On October 26, 2020, at the 7th Old Friends Reunion IOSG DeFi Summit, We invited Opyn CEO and co-founder Zubin Singh Koticha, MakerDAO founder Rune Christensen, UMA co-founder Hart Lambur and Thesis and KEEP CEO Matt Luongo, on the theme “Musing on DeFi: Assets, Collateral, and Users”, to discuss the reasons behind the wave of DeFi 2020, real-world demand for the role of a stable currency, synthetic assets assumed, incentive mechanism, the limitations of decentralized governance, capital efficiency and the concept of unsecured lending and other topics.
Moderator-Zubin Singh Koticha (Opyn):
Hi guys. I’m so excited to be here. I’m Zubin from Opyn. I’m working on an options protocol for DeFi. We’re super excited to be working on a new version of the protocol and super excited to have Hart and Rune who’ve been thinking on some very similar questions as well. It might be helpful for you guys to introduce yourselves. Hart first. I would love to hear more about your background.
Hart Lambur (UMA):
My name is Hart Lambur. I’m one of the co-founders of UMA, which is a protocol to build synthetic assets on Ethereum and other blockchains. I actually come from a traditional finance background before getting into crypto a while ago.
Rune Christensen (MakerDAO):
I’m Rune Christensen. I’m the CEO of the Maker Foundation and co-founder of the MakerDAO protocol. MakerDAO is a decentralized stablecoin protocol on Ethereum. The main product is the Dai stablecoin which is a decentralized currency that’s soft-pegged to the US dollar. But the protocol also offers a peer-to-peer credit platform, where people are able to lock assets like Ethereum (and) Bitcoin and then borrow from the protocol. And that’s how the Dai stablecoin is generated.
Moderator-Zubin Singh Koticha (Opyn):
Along these lines, one thing that you had mentioned was that this community doing useful things for the protocol and both UMA and Maker have protocol tokens and decentralized governance to a certain extent. ***What has been your experience as developers of decentralized protocols in trying to get a decentralized community excited, getting them kind of using these applications, getting them to have new updates to governance? ***It seems like a very interesting transition period for the DeFi protocol, what have you learned in this process? What do you wish you had known earlier and what has kind of surprised you?
Rune Christensen (MakerDAO):
I think for the Maker community, it has definitely been a pretty intense learning curve. It’s completely transformed itself multiple times almost six years now that the project has existed. I think that one of the most key takeaways is that if you get it wrong, you’re never going to (make it). It’s one of the most important things to deal with when you’re actually founding and you’re bootstrapping at the very early stages of the community. That is kind of like a sweet spot in terms of the complexity that you can control through decentralized governance. So, if you don’t give people enough control, if you don’t give the governance community enough levers to manage, then, the community is not really exciting and you’re not going see any engagement.
But it actually also exists on the flipped side. If you create too much complexity, too many levers, too much stuff, then, you create analysis paralysis, and you will completely overload your community. Thus, wherever you want to go with the decentralized governance in the decentralized communities, you should always be towards an equilibrium where you have a self-sustainable community that can operate a protocol and maintain that stable equilibrium of sustainable growth. I would say it is really like a variation of art form to shepherd a community towards that equilibrium where you have to make sure that the community’s understanding of what they’re governing evolves alongside those levers being exposed to them and the complexity being opened up.
So, that’s something that we’ve constantly grappled with in the Maker community where we might see there’s not enough stuff to decide. For token, for instance, they (community members) start playing around with the rates because it’s almost like they’re bored, so they got to do something. And they’re going to start messing with the rates. And on the other hand, there has also been plenty of situations where the development team has dropped some super advanced (questions). Like here are 2000 decisions, what do you think? And then the answer is there’s just no answer. You don’t even get a response in that kind of situation. It takes a lot of intuition to try to hit that proper sweet spot where the thing like the knowledge of the community is able to expertly and properly control the governance levers, so that you get the a positive outcome that you want.
Moderator-Zubin Singh Koticha (Opyn):
So, it’s about not changing too many variables at once but also making the decision space sufficiently interesting and meaningful.
Rune Christensen (MakerDAO):
Something that’s super important to also mention is that there has to be the clear direction towards total decentralization. I really think that you can’t have any sort of compromise in terms of a self-sustainable equilibrium for decentralized governance. It means that you must have true decentralization. But the thing is that we have to take it step by step on the way to get to true decentralization because otherwise you will create things like power vacuums and get stuck in.
Just to come with a recent example, if you look at SushiSwap which is an example out of the Yield Farming age I think that shows this kind of dynamic where you could almost have what’s supposed to be a decentralized community get stuck in a particular state. And another example is there was a literal governance Takeover of the Steam platform which I think is another crazy dynamic of the decentralized governance when you navigate those.
Hart Lambur (UMA):
Rune and Maker very much are the OGs that really helped make out or provided a roadmap or at least some data points for how this stuff should work. I think it’s really interesting the governance token has become a buzzword too this year, and it’s supposed to capture value. Academically, the way I look at this (governance token) I think it is kind of interesting. If you think about a traditional corporation, a corporation has shareholders who elect a board of directors who then elect a CEO. And you have this principal agent problem: the shareholders hope the board of directors do what’s in their best interest and the board of directors hopes the CEO does what’s in their best interest. This incentive structure gets constructed this way. If you think about it that a traditional corporation is structured that way because there’s no way you could (do it in decentralized way). For example, if Amazon be run by its shareholders as like decentralized (which means) Jeff Bezos could not have all the decisions he has to make.
I think we can all kind of agree that (it) would be a disaster because there’s so much complexity here. What I think we’re trying to do with these protocols is we’re trying to keep them simple enough. So, you don’t have a CEO, and you don’t have even a board of directors. You actually have a simple enough protocol that the shareholders who are not shareholders to be clear these aren’t securities and all that kind of stuff. But the shareholders, the token holders, the community can actually govern, can make the decisions that are going to point the protocol in the right direction. I think Rune’s point about a sweet spot of complexity is intuitively spot on right. I haven’t thought about it this way. Intuitively, you want the shareholders, the token holders here to be able to do their job. If they have to make super complex decisions, they’ll just fall flat on their faces. But you also need to keep them engaged in what they’re doing. Lastly, the other thing that would be interesting to get Rune’s perspective on is I think you have to incentivize your token holders here too. So, one of the things that I don’t know if Rune would agree or not, but you know MKR holders aren’t incentivized to vote outside of the fact that they’re supposed to. I think that’s a kind of interesting point here too, it’s too bad MKR holders don’t vote in higher participation rates right now.
Rune Christensen (MakerDAO):
That is a key issue that is constantly fought about in the community and discussed. My perspective is that you want maximal voter participation, and you don’t want voter apathy and so on. But I always argue that the problem is that in most cases what gets offered up as a solution in my opinion is a naive solution. A Maker protocol is (not) just because of how the voting system works, it’s designed around maximal resilience in terms of the flexibility. If the front ends get turned off, or if all the people who trust it turn out to be corrupt, and you have to just go. And (if) the websites get taken down, (facing) that kind of stuff, what kind of dynamics then unfold in the governance system? This is one of the reasons why it has a very dynamic quorum that’s based around how many people are actively voting. All protocols have some form of dynamic quorums or at least the concept that the more people vote doesn’t necessarily mean you’re going to pass more proposals. That’s right because you might have people that vote against proposals as well. So, the basic issue that I think you run into when you try to incentivize voting is do people actually care enough to vote in the best protocol? If they don’t, then if you start paying them, that’s stopping them from voting something random. Specifically, in the Maker protocol, the protocol assumes that if you’re not really being very specific about how you vote, then typically what ends up happening is you should just vote no. You just end up facing the quorum essentially in governance. So that gives the Maker protocol specifically, the direct confrontation with this issue — are people actually voting altruistically? Are they really trying to vote in order to improve the protocol? Or they’re just trying to capture this incentive. Because the problem is if most people are just voting to capture the incentive then having a voting incentive will actually slow down the governance and will actually make it harder to pass proposals. I really hope that this is an issue that can be solved. I just think it’s very complex. I think the solution necessarily has to be complex to some (extent), at least it has to sort of involve the concept around figuring out what’s a good vote and what’s a bad vote, what’s a good decision and what’s a bad decision. You have to actually be able to create a broader framework for that in order to answer the question of who to and when to pay out an incentive and when is someone just voting randomly. I think another really important innovation that pretty much everyone else has, but Maker doesn’t have it yet is delegation. I think that’s kind of like the saving grace. At least I have very high hopes for delegation in the Maker protocols. It being the thing I can inject some new dynamics in how Maker governance works.
Moderator-Zubin Singh Koticha (Opyn):
It seems like a pretty difficult problem but one that is so fundamental for keeping DeFi decentralized it’s so interesting to see the work that you guys have both done from that perspective. Let’s shift(ing) gears a little bit, it seems like a big difference between DeFi and CeFi or traditional finance, (which) is just capital efficiency. When two counterparties in the traditional markets do a large position, either a loan or a derivatives position with each other, usually they don’t have to put up full collateralization or even overcollateralization. And so, you have these fractional reserve systems, you have margin, trades, you have leverage that’s baked into most of these kinds of positions. To me that seems like a big unlock for DeFi, it’s like solving this exact problem. You guys have been working on systems that have fuller overcollateralization. But I’m sure you’ve thought about these questions a lot. What is the unlock here? How do you see DeFi evolving to answer these questions: Is it identity? Is it using real world assets which are themselves leveraged?
Hart Lambur (UMA):
I might reframe your questions slightly here. I think there’s a difference between secured lend ing and unsecured lending. I’ll also say there’s a difference between margining. So, most loans in this world are secured loans where you are borrowing against an asset. And you are not borrowing more money than that assets’ worth. Just take a house for example, I have a million-dollar house and I can borrow less than a million dollars against the value of that house. This makes perfect sense what finance is acting here, what it’s doing is transforming illiquid assets to liquid assets or vice versa. And Rune can talk a lot more about this. But I look at Maker as very much being this credit facility where you take an asset, and you borrow against it. So, it’s secured. To me this (secured loan) is a crypto native concept because everything is secured by the blockchain. You can observe the collateralization ratios. You can observe and make sure that all these loans are correctly collateralized.
The other type of loan that exists in the traditional finance world is unsecured lending where you’re actually borrowing against someone’s credit. So, I’m going to lend Zubin money on a credit card because I know Zubin and I have something (trust) about him. To me this is really interesting, but it’s far less crypto native because it fundamentally involves the concept of identity. In order to do an unsecured loan, I need to know who that person is or something about them. And maybe there’s like proxies for identity, maybe I can look at some history of their activity. Or I can have people vouch for them. But to me unsecured lending is just actually a very different problem in a crypto setting because of this fundamental tie to identity.
Moderator-Zubin Singh Koticha (Opyn):
Here’s a follow-up question on that. The house or the car that you’re lending against when you take an auto loan or a mortgage, you can use them, and they fundamentally have a lot of utility as assets***. Do you see some kind of interesting unlocks there? Does a cToken, which represents compound loans, already look like that? What do you see as the equivalent here?*** Because when you lock up Eth to borrow Dai, you’re not able to use that Eth anymore. In the same way that you can still live in the house. So, what do you see is the unlocks there? Do you see this as potentially? DeFi is good enough without unsecured lending. This crypto native financial system can become really large without even necessarily needing identity and leverage in that sense.
Hart Lambur (UMA):
I think of these are slightly different things. So, if you think what DeFi exists right now is within crypto assets I can borrow and lend against them. I’ll be able to transform one crypto asset to another, it’s like this transformation thing. What I think is really interesting in terms of making it usable from like a real-world perspective is it would like one of the assets. I could borrow against to my real-world property or my auto or whatever, and that then requires some tokenization of that real-world thing. If I want to bring in a DeFi and I think Rune can actually talk a lot more about this about what that pathway might look like. And then Zubin I can also talk to you about how with derivatives. I actually think derivatives came on this idea of real-time margining. This is a completely separate concept, but we can come back to that later about how I think derivatives in DeFi can be hyper capital efficient.
Moderator-Zubin Singh Koticha (Opyn):
I absolutely agree. That is the thing that I’ve been working on in research at Opyn with a couple of other folks as well. It seems like derivatives are a big unlock here for DeFi as well. Specifically, we’ve been talking about the capital efficiency of DeFi assets. And tell me if I’m wrong, tBTC is kind of an over collateralized asset as well. What do you see as some interesting future directions to make DeFi more capital efficient and what gets you excited when it comes to the question of capital efficiency as well?
Matt Luongo (Thesis / KEEP):
I’ve got a pretty probably different perspective than these guys. The reason that were inefficient, the reason that we require so much extra collateral is to ensure good behavior rather than to enable some sort of like synthetic or derivative. So, I think the way that I look at this is that right now capital in the crypto space is not scarce, but that will not remain the case forever. So, for tBTC we’re definitely looking at how can we lower the collateral requirements as much as possible. I’d be really interested to hear what Rune thinks for example when suddenly we’re all competing for ETH with ETH 2.0.
