DeFi and Self-Driving Banks

Jeremy Allaire: I'm Jeremy Allaire. Welcome to The Money Movement, a show where we explore the issues and ideas in this brave new world of digital currency and blockchains. 2020 was an enormous year in crypto and there was obviously incredible growth across the board and in crypto assets on the whole. I think a defining theme that drove so much of both the innovation and economic opportunity in this space was driven by DeFi. In particular, what sparked a lot of this was the rapid growth in compound protocol and the launch of the COMP governance token which sparked enormous growth in the entire category of decentralized protocols for financial market infrastructure.

I think a year ago, I took a look and there was about $800 million in value locked in DeFi smart contracts, and today, that's mushroomed to around $23 billion give or take. This is obviously just the start of a megatrend that will eventually append, I think multi-trillion-dollar capital markets. I previously had Robert Leshner on the show and we talked about DeFi. We talked about the concepts of programmable money and what people can do with stablecoins and smart contracts. Recently, there was a really cool editorial from a former OCC interim head, Brian Brooks, who talked about this idea of self-driving banks.

I really liked it, so part of the title this week is DeFi and self-driving banks, but this concept here is like self-driving cars where automotive machines are governed and operated entirely in software, and how that poses new challenges for how people think about risk, how people think about rules, how people think about a lot of things. DeFi in many ways really is this similar thing but in what banks do, self-driving banks. Fundamental building blocks of finance are governed and operated entirely by machines and software on the internet which is super exciting. Joining us this week is repeat guest Compound founder and CEO, Robert Leshner.

We're going to explore this megatrend, what 2021 will bring in DeFi. Then I think a long-term view, I want to talk a little bit about the transformation of capital markets that's on our doorstep as we explore this. Welcome back, Robert.

Robert Leshner: Thanks, Jeremy. Great to be here.

Jeremy: Awesome. Maybe we can just start and I'm sure you're sick of this but just talk about the last year for Compound. It's just been incredible but maybe just give it a couple of minutes on what's happened. It's staggering and super impressive so congrats as well.

Robert: Thank you. Compounds is a protocol for users starting interest rate on crypto assets and borrow crypto assets against them. Most popularly, users are using crypto collateral to borrow stablecoins. The system has been running for a couple of years, but in 2020, what we did is we transitioned Compound from being a series of smart contracts that are overseen by humans and centralized humans, our company being the ones responsible to oversee the protocol to transition the protocol to decentralized community governance.

Allowing the entire crypto community at large, the opportunity to upgrade the computer code that runs Compound, and to be able to collaboratively set the different parameters of the markets. Compound is now running as a decentralized protocol on a blockchain. It's a financial market that doesn't need any one person in order to run it. It can run independent of the original team and it can ideally run for the next century. It's autonomous and these markets will hopefully operate forever.

Jeremy: Unlike self-driving cars that maybe need to go to a charging station, as long as the internet is running and I guess people are validating blocks on these computer engines, it will just run as a self-driving bank.

Robert: Yes. The best way to explain it is this started off as a mostly autonomous vehicle where there was a driver behind the wheel just in case something went wrong. That was Compound Labs, the team that built it. Nothing ever really went wrong. It didn't require too much touching of the steering wheel and we've replaced the human driver with internet hive mind.

Jeremy: There's that people in Phoenix yet?

Robert: Not yet. [chuckles] Now, there's essentially a self-driving vehicle when the steering wheel needs to be touched, there's an internet collective. With this structure, it is essentially autonomous. All of the logic of how the markets work and how balances function, it's all computed using these computer programs.

Jeremy: There's a lot going on there. I think to the uninitiated, it's like, "What? What are you talking about? What do you mean?" There's a lot I want to explore here and maybe first actually is the functions that a bank might provide traditional intermediation in borrowing and lending as an example. Just talk through a little bit of where Compound is today. What are all of the different dimensions of intermediation that you think can be self-driving, that can be entirely machine operated and governed in this space?

Obviously, we see a lot of building blocks composability et cetera, et cetera, that's out there today. When you think about the primitives that you're working on and compare that against interest rate markets or other things, how do you see that?

Robert: It's a great question. Fundamentally, Compound allows two different counterparties to participate in a financial market. One side is earning an interest rate and the other is using collateral to borrow assets and paying an interest rate. There's so many functions that would exist in a traditional bank just to create this economic function. You would have people building systems to accept funds from one group of users. This is like a liability side of where you're building--

Jeremy: Payment systems that go into that, and so on, you just ride on stablecoins and crypto-assets.

