GM everybody, I'm excited to be here. I'm going to talk to you about borrowing across blockchains, also the DeFi capital markets on Ethereum and across blockchains. My hope is at the end of this presentation you are 0.1% smarter about how to think about the capital markets, how to trade, and how to look at the future of where DeFi is going, not just on Ethereum but hopefully on other blockchains and how it all balances out.
So for quick background, I'm Robert Leshner. I'm the founder and CEO of Compound Labs. We launched the Compound protocol in 2018. This is before anyone was using the phrase DeFi. At the time we just called it smart contracts doing finance internally. And over the last couple years it's really become DeFi and it's been exciting to watch this really nascent industry mature extremely quickly with everybody here.
So quick agenda: I'm going to show you what DeFi is right now, then I'm going to go into sort of the academics of where interest rates actually come from. Then we're going to start looking at what happens when you add more blockchains and interest rates have to unfold across a much broader ecosystem.
So to recap, I just want to show you what Compound is because if you understand this you'll probably understand a lot of what DeFi is and how it works. So Compound is an interest rate marketplace that allows you to borrow crypto using other crypto assets as collateral. And the protocol in the aggregate is basically this giant risk management engine that calculates how much you can borrow at any given moment. And if you're borrowing too much it liquidates you, and if you're not borrowing too much it allows you to borrow more. This is very simple.
Over time Compound has scaled from a couple million dollars of assets to up to $20 billion of assets. There's been about $300 billion of transactions, which for a piece of computer code running on a blockchain, it's pretty cool.
So at the heart of this is actually something extremely simple. It's basically just a single math formula that says liquidity, which is how much can you borrow, is what collateral assets are you using, how valuable are they, we fetch this price from Chainlink, and what is the collateral factor. What this means is how useful is the collateral to borrow against.
So in the traditional financial world, houses are actually incredible collateral because they don't really go up and down in price very often. So everyone uses houses as collateral. In crypto, you have large liquid assets like Ether and Bitcoin. They're actually amazing collateral because you can trade billions of dollars at any given moment. They're super good collateral. Other assets are less good collateral if they're small, if they're volatile, if they're newer.
So to look at what happens when you bring this together, a single asset on Compound, let's just take USD Coin as an example because USD Coin is a stablecoin, it's super popular right now. There's $945 million of USD Coin earning 1.31% on Compound and there's 26,800 total users earning interest on USD Coin.
So this is what happens when you bring together all of this. You have an asset that has an interest rate because people are borrowing it on the other side. And in total this is basically a billion dollar asset in one smart contract running on one blockchain.
So where does that 1.31% figure come from? Like why would anyone earn money by providing USD Coin to a smart contract? If we figure this out we'll figure out a lot of other stuff.
Fundamentally, it comes from borrowers. And there's really two things that go into how an interest rate is created: what the cost of money is, and this is just the supply and demand for money. Now this comes down to the fact that people have opportunities to deploy capital, whether it's an investment strategy, whether it's buying treasury bonds in the US, whether it's staking Ether. There's always an opportunity for money and you're always willing to pay pretty much up to the cost of that investment opportunity to borrow funds if you have a good strategy.
Then you have other liquidity providers who are able to basically finance and lend money to those that have a better use case for it, that are willing to take risk, that are willing to deploy capital. And this interest rate really just comes from matching off those who want more money to do cool things with it and those who have it and don't want to necessarily do cool things with it.
Now the people who have it aren't necessarily individuals. It can also be smart contracts, it can be any source of idle funds, it can be an entirely programmable composable ecosystem. The two best examples of why people want to borrow right now are: one, we'll use USD Coin, there's a bajillion investment opportunities right now. The sort of spot rate in the US is closer to 3% and Ether has a four plus percent staking yield. So a lot of people want to borrow Ether up to the point that they can deploy it in staking in the proof of stake system.
So here's a number of conditions that actually change interest rates and if we think how these work we'll start to understand this in a little bit more granular detail. So the first is utilization rate. As liquidity becomes scarce it becomes more expensive. So if less people have funds to deploy the cost of money is higher. This is like the biggest driver behind interest rates.
The second is actually collateral factors in a protocol like Compound. So the more useful collateral is, the more you're able to borrow against it, the more investment use case there is, and the more borrowing demand leads to higher interest rates. So the more assets we can bring into DeFi that are high quality collateral, the more aggregate borrowing capacity there is.
So eventually as people start talking about real world assets and tokenized assets, this will most likely increase the utility of them as collateral in DeFi and increase interest rates. And that's also the choice of collateral. The more assets the better, even like longtail assets, fringe assets, the more stuff that's out there, the higher base interest rates are.
Next comes down to really security. The more secure something is the lower interest rates have to be. If something is incredibly secure from a technical perspective, from a confidence that you're not going to lose your money by using a system, the lower interest rates can be. And in conjunction with this, the more reserves there are or insurance there is in the system, the lower interest rates are.
So what happens when you go from having a single market like Compound on Ethereum to having multiple blockchains or multiple markets and how do they play together? So let's say, hold on, I don't know if this worked. So let's say you want to just copy paste a market like Compound. Right now all DeFi applications are just open source computer code. It's trivially easy to generally copy paste and deploy something anywhere else. So what do we expect is going to happen? Do we think it's going to be the same Compound? If you take a Compound on Ethereum and a Compound on Avalanche, do we think it's going to be different? Let's find out.
So here's a theory: no two deployments can be the same. Even if it's the exact same code running on multiple chains, the markets that exist are going to be fundamentally different. Now why? The assets that you use in this second deployment are guaranteed different. Ether on Ethereum, which is a native asset, is different than Ether on OP or Ether on Arbitrum or Ether on Avalanche that's bridged or wrapped. It has different security characteristics, it has different functional differences to it. Ether is not the same even if it's on two different deployments, just like pretty much any asset that gets wrapped.
And the opportunities are different. So Uniswap on Ethereum is different than Uniswap on Optimism. The liquidity is different, the assets that are there are different, what you can do on the different blockchains is fundamentally different. And when we go back to where interest rates come from, it comes from the opportunities to deploy capital. And so even a market deployed on somewhere else, there's a different opportunity cost.
And then there's this really annoying thing about different blockchains that has to do with time. And interest rates are really the time value of money. When you start having multiple deployments of multiple different DeFi applications, the time it takes to bridge assets or bring them back through a challenge period, there's actually a time component that doesn't exist when you're just using an L1 like Ethereum directly.
And lastly, security is different. This goes back to the risk inherent of a system. The security assumptions of different deployments can't be identical. There's bridge risk, there's governance risk, and there's oracle risk that are all going to be a little bit different even with a similar deployment.
So I'll call this Leshner's Law: the interest rates for the exact same asset on two different blockchains can't be identical. They just can't be. We can do things to bring them similarly, so to try to make USDC on Ethereum in Compound and USDC on Avalanche in Compound be equivalent, but they're never going to be the same because they're fundamentally different deployments and different systems.
So let's take one simple idea and this is one of distance. So back in the day, you used to have people doing business in Europe and bringing their business to America. There's this concept that as distance increases, interest rates also increase. We can think about this in the context of blockchains and deployments as well. So the farther an L2 is from the native market, the higher the interest rate differential will be. If it takes one second to go back and forth, the interest rates might be more similar than if it takes a week to go back and forth or if it takes a month to go back and forth between these systems. Just a really simple curve to keep in mind.
So let's say we have multiple different blockchains. Capital is going to want to move between them, ideally to find the highest yield. Take USD Coin as an example. If we have it on multiple blockchains, capital is going to want to move between them. But this is not as easy as it sounds.
So if you just take the assumption that there's transaction fees or withdrawal fees to skip a challenge period of withdrawing from an L2 or a foreign blockchain back to the native blockchain, there's transaction costs to move back and forth between them. This is a simple piece of math, but if you took a 10 basis point fee and did it once a week, it would cost you 5.1% over a year. I would expect all things equal, you would see an interest rate on a foreign deployment of a system have to be higher than the native market.
Let's go back to that really cool slide before, because of these transaction costs and the time to go back and forth and there's a time value of money. And if you want to skip that period you're going to have to pay a cost. And so interest rates on foreign deployments must naturally be generally higher than they are locally on the first market like Ethereum.
So how do we get these different markets to actually look the same? Well going back to the beginning, what we can do is we can try to have as similar assets as possible. So instead of having two very different versions of Ether, we could try to have the same version of Ether. Instead of having two different versions of USD Coin, USD Coin and wrapped USD Coin or bridged USD Coin, Circle just announced yesterday they're actually coming up with a system to natively migrate USD Coin onto every single blockchain instead of having wrapped USD Coin. Things like this are really cool.
You can have the same price feed, Chainlink, on every single blockchain and you can try to get the parameters of all of these markets as simple as possible through a little bit of engineering. So one of the things that Compound Labs, we've recently built and released was Compound 3. Compound is actually a system designed so that the markets operating on different blockchains are as similar as possible so that you can use as few assets as necessary to build a borrowing collateralized marketplace.
And so it can be deployed with the minimum differences of governance or computation on different blockchains and to create this extreme similarity between different deployments, the protocol only allows you to borrow one asset and to create an interest rate around a single asset. So in the case of Compound 3 on Ethereum and Compound 3 that will soon be deployed on other blockchains, it's USD Coin. There's not interest rates on 18 different assets like there are on the Compound that you use today. It's not like Compound or Aave or any of these systems where there's a ton of assets that all can be borrowed and all have interest rates. It actually isolates the entire protocol down to just creating a single interest rate for a single asset. The only asset is for now USD Coin.
And what this does is it allows there to be a much more similar base interest rate across every different deployment of the protocol. Now this enables some really cool stuff. So when you have multiple different markets on multiple different blockchains all that are behaving similarly, where the cost of capital is the same, where you have a market where USD Coin yields the same thing on Ethereum as it does on Arbitrum, as Optimism, as Avalanche, as somewhere else, you can start to move money between the chains incredibly easily.
And you can start to do really cool things. So a great example is if you don't have a big differential between all of these different systems, you can have a simple load balancer or cross-chain mechanism to move positions between these different blockchains easily. If the positions on Ethereum and Avalanche are more fungible, the cost of moving them is lower and you can almost allow the user to share a position across blockchains in really cool ways.
The more fungible it becomes, the more unique interest rates, the more unique the assets, the harder it is for there to be this fungibility. And what this leads to is a really interesting end state for applications that run across different blockchains. The end state is one in which you can begin to abstract away the blockchain completely for the user.
So if I'm a user of Compound in a year or at some point in the future, I might not even know which blockchain my assets are going to. I might not care. I might say I want to earn an interest rate on USD Coin and that USD Coin travels to whichever blockchain it needs to, wherever the highest yield is incrementally by a couple basis points to load balance the system.
There's really two ways to do this. One is to have a smart contract that actually routes the asset across blockchains intelligently. A great example of how to do this is to have an interest rate oracle, to use something like Chainlink to say what right now are the interest rates across different deployments, where's the highest, where's the lowest, and route the asset accordingly.
And the second is a sort of synthetic L2 system, like an exchange or a business built on top of these different deployments that can make that decision off-chain and route funds accordingly. But for the user, the end state is actually really cool. You're going to be able to use an application and under the hood it will take advantage of deployments on multiple different blockchains without you knowing, without you having to configure the network toggle in MetaMask and say oh I want to switch to Arbitrum now and then press a button or I want to switch back to Optimism and then press a button.
You'll just be able to do a transaction on one blockchain and the capital markets across many different blockchains will all be in sync in real time. And for users this has an incredible UX advantage. Eventually we're going to be able to stop thinking about what blockchain am I using and why. You're just going to be able to think about what are the economics of using this system, just like we don't care about what technology or programming language any of the apps in our phone are written in. We just care that they work and they're fast and they're efficient.
And so the future for users is going to be incredible as we can begin to abstract away the underlying networks by making the economics across the networks completely uniform. If you have any questions, find me afterwards. I always love to talk about the academic side of this stuff. Try Compound 3, it's pretty sick. And thank you, Chainlink, and thank you everybody.
