Recording:

Transcript

Thank you so much for joining us. Of course. Thanks for having me. No, no, thank you. Thanks to both of you. I am extremely, extremely excited about this. Twitter spaces because you're both. I've gotten to know you separately in separate two separate interactions. But I think I mean, you are both very thoughtful people and you're deeply interested in, I would say, money at the molecular level, and therefore you are interested in algorithmic currencies.

And we are right now at a really interesting juncture in in defi where the collapse of their aluna is resulted in calls from many thought leaders, including vitalik, to say all experimentation on under collateralized or credit based stable to just stop because you know they can hurt a lot of people. At the same time, I find myself in the gap that says we're on to something.

There's something really big happening here. We can find levels of capital efficiency we've previously never seen, and there's no scientific reason like a gravitational constant or some law of thermodynamics that says, you know, this can't be done. And that's something I heard John briefly talk about in his interview with Laura Shinn. And I really like the way he framed it and that made me think, you know, it could be a really fun discussion to have with both John and Publius.

Now, I know both of you well, how about I don't think you guys have spoken to each other, so. That's correct. Just to me, I'll just request maybe John give an intro. I know John, for you, privacy is a key thing to talk about because of Aztec. So give us an intro, John. Tell us a bit about yourself and then, you know, Publius will do the same and then we'll get into it.

Yeah. Great to meet you all. And great to meet you. Publius, I have to ask, is the name from the The Federalist, the pseudonymous name that the Federalists used to write about the US Constitution. Exactly right, sir. Exactly right. So we're kind of aligned on these principles of of freedom and self sovereignty. So I'm I'm John Woo, head of growth at Aztec, which is a privacy first XLK roll up on Ethereum.

I grew up in the travel universe. I was a consultant. I worked for a large private equity firm. I went to Harvard Business School. Not the the typical profile of somebody who is in crypto these days. So kind of grew up as a suit, but I discovered crypto really formally like early last year, and my entry point was uniswap v2.

I was like, Wow, these permissionless marketplaces that don't have intermediaries are kind of crazy. And, you know, then it just went from there. I had spent a brief stint at Uniswap itself and then realized that the Capes were going to transform the space and spent some time getting to know some of the ZEEK teams and ultimately settled that asset where we're building the privacy infrastructure for the next cycle.

And yeah, thrilled to be here. Thanks for having me. Awesome. Thank you. Publius, please Intro. How is your brother? That is awesome stuff and I tend to agree that we're highly aligned on financial freedom and the importance of it and also on how privacy complements that. So very nice to be chatting today. The maybe not even compliments, but how privacy is fundamental to it.

The what? What are we doing here? So Publius, there's three of us or three people behind the pseudonym. I'm one of them. We met at the University of Chicago where we were studying economics and computer science, uh, individually in some capacity. And between the three of us have kind of a, a diverse set of technical and economics interests, but at a high level, just speaking on myself have always been highly interested in economic systems and felt that economics are really the core driver behind behavior.

And in 2017, I had an internship, uh, after my freshman year of college trading affects global macro, and it just so happened that my first day on the job was the same day as the firm's newly hired crypto guy, and two weeks later he pulled me into crypto to come work with him and one of my favorite stories to tell is that that evening I got dinner with a very good friend of mine and I.

I was fucking livid. Like, I can't believe I have to now go work on fake internet money. I have my dream internship now. What is this nonsense? And within 48 or 72 hours, I had like a eureka type of realization of this. This technology is going to be fundamental for the rest of my life and presumably a significant period of time beyond that.

And it was it was immediately evident how this technology would facilitate an ownership based economy as opposed to a rent based economy. And that was something that was particularly compelling to me as the 19 year old growing up in a world that seemed particularly unaffordable. And youth transaction fees are one obvious answer. But the other answer that we came to was the high carrying costs of stablecoins where in practice, if you have to pay ten, 12, 14% a year to borrow a stablecoin to make a market, or even if you own those assets, there's an opportunity cost where you could just one that out for that amount.

The concept is it's very hard to compete with traditional markets, compete with Vegas, with sports betting, for example, it's too expensive. So in addition to just blockchain transaction fees, the carrying costs of stablecoins were to us a major thing holding back not just decentralized prediction markets, but it's very obvious to see how high carrying costs hold back adoption of lots of different technology, or make it impractical to use.

And so then the question becomes, well, why are there such high carrying costs for stable? And just to interrupting here, for those who don't know if you can just want to define guiding costs, I don't think it's obvious to everyone. Oh, sure. So carrying costs can be thought of as the cost to hold a position. So in practice, every stablecoin either has an implicit or an explicit carrying cost, or both, sometimes in the form of either a fee for holding the stablecoin that is explicitly enforced or an implicit cost in the sense of if you're holding a stablecoin and not doing something productive with it, whereas you could just lend it out and collect passive

interest, there's an opportunity cost there as well. And so people have to use their stablecoins for things that they expect will return them on a risk adjusted basis more than whatever that risk free or close to risk free rate of return is for lending out your assets on or compound, for example. And so the question then becomes, well, what can you do with your stablecoins that's more more higher expected value than ten, 12, 14% a year?