Robert: Exactly. There'd be so many pieces of work in order just to accept funds from users. With Compound, it's purely crypto there. All of the rules for the transfer and settlement of cryptographic assets are defined on a blockchain. That part is completely abstracted away. There's no humans necessary. Users administer their own transfer of funds and the settlement. The record-keeping for that is also autonomous and stripped away. The entire history of transactions is transparent and accessible on the blockchain. You don't need all sorts of bookkeepers to reconcile it, always reconciled basically in real-time as users submit transactions.

Anyone can audit or inspect the state of the records and can see how we got to the current books. That's completely abstracted away. Then you have the calculation of balances and if users are earning an interest rate, traditionally, you would also have entire systems of record. You would have entire banking databases and software packages for that economic behavior. With a system like Compound, it's all encapsulated just in very simple code running on a blockchain and there's no humans necessary. When it comes to the other side which has people borrowing assets, this is where you have huge quantities of operations, systems, and people involved in a traditional organization.

You would have the evaluation of collateral. You would have the communication with customers. You would have the transfer of collateral and the transfer of borrowed funds. You would have non-stop reconciliation. You would have massive systems, let alone when market conditions change and the value of the collateral changes. In a system running on a blockchain like Compound, all of that is replaced with smart contracts where it basically strips away all the steps that would fail if the humans mess up or if the software messes up, or if people don't show up to work one day. All these systems of analyzing collateral, monitoring the market risk of them, enforcing liquidity requirements and risk requirements, all of that, that would take teams of humans is replaced with a computer program.

We call them smart contracts but a computer program running on a blockchain that anyone can inspect, everyone can see the rules of the program, how it operates, that its operating, that it's working correctly, and can see that instead of needing scores of people and databases and records and spreadsheets, it can all be done with a computer program.

Jeremy: It's obviously super exciting. If you look at traditional banks there's some fundamental differences, you've outlined a lot of those that have to do with human intermediation and how the counterparties themselves, the experience that they're going to have. I think one of the things that comes to mind is, affectively risk management, is such a huge part of what a bank does, and bank examiners, if you go get the OCC bank examiner's manual, and you look at what that's about. An enormous amount of it is capital, liquidity, risk management, underwriting, all this stuff, these incredible amounts of things there.

In this world one, obviously, major difference is, this is all management software. If I were a bank examiner in the, "Self-driving bank world," I'm going to basically be examining the smart contracts and examining how effective they are and they're incontrovertible by definition. It seems like another major difference here is, Compound can't create money. It does not have the ability to create money. Banks do create money, they say, "I have this collateral, I have these assets, and I'm going to create commercial deposits with it. I'm going to basically create money." Crypto, obviously, is a full reserve system.

Do you think that, that particular dimension, credit creation, how do you think about that, relative to the way that these nascent capital markets exist in crypto? Is there an equivalency there somewhere that can take place? What does that look like?

Robert: It's a great question. Traditionally banks in effect increase the money supply by lending out money, which then becomes a deposit in another bank or their own bank. There are similar principles, even with crypto assets where there's a finite supply. Take Bitcoin as an example, everyone knows that there's going to be 21 million Bitcoin, but Bitcoin also can be moved on to other blockchains and tokenized in new ways. A great example of this is an asset called Wrapped Bitcoin. The custodian BitGo holds a bunch of Bitcoin, I think it's like 110,000 Bitcoins, it's a lot at this price.

They issue a token on the Ethereum blockchain called Wrapped Bitcoin so that users on Ethereum can send Bitcoin, interact with DeFi protocols like Compound, add programmable logic to Bitcoin on Ethereum. In effect, you can see the ledger of Bitcoin, that there's only ever going to be 21 million and you can see the location of all of those Bitcoin, but 110,000 of them are held in BitGo and they've issued a token on Ethereum against it. In effect, you could say it increased the total quantity of Bitcoin with this transformation. Although, the original 110,000-

Jeremy: Is it now, it's always there.

Robert: Yes, it's always there and they're not moving. If it were to move everyone would scream, "Bloody murder."

Jeremy: Hair on fire.