Rune Christensen (MakerDAO):
I mean that’s a really great point and I really think we can completely zoom out. I don’t have a financial background, so the way I view this whole question is that secured lending and unsecured lending, those are kind of like these specific terms in the financial world. That means whether you put up let’s say something physical, something valuable, or whether you’re just borrowing like against your character your credit or whatever. But from the DeFi perspective, or sort of from a more generalized perspective, in my opinion, there’s no such thing as an under-collateralized loan like the concept doesn’t even make sense. Because a loan is always based on the concept of recourse. So, no matter what, any sort of credit always has to be over-collateralized because otherwise you’re not going to borrow money from someone. If whatever that person gets in return, whether it’s your promise or your house or whatever it is, they would never ever give you a loan of higher value than what you’ll give them in return to ensure that they can get that loan back. So, the way you can think about unsecured lending on the blockchain is simply if you can represent someone’s trust, essentially the credit rating. You could represent that (trust) as a token with some sort of value, then now you basically put into the classic DeFi paradigm of overcollateralization. So, where someone who can be really trusted a lot and (there is) a legal claim against them. (The amount of the legal claim) could easily be of a million dollars and you could feel safe with such a huge legal claim because they’re totally trusted, they have a great reputation they have tons of assets and so on. And on the other hand, if someone can’t really be trusted, then the value of a legal claim against them is just not going to be much greater. So, in practice the way you would implement something like consumer credit in Maker would be (like) for instance a company which is financed by a lot of credit cards. And then that company aggregates all of these lines of credit and has into a single pool of aggregated value. You could say ultimately that business can be tokenized and can be used as collateral in Maker. And if you know that too many of the loans start defaulting, the overall value of that company might fall. And then that is where the Maker protocol might liquidate it. And ultimately the point is that you can always reduce any kind of financial relationship to this paradigm of an over-collateralized loan.
But then to the other point of ETH staking, that’s a really exciting topic. This is where the amazing nature of blockchain and programmable smart contracts really starts shining. First of all, you can actually wrap ETH that is staking. There is a thing called Rocket Pool for instance, that’s an example of wrapped staking ETH. And I’m sure there’s going to be a lot of other services like that right. What is cool is that you can actually use those as collateral as well in DeFi. This is like the key concept that actively staking ETH is always going to be riskier than pure ETH that’s just like highly over-collateralized. So ultimately, no matter what you choose to put as collateral into DeFi, you’re going to have to pay. The riskier the asset you want to put in, the higher the rate and the stability fee. And obviously the goal of the builders of DeFi should be to try to offer as many options as possible. So, users can pick whatever they prefer.
Moderator-Zubin Singh Koticha (Opyn):
I would love to get Matt your thoughts. Because as well about specifically ETH2.0 and you have mentioned competing for yield. How do you solve that issue? Do you use wrapped ETH that represents like a staking ETH? How far away do you see ETH2.0 really being for this to be a real issue.
Matt Luongo (Thesis / KEEP):
You know what’s interesting is I haven’t been quite an ETH2.0 skeptic. Not in the sense of like the technological approach just in Ethereum we have an ecosystem that’s premised on over-promising and then taking an extra five years or so to deliver. So normally, I would say we can wait it’s not a big deal and I wouldn’t worry and if you look at how we’ve built our Random Beacon under tBTC. We did that because I was confident that ETH2.0 wouldn’t ship in time relevant to our project and that’s what happened. That’s said if you’re going to be able to lock up ETH in a month whether or not the rest of ETH2.0 is functional that’s still competing for yield. So, I’m in the unfortunate situation where we have kind of handcuffed ourselves where we’ve taken the opposite approach to governance as Maker. So, we don’t have a way to onboard new collateral and I’m really optimistic about staking derivatives. I generally think to Rune’s points earlier finding a way to properly represent and tokenize cash flow in addition to hard unproductive collateral is a great way forward in this space. But I think it’s going to be tricky for our team and I think that we’re going to have to ship to (Ethereum) v2 sooner rather than later. If ETH2.0 is popular, many protocols who haven’t thought about staking derivatives are going to be hurting because right now it’s really hard to overstake ETH. It’s just slashing around.
Moderator-Zubin Singh Koticha (Opyn):
Actually, what might be helpful for everyone is just a brief high-level overview of your background and what you’ve been working on as well.
Matt Luongo (Thesis / KEEP):
Well, hi guys I’m Matt Luongo. So, my background is I’ve built a Bitcoin company in 2014. I since (then) started working on Ethereum stuff, in particular a bitcoin-Ethereum trust minimized bridge called tBTC. I think what’s probably interesting about my perspective that is typically we’re governance minimalists. Not because I think that it’s bad but just because a lot of Bitcoiners would prefer to not have majority governance. And also, just a little bit more skeptical I think of the Ethereum ecosystem. I love the tech optimism, but I think a lot of the problems that we’re going solve in the space are going be not just solved by technology.
Moderator-Zubin Singh Koticha (Opyn):
Are you scared of consensus instability and MEV and the future for Ethereum at least in the Proof of Work?
Matt Luongo (Thesis / KEEP):
I love it frankly. I think as a protocol designer MEV is fascinating. It does not scream that this platform will become unusable to me. It means that protocols need to be cleverer. So that’s kind of been my takeaway so far. I think it’s going to be really interesting. I think that Ethereum right now has kind of accidentally landed in this really interesting spot of being Proof of Work as well as obviously being expressive enough to build interesting gaps. And so, I hope that it can maintain that as its transitions to Proof of Stake.
Moderator-Zubin Singh Koticha (Opyn):
All right I think we’re just about done here. What might be helpful is just a little bit of a wrap up what makes you optimistic about the space or just a little tidbit on something cool. I think we can go Hart, Rune and then Matt to end it.
Hart Lambur (UMA):
Collectively, DeFi is all about designing incentive structures, even Rune’s last point I like that. Unsecured credit doesn’t make any sense. There’s still an incentive structure. In traditional finance unsecured credit has some value it’s your credit there’s a credit limit. Everything we do is designing incentive structures, and the thing that is really powerful is that when you get those incentive structures. They are super powerful. They’re kind of like overnight exponential explosive growth and like that’s kind of what we saw with the Yield Farming bubble. So, the part that gets me optimistic and excited is that DeFi as a group of mostly good-natured or good-hearted people. They are actually pretty excited about trying to innovate in financial concepts building incentive structures that I think could be really powerful and that’s quite fun.
Rune Christensen (MakerDAO):
I think this is what I always talk about whenever people ask me what’s the next exciting thing. It’s the coming intersection of real-world finance and real-world assets and then DeFi. We’ve been talking about that in the Maker community for two years or more at this point right as like the holy grail that is coming somewhere far out in the future. But now actually this month the Maker community is onboarding. It’s the very first couple of real-world financing projects and it’s funny to actually make these real examples. Because basically the Maker protocol is right now considering whether to finance the construction of a number of auto parts shops in the northeast of US as well as some trade finance companies. And it’s funny that now we actually get to see what this actually means, what does it look like in real life as it actually happens. I mean anyone who wants to follow along and just go to forum.MakerDAO.com. All of this stuff is actually just happening there, completely in the community. Personally, I haven’t really been involved in this at all. It has all been just unfolding because of the framework that has been put in place. So, if we can prove that it (putting real-world financing on-chain) can be done, then from there it’s going to start to happen everywhere in all of DeFi and all of these protocols and systems. Interactions and incentives will start to take on a new dimension of actually being used in the real world to do real stuff at scale. That’s going to be a pretty epic I hope.
Matt Luongo (Thesis / KEEP):
I think I’m the most excited that we’re still here. After 2017, after this very softly popped FoodCoin bubble, we’re still doing our thing. And I think every day that crypto is alive is another day that we built legitimacy. You see it in Rune’s work, you see it with the latest exchange where founders get arrested and it barely impacts prices. Those are maturity that we’re starting to see. I hope that we’re also going to start seeing it. I’d say I’m really excited about that. I’m also excited to see a lot of the promises that I’ve been showed on for a while some people are delivering. We are starting to see actual trust minimized solutions. We are starting to see DeFi that is decentralized and not based on admin keys. I’m a censorship resistant maximalist and I’m glad to hear that that’s where we’re going.
Moderator-Zubin Singh Koticha (Opyn):
Just to kind of wrap it up what gets me most excited is how the caliber of people having people like you guys (who are) super hungry, super curious and super smart that are working in DeFi and solving some of the hardest problems. And I’m optimistic that we’re going to figure it out.
🦄 About IOSG
Founded in 2017, IOSG Ventures is research and community-driven with offices across China, US, Singapore and Germany. We focus on Open Finance, Web3.0 and cross-chain ecosystems, investing in teams with top potential worldwide. Our portfolio covers more than 60 projects, including Layer-1 blockchains (Near, Polkadot, Cosmos), middleware (Celer, Raiden, Reach) and applications including Defi (MakerDAO, Synthetix, UMA). We have been actively involved in various developer & DAO communities. We believe in long-term partnership and we work closely with our portfolios to advise and support them along their journey of entrepreneurship.
If you would like IOSG Ventures to consider your project, please send a summary of your project along with a pitch deck and/or white paper to [email protected]
Follow us on Twitter.