It's highly limited. And so the question then we were asking ourselves is what is the nature of these carrying costs? Why is it that stablecoins are so expensive to borrow on chain? Another way to think of it is why is it that despite the borrowing costs often often being incredibly inexpensive in the grand scheme of things, at least until recently, why is it that borrowing dollars on trains is so expensive?

And the answer that we came to was a supply shortage where even though the various Stablecoin implementations had strong peg maintenance models to a large extent, where by their price was at a dollar, the supply and demand imbalance that existed in the market manifested itself in high borrowing costs and those high borrowing cost again was what we thought was holding back the adoption of lots of this technology.

And so the starting point, just to state it again is that our opinion is that the carrying cost imposed by collateral requirements, the requirement to lock up some collateral in some form that's just there to create the stablecoin the opportunity cost associated with that action was largely holding back the adoption of Defi. And so that's, as you were putting it, from more of a macro perspective, drawn of what is the the attraction of a decentralized, stable, credit based currency.

And there's a lot of attraction there. But why we were compelled to start working on it was more of a practical issue of all of this cool tap we we want to see adopted seems to be actively being held back by these carrying costs. And from our perspective, the only way to remove those carrying costs is to remove the collateral requirements.

Or at least that's how it appears to us at the moment. And so thus begins the question and I'm sure we'll get into it in depth of can you actually solve is this a solvable problem? And I would actually push back on something that you said at the beginning ask me, which was that there is no gravitational constant or law of thermodynamics here that prevents this from being the case.

I actually disagree. The fact that there is a difference between an on chain stablecoin and a the value that it's supposed to represent. If you don't have collateral 100% collateral and convertibility that you trust. So if you have no collateral behind it, I would make the argument that it's basically impossible to actually have the value of the stablecoin be the same as the value of the the value that it's pegged to.

I think that is technically impossible. And that is in fact one of the main tradeoffs that Bienstock makes, where it doesn't try to maintain a hard peg in any capacity. And it's very comfortable with the price of a being oscillating above and below the peg with unknown volatility. And so I would just push back on the problem, and it is a very hard problem to solve The true theoretical, you have a perfectly stable artificial asset that's pegged to something else that that probably isn't possible to do.

Instead, what might be possible is to create a value, a token that oscillates above and below a value, and the incentives of the protocol create this oscillation and this expectation that any time the price is too high, it will at some point oscillate back to the peg and below. And any time it's too low, the price will once again oscillate back above the peg.

And these oscillations above and below the peg are ultimately where people can derive confidence to sell high and buy low and profit from participation in the peg maintenance. But the fact that you have this true separation between the stable asset and the value that it's paid to make it, we would make the argument impossible that you're always going to have this 1 to 1.

There is some trade off here. And so then the question becomes what are acceptable tradeoffs? Awesome. No, thanks for clarifying that. I mean, I think meaning I was kind of getting at was that that, you know, at least in finance and economics, I've not seen like the, you know, the same kind of laws that you find in hard sciences.

But yeah, I totally get I totally get your point. And we are going with that and I'm looking forward to getting a bit deeper into it. I think one thing I'm assuming, John, you're not that familiar, but you know, even wouldn't think the one there's one very interesting overlap do with being and privacy and that this that's that's that some decisions that these guys took around the launch which is that you know it was it was truly a fair launch like there was there were there was no external capital.

They just literally, you know, let out the program into the wild. And I don't know I don't know how familiar you are, John, with that history. And I'm just I just want to make sure you have some context before we start diving in. So do share how much you've studied this before this conversation. Yeah, I think I'm pretty up to date on how the protocol mechanism works, but did not study the fair launch.

So yeah, happy to hear how that went. And maybe for audience members who aren't familiar either. Yeah, I, I mean, you can maybe talk about it, you know, but, but I mean for me that the reason I'm bringing it up is, you know, there's, you know, I think I think there's I think that I think both topics, you know, I think, you know, I mean, privacy, it's almost like, you know, it's going to be really hard to have privacy without without algorithmic guarantees on.

I think you're leaving out the punch line to ask me, which is that until the attack a month ago, we were totally anonymous and. Right, Right. Yes, please. No, you you please tell us the right punch lines at the right order and then we can kind of get into you know, the the the overlap between the two topics, privacy and algorithmic currencies, fun stuff.

So the the the punch line, as we were alluding to, was that we deployed Bienstock in August of 2021 as cold turkey is as humanly possible. So we deployed the software. There was no pre minor pre-sale or pre-launch or team allocation or anything of the sort. The first 100 beans, the stablecoin were minted when the protocol was deployed and then the protocol started to function as normal and early participation in the system was the way that anyone that worked on the project in any capacity from the from us to the people that designed the beta logo, everyone still had to buy into the system.

There was no team allocation whatsoever. And then the other point was that we did this totally anonymously. So there's a couple of different reasons behind that. The main one of which is we do want and do think that it's possible for Beanstalk to become a global issuer of money over time. And we really don't want to be at the at the head of that.