Robert: They basically, in effect said there's still the same 21 million Bitcoin, but they're usable on Ethereum. On Ethereum, in effect, through a system like Compound, the number of people who believe that they hold Bitcoin can also increase. A user might borrow Wrapped Bitcoin from a system like Compound and then the user has Wrapped Bitcoin and somebody else has supplied it and believes own the claims to that Wrapped Bitcoin. Through decentralized finance, you can see how an asset can spread to an ecosystem or is in effect a little bit of this multiplicative effect on assets, but that's through finance in general.

If I were to lend Jeremy my car, I would say I own one car and Jeremy would have a car even if later he's going to return my car to me after the weekend.

Jeremy: Obviously, there's a lot of dimensions to this and I'm really curious about how you think about, and I'll come back later a little bit to tokenization of assets more broadly and how that flows to DeFi. We're already starting to see that, and we'll see a lot more of it, but very specifically, Compound protocol today is secured lines of credit, [chuckles] essentially. It's entirely secured, it's over collateralized and so it's safer. We have a lot more comfort in terms of the risk management that goes along with that and so I think the value at risk models are safer, just, but do you see a world of unsecured lending?

In the crypto, in DeFi specifically, it's conceptually challenging, but essentially you have people showing up and basically say, "I'll give you some USDC," or, "I'll give you some Wrapped BTC and you're not giving me anything." I guess the overlay question on that is, and I want to come back to this from a regulatory perspective, too but identity, reputation? How do you think about identity and reputation emerging and interacting with DeFi protocols? That really seems like the only way it gets us to unsecured lending.

Robert: That's a great question. There's a lot there. I'll start by going back to the concept of autonomous banking. The reason why a system like Compound works, is because its replacing, what otherwise is a relatively simple banking function with code. It's using fungible liquid assets as collateral, to borrow against. Fungible assets are ones in which they all look and feel the same. There's no nuance between one Wrapped Bitcoin and another Wrapped Bitcoin or one Ether and another Ether. They'd be the exact same. Unlike two houses where each house is fundamentally different, or two borrowers that are fundamentally different.

Compound relies on fungible collateral that's liquid. It's very easy to build a system that knows how to value that collateral and allow for the liquidation of that collateral. It's very easy to build a system like Compound that requires excess collateral and sure, liquid fungible collateral to borrow against. It'll build a computer program that can administer this safely. It's why the very first DeFi applications are probably the simplest and the safest and probably perceived as the most conservative. What Compound creates is not very dangerous and it's easy to administer.

Eventually, teams will figure out how to replace human judgment with code running on a blockchain, but that's a significantly more challenging problem to replicate and replace. It makes sense why the very first protocols like Compound are just taking the easiest lowest hanging fruit.

Jeremy: We're seeing that, obviously, today, like Max Levchin and Affirm just IPO'd today or it was yesterday, I don't know what it was, it's doubled in price. It's an incredibly valuable company, but that's basically machine-governed. It's AI. Obviously, deploying and computing AI on a blockchain is hard, but you can imagine through Oracle's and other means, ways to have highly-automated, machine-governed, risk-decisioning right, but that still sit inside of an underwriting smart contract that is interacting with some form of verifiable, attested identity that's interacting with a smart contract.

Robert: Exactly. As you start to look at how do systems manage to take on credit risk, Compound starts with the lowest credit. You're using a lot of liquid collateral at the maximum, your risk is extremely low but as you start to move into, hey, this is an extension of credit based on someone's reputation, or other data points or other parameters, it starts to increase risk. Really intelligent systems like Affirm or other can manage those risks, can be extremely good at it. As you start to extend the amount of risk, that's where I think you need to start to extend oversight and regulation of systems, because it's very possible that you have a bad parameter and you extend money to a huge swath of people that can never repay.

Jeremy: Yes, totally. I guess a follow-on question to that is, how do you think about and it ties into regulation, again, it's a theme I want to come back to. How do you think about decentralized identity models interacting with smart contract, protocols like Compound? Do you see that emerging? What do you think that looks like?

Robert: I think it's going to be slow to evolve because I think it's one of the harder problems in the space. I do think eventually, the market structure is going to evolve in general. It's going to be in which there's large products and institutions that are built on top of DeFi protocols and offer and are the on and off ramps to them. Those types of institutions are phenomenal at standardizing and aggregating information in a good way and a bad way about their users. I could actually identify [crosstalk]--

Jeremy: Well, they're actually required by law to do so as well. Some of it.

