Behind Crypto's Builders: Past, Present, Future
This is the first of a new Q&A interview series at Standard Crypto, titled Behind Crypto’s Builders: Past, Present, Future. In this series, our co-founder Adam Goldberg will be speaking to founders about some of the most important inflection points in their careers and their projects or protocols.
Here we speak with Rune Christensen, co-founder of MakerDAO.
This transcript has been condensed and edited for brevity and clarity.
Adam: Startups are a series of really unknowable choices. All great startups have inflection points where builders are faced with difficult decisions. These decisions compound for many, many years, for better or for worse. Today, we're really fortunate to be joined by Rune Christensen, the co-founder of MakerDAO. Maker needs no introductions, but in a few words, it has created the largest decentralized stablecoin DAI, which has over $5B in issuance today as measured by makerburn.com.
This is a rare opportunity to do a post-mortem inflection analysis for such an important project. I want to start off today's discussion by talking about some of the prior inflections that inform Maker today. Let’s begin with the history of mechanism design. You started the DAO a decade ago — and in many ways it was the first of its kind. DeFi was really just a concept and the mechanisms were almost entirely theoretical. What was it like designing mechanisms for a system with no analogues?
Rune: Well, so first, I definitely can’t take credit for being the first to do this kind of crypto economic design. Maker effectively started off as a clone of the earlier stablecoin system called BitShares. So, that's really where I got my first taste of this whole new world of what's possible with blockchain, coming from basically being an OG libertarian Bitcoin maxi.
BitShares was definitely the first that had this clear idea of how you can use blockchain to build a real business. You’re gonna have a product, and you can manipulate stuff on the blockchain through financial engineering that actually creates real and useful financial products. And to me, that was just as big of a shock as encountering Bitcoin itself. It’s amazing that BitShares never really took off. Right? It’s amazing to think that one day we'll have all this amazing stuff we were dreaming about back then. And a lot does exist nowadays, which is awesome to look back at.
Adam: What did mechanism design look like back in the early Maker era? And how has that shifted now when you fast forward to 2023?
Rune: Funnily enough, it really hasn't really changed that much. I'm a person who says — and I think a lot of people will tell you this — that stablecoins are pretty much the only business model that still works. Even to this day, this is the business model that makes sense in crypto right now, with a few additional twists. There are pure lending markets that don't have a built-in stablecoin. But ultimately, arguably, they're also very, very close. You could argue that something like a USDC deposit into Aave is really also a stablecoin. And as a result, Aave is sort of actually also the same as the original BitShares model, with just a few modifications.
Adam: Synthesizing what you're saying, the core of the mechanism hasn't changed that much, even if things on the edge have. One really interesting thing to unpack that felt pretty controversial in the past was the transition from SAI to DAI — from single collateral to multi collateral DAI. That was a really big shift. And in some ways even in part enabled the incredible accrual of value that you've been seeing in Maker through RWA under its shared stable asset. Can you talk through some of the major factors in that decision? What was the case against it that was articulated?
Rune: I think the case against it was some kind of ideological vision, some ideological dogma, I guess. But one of the big changes that we applied on top of the BitShares model was the diversification of collateral — because it logically follows that this is required to make something that actually can scale. So SAI fundamentally wasn't itself sustainable. And it wasn't meant to be sustainable. It was like an integrated product.
[When it comes to a DAO], originally I thought, you write a white paper and then the free market takes care of it. And we continuously realize this is not how it works — all these mechanisms like a stablecoin and stable assets are a tiny fraction of what it takes to achieve success.
With Maker, we were just basically lucky that we were so early, as we've made so many mistakes. And we were basically really, really bad at executing compared to even the general average standard in crypto today, which is still very low. And especially compared to what the teams at Maker and its users today are capable of doing. But this gradual transformation throughout the years of the project was a kind of realization that increasingly it matters less what exactly the numbers say, or whatever the fancy code or the fancy economic design may be. What matters more is how to make humans act, right? Like the behavioral economics rather than the math behind it, in a sense. And that's what it took for us to even launch anything — we had to originally create the Maker Foundation. That enabled us to build SAI, and then after that multi-collateral DAI. We then dissolved the Foundation, and ran into the same issue — people problems in the DAO. And that brought us even further down this path of how to overcome the problem of organizing humans.
Adam: That's a perfect segue to talking about some more recent things with Maker. A really prescient one is what recently happened around the Maker Burn, the smart burn engine. We just talked about mechanisms in detail — in some ways we’re talking about this intersection of market forces and organizational forces. It feels like one of the most novel things that Maker has done is borrowing from the traditional finance concept of share buybacks. A company can rebuy and retire shares to increase share price. In many ways, this is economically isomorphic to a dividend. But the best criticism of the burn model is similar to why we generally as venture capitalists think that startups shouldn't issue dividends. In many cases, you're better off reinvesting that cash flow, in research and development and growth rather than paying it out. How do you think about that trade off? Especially given that Makers’ income revenue, so to speak, was rapidly approaching an all-time high?
Rune: Well, I totally agree with the classic logic, that when you're trying to grow, you should reinvest. I mean, you shouldn't just reinvest your earnings, you should borrow money and raise money and go completely all-out with doing things that make you grow. The problem is, it's not as simple as putting money into one end of the machine and growth coming out the other end. In fact, this is almost impossible. The closest we have to that is yield farming, which sort of works at a different level. And it's much more primal and is very driven by market forces.
In the [DAOs] where we see good growth and good income and sustainable business — and Maker is one of them — it's always like they were first or something, right? Organic growth of some kind. But if you go and ask them, “How did you grow this far / And how do you intend to continue growing?” And, actually, since all these projects are DAOs, right — this is supposed to be written about, supposedly this is the kind of stuff that the DAO should be discussing in the open — it should be looking at the data. But that's not happening at this stage. There doesn’t seem to even be the sort of the thought process of, “Let's think about what made us grow in the past, and how do we translate that into reinvesting into growth in the future.”
We’re trying to do that with Maker, and the number one thing we're tapping into is the yield farming mechanism. Rather than doing marketing, integrations, this kind of classic stuff that we did try in the past, and we just couldn't really tell what worked and what didn't work — instead, you focus on gamification. Gamification tends to attract users and drives some second-order, financial speculation, or something that actually accelerates a kind of a dynamic that appears to result in the best kind of engineered growth we've seen in the space so far.
But we also do real marketing, real branding, the way you're supposed to grow in the real world. But because we're a DAO, we have to figure out how to make those decisions as a DAO. So we have to solve governance in order to be able to solve the question of growth, which is then what we're basically entirely trying to do with what we call Endgame.
Adam: Your point is that [Maker] is a trailblazer in the sense that there haven't been many other crypto projects operating at this scale with profitable unit economics. Let's unpack what that decision process looks like. Because I think that sometimes the best way to address these inflections is to run an experiment, [especially as it pertains to] shepherding the collective Maker community to go on a journey that has a beginning and potentially an end. When you think about burns, there were decisions that had to be made. You had to think about the buffer size, what percent of excess capital is to be used for burns — all sorts of parameters. How did you think about how much comfort you and the community needed to start that experiment — and how much discomfort would you have needed to stop that experiment?
Rune: The way I used to think of how decentralized data governance would work and how I think most people think of DAOs, is that a bunch of people are holding a token, they come up with some ideas, and then decide if they are going to do “this” — or something else. And then if that doesn't work, they change themselves and do something else.
I entirely believe that to be impossible now: There’s no way you can make any sort of substantially fundamental change through DAO governance in a system at scale. It would not be possible to have the option to change something such as — instead of a burn — we’re going to donate all the money to charity. That as an option would be impossible to ever [achieve] through decentralized governance with any sort of meaningful shape or form.
The only way to realistically get autonomous governance is to have simple enough decisions that have a constrained downside, if the decisions are very, very bad. And then you can have decentralized governance. But if you have serious downsides for bad decisions, then it's not going to work, because there's so many things that can go wrong, basically.
Adam: I feel like Endgame has been a really thoughtful way to bring more resilience to the system. I also wanted to explore the messaging around it. I think that there was a period of time where it felt like Endgame was one of the most controversial things overall in the onchain world. That's to be expected, right? When you're shaping something as deeply important as Maker, you're going to elicit quite a range of responses. But I'm curious from the seat you sit in, what decisions regarding messaging were a learning experience for you, and which ones do you feel like were correct in retrospect?
Rune: Endgame was born out of hours just talking about corruption and all this insanity of politics — I don’t think most people realize [the realities]. It’s very hard to even accept that the reality is very crappy human politics is what drives how DAOs end up working in the wild. So I was actually convinced that the whole idea of DAOs were doomed.
The problem is that it’s like a game that begs to be exploited — a system that expects everyone to choose what's worst for themselves. This is like the worst possible prisoner's dilemma, right? Where there's no upside at all for cooperating and the huge upside if you’re in for yourself — if you defect. But it turns out that it’s completely possible to flip that around and make the opposite system, where you have all of these positive incentives for making the right decisions. And that finally means you can drive growth, because this sort of positive incentive to drive a good decision is also like a growth funnel. That's a way to connect people into the ecosystem and into the network.
That then led us to things like the SubDAOs, which again, is not something we invented. This was an existing concept that we took and supercharged. But they are one of the most unbelievably powerful tools that you can use to deal with these kinds of issues. You compartmentalize the risk and reward. You get to this point where you can make reversible decisions, so if experiments fail, it doesn't actually cause that much damage.
And once you get to that point, you can grow your user base and recruit your ecosystem at the same time. And it all comes down to this fundamental question: Can you create real incentives in the real world that cause external actors to interact with you in a way that creates net value?
Adam: That’s a nice segue into talking more about future-looking considerations, especially around RWAs. I think we certainly would agree that much of the future of Maker [involves] RWAs. Let's talk about some of the arrangers as an example. There's been a really interesting sort of live case study on how all these things interact and how incentives and trust and modularity interact. And in some ways, there's been a new type of paper API template made here — a trust structure that essentially only can engage in a certain set of pre-approved operations. Now you're letting someone run a strategy to generate yield, like using T-bills. How do you trade off centralization versus agility in executing strategies? It’s a blend of thinking about counterparty risks and better and more up-to-date live reporting, versus just moving fast and taking it to market and getting market share. How do you think about that?
Rune: It's yet another aspect where it turns out the hot thing in crypto and DAOs [involves] people and — even worse — legal stuff. You know, the worst and most boring [elements]. But it turns out to be the most important one. What is cool about the arranger model is this breakthrough that allows the decentralized crypto world to reach into the real world and have something that acts as if there's a real legal claim to real world assets — to things like treasuries. How do you structure this stuff legally? You actually draw inspiration from how you make smart contracts. Which, funnily enough, allowed us in the Maker community to design structures that are probably more robust than most legal structures built around gigantic financial asset classes.
Usually, the way you do legal work is you put all this trust in one guy, and then if this person rips off the entire market, then you expect the government to step in and deal with it. This is very often the general attitude in traditional finance and traditional legal setups — you don't worry that much about the edge cases. And in crypto on smart contracts, you have this obsession with edge cases. In the end you can't cover every edge case in the real world. It's just not possible. You can make it good enough in the sense that the likelihood of failure is just extremely low.
Also worth mentioning, is that all this is going back onchain. So the next step for the arranger model and the RWA model — and Maker will probably spearhead this — is getting this done at scale. There’s a bit of a chicken-and-egg problem. But if it really happens, everyone will benefit and everyone will want to use it.
Basically you want to simplify the legal claim to a common standard. You’ll still need these things that we call arranger structures in Maker, that are like legal entities that can enforce legal claims. But what actually matters in terms of making decisions and getting payments and so on is that it all can be run onchain. So you have this technical, onchain or physical setup, and then it's mirrored in the real world through a legal setup. If that can actually be bootstrapped, then it will be a lot more robust and have a bunch of upsides like liquidity. We definitely want to be first with it.