Join Delphi Research today and immediately get access to our full Member Portal!
Join Delphi Research today and immediately get access to our full Member Portal!

Jose & Yan discuss Delphi Labs, Incubating Mars/Astroport, & Token Economic Design

Nov 23, 2021 · 71 min media

By Jose Maria Macedo, and Yan Liberman, CFA, CAIA

Head of Delphi Labs Jose Maria Macedo sits down with Managing Partner of Delphi Ventures Yan Liberman to discuss all things Delphi Labs. The two discuss the origin and goal of Delphi Labs, design considerations around Mars and Astroport token economics and lockdrop mechanisms, and much more!

Social links: 



Show Notes: 

(00:00:00) – Introduction.

(00:01:21) – The genesis of Delphi Labs.  

(00:03:10) – Working on Axie Infinity. 

(00:04:30) – What is token economics? 

(00:07:27) – Moving from consulting to building. 

(00:08:50) – The Terra Luna story.

(00:19:03) – Mars as a generalized credit protocol.

(00:23:04) – $MARS token economics.  

(00:28:14) – Mars’ dynamic interest rate model.  

(00:30:49) – Overview of Astroport.

(00:31:40) – $ASTRO token design. 

(00:40:02) – Astroport’s launch strategy.

(00:56:50) – Other incubator projects. 

(00:59:34) – The goal of Delphi Labs. 

(01:05:04) – Closing thoughts.

00:00 • Jose
Let’s do it. Intro me, intro me b*tch. 

00:07 • Yan
So, we’re here in sunny Portugal where we’re currently, rugging the rest of our retreat out of full access to record this podcast to really go into kind of labs the Genesis, what we do there and, yeah, everything that kind of led up to that. And, yeah, with me is a Josie Macedo. 

00:24 • Jose
I believe he did the sunny Portugal thing as well. Yeah, that’s announcer. 

00:30 • Yan
Yeah, we’re here with Josie Macedo. Who’s head of labs, some managing partner on Delphi ventures, but we spend a lot of time working together in both areas and, yeah. Thinking about token econ and token design and, how to really allocate in the space. Yeah. 

00:45 • Jose
Yeah. That was one of the things we get the most questions about. Right. Even at the, even at the retreat, the questions we get from the team and from outside is where I was like, what do you guys do at labs? Or how does it work? ? And so w I think we just wanted to dig into it today. Like what labs is, how it started, what we’re working on now and what, like the future plans are as well. So it should be a fun chat. Do we need to do intros or did you kind of intro us already for like, 

01:08 • Yan
It does good enough of an intro. All right. Yeah. That’s fair. Are we just going to keep doing this and like, they’re going to have to edit the shit out of it? 

01:14 • Jose
No, I think this is just it. This is just fair enough. We just kind of suck. It’s fine. We’ll just do it. And so, yeah, the Genesis a lot. Did you want, do you want to start with that because you want to say hijack your lines. So. 

01:27 • Yan
That’s fair. I mean, you do it pretty often. Habitual Hijacker. Yeah, the kind of the start was consulting. We started Delphi in 2018 and, we didn’t really bring any outside capital. We had to really get into consulting just because it’s really tough to scale a subscription business just to eat literally just to eat. So, the, originally the idea was, based on our skill set, we’d kind of be an outsource due diligence team for funds, but, and the process of really building out research and really trying to understand what drives value cruel tokens. We started to understand that, realistically are bigger, the bigger value add and what makes more sense is to really work with projects rather than kind of with other funds. So, they quickly transitioned into a consulting model where we would work with projects, from the ground up or the ones that were already live, where we would do a lot of the token econ work. 

02:22 • Yan
So, that was, designing entire models, or just thinking about kind of sub areas where it makes sense to make some modifications. So, that was kind of the Genesis. Realistically, we need literally needed it to eat and over time, as we got a bit more selective, or basically as we got a bit more sustainable, were able to be a bit more selective. And, and through that, we shifted it to just a pure kind of mercenary-esque consulting model to, working, partially for cash to keep the lights on. And then also, basically for tokens. The idea was, we didn’t have a fund and we had no way of really investing. This was kind of a way for us to invest our time rather than capital, and then kind of replicate that exposure. 

03:09 • Jose
Yeah, it was dope for us as well. Cause we got to learn from, and work with really top teams in the space. Obviously we worked with Axie design, the act infinity token got really deep into it, into the Axie model there. Maybe you want to, I mean, don’t know how much you want to say. 

03:23 • Yan
Yeah, no, that was a, that was probably like the first full circle engagement where we, found out about them. We were looking to gaming in late 2018 and, there were a lot of games and most of them were touting fake volume, and most of them were pretty bad as well, but that the standard at the time was Axie. That was kind of our takeaway, but there wasn’t really much that we can do or any kind of way to get involved. We got close to the team through the due diligence process cause we reached out to every single team to talk to them and really understand what they were building. So, within, in 2019, once they were trying to, begin to, thinking about launching a token, they engage in and we worked hand in hand with them to help them on that design. In 2020, our goal is, at the same time we didn’t really have a fund in 2019. 

04:11 • Yan
So, the idea was, once it goes live, what we’re certainly gonna ape, but, thankfully just as they were going around for their seed raise, we happen to launch right. Were fortunate enough to be able to kind of lead that. That was, the first time we found it through research, worked on it with consulting and then we’re able to really invest in the venture side. And, 

04:31 • Jose
That’s like, people ask us a lot, like, what is token he gone mean? Like, what do you actually do? And it really varies depending on the project, right. With actually it was very much like designing the whole economy from top to bottom with something like were focused like on the token incentive bit to how to design that. Also like restructuring the safety module. How to make that kind of a better insurance fund design, for a lot of platforms that involve leverage. We do a lot of work on like the insurance fund and like the token econ structure or skin in the game. We do a lot of work on planning out like supply schedules. I think it all starts with understanding what the project does, right? Like where’s value captured, who are the supply side is who are the demand siders? When do we need like, to reward who? And so it’s like, it’s kinda tough to generalize, right? Yeah. 

05:16 • Jose
Always a struggle. 

05:17 • Yan
It helps to kind of think through how tokens kind of got to where they came to be, where in 2017, tokens are just a way to, quickly raise capital without, too many rules. And, and so that was, the cash grab that kind of happened at that period. A lot of interesting ideas came about then in the bear market in 2018, you had all these tokens and everyone’s like, all right, we’ve got to figure out how to make these things actually valuable. That’s when fundamentals and kind of more token econ really kinda came to be. And then, soon after that, 

05:49 • Jose
Well, I was doing token econ in 2017, but it was mostly just like begging projects. Like, please don’t do this. So you’re scabbing your users, like yeah. A lot of a medium of exchange tokens, try to talk projects out of that. Yeah. People just saw it as a source of like non-diluted financing. It’s kind of hard to get people to do anything. But yeah, 

06:08 • Yan
Yeah, exactly. Like soon after that, it kind of got to the point where now it’s all right. It’s actually a massive asset and tokens are projects that didn’t really have a token were at a disadvantage. I guess that kind of gets to the crux of what token econ is and why it’s valuable in the whole, the high level view is basically you have a lot of users that are all self interested and the token is a kind of coordination mechanism that’s, that can be really useful. And, the whole idea is if you think through, every, if every user acts in their own best interest, how do you put a token in between all of them to organize this in a productive way? 

06:45 • Jose
Exactly. That’s it. If you’re trying to like, run the simulation in your head, right. To see like, all right, given the incentives we’ve designed, how is everyone going to act? And then you’re trying to figure out, is it breakable? And then you’re just figuring out how it’s breakable trying to fix it, going back, trying to break it again until it’s as I’m breakable as you can make it. Right. That’s the design process and it depends a lot on the project. And yeah. 

07:09 • Yan
Yeah. Like, it really helps to, look at almost every project or every meaningful project under the hood, understand, why things worked, why things didn’t and start to, pick and choose different components and piece them together and then think through how everything would work with these combinations of kind of company. 

07:28 • Jose
Absolutely. Then, so as we started doing this work to do token, you got, and you really have to understand the project super deeply. Like you have to understand everything about it, obviously how the mechanism works, who the supplies that as are, who the demand side is, are what the growth plans are. We just started naturally kind of understanding the projects and also figuring out more ways that we could help. Right. Like with THORchain started off one way, and then by the end of it, were kind of building the whole THORchain, mid guard API. We started hiring developers. We started kind of expanding what we did with, labs, with the consulting, which eventually became labs and helping projects in a bunch of ways. Like with Perp, we helped a lot with like the mechanism design for perpetual protocol itself and trying to fix the long, short skew with Aave. 

08:05 • Jose
As I mentioned to the safety module balancer, we did a bunch of layer, two research compound we helped them with like a new product design. We helped them with like a new product design. We just started gaining a lot of IP from working with like really closely with all these teams. We started figuring out that we had like, a lot of IP internally, and we had a lot of ideas about how things, how we thought things should be built. The only way for us to really implement those was to kind of go earlier and start building stuff ourselves, and like expressing kind of, how we think the space should be. It was the natural next step in evolution of labs. 

08:37 • Yan
Yeah. Then, soon after that, we kind of closed down the consulting side as well, just realizing that it makes a lot more sense to spend the time here. This is where we can make a larger impact and really, get a lot more involved. 

08:49 • Jose
Yeah. I guess maybe it’s worth telling the Luna story because a lot of our labs, incubations are focused on Luna and Solana, right. Tara and Solana. That’s because a year ago, when we started labs, we identified Luna and Solana as the two layer ones. Like other than that, were kind of most excited about, but maybe it’s worth going through like how that came about. Right. Cause it was, we, I mean, I think stable coins go back, like our interest in stable coins goes back a long way. Maybe you can talk about like the maker reports and the ESD, yeah. Maybe gets an exit liquidity or something, please. 

09:27 • Yan
So, yeah, we’re massively interested in stable coins. It’s, it’s a huge market. We’re not particularly sold on kind of Bitcoin or there being this universal, might be. 

09:39 • Jose
Italo jobs. Yeah. You’re getting a bit shiny. That’s. 

09:41 • Yan
Why my glasses keep sliding down. So greasy, 

09:46 • Jose
It happens outdoor podcast. Am I shiny a little bit? Alright. 

09:51 • Yan
At the time I was here yeah. That’s useful. Yeah, so, we kind of got into defy early just because were really looking into, when were, when we started publishing research in late 2018, early 2019, the idea was we needed to provide something actionable. The actionable idea was all right, what tokens can be valuable based on some level of adoption. Some of the early reports we look, we worked on, well, there’s a maker one. Part of that was because there’s a pretty, explicit value accrual model, whether or not, it’s the best one, like, we can certainly disagree there. The idea was the stable coin component was really interesting. And, and we knew that, these would have to be massive. Kind of came the idea of a decentralized stable coin or one that’s not really, controlled by whether it’s a centralized entity or custody somewhere by, a centralized entity. 

10:52 • Yan
So, the idea was we started looking to make her, we got into so looked into dye and that was when, we really started to kind of explore these models, make her obviously, holds his peg very well, but, there are some kind of, 

11:06 • Jose
I remember actually when we first met in like, well, when we first met in real life in 2019, this was one of the things we spent some time talking about, right. Maker, like, w we all liked the value capture of it. Cause it’s token econ people. It was like one of the only tokens that had like fundamentals, but the problem was always like the demand for credit is like completely uncorrelated with the demand for the stable coin. Right, right. It’s very hard to grow the stable coin supply because you have to like grow demand for. 

11:32 • Yan
Yeah, exactly. It’s an, it’s an output. Like the product is credit. The part of what’s needed to make the product work is the stable coin. It’s. 

11:40 • Jose
Like an offshoot right. Of the credit. 

11:41 • Yan
Process. Yeah. It, it’s done a great job of holding its peg, but the, some of the issues there are kind of some of the collateral that’s used to back it w namely being kind of, USDC the idea that, realistically these wallets and these tokens can be frozen. And so it’s not no, truly decentralize. 

12:00 • Jose
And there’s worse. There’s like, security’s right there as well. There’s like, off-chain, assets as well. Like it’s not a fully decentralized, like censorship resistance call it. I think part of that is the problem that we identified. It’s just difficult to grow credit. You need to expand the collateral that you accept. As you accept, like sensible collateral, you kind of end up inheriting the problems of the centralized collateral that you’re trying to have the centralized stable coins that you’re trying to replace in the first place. Yeah. 

12:25 • Yan
Yeah, exactly. So, as time passed, and other, and, there’s been a lot like that was obviously 2019, then we kind of gradually got into, algo stable coins. And, and the idea there was now you actually have like a truly decentralized, token where there isn’t any collateral. And, and so I think that helped us, the kind of the upside there was really interesting because you were kind of creating as many incentives and mechanisms to try and get this peg to actually hold, which is, a bit of a fun game in and of itself. But, it was also like a really useful learning experience because we started to realize what makes some of these work and some of these doesn’t, some of these not working. So, with the, going back to the maker example, the supply was a product of demand for leverage. With, with the ESD, we kind of realized, and just those kinds of algos stables in general, realize that there’s like, there’s, it’s tough to do these two things simultaneously. 

13:23 • Yan
One it’s kind of incentivize, supply expansion while also maintaining collateral. If you need to offer up some type of incentives to expand supply, which is what we saw with USD in terms of like the reflexivity, the kind of the sacrificial lamb is that you can’t, you won’t have it, can’t be fully collateralized because you have to offer up some of this upside. That usually at the expense of collateralization. And, and so the other thing was, it was a really good experience in terms of trying to work through the issues, but part of it, you kind of noticed the, I guess, the issues with working with the decentralized organization and kind of pushing certain things through and kind of getting, 

14:03 • Jose
Yeah. Well, for me, the big thing, I think you explained it really well for me, the big thing is like, you can use Ponzi incentives to grow because as they would call need supply and liquidity. Right. So, yeah, liquidity mining, like ESD and stuff was like, okay, let’s use liquidity mining to bootstrap this like a stable coin supply and like the liquidity for it. The problem is that the supplier is like, it’s a liability of the share token, right? So you can use these funds incentives to like bootstrap supply, but at a certain point, once the Ponzi incentive stop or like animal spirits turn, you have the situation where that there’s like a huge, outstanding supply of stable coin, which is the liability of the share token. There’s no like actual non-speculative demand to that stays. Right. You have like this huge supply, stable coin with no demand for it. 

14:46 • Jose
The only thing that can happen is like, you redeem it for the share token and you get that spiral situation. Yeah. I think we spent a lot of time on the incentive side. I do think our, one of our main conclusions was you need like, demand for these things, right? Like you, like, you can do what you want on the incentives, but eventually you need some people to actually be using it as a stable coin because you need that demand. That’s that stays there no matter what the cycle is, no matter what the incentives are. Right. Right. That’s, I think what meant that when we came across Luna, after having the experience of yeah. Losing our shirts on, and yesterday we came across Luna, were like, this is, this is the approach that makes sense. Right. Like the mechanism for first of all makes sense, like they have smart mechanism design, but the biggest thing is he’s actually trying to, like those actions and Tara is actually trying to create demand for UST that’s uncorrelated to just like Ponzi incentives on the share token. 

15:36 • Jose

15:37 • Yan
Right. Whereas like, yeah, w with the, like a flight to quality is actually getting out of the token, we’re here, ideally, you want flight to quality to be into the stable. And, and part of it is just getting it ingrained in really building out the kind of the products around it. Okay. 

15:52 • Jose
Yeah, exactly. So, yeah, I think at that point it was like, or this year, we ape Luna right. With the, with the fund after the report. And, and I think one thing that’s important to say about Luna is like, it isn’t an experiment, right? Like all of us, building on it, like, and we’re heavily invested both building and investing. Like, I’ll go, stables are an experiment, but for us, it’s like the upside is so large in terms of a decentralized stable coin. I think there’s all sorts of tailwinds that we can, we’ve discussed in reports and stuff. We’re not gonna kind of go over that too much here, but like, CBCs and centralized, stable corn regulation, which I think is inevitable, that is like tailwinds for Luna. We just see, although there’s a with algo stables, there’s this higher risk of bank run and kind of an insolvency. 

16:38 • Jose
There’s also like a huge upside, which is having this capital efficient, stable coin, which if you can actually build demand for this decentralized stable coin, at some point it becomes like countercyclical, right. Which is like the magic situation where, as there’s a bear market in crypto, which causes the price of like Luna, the share token to go down, at bear market and crypto is just like a bull market and stable coins. Right. So there’s more UST team. There’s more demand for UST which leads to more Luna Brian’s, which kind of makes the counter cyclical. That’s when he gets to that like magic point, right. Where, where you have like super hyper capital-efficient, fully decentralized censorship resistant, stable coin, which is like the holy grail. That’s like, the upside for us is so big in terms of like USD T supplies is like 90 bill, right? Like stable coin market is arguably the biggest market in crypto. 

17:24 • Jose
Our UST supply is still only three bills that we kind of accept the potential downside risk. So, yeah, we, so we took a position and we looked around loony ecosystem. We saw some cool stuff. I think mirror really helped us get the potential of this thing, but we realized there was like some core primitives missing, right. There was a, it was missing a good AMM, like, sorry, like Terra swap kind of sucks. It was missing a generalized money market in that, anchor is a very specific kind of money market optimized for a certain thing, and specifically like a credit protocol. We thought instead of like waiting around and to kind of invest in these things, we can just like contribute to help actually making them happen and kind of building them. That’s when, what we started doing with Mars and Astro port, yeah. Anything you want to stay on that for. 

18:12 • Yan
Luna yet, part of it is as this platform grows you, I think the kind of the risk is also diminished, as more and more platforms or, more products are built out and there’s more utility built around it. You kind of really reduced it to downside risk and the unwind and, seeing kind of it surviving and kind of, thrive after the may sell off was really, I think important in terms of, validating whether or not because everyone understood, this can do well in a bull market, but the risk with all of these is what happens in the sell off. I think a lot of people gain a lot of confidence after watching, how that played out in the fact that it survived considering, the drawdown that you saw across the board. That, it just further bolstered our conviction cause were already building on it. But, it just kind of helped us get even more excited. 

19:03 • Yan

19:03 • Jose
Yeah. We had a lot of ideas about how to improve both, money markets and AMMS. We started with Mars, which was a cause that was the we’ll start with that. Cause chronologically, that was first the two in the core innovations on Mars, our smart contract lending, which is kind of a new primitive because it enables trustless lending in that you don’t need to know anything about the person or what they’re going to do other than what’s written in the smart contract in order to lend to it. The second thing is tokenized collateral, which is as more and more collateral is tokenized more and more of it can be put into smart contracts with like liquidation conditions, which enables you to do kind of more and more different types of trustless lending. I think, I’m going to run your land. I think a lot of it’s fine. 

19:45 • Jose
Uncollateralized lending is like, the big thing in defy is like, it’s not capital efficient people say. The idea is you have to move to this like uncollateralized lending, right? So people do stuff like DAO credit histories or, like, and people doing really cool stuff with it, like, Carmen and arc X’s identity score and stuff. From our perspective, there’s like another alternative, which is, smart contract lending and tokenizing everything, right? Because it’s a claim. The, if you’re talking about the more stuff is tokenized, when you do smart contract lending, it’s very easy for someone to just write down what they’re going to do. If there’s tokenized, collateral and liquidation conditions, you can be sure that, the lending is basically trustless as much as it would be in a money market. If you think about like a traditional money market, how it works, like something like, are they, or compound, or many of the ones in other chains, your customer is the borrower, but actually you can only be a borrower if you’re also a depositor, right? So you’re like limiting the customer base and the demand side that can actually borrow, whereas on something like Mars, and I think the first example of this in defy, someone doing it, the big innovation was kind of alpha would leverage the farming. 

20:53 • Jose
This is kind of a generalization of that is you have, the supply side one side that come in and deposit capital. Like, in Mars’s case, UST, and then on the other side, you have farmers who come in and they just want a farm and they want it, they want to get leveraged, exposure to the stuff they’re farming. They’re borrowing UST from the money market and using the farming token as collateral. To go through an example, you could have like a mirror UST farm, right. Someone can come in the deposit mirror, like w a hundred dollars worth of mirror, and they borrow a hundred dollars worth of stable coin. Now they have they’re farming on two X, the notional and $200. There are only two X the yield. At the same time, the contractors has a liquidation condition where if mayor drops to a certain amount, the Lord’s somatically get liquidated. 

21:35 • Jose
Right. Which just means that the LP share gets unwrapped. The UST gets used to pay back their debt, and then the mirror gets sold in order to pay back the remainder of their debt. The, and the Delta gets handed back to them. You end up having the situation where, the person farming can actually get leverage, right. They can actually get like two X leverage on mirror, but on something like the M assets on Terra, which w which is something that we won’t support to start with, but others can build on top, you’ll be able to get much higher leverage, right. Cause something like gold is like a few standard deviations, less volatile than crypto assets. You’ll be able to kind of get much higher leverage, which makes the lending the borrowing much more efficient as well, because you’re not just relying on this deposit or demand for, to borrow you’re relying on actual demand from farmers. 

22:20 • Jose
And that’s just the first use case. You can, you can extend this to yield aggregation, right. You could do, anchor farming strategy, for example, and just, and lever that, right. You could be depositing B Luna into anchor farming, a and C, and then you could just use the balloon as collateral to borrow at UST by more be Luna and do the whole strategy like levered. Right. And, and you can add leverage to basically anything that can be put into a smart contract. That has tokenized collateral, which is kind of those two conditions that I mentioned before. Do you want to yeah. Anything, I don’t think they’re on the explanation side. No, no. I think that was really good. Yeah, that was okay. Yeah. That’s, I think it’s like, it’s a pretty cool thing. We’re already seeing a lot of projects that are, want to build on top of this like generalized kind of credit line. 

23:03 • Jose
I think one of the really key things with doing this, and one of the big innovations of Mars and I spent a lot of time on is also that the token economics, right. Which is obviously our specialty and something we spend a lot of time on. One thing that we really liked about RVs is this safety module idea. Like, I think makers was kind of the overview of it actually, which is just the idea that if token holders are making risk decisions that end up affecting third parties, right? Like the users of dye and the users of the protocol, they should have some skin in the game for the decisions they make, right. Because token holders get the upside so that they should also have to suffer the downside of kind of bad decisions they make. We really wanted to build that in. So with maker, that’s the mint, right. 

23:41 • Jose
Which ultimately means that, the token backs the, any bad debt in the system with our beds, it’s the safety module. Yeah. But with. 

23:49 • Yan
Bad, bad risk decisions, basically for maker, or, allowing a certain type of collateral that might be too volatile or miss judging the level of overcollateralization, you should enable for a certain type of collateral based on his volatility where, if it dumps too quickly and people can’t be made whole you’re on the hook. 

24:06 • Jose
Exactly. It’s like almost a trade-off right between capital efficiency and risk. Like you want to be as capital efficient, as possible, give people as much leverage as you can. As good terms on the loan, as long as you can, without going so far that in a volatility, you end up with like bad debt in the system. So, with what we wanted to do was kind of tight make it so that the people have the upside in the system are also the ones who are taking the risk directly. Right. Because with it’s like really good, but it’s like indirect because not everyone has a stake in the safety module in order to vote. You can have like token holders who aren’t staking, who have a vote, but who actually like, can have less skin in the game than someone who’s taken the safety module and has a 14 day cool-down. 

24:48 • Jose
Right. They can sell earlier or whatever it might be. Also the mint again, is like, can get front run. Right. We, we want it to have this architecture where, do you want to talk through X Mars or shall I, 

24:59 • Yan
You can talk to that tomorrow and then I’ll go into kind of the next situation with Astro and cool. Yeah. 

25:04 • Jose
Yeah. So, yeah. W where Mara, so you stake your Mars into this pool, which is the X Mars pool. Like you can think of it as similar to the X sushi. Up to 30% of that, Mars can get slashed in case of a shortfall event, which is an event of bad debt in the system, and only X Mars holders can vote. That’s a key distinction where if you just hold Mars and you’re not staking, you’re not showing skin in the game, you both don’t get the upside. So you don’t collect any fees. Also you don’t get to vote because we want only people with skin in the game to vote. This is really important in a system like Mars, where there’s these credit lines, because, there’s a lot of risk, right. In terms of, you need to make sure that you’re auditing the smart contract that you’re giving credit lines to. 

25:45 • Jose
You need to make sure that the collateral is high quality, that it’s not that, the risk parameters that are coded into the collateral are high quality. That’s why we spend a lot of time, labs like helping advise Mars on this in terms of building out a risk team, building out a risk framework for both, which has never been done before, right. Because of a pioneered like their risk grammar. Cause it’s super dope, was a huge inspiration for us for the money market side. Looking at different assets, grading them on their counterparty risk on their economic risk. And, and figuring out like what the risk parameters should be for an asset given that, and we’ve not had to do the same thing, but on the leverage yield farming side. Right. Figuring out how much leverage can we extend to this asset? how volatile is this asset? How liquid is it? Is there a counterparty risk? How much exposure do we want to have to it? And so we’ve kind of put that together as a guide for governors to be able to vote on that stuff. 

26:33 • Jose
Would you want to yeah. 

26:34 • Yan
Like basically in order to kind of facilitate some of the functionality on Mars, you really needed a, a really strong AMM, which didn’t necessarily exist on Tara at the time. So, there were actually, like, I’d say a combination of factors that kind of led to, the building of Ash report. And, part of it is we’ve always had an interest in AMMS, both in terms of just like interesting products, going back to kind of unit swap and Kyber, and then, the first ones that really came about. So, there’s a lot of interest there AMMS are massively important to just facilitate any kind of, or DEXs in general are really important for, all off-chain transactions. So, it was something we’ve been tracking for quite some time, something, 

27:21 • Jose
It goes worrying for us, we needed to up articles for the leverage, do farming, especially, and yeah, we needed enough liquidity for liquidations, like, 

27:29 • Yan
Yeah. Yeah. And, and so that didn’t necessarily exist. And, we spent a lot of time, both like researching, and kind of, working on indexes. So, for us, it really felt like something we wanted to spend quite some time on. I think Dex is part of w the existing, what was missing with existing designs is really aligning, token ownership with the users of the product. We, we had a lot of ideas in terms of what we liked with existing ones, what we didn’t. And, and so that was kind of the Genesis of Astro port where, the G. 

28:06 • Jose
It worth just before we go into Astroport, it’s worth talking through the dynamic interest rate model from Mars. Do you wanna, yeah, you can, yeah, go ahead. It’s just going to cut the, yeah, but w just quickly, I guess the dynamic, cause that’s the last innovation tomorrow is I guess it’s like the first one is generalized credit protocol. Second one is kind of the token economy. The third one is I think dynamic interest rate model, which is just, if you think about how interest rates are calculated in traditional money markets, you have like this utilization curve, right? Because, you want to punish, like you want to make sure that the pool is utilized so that the positives are getting a return, but you don’t want super high utilization because at least the illiquidity, right. If it’s a hundred percent utilization, it just means that the pool is illiquid. 

28:44 • Jose
Like it’s solvent because there’s enough like assets to cover the liabilities, but it’s a liquid and that no one can withdraw. What it does is you target an optimal utilization, right? Which varies based on the riskiness of the asset. You have a curve that kind of looks like this, right? It’s like kinked. At the optimal utilization, the interest rate, like sloped sharply upwards in order to punish utilization at those high levels, which encourages people to pay back their debt. They’re not paying high interest or more depositers to come in and supply and collect the interest rate. Now that model works well. It’s worked super well for fraud and combined and stuff, but it also has some bugs in that it kind of doesn’t react to like external market conditions. Right. Like one, there’s been a few examples where we’ve seen this, or like a hundred percent utilization has also pools with very low utilization. 

29:28 • Jose
What w what happens is for example, when alpha launched, like V2 and they had super high yields on their purpose, STC pools and stuff, you had a situation where Cream was a hundred percent utilized. And, because you could such high yields on alpha. Like alpha yields were like 500%. You, people were willing to pay like a hundred percent on their debt on cream, because they could earn that yield on alpha. You just had like this a hundred percent utilization. What we wanted to do instead is have this dynamic interest rate model, which just, it uses control theory and a PID controller actually, which we wrote a report about it. You can check on our porch. I think we’ve actually opened source that one as well, but the idea is it just targets an optimal utilization and the interest rate keeps shifting every time period, which right now is every block in order to target that optimum utilization. 

30:16 • Jose
It means that if, there’s no like, alright, it’s at a hundred percent utilization, the interest rate is 70% because if you thought it sections a hundred percent and the optimal is 90, the interest rate will keep going up until utilization goes down to 90 and vice versa on the way down. We think this will end up being a lot more reactive to market conditions and a lot more accurate in terms of, being, yeah. Being able to just reflect like actual supply and demand. Yeah. Those are the three main things about by Mars, but yeah, we’d love to go into go and task report a bit now. Yeah. So, 

30:47 • Yan
In order to really enable what was needed on Mars, we asked report had to exist. There were a combination of factors that really, led to us getting involved on in helping build Astro port. And, we’ve always had a lot of interest indexes going back to, Kyber swap when they first came out in the, some of the earliest reports. And, just in terms of like their design, their utility, they’re kind of, they’re massively important primitive for any of these decentralized finance. So, it was an area we’ve always, followed closely and, looked at a lot of different token design models. You know, there’s certain components. We liked certain we didn’t, and, this was the kind of an ideal way for us to implement what we thought would be a bit of a best in class, like token econ design on top of a really good, DEX. 

31:37 • Yan
I think one of the bigger gaps that you see with DEXs is that you don’t really have a lot of alignment between, users of the decks and token holders. So, the token accrues value in some indirect way. You said you were gonna kill these as a problem. 

31:54 • Jose
With bees. Yeah. 

31:55 • Yan
You’re going to kill the bee and we’ll get canceled. And then, 

32:03 • Jose
We don’t kill bees at five level bees and all animals we love. 

32:08 • Yan
Yeah, it was a really good opportunity for us to kind of implement, some ideas that we had. So, part of it is you see a disconnect where there’s some value accrual the token, but typically the token that the value of cruel is basically, a way. So, the token ends up being used to incentivize a lot of liquidity and bring liquidity to a platform. The value of cruel is what allows the token to continue to function as this incentive mechanism. That ends up being kind of the design where, portion of the fees goes to the token. The token issued as a way to incentivize additional kind of liquidity, which would in theory, lead to more volume because there’s better execution. That’s kind of how the flywheel works, but there’s a bit of a, value cruel and use case for the token, I think are a bit removed where we think there’s a better way for users of the platform to interact with the platform using the token, rather than disconnect between like token holders and LPs. 

33:06 • Yan
Right. Right. The idea here is, you’ll have the kind of ex sushi model, where, so, if our pool has 30 bips, 20 bips, we’ll go to the LPs instead of 25. The idea is five bips goes to, the standard kind of ex Astro model. There’s also going to be like a VX Astro. Basically they explained the standard X Astro for people to donate. Right? Yeah. The idea is, every transaction costs 30 bips, and that model five bips goes to this pool. What happens is you deposit, Astro into this pool, you get X Astro. The idea is, let’s say, one X Astro was deposited. And, over time during the year, there was like 50% worth of fees. The idea is your one X Astro is now worth one and a half Astro. What happens is that fee is used to purchase a Astro off the open market and deposit into this pool. 

34:03 • Yan
Over time, your ex Astro is worth more Astro. It’s just like a product of kind of the buyback from that five basis point fee. That’s a bit of the standard model, but ideally you get, another level of kind of interaction. You can take this X Astro and you lock it up. The idea is you lock it up in, and based on how long you lock you get a certain amount of VX Astro, very similar to the kind of the curve model. The idea is you deposit, one X Astro and you get a voting share basically. The quantity of your voting shares as a product of, as a function of how long you lock up the X Astro for. This gives, those who have more, skin in the game in the sense that they’re locked up for longer, have more voting power. The thing is, all right, so what are you actually voting for? And so the idea here is we have, and then also, so the standard design is 30 basis point fee five basis points goes to X Astro, and then, 20 basis points goes to LPs and then the other five basis points goes to this voting ex Astro. 

35:06 • Yan
You get an additional fee and you get, the voting rights. What are the voting rights really for? And the idea is, you, we will have like a, basically a kind of a list and that’ll be, grown over time as a function of kind of the amount of projects on Terra that grow. The idea is you can vote yourself similar to curve, inflationary rewards, and rather than doing a pool your pool, right? So if you’re a token holder of a certain project you’re in and you’re providing liquidity, there’s a lot of incentive for you to actually own. If you’re providing liquidity on Astro, there’s incentive for you to own Astro and vote with it because you can increase the kind of rewards that you’re getting for being a liquidity provider. This really gets all of the projects on, Tara to be, natural X Astro holders, or, VX Astro holders as well. 

35:55 • Yan
But, to, in order to vote and kind of direct rewards to themselves, it means that. 

36:00 • Jose
Anyone that’s an LP or, is farming up here, whatever has an interest in holding like Astro, right. In order to maximize their rewards, not just for the issuance, but also maybe the boosters, which you can talk about. There a B again, I’m just a fly. Did the sound the. 

36:16 • Yan

36:17 • Jose

36:17 • Yan
Not going to hurt you that, , 

36:18 • Jose
It’s just your hair. Look at that. 

36:26 • Yan
We w what do you want Boothies, we already kind of talking about though. 

36:29 • Jose
Yeah, no, you talked about the issuance, but not the boosters, did you, 

36:34 • Yan
You’re locking up for wars and like, that’s the boost, these, 

36:36 • Jose
Yeah. You’re looking for rewards to get more issuance to the pool. Also you get like, basically a fee discount. Right. Cause you get more. Yeah. I mean, it’s like two different mechanisms to get the more work you can. 

36:44 • Yan
Do that one then. 

36:45 • Jose
Yeah. The, the idea is that you’re incentivized to lock up in order to, if you’re an LP as a sense of that, still a cup, cause it’s the way to maximize your rewards. Right. I think the flywheel, yeah, we can maybe talk about the, we don’t need to go into the boosters too much cause we’ve been in the weeds, but the general idea is that if you’re an LP in order to maximize your rewards from LP-ing, you need to be a token holder and ideally max lock, right? That’s the way to maximize your words at the same time. If you were a token holder, the way to maximize your return is to LP because you’re not making the best use of your token. Cause you’re not getting the issuance and you’re not getting the boosties. You create the situation where your LPs are incentivized to become token holders and vice versa, which just makes sense. 

37:24 • Jose
First of all, from a governance perspective, you want the people that govern the protocol to be the users and in the, in a, in the case of an AMM, the users or the liquidity providers. Right. It also makes sense just from a skin in the game perspective, it becomes much harder to vampire attack. Right? you can’t vampire, you can’t vampire attack liquidity from a protocol where all the LPs are also like token holders locked for four years, right there, they have too much skin in the game. I think that’s what you see with curve, is that it’s been very difficult to vampire attack. It’s also like amassed a lot of liquidity. While it’s not a token, that’s very widely held by like pure investors because of the high issuance. Because of the fact that it did just don’t get the most benefit from it, the people who, the strong hands and the big LPs who need to own the token, they all own it. 

38:05 • Jose
Their max locking, it’s working as intended. Right. Right. 

38:08 • Yan
The other kind of component we’re adding as well as, basically kind of creating a, a small list and that list will kind of expand through governance and as more projects launch on Terra. The idea is, it’s like your boost is a product of, or rather is conditional on how much you kind of deposit, but rather than it being purely, your share of vote means like reflects your share boost. We have this kind of a, basically like a top, so it’ll start as a top eight. The idea is, if you’re, if it’s ranked based on vote and then, so if you’re ninth, you effectively get no votes or no known inflationary rewards, but if you’re an eighth, then you have that model where it’s your share of the vote is your share of the, boosts and, or your share of the yeah, your share of the vote is your share of the inflationary rewards. 

39:00 • Yan
We think this ends up creating a lot more competition on the lower end of the kind of the list. And, and that’s important because you got these communities are very tribalistic and are always trying to, do what’s best for their token. This creates a really come kind of competitive situation where if you just allowed, 7, 8, 9, and 10 to all get rewards, and that’s just like a function of what they get. If, you’re getting, if you’re two and a half percent versus 2%, there’s not a lot of incentive to try and get above that two and a half percent because, it’s incremental. It’s kind of like scarcity. The cutoff means now if you’re at the bottom and you start creating a lot of competition, and so you can get votes by not locking up for, a long period of time, but obviously you won’t get too many. 

39:46 • Yan
We kind of envisioned that initially you’ll start get some people that are, trying to test it out and won’t lock it up for much. Once they see what’s really happening and the level of rewards they’re going to get, there’s a lot more incentive to lock up for, a significant portion of time. We’re excited about the engagement there and that kind of leads into, the launch strategy as well. And, and, we think that those are massively important in terms of, launch strategy really matters in the long run, a poor one can set yourself up to can just create a lot of unnecessary difficulty and kind of headache to getting a protocol, functioning well. There were, like kind of a couple of things that we really, wanted to when we kind of came up with a design to incorporate. There’s kind of a, a legal component where, we’ll get into that as well. 

40:32 • Yan
It’s a strategy that we think can be a bit more broadly applied and it kind of happens in multiple phases. It’s the second phase that we think is. 

40:43 • Jose
Maybe go into like the design considerations for it a bit for us, like a bit more like how we landed at this. Cause it kind of, it was a bit of a process, right? Like, I think there’s some problems that we have with tokens right now. Like, like John said, we think the token launch is really important. Not because in like public sales and crypto are pretty important, not because like projects really need money because no one needs money right now it’s a bull market. But, mainly because of the distribution, right? Like your cap table and crypto ends up being a big moat. That’s like why the curve model is so good because your LPs end up being your token holders. You want to make sure that you distribute your tokens to the right people. Also that you have, like one thing that we’re kind of, that we think is a problem right now is you need to have enough of a float when you launch to absorb like demand, because I think demand below a certain point is just kind of inelastic. 

41:28 • Jose
If you have a very low float, it just ends up leading to like an inflated for that loot evaluation. Just like token slowly bleeds out over time as unlocks happen and the kind of supply heads, the market, and that’s really bad for community building, right? You want your community to kind of succeed with you and to be involved from the beginning. In order to do that, you need to have like, an afloat for that to be the, 

41:47 • Yan
Part of it is if your token design to have a lot of interaction with its users, then you also just need to flow in order to facilitate a larger community and a lot more kind of engagement. So, w w we’ll kind of go into the phases, but the idea is, again, who are the token holders who are, who would be the ideal token holders here? It’s the people that are providing liquidity, the LPs. In our design, it basically it happens in two phases so that. 

42:15 • Jose
The latter of them, right? Like, that’s the key bit like, yeah, the ideal users are LPs, but you also want to wait for like other people that aren’t LPs to be able to, acquire the token. That’s not gameable right. So that’s like a few issues. Sorry to keep it. I know you want to go into the mechanism, but like, just a bit more context, like the, like 1, 1, 1, 1 thing we could do is we could just hand it out to suppliers, right? We could just hand it out to liquidity providers and just like, wait for an M to form and stuff. There’s a reason we didn’t do that. Right. Because first of all, you end up with a really low float. If you just do that, right. If you’re just slowly handed out, hadn’t handed out to liquidity providers, you have to start with a really low float necessarily. 

42:49 • Jose
Then, second of all, if you don’t initialize the AMM yourself, you end up with these issues where liquidity is really thin. First of all, there’s no way to find a price, right? Someone has to start the AMM at a certain price. It’s almost going to be wrong. They’re going to get wrecked. It’s going to be thin liquidity for awhile. Like people are going to get wrecked. So we wanted to avoid that. Same time we can do an, like auctions are tough because of like regulatory issues doing public sales, even if you like take the auction proceeds and put them in the DAO, it’s still like the effectively, the inside is control the doubt to begin with. It still looks very much like a public sale. So, 

43:26 • Yan
And, and VNL BP. The, the risk is that you’re still setting the initial price, even if there’s a lot of price discovery. We don’t have to necessarily go into VPs, but, 

43:35 • Jose
These five things we want to sort right, there’s like distribution want to get it in enough people. We want to have price discovery, so want to have some method for supply and demand to come together and set a price without some guy having to like initialize an Imam and get wrecked. We want it to have a sufficient float, just like enough float to absorb demand. We want it to have like liquidity from the get-go so that there isn’t those volatile time of low liquidity. We wanted to minimize like regulatory issues and that, 

43:59 • Yan
Yeah. Some of those are solved through the generalizable, second phase. Then, the first half, it ends up being more protocol specific because that’s where the distribution to, your key product users kind of takes place. And so we, 

44:15 • Jose
We think this is a token, launch mechanism. That’s going to be pretty well. We’re pretty happy with it. I think we had to design around a lot of things we can, yeah, we can get into like particularly. 

44:24 • Yan
The second day, the second half. So, the initial launch strategy is basically, you have, who, the idea is how do we get the tokens into the hands of the LPs? And there’s going to be like this, basically a large migration from Terra swap into Astro report. The ideas LPs can take their capital deposited into Astro port for a certain amount of time. In response in return, they get Astro port tokens. Now you immediately give tokens to the communities that are going to be the ones using those tokens to give of themselves liquidity. Okay. 

44:59 • Jose
We’re doing 75% of supplies so that we get a lot of flood out there. It’s not like they get it on an ongoing basis. It’s like, think about an airdrop is like, you get rewarded for past behavior, right? You use this app, you get rewarded. I locked her up, which is what we’re calling. This is you get rewarded for future behavior for committing to be a future user of the protocol for certain, 

45:15 • Yan
And your liquidity is locked up. It’s not as if there’s, it’s a, a temporary or mercenary capital situation. At the same time, now there are token holders. With the evidence endive to lock up these newly found tokens, to some extent to kind of benefit the locked up liquidity they have. Realistically, what we want it to offer. 

45:34 • Jose
Was like a week long period. Right. Rick lung period, anyone can come in and deposit. Right. So shares and then yeah. 

45:39 • Yan
Yeah. You have, everyone gets kind of a pro rata share based on kind of their, the deposit length and deposit amount. The next phase is basically, so that happens in the first week and that’s again, to get distribution to your token holders. We think, that part will be, I guess, protocol specific where, it just makes more sense in certain situations to have, yeah, it’ll vary. The distribution will kind of vary based on the protocol and who you want to give it to. It’s in the second phase, the idea is so, we’ll have two and a half percent will be airdropped to, Luna stakers. 75% will be to the individuals who provide liquidity. Now you have this 10% and how do you distribute it? How do you basically enable both prices? So this is where you deal with price discovery and, avoiding low liquidity and avoiding setting a price yourself from the regulatory side. 

46:28 • Yan
So, part of it is you can have this LBP idea, but again, there’s a lot of kind of gaps. They’re trying to go from the initial distribution to something. 

46:37 • Jose
Like that. You need some hard coding, right? Know, it’s a hard code, like a price where it’s, 

46:41 • Yan
It’s like, there’s a lot of differences engaged in plus, you have to put in a certain amount of capital yourself. That kind of issue, and that’s where the legal side comes into play. So here are the idea is, 

46:51 • Jose
Were trying to find ways to like, how do we decentralize it so that like, it’s an LVP that can be initialized by the community, right? Like a self run, like bottom up LEP in there. 

46:59 • Yan
The idea is one side, the individuals who received the airdrop can either choose to, hold, it can choose to, w with the idea of maybe selling it or with the idea of participating in a protocol. There’s some people that can take that the airdrop they received and basically seed what will be the pool too. The way that happens is so one hand, everyone who wants to participate, deposits a bunch of Astro. On the other hand, you have the individuals that maybe didn’t participate in the airdrop, or would like to get more astrobiology, 

47:33 • Jose
Not like, cause like liquidity providers are definitely the heart of anyone, but like community is a, is makes up a lot of different people, right? Like people who are just passionate about it, like growth people who are talking about it, who were writing explainer videos and like all those people want a way to have skin in the game without getting wrecked. It makes sense to have some way for non LPs to also be able to treat. 

47:52 • Yan
The projects that are looking to launch as well. There’s a lot of value in owning a share of this voting asset that can, you can direct use to direct, rewards. The idea is, similar kind of designs like this have happened before with, the most notable being probably mango and the issue that happened there was, so the way that the design would have happened is, so, one side you have people deposit the, Astro token. On the other side, you have individuals who will deposit UST and the kind of the price that settles is basically the amount of Asher that’s deposited is one side, the amount of USDA deposit on the other, the more USC that’s deposited since the 50 pool basically prices Astro higher. So, if there’s 100 Astro and 200 UST, that means each Astro’s with two bucks, if there’s 300 UST, each as was worth $3, so on and so forth. 

48:43 • Yan
The price discovery is basically a product of how much UST gets deposited and, something like this happened with mango, but the gap there was, because you can deposit a bunch of this liquidity, you can effectively inflate the. 

49:00 • Jose
Microwave to liquidity pool right there weren’t, it was just like an. 

49:02 • Yan
Auction. It was just an auction. So, yeah, the kind of the, benefit that we’re adding there is like, there’s was an auction ours. This will once the settles we’ll seed pool two. Going back to kind of the issue with mango is basically, you, if you deposit a lot of USDC, I think they used, what happens is, the more you deposit the higher, the price of mango. You had a few large players, if they were looking to deposit, X, they instead deposited five X, and what happens then is mango looks, the price of mango looks massively. It just looks very high, which discourages a lot of would be depositers because they’re not comfortable paying that price. And they can also be. 

49:44 • Jose
We’ll pull out their liquidity. Right. Okay, sir. 

49:49 • Yan
The idea is, people can deposit, USDC and as a result, like, yeah, the more they deposit the higher, the effective price, and what happens is you kind of, for, would be deposited, or they’re just less incentivized to deposit at that point, because, the price is a lot higher than they want it to be. And, maybe the expectation is that it’ll kind of sell off, but the problem is, the people that are depositing the deposit, if they actually wanted deposit X, now the deposit is five X. The idea is before the auction ends, they’re going to withdraw Forex. And, and the whole idea is we’re just going to scare off a lot of scare off. A lot of would be depositors, withdraw our liquidity, and then realistically get a much better settling price. And, and like, this was kind of the objective going in, and there’s no reason they shouldn’t do it because the mechanism allows it. 

50:36 • Yan
Like, that’s the kind of the thought process. A lot of our design is always, if it can be broken, it most likely will be. So, like, we have to say, sit there and think through how as well. Right? Like how people will like yeah. Basically cancel and withdraw. Agreed. Yeah. I guess other depositors that might’ve deposited, we’ll also kind of withdraw, but there’s a few kind of simple ways to solve that. So, the issue is, again, people can deposit a bunch and withdraw, and the last block with no recourse and the, like the whole idea here is you want to know, 

51:08 • Jose
You deposit a bunch, right? Everyone else did not deposit or withdraw theirs. You withdraw yours in the last block and give yourself a much better price, right? 

51:14 • Yan
Like the whole, this whole, the whole point of this mechanism is for price discovery. And, this, you’re not accomplishing that at all because it’s a very ineffective way to actually find what the market perceives the price to because you can really easily manipulate it. For us, the idea was, so how do you start to address that? Well, on the tail end, you can kind of curb the amount of withdrawals you can do as time expands. Early on, you afford a lot of flexibility on the deposit and withdrawal side, and then as time passes, you limit the level of withdrawals and it gradually declines over time. That as you get closer to the finish line, you’re setting up for a much more, the, basically the ending price volatility is much lower where, the amount you can withdraw is limited. That also has like the knock on effect of if I can only, potentially deposit X or deposit two X with the idea with the goal of actually only depositing X, my, the impact of my kind of manipulation is also, diminished. 

52:12 • Yan
So, like, it’s, it just becomes a much lower Evie play to manipulate. So, then you have the direct and the knock on effect where it’s just not as attractive to game. Perhaps they’re just being less natural gaming of it. So, this will ideally, create a much more effective form of price discovery. As additional incentive will provide a 1% of the supply as a reward for those who want to participate. Just going back to that, people’s, deposit Astro people deposit UST, price settles. This, that liquidity is used to seed pool too. That’s going to be locked for, three months. 

52:47 • Jose
All the participants get, rather than getting like the people who put in USC getting asked for, and the people who put on Astro getting UST, they just get LP shares, LP, shares that vest lending over three months, which means that you have instant, super deep liquidity at the price that’s discovered during this like price discovery phase. 

53:04 • Yan
Yeah. It is like if there was a hundred, 100 Astro, 100 USD deposited and nationalist price at a dollar, if you deposited 10 Astro, your exposure is now five Astro five UST. It’s a way for you to kind of realize some gains, but also, see the liquidity pool. The kind of the question becomes, well, wouldn’t, I just wait until it launches and then I can dump my entire stack. The the way to address that is we provide additional center for those who want to participate in liquidity pools. For depositing Astro or depositing UST, you there’s 1% that split them across everyone. What that means is those depositing Astro sell half of their Astro at a premium, because they’re getting additional Astro as a result. Those that are depositing, UST are buying half the Astro at a discount because they’re getting additional Astro for their, every dollar of USD spent. 

53:56 • Jose
Yeah. So, like, you can think of it the way I think of this phase one and phase two is like phase one is the distribution phase, right? You’re getting these tokens in the hands of your users. People can commit to being used as a protocol and they get this big flood of tokens up front. Phase two is the price discovery phase where instead of the project conducting a sale, it you’re actually giving users the ability to sell their tokens to other users so that there are no proceeds for the project. It’s not like a public sale. It’s just that the users themselves are committed to using the protocol, can sell their tokens to other users. You get price discovery, you get liquidity because that gets deposited into, and total liquidity pool afterwards. It kind of addresses like all those issues that we laid out like at the start. 

54:35 • Yan
It’s that second part where, deposit Astro deposit, USC, that we think is the super applicable side. The first one is where you kind of tinker a bit. We also applied this model for Mars. And, and so the idea of for Mars is again, the product is kind of leverage. What do you need in order for, people to be able to borrow USC, you need people to deposit USD. The phase one, like with Astro people deposited, into their LP tokens here, people would deposit UST and the longer you deposit UST, the more rewards you get. There’s a similar lock drop component where everyone deposits UST and Tamar’s, as a result, they get Mars tokens with that. It, you begin phase two where those who want to, hold their Mars, Candace, we want to provide liquidity and for an additional bonus, can they deposit Mars? Additional people will deposit UST to get access to the other half of the pool. 

55:31 • Yan
You kind of get the same result where you get distribution, you get price discovery, you avoid price setting. And, and like, and it’s pretty, resistant at least, we’ve tried to think of every way to kind of game it. But, I’m sure there will always be random gaps in learning experiences. So, yeah, that’s kind of the fun part of all of it is seeing it live. Because you assume, everyone there’s such a, a big bounty to figure out how to game this stuff. People are very savvy. We’re excited to see how all of this goes, 

56:05 • Jose
You can kind of see, like, we tried to kind of take you through the thinking, cause you can see all the problems that we had to deal with and how we like designed around them. Right. Whether it’s the mango markets thing, or even just the, like why we subsidize the phase two, right. Because maybe you have the situation where, okay, now everyone who wanted to buy his boat. Once the LP goes live, there’s only sellers left. You have the situation where like LPs get dumped on. You need to subsidize it and stuff like, so we’re just trying to go step by step and we kind of figure out how would we play this if were trying to exploit it, right. Like what’s the way that this is exploitable. That the optimal behavior for the protocol? If not, we need to like change around the incentives to do that. 

56:39 • Jose
We think, I think we’ve done a good job with that, but if you spot any floors message us and we’ll definitely give you a nice bounty for that and probably hire you as well. Yeah, so we have like a few, three topics did it that I thought we could potentially touch on for labs. One is like some of the other incubator projects. The other one is governance 2.0, and then like just a future vision for labs, I think other incubator project. We’re working on some cool stuff on terror. There’s going to be Levana, which is going to be, initially like a leveraged token protocol, built on top of Mars and eventually like a, a perpetuals, platform on Tara, which is going to be pretty cool. Kind of working very closely with Astro port and, Astra port V2, which will feature, like a version of concentrated liquidity. 

57:29 • Jose
We’re really excited about Levana. The founder is one of the, he was the head of, his debit stark where really smart guy, Jonathan. Yeah, I had a lot of fun with him, at the Levana at our party. Were doing, Ryan, which one, the hackathon that we organized on Terra, which is kind of right now, like basically a cross chain, anchor. Just trying to make anchor question and, front end and smart contracts rank or another chains. We’re doing like, angel protocol and Saberra which we’ll be announcing very soon, which is gonna enable kind of UST subscription payments. Our Delphi is going to be the first to trial that which should be pretty cool. And like we’re always, yeah. 

58:10 • Yan
Yeah, the goal there is to really, part of going back to what makes terror work and kind of USD valuables is adding additional use cases and just additional utility and really improving the, so like the product for Tara is UST and basically improving the ways you can use UST, whether that’s from an investing perspective or like a, just a general yield side and kind of all of this really basically creates a massive floor of support and utility. And, and so, the risk is always this kind of deep debugging, but there’s a lot of stuff really happening on that front to kind of reduce that risk. We think, continuous product growth and utilities is really important on that front. 

58:54 • Jose
Yeah. We’re doing an Everland on Solana, which we’re really excited about. We only want to build things really that, could only be built on this chain or at least are like completely unique. And, Everland’s certainly one of them, it’s like a money market aggregator, which on the deposit side kind of looks like Yearn, right? It’s just like, allocating your deposits to the money market where it earns the most yield, but on the borrow side is where it’s really interesting because it’s like rebalancing your loans, constantly turning to ensure you get the best rate, which also allows you to basically margin across protocols. Starting off on Solana, but eventually this could be kind of a cross chain effectively like a Metabank, right? The aggregates underlying banks, and make sure users are getting the best rates for it. We’re in terms of how we’re thinking about labs. 

59:35 • Jose
I think that very high level that kind of emerges is we want to be the best place for smart people to come together and build cool stuff. Right. And, we’re not really interested in scaling up the incubation model super fast and like incubating like 200 projects next year or anything like that. We’re interested in working really closely with super smart teams on projects and problems that we’re interested in. We have a really good team, about 25, including like devs quants, lawyers now with game para and some others. Our goal is just to help these projects succeed and also to help them like get these centralized. I think that’s a big focus of ours because for us, the real innovation of crypto is, when you can create these incentive systems and that work in a way that you don’t need a central authority to recreate like services provided by a central authority. 

01:00:24 • Jose
Right. That means nailing the incentive system, but on the token you can level, but it also means new governance structures. That’s something that we’re going to spend a lot of time on. Like, were super inspired by Yearn’s governance, 2.0, I think we all, we’re all a bit suspicious of the idea that like, a DAO was just going to be every single person voting on everything. I think it requires specialization in terms of having specific like committees or groups of people that are specialized and empowered to make decisions in a certain realm, and then report back to the rest of the DAO and half are accountable to the rest of the DAO in some way. We’re working on a lot of systems for that. 

01:00:58 • Yan
Yeah, I dunno. Yeah, just going back, a lot of everything kind of really boils down to incentives and that’s something I think we understand decently well and also really applies to just labs as a business as well, because we kind of the incentive for smart builders to come join us and get involved in Jesus Christ. Especially I haven’t run off yet. I I’m about to, he runs from bees normally right after you killed them right now. Don’t f*ck us a lot. He’ll get angry. He’ll bring his other friends. I think that too. Do they normally show up after you kill them or what happens? 

01:01:36 • Jose

01:01:38 • Yan
They don’t have like some sixth sense. Like you’re dead on Twitter. You’re just gonna be known as the bee killer. You canceled it. Then, so yeah, the, going back incentives are just massively important everywhere. We think, with labs as well, the model is really well-designed to partner with really strong builders and provide, massive upside for them. I kind of, an entity in a system that can really help them achieve that where, we’re, we have certain areas we’re really strong in, but we want to partner with kind of some of the smartest builders and get as involved as possible. Like, like there’s, I mentioned it’s not about launching 200 projects, it’s working really and intensely on a handful of them. We think that design is also really appealing for builders because, if you’re a builder showing up and we’re doing 200 other projects, it’s just the incremental benefit to you. 

01:02:30 • Yan
Isn’t really there. But, if it’s clear that this is just going to be priority, number one for us. We think that this model will also just help us continue to build with the best. 

01:02:40 • Jose
Yeah, exactly. Like, and when you get to like, so the way we’re thinking about labs is, there’s going to be like some stuff that we offer, like a lab so that there are services kind of old projects need, like token. Econ is something that we’re going to want to help other projects with. We have a risk and like quantitative team that’s doing like agent-based modeling stuff. We have like, a legal team that’s advising on basically like governance hacking and how to get to decentralization. Cause that looks different for every project. We have a developer team, obviously that it’s doing a lot of that work and we’re kind of going to be open sourcing. A lot of, a lot of the stuff we’ve done. We’ve done recently. We’re going to be open-sourcing LBPs on Terra. We’re going to be open-sourcing this sales structure. We want to kind of, work with teams that have their own teams and that we support and really like take a very hands-on role with helping in these areas. 

01:03:29 • Jose
Like when you join labs, you have access to not only all the labs resources, but all the like Delphi resources. Right. We’re like a seventy-five person, a hive mine here, there’s a, and so this, you get the benefit of the research analysts of like the ventures, analysts of all the specialists that we have in different areas of the deal flow of the operational stuff. I think it’s like a, it’s a pretty attractive kind of a value prop for builders. And, and for us it’s like, yeah, it’s a, it’s like a studio slash like incubator, I guess, like a, a venture studio slash incubator. Like we’re not trying to, scale too quick. We just want to do, we’re just really, we just really enjoy building, and thinking about this stuff and building stuff that’s interesting to us and that we think is gonna make a dent in the space. 

01:04:15 • Jose
I mean, that’s what we want to keep doing now, with labs and just like, I think the main limiting factor for us right now is just great founders and entrepreneurs. Like initially it was developers, but I think we found a pretty good pