generating yield and things like that. There are people out there who would see this very much as a money-making opportunity.

Hayden: One thing that's kind of important is, you know, what infrastructure is the global economy run on? And right now it's sort of run on these kind of siloed centralized systems that people are bridging across. And there's these inherent benefits to running everything on this sort of shared globally accessible infrastructure. And, you know, there's still benefits... I'd say that DeFi in the long run, not everyone wants to self custody, their own funds. Right? Plenty of people don't want to think about private key management, all this stuff. Many people want to use a bank, but what are banks using as their rails? Are they using kind of these siloed closed off systems, or are they using this sort of global infrastructure that is kind of inherently accessible and shared.

And I think that's part of the idea here. There's these inherent benefits, right? So one thing to think about is downtime, another one is centralized exchange hacks, right. A lot of the problems that DeFi solves users don't really care about until it affects them personally. But when it does it's like catastrophic, right? And so a lot of people have held funds and not been able to withdraw them or your bank shut you out, or all these different things happen, right. So in the long run if DeFi is successful these types of events should be much more rare and users should have a much better kind of experience using financial tools in a way thatthey might not even think about it in the moment, but they're sort of benefiting from it indirectly.

And so, yeah, I do agree that it's incredibly important as we build DeFi.  To some extent, it's a chance to correct some of the wrongs that were made the first time in the kind of creation of the existing financial system. One really interesting example of that is incentive alignment. There's sort of this principal-agent problem that we see. If you think about even the subprime mortgage crisis, a lot of the people who were kind of creating these risky mortgages, they didn't have personal risk and personal accountability if something went wrong. And our hope in building out DeFi as an ecosystem is essentially the people who are making the decisions, they're directly accountable financially if something goes wrong. And the hope is that that kind of benefits everyone. Yeah, there'll be a lot of people kind of making money and in it for themselves. And, you know, that's just part of how capitalism works, but I think that there's these sort of inherent benefits that can be built into the systems, if the people who are building them care to do so. And that's sort of my hope.

Joe: So on these big, you talk about some of the benefits of a decentralization. So on these, like the big blockchains like Bitcoin and Ethereum, it would be extremely hard for anyone to hack them or take control of them or block transactions, but conversely, they're computationally expensive. And that translates into monetarily expensive fees for if you want to send a Bitcoin transaction are fairly high. If you want to send an Ethereum transaction are fairly high. They're not particularly fast. There are other decentralized exchanges that exist. The second biggest one right now is I think something called PancakeSwap, which runs on the Binance smart chain, which has its own blockchain, and their selling point is from what I understand, way faster, way lower fees, price somewhat less decentralized. But if you're just sort of a random retail trader or you're someone who wants to spin up a project, why go to Uniswap where a trade or a connection could cost like $50 in Ethereum when you can go way cheaper and faster on something like a Serum or PancakeSwap.

Hayden: Something that I said before is people don't care about the benefits of decentralization until it affects them. PancakeSwap is essentially saying, yeah, you get all the benefits of Uniswap except it's run on 12 servers run by Binance. Right. And that sort of gets at this, what is the fundamental value of Uniswap? The fundamental value of Uniswap is that it's decentralized. And there are hundreds of people around the world working really hard to kind of scale up decentralized protocols, but it's a sort of long, and I think that some of them are kind of going to be coming to the market very soon, but it's still like a very kind of intense technical problem. And if you're kind of just saying, Oh, well, you know, we can centralize it and everything will be fine, some people don’t care and some centralized exchanges probably will always have market share.

But in the long run, they still have this same problem. Right. Which is what happens if Binance decides to kind of change everyone’s balances. They kind of change their 12 nodes to running … they can kind of rewrite history, whatever. Right. And so you sort of lose these kinds of fundamental value. And so yeah, you can always scale up to by sacrificing decentralization, but what is really hard is scaling without sacrificing decentralization and the main kind of solutions there, which are in the works and are coming, you know, Ethereum 2.0 with sharding, but also  Ethereum layer two solutions, such as roll-ups, those are kind of the real solutions that are much more decentralized and retain more of the properties that I think are fundamentally valuable about DeFi.