Robert: I think as DeFi evolves, the structure that will be in place is, you'll have DeFi protocols, which themselves are relatively data. It's hard to get lots of data into them. It's hard for them to do complex decision-making. They're best equipped for very simple things. Things like we'll start to have a whole market structure built on top of them. Exchanges that integrate these protocols, banks that integrate these protocols, institutions, and wallets and prime brokers and hedge funds and all of these things on top of it, that over time will be able to provide identity and either do it on behalf of their customers, where they're willing to take on risks of their users or provide information to the rest of the network. I think it's not going to take place when we're leaving this DeFi structure where users interact with contracts directly. It's not ideal.

Jeremy: It’s one of the places where the rubber meets the road in that, obviously, we saw the news today of course, that the FinCEN has extended the notice, and comment period for some of their rules around institutions and their users interacting with self-hosted wallets. One of the critical issues in that, which I commented on publicly, lots of other people did as well, which is that, as stated it would make it impossible for a bank or a financial institution through its own software to enable its users to interact with DeFi protocols, because as I use as an example, what's the name and physical address of Compound protocol?

There's no name and physical address. There’s a contract on the Ethereum blockchain, is that the address? Put that in my CTR. In any case, clearly, this is becoming an area that regulators are going to care about, and obviously the OCC guidance that came out last week and giving banks permission to use stablecoins as a payment infrastructure, interact with public blockchains. We’re just a step away from regulated financial intermediaries embedding DeFi into their infrastructure and enabling entities, i.e. individuals, or firms to interact with that.

It does seem like between the way that regulators are trying to think about, how do you deal with record keeping and identity? Then similarly, how does mainstream finance plug into this infrastructure? We're triangulating on this and it seems like those issues are going to come to the forward faster than we realize maybe.

Robert: They will. I personally think that there's a lot of properties of DeFi that make it extremely well suited for institutions and for regulators. There's a lot of virtues. The first is that it's fair. The second is that it's transparent. The third is that there's less room for human error. These are all virtues that I think both the institutions and the regulators want and see. It’s better primitives than what currently exists. Right now, most financial institutions keep their own records. When you're keeping your own records that's where you can get into trouble of sending money to the wrong place or the wrong person or allowing someone to use your institution nefariously.

Every institution maintaining its own ledger and then interacting with other institutions, it's a structure that's right for disaster and intentional or inadvertent money laundering or abuse. Which is why it has to be regulated so carefully versus an open ledger, like a blockchain to conduct transactions and store value. It's a huge blast of sunshine, and you won't necessarily have to regulate it in the exact same way. You won't need to regulate it in the exact same way, because it's almost going to be 1,000 times harder to use a DeFi system to launder money to use it nefariously. Just because you have a shared ledger where everybody sees all of the transactions and the state of the world. It’s very hard to lie about.

Jeremy: Like I said too, an institution facing another institution through an intermediary. What is the counterparty risk? It's just the fundamental, inherent problem in the way that much of the financial system is organized. It’s just fundamentally about that. Increasing transparency, reducing counterparty risk. What institution doesn't want that? Who doesn't want to face a market where you're just systematically eliminating counterparty risk?

Robert: In the existing market structure, every single institution has its own books and records. Of course, they have to be inspected and audited, and analyzed because something can go wrong at any institution. It does go wrong at many institutions constantly versus using financial infrastructure, but there's one set of books and records that everybody can see transparently and clearly, it's in order of magnitude better.

Jeremy: Obviously, zero-knowledge proofs and privacy-preserving technology will make it possible for individuals and corporations to be able to transact and not have their books and records be visible to everyone, but still actually have it be provable on-chain and still have an auditor or whoever it is be able to look at that in an automated way.

Robert: Exactly.

Jeremy: The other thing just that came to mind is there's no such thing as a drunk self-driving car. That’s like, "Who's at the wheel?" If you have self-driving banks that have this automation and transparency, you don't have to worry about the human operator, which is maybe a good segue into this, I think just enormous transformative shift that you made to Comp, the governance token, to really turning over quite literally the keys to what is the administration and operation, and upgrading and ultimately governance of risk management, just parameters. It’s just really tremendous.

One of the themes that I've been exploring, we did last week with biology as well was these new corporate forms that are coming into existence. I really look at the work that you guys have been pioneering. There's many others too, but you guys have been pioneering which is this is a new corporate form that exists entirely on the internet entirely in software. Just talk a little bit about the inspiration for that and then how it's operating and maybe I'll have some follow-ups.

Robert: It’s a great question and thank you. We chose to transition Compound from a system in which our team was able to upgrade the contracts as needed to one in which anyone with the buy-in of the broader community can upgrade the contracts as needed. This makes it so that one, our team can't mess it up accidentally or deliberately. I no longer worry about getting kidnapped because I have private keys. Two, by increasing the amount of people that are able to upgrade the system and maintain the system or the parameters, it leads to more and better ideas and a larger throughput of intelligence and work able to maintain this system.

Instead of one person with their hands on the wheel, now there's a lot of people with their hands on the wheel. This was a complex transformation where one of the first projects to really go from one party has decision-making to everybody has decision-making. I think it's the arc that many projects in the space will have where when it's small and when it's not widely used and widely understood, there should probably be only one team working on something. As it starts to become widely used and you start to have a variety of businesses built on top of it, that depend on it and know how it works it really doesn't function well to have one party still in control when there's many stakeholders.

You need a higher standard of reliability and permanence that by removing a single stakeholder from control, I think you can give to the community. At this point, everyone who uses Compounds can have confidence knowing that if they need or want to make a change to it, there's a process. The community can make a change to the protocol. It still requires the buy-in and the support of the broad ecosystem. It's much more of federated decision making. I think that's a really powerful model because at this point, we can step away and the protocol will continue to function in perpetuity.

Jeremy: I think, there's so many interesting analogies like credit unions or community home banks, and I'm coming back to this self-driving bank concept for a minute, which is, community ownership and governance is so powerful, and it's truly global. It doesn't require some really hairy, multinational corporate thing with all international jurisprudence. It's just software and participants that can participate in that tokenized corporate form. It's a pretty radical shift, though, to go from, there's a board of directors and a single corporation that makes decisions to the community as a whole. This is one of the great laboratories of experimentation. What has it been like to be community governed for whatever it's now been? I think it's over six months.

Robert: Yes, it's been quite an exciting six months. One of the things that we did before the six months, is we actually tested the mechanisms publicly with a limited stakeholder group. Instead of there being one party, there was 20. Now there's 20,000 but over the last six months, we've started to see some really exciting things. I mean, community to developers, just folks who love Compound, saying, "I have an idea to make it slightly more efficient," or to "Lower the class costs of interacting with it." Or "To add a new feature, which will increase the safety of it." I think that it's opened up ideas that weren't possible, just putting one team was overseeing a protocol.

I think that's a real virtual used case. The downside is that there's not one party in control. There's not one person leading it. It's been awkward that everyone always constantly looks back to me and says, well, "Robert, what do you think we should do? I say, "I'm number six, on the list of stakeholders, less of a voice than most was in the community at this point."

Jeremy: Open source has a wide variety of modes of governance. Some formal, some informal standard setting, same thing. You're in that pool of experimentation it feels like.

Robert: We are and most organizational structure over last couple 100 years has evolved up to a pyramid structure where there's one key decision maker that drives organizations and systems fully. In crypto, we're really saying, well, it's not a corporation, it's a financial market that's going to run forever, that probably doesn't need too many modifications to it. What's the best way to continue to drive it? We really stepped back from the single decision maker approach. It's too early to say if it works, or it doesn't work. Our original team is not just one developer amongst many, and we're actually working on a separate project, which can hopefully connect back to Compound but it's really been interesting to just see how folks are stepping up to become involved.

Jeremy: Yes, it's amazing. The other piece to these new corporate forms, and obviously Comp is related to this, which is economic incentives. If I think of traditional equity, it's got votes, and it's got dividends, or whatever other features that may have various coupons, or whatever it is. These decentralized, community governed protocols that it actually interacts with the underlying economics of the protocol itself, which is really a breakthrough to be able to effectively enable that economic participation. How has that been going? How's it been going for Comp holders themselves and this deployment of the fee structure into that? It'd be great to hear you talk a little bit about some of those things.

Robert: Yes, to start off with one of the biggest experiments that we conducted and now emerged was a trend within DeFi and crypto but it really wasn't all of six months ago is we've been distributing ownership and governance of the Compound protocol to the users. It would be if Nike gave a share of Nike stock to everyone who bought a pair of shoes and everybody was able to set the future direction of Nike. The reason why that isn't technologically possible is Nike doesn't have a way of officially transferring ownership to its customer base. With a DeFi protocol like Compound that's already built on a blockchain, by definition, all of the users of it have crypto wallets and they're interacting with a piece of software.

It's very straightforward and technologically possible for the first time for the protocol to very directly distribute ownership of itself to its users. It wasn't technologically possible up until very recently. The things that are even possible for in a modern organization might evolve. Maybe Nike would have made every one of its shoe purchasers-

Jeremy: Not too late?

Robert: It's not too late. Maybe buying it on the blockchain.

Jeremy: Go ahead, sorry.

Robert: It wasn't even technologically possible, now it is. It's a very early experiment. We're one of the first to try this. We're starting, there's still a couple of years left of distributing this ownership and control to users. Just like you can distribute the ownership directly to the users, the governance tokens can interact with the protocol in new and interesting ways. Users can upgrade the protocol, they could change the economics of it, they could change the cash flows. I've actually been surprised and this is actually one of the things I didn't anticipate. Since launching the Comp distribution, there actually haven't been any proposals to send cash flows to the users in any way.

This actually surprised me, because I actually assume that the second that the users had governance, they'd say, "The first thing we're going to do is pay ourselves." It actually hasn't happened. I'm actually delighted by that but I'd assume that the very first thing that people would say is we're going to pay ourselves from the protocol.

Jeremy: Yes fascinating, really fascinating. A couple things that this connected to that I wanted to talk about, too was the Nike example is an interesting one. We own and operate SeedInvest which is an equity crowdfunding platform, lets people directly sell equity to individuals on the Internet. The rules for that are really expanding and we had the SCC, at the end of the year finally issue some clarity and guidance around how broker dealers can clear custody, settle digital assets. You are seeing this opening where the between crowdfunding rules and SCC clarity on digital assets where effectively tokenized assets, and when we think of financial assets, I'm talking about equity in a corporation or a debt instrument, or a physical property or other things.

Really, there's an opening for those things to happen now. What role do you see for tokenized real-world assets, like if someone had a valid form of tokenized equity in a Nike or whatever? Do you see Compound interacting with liquid markets in equity tokens, in physical asset tokens in other things?

Robert: The answer is, I think, obviously, eventually, there's going to be lots of digital assets that represent more and more often real-world things. They represent stock in a business, or they represent a bond, or they represent currency, or they represent a house.

Jeremy: There's this currency thing called USDC, I heard about.

Robert: Yes exactly, it's great. I love USDC. I think there's going to be more and more assets like this administered on blockchains. Blockchains are a better place to administer any asset versus just in a database or using paper. It has tremendous advantages for transparency and settlement and all these great things. Every asset will wind up on a blockchain. Whether or not all of those assets interact with something like Compound is a broad question and it's hard to understand, but will they interact with DeFi, absolute right in general. Compound specifically relies on very large liquid fungible assets where an early-stage company stock probably isn't that. The dollar tokenized is or the Yen tokenized is.

Jeremy: Yes, it's interesting, you look at the liquidity pools and MMs in various decentralized markets. I mean, markets exist for highly illiquid tokens and in some ways there's ways to parameterize those markets to deal with the fact that these are highly-- Like when you get someone who can stand in there and say, "No, I'm actually going to stand in these trades to provide whatever that baseline or liquidity is." I've always been interested in this idea that in particular DeFi and blockchains will enable long tail capital markets.

We have long tail advertising. We have long tail content, long tail everything, but long tail capital markets where actually a liquid quote, "Private company stock." Or a debt obligation of a farmer in a market somewhere in the world. All these things could actually have the ability to participate in markets like these.

Robert: Absolutely. I think the quantity and type characteristics of different DeFi products that emerge will start to cater to that long tail, because it's all possible. When you replace these systems with programmable ones in which it's really the creativity of developers, the ingenuity of them to create new financial services and products, or the barrier anything is possible. I think eventually there will be assets like seed invest companies tokenized living on a blockchain that are able to interact with all new financial systems.

Jeremy: Yes. It ties into another question I wanted to hear your thoughts about which was, this is inherently global. Anyone connected to the internet can participate. That's causing a lot of heartburn [chuckles] for national governments who are like, "Wait a minute." We look over some of these things. Is this going to be a market, and I don't just mean about Compound, but I think about this more broadly. Is this going to be a market where it's 10X better and society just says "We want this. We want this, and so you guys need to adapt versus technology. You need to adapt to governments." Does that happen?

I hesitate to use these examples, but Uber and Airbnb or DD or Tencent or these companies that make these 10X better products. Then society's just like that's what I want, and the government needs to adapt, or what do you see happening?

Robert: Well, I think we're already seeing where you're starting to see massive volume and massive interest in open financial products and services. They truly are in some ways just categorically 10X better than existing financial market analogs. That comes with dangers and risks. Like is a self-driving bank a good thing? Yes. Is it dangerous? Of course. Is the very first one even more dangerous? Absolutely, but over time would we rather have extremely advanced autonomous financial markets that are open to everybody, that are transparent, that are cheap, that are fast, that are 100% gross margin? Absolutely.

I think society will pull us forth. I think at the end of the day the benefits of enabling this technology will shine through. I think regulators will over time embrace this for the amount of innovation that it can unlock. There's obviously risks and there's obviously dangerous. At the end of the day, everybody wants to protect retail investors which is a good thing, and everyone wants to prevent nefarious activity. DeFi done correctly is a massive, massive net benefit for retail investors. From the cost of banking to the cost of investment to everything else, massive net benefit for retail--

Jeremy: I see that with stablecoins obviously. Our ultimate mission here is ubiquitous value exchange had no cost on the internet. Who doesn't want that any person, any business anywhere? We're certainly moving there.

Robert: They want it for good reason. It's like it's extremely important. It's just so much better, and easier than what we're used to. I think it's like a market that's a virtuous one for the public and it's not a great system for abuse. Traditional financial markets are so much better for nefarious activity. Cash is a nefarious activity.

Jeremy: Opacity and all that. I'm going to end with two-part question. The first is obviously stablecoins and USDC play a significant role in the DeFi ecosystem and certainly with Compound. We've enjoyed the collaboration and seeing that, so just first like high level comments on what you see in terms of the role of stablecoins in DeFi. Then I have a follow-up question.

Robert: Yes. Well, stablecoins are one of the first real killer apps in DeFi in my opinion. The reason being that DeFi it's two words combined, decentralized and finance, but the real word is finance. People like to conduct financial transactions in a currency they're familiar with and they understand which is why stablecoins has become the bedrock of decentralized finance. It's much easier to conduct a financing activity or a trade or anything when it comes with predictable value. Stablecoins are a killer used case there. I think it's obvious why, and I think they'll continue to be.

I think it's possible that what is a stablecoin starts to evolve we're going to possibly see other currencies. In crypto we might see Central Bank issued currencies. There's a lot of different directions that this might evolve, but I think at the end of the day people continually mentally rely on stablecoins for finance.

Jeremy: Yes, we agree. [chuckles] Last question. I like to ask guests various predictions, so I'm going to ask you. Obviously USDC grew 800% last year and ended the year literally on December 31st around midnight at 4 billion USDC. It's already increased by 900 million in the first two weeks, so it's growing. TDL, 800 million to 23 billion USDC, what do you think end of 2021 USDC in circulation and total value locked in DeFi?

Robert: I'm bad at these games. Whatever I say just discount it heavily. I would say we'll probably end 2021 at about $60 to $70 billion in DeFi. I think the rate of growth will decline a little bit, but the numbers will be staggering still.

Jeremy: That's big.

Robert: I'll put my money at $60, and I think USDC will probably be at about $15.

Jeremy: $15? All right. Consensus seems to be consolidating around $15 to $20. [laughs]

Robert: I could be way off, so it'll be fun. Maybe I'm getting anchored there, but-

Jeremy: Awesome. Robert, thank you so much for joining us today. Great to have you. We'll certainly have you again, and look forward to the continued collaboration.

Robert: Thanks for having me on. Thanks, Jeremy.

Jeremy: Cheers. Bye-bye. Exciting times in stablecoin land, DeFi land, self-driving banks, things are moving along very, very quickly here. Really exciting to see and looking forward to next week. Until next time stay well, stay safe and stay informed.

[music]

[00:48:32] [END OF AUDIO]