We don't want to be the leaders of of that system. And we really do believe in decentralization and permissionless business and don't want to feel like to some extent, the expectation of having someone who's a leader, even if they have no control, is not is not optimal for a money system. So we we've tried and since the attack we did feel compelled to dox ourselves because a lot of people thought we were involved in the attack and wanted to make it clear that we were not.

But generally the the concept is we we don't view ourselves as anything other than the people that deployed the protocol and made it and are now active participants in the community. And while we are now docked and I'm not talking through a voice modifier anymore, the we continue to operate through our pseudonym because the goal is for at some point have to still disappear and we don't want to have this role in perpetuity.

So recognize it's still early and there's still a community forming and people look look for guidance, particularly in tenuous situations. But generally we don't we don't we think that there's elegance in the separation from ourselves and the student pseudonym through which we communicate because we don't really want to, you know, we don't want to be leaders of the system in the long run.

We want the system to speak for itself. John, when you tweeted about an ICO, tweeted for the spaces, you talked you touched upon some of the overlaps between algorithmic currencies like this and privacy. And I know that you've been deep with Aztec for almost a year now, or maybe less than a year. Tell me a bit about what you meant and elaborate a bit on that, on that code tweet that you did for the starting of the session.

Yeah, I think I'm in the beginning of kind of thinking about this paradigm because to be quite honest, we've all been trained on a public blockchain paradigm, and so it's almost hard for anyone who's been in crypto for any amount of time to consider the role that private state can play right in smart contract design, but something to offer.

And maybe this is more of a brainstorm with you and Publius is this notion that, like you can still write a deterministic contract that has private functions or variables, or even for the whole contract to be to be private. Now, why is this like valuable people immediately kind of like push back people immediately. They kind of they kind of push back because they're like, well, why would you want to obscure anything?

It isn't full transparency. The goal and I think that's true in a public state paradigm, but I think there's many examples in the real world where all you need to know is have faith that the process is working effectively. You don't actually need to have full transparency. And there's many, many examples of this, right? You know, private contracts are like a really good example where you can there's an element of trust there.

But like you trust that the two people have agreed to a process that works efficiently and you don't necessarily need to know the details. You just know that the two parties are acting in a way that's, you know, conforming to expectations. Well, I think the cool thing about KPIs and like on chain privacy is this notion that you can create a contract that's deterministic, that is provable mathematically.

It's basically impossible for anyone to cheat as long as the contract is written effectively. And of course the cryptography works and is audited correctly. But at the same time, you don't necessarily have to reveal the current system state. And this sounds really scary to people, really scary to people. And I'm not saying that this is definitely something that has to happen going forward, but you can imagine a world in which, for instance, you know, there's something about the system state that is obscured, and I'm not exactly sure how that would work with Bienstock.

My understanding of the system, for instance, is that it's very important to know where you are in line in terms of like the seniority of your credit position. And it's of course, very important for everyone to know where the protocol sits in terms of the peg. But I think this is just we're in such early innings thinking about private state and how it can inform game and mechanism design that it's worth considering.

Like is there something that can be obscured while, you know, one one thing that comes to my mind and and that's there was a there was a really great discussion that both the Publius that in terms of just, you know, their dissection of what happened with data and, you know, I mean, I'll I'll ask you to summarize it as well.

My takeaway was that, like, you know, when that that going to the collateral part was it was less problematic because it's one that was a bit of an acknowledgment that, you know, the system wasn't as as robust. And then the second thing was that the moment market players figured out some of these parameters, there's a way to start exploiting them because you kind of know exactly where the weakest link could be.

And in the same discussion, you know, there was there was also quite a bit of talk about this convert feature. And if you're John, if you're not familiar, there's a and there's a feature that got launched in the Beanstalk protocol that allows the bean holders to go in and out of an LP position and so the big point is that, you know, unlike a lot of other protocols, there's a whole lot of people that that are being holders who are involved in doing bank maintenance.

Now, some of those actions, you know, if made private, I think could be interesting. I don't know. But I'd also then now Chip the same conversation to probably as public as I'd ask you to do two things. One, explain the convert a little bit, you know, and then talk about is that is it would there be a benefit to, you know, have greater privacy in terms of, you know, who's in terms of convert in and out of liquidity pools?

Well, the the the high level answer is there's a ton of value in introducing privacy to most systems, but certainly in stock, which will answer your second question in longer form in a second. But to talk about convert, so the beanstalk Dow, the silo, uses the stock system to attract long term depositors and stock is the governance token.

The incentive structure of the stock system is that you receive stock when you deposit beans or LP tokens that are whitelisted into the silo and then you receive more stock over time linearly. And so there's an opportunity cost when you withdraw from the silo because you have to forfeit all of your stock. So if you're going to withdraw and come back, there's opportunity cost in the form of all of that grown stock, that extra stock you've received for the time that you've already spent in the silo.

And so the originally, when being stock launched, there was a weird tension between the being seigniorage that was paid to silo members and this auto compounding and their desire to sell the beans, but their lack of desire to withdraw from the silo, have a withdrawal, freeze, forfeit their stock and then sell into LP, add the LP and then deposit.