Tracy: So there's one other big thing, which we haven't really mentioned. And I guess it's, again, a sort of tension or risk that's inherent in DeFi. And this is the response of the regulators. Do you worry at all that someone like the Securities and Exchange Commission is going come in and take a hard look at the tokens trading on your platform and decide that you are are in fact dealing unregistered securities?

Hayden: Yeah. I think that one interesting thing here, right, is you can almost look at Bitcoin and you can look at Ethereum as leading examples, where I think it took time for regulators, but I think over time, regulators have slowly become more comfortable with Bitcoin and with Ethereum as these infrastructure layer, decentralized infrastructure layer platforms, and sort of recognize that people can build regulatorily compliant or noncompliant applications on top of them, right. You can build a compliant token on Ethereum, and you can build a noncompliant token on Ethereum and you can build. And Ethereum is sort of this lower level infrastructure underlying it and Uniswap while there are stuff like interfaces to it that exist on top of it and those interfaces kind of feel more like traditional exchanges in terms of what they look like visually, the infrastructure underlying it is still this sort of low-level infrastructure and platform that supports regulated ... that can kind of ... people can spin up a regulated compliant market on top of Uniswap, right?

And we've seen that happen and there's still value in DeFi even with that being the case. And so I think that as, the same way that sort of regulators became comfortable with Ethereum over time, seeing it as this infrastructure layer, I think that all the same things, Uniswap kind of hits the same bars of decentralization that Ethereum itself does, because it runs entirely on Ethereum and is entirely composed with Ethereum smart contracts. Right. And so, the hope is that in much the same way that Ethereum is understood to be infrastructure, it will be recognized that Uniswap is also.

Joe: You say it's infrastructure, but is the argument therefore that in the end there's no securities being traded? It's just infrastructure all the way down or up? I mean, I don't know how many, apparently you have 2,000 coins, are some of them — what are they? Are they securities?

Hayden: It’s tens of thousands? It was about 200 per day. It's not that. The idea is basically that much like the internet, people can do illegal or non-illegal things on the internet. Right. People can do illegal or non-illegal things on Ethereum. People can do illegal or non-illegal things on Bitcoin in that same way. People can do illegal or non-illegal things on Uniswap. And it's a decentralized protocol that no one controls. Right. And so the idea is basically that maybe, and I don't want to speculate where regulation should occur. But it could be the issuers of assets, right. It could be, you can build a token that's compliant. Right. You can, it could be the kind of, it could be interfaces built on top of it. It could be specific ... who knows. Right. But Uniswap is kind of fundamentally decentralized. Someone could come to me, send me a letter that says ‘Hayden, shut off Uniswap.’ And I'd say, I can't do that. Right. I have no ability to stop it.

Joe: You could shut off Uniswap.org., right?

Hayden: I can shut off Uniswap.org, which is a, well Uniswap.org is a specific training interface that exists on top, but there are also something like 30, 40 other trading interfaces that let you access Uniswap liquidity. And so the portion of trading volume that comes through the one interface is actually a relatively small portion — something like 75%, 80% of trading happens on other interfaces or in other smart contracts.

Something we didn't really get into is that Uniswap, because all the liquidity exists on-chain, it can be used very easily within other applications. And so a lot of trading on Uniswap doesn't even happen on any interface, but happens directly on-chain in other applications. And so, yeah, maybe Uniswap.org could be shut down, or it could be kind of run in a way where it kind of restricts assets, specific assets. In fact, it already kind of does limit what you see in a way that is kind of designed to protect users, but there are alternate interfaces and the code is all open-source. And so people have even forked the interface and spun up their own versions that do other things. And, you know, we don't really have any control over that.

Joe: Hayden, is there anything else that you want to make sure our listeners understand before we go?

Hayden: One really interesting aspect is what I would call almost like permissionless innovation and sort of the ease of building applications. Traditionally, right, if you want it to build, let's say you wanted to build, I dunno like a margin trading platform. The kind of amount of things that you would need to build to do that, you would need to build an exchange. You would need to build a lending system or integrate with one. There's sort of all these different things that you would need to build. But in crypto there's sort of just like, what we call interoperability between applications. There's, you might've heard the phrase money-like, and essentially what that means is that in a single day, someone could basically spin up a new application that -- and they don't need, they could source liquidity from Uniswap. They don't need that application. It might need exchange as a component, but they don't need to build their own, they can just access liquidity from Uniswap. 

And, you know, there may be that you spin up and they use compound for lending, right. And you can kind of build a new system that lends an asset on compound and trades it for another asset and you can swap, and then you can kind of create these sort of like, you can kind of compose, applications in this sort of really powerful way where things that take years to build outside of DeFi can be built in days. And one of the reasons that I'm we've seen so many different things and it can be almost hard to even keep up with what's happening, is how much is getting built and how quick people are building it. And there's kind of a lot of noise, but beneath that noise is like a lot of value being built. And I think that the really valuable applications will become more apparent over time. I guess that's the one other thing.

Joe: Awesome. Well, I really appreciate it. I learned a lot from this and I know there's a lot more to DeFi. We didn't really get into like compound and lending and yield farming and all of that, but this was great.

Hayden: Yup. Appreciate you having me on.

Joe: Thanks a lot, Hayden.

Tracy: Thanks so much Hayden. That was great.

Joe: So Tracy, I found that conversation to be really fascinating. I find Uniswap to be fascinating. I find the whole like rethinking of market structure and this automated market maker model to be super fascinating. I have to say though, the thing I'm like still hung up on is, like, it's great to trade and it's great to have liquidity, but I'm still hung up on the, like, yes, but liquidity for what? On a stock exchange, at least in theory, you go raise money and you do something with it. With DeFi it just seems like liquidity for the sake of liquidity or trading for the sake of trading.

Tracy: Yeah. I agree. That's still, I guess, a big hurdle to the market really getting accepted or incorporated into traditional finance. The other thing I was thinking about was this idea of Uniswap, basically having to completely — or completely is unfair -- but, you know, reinvent itself every once in a while and evolve and respond to new developments in the market and things like the Sushiswap vampire attack. And I think in technology, that kind of experimentation or adaptation, is the usual thing. Like people change their products and they kind of experiment with what they're doing. But I think the difference with crypto is that you're always doing that with money. And that seems to make it so much more difficult, right? You're experimenting with people's money and for a lot of people at the moment, crypto is still this fun thing, it's just something they kind of do on the side. But it is more sensitive, and I think that makes it more difficult over the long run.

Joe: Yeah. And it's a lot of money. Like all the coins added together,  I don't know, as of right now it's somewhere around $2 trillion. And so a) it's a lot of money. And two, you have a lot of people who've become insanely wealthy by buying some coin, and then it goes up a hundred X over the next year, which is great. But it's a little weird when it's that much money. And as Hayden basically seemed to admit, there is no one like at the other end ... it hasn't actually found its end market for what DeFi serves. And so it's like, okay, maybe it's going to be something in emerging markets and people have access to something or other, but it's kind of weird. Like, a) how much money there is and b) how much money people have already made, when there's still this like ambiguity of, okay, well, what's it going to be used for?

Joe: Yeah. I mean, DeFi contains multitudes, I think, and means different things to different people. For sure. It's kind of a shame that we didn't touch more on the yield farming aspect of it, on that note…

Joe: Let's do a yield farming one too, because I know that's like a big thing and you like put your money in here and then you get a coin and then you put the coin in another thing, and collect more yield. I know a yield farmer. Maybewe’ll have a yield firmer on the show.

Tracy: Yeah. Let's do it. Let's do like opposing ends the DeFi movement. Let's do that.

Joe: Yeah. Like there's something for super interesting there. And like the whole, it is very inventive. So I'm super interested. It's fun. I went on Uniswap. It's fun to look at how these things work and the ease, you know, you don't have to set up an account. You just start like dragging down things. You could theoretically trade, but yeah, it's fascinating to see where it's going.

Tracy: Should we leave it there?

Joe: Yeah, let's leave it there.

You can follow Hayden Adams on Twitter at @haydenzadams.

©2021 Bloomberg L.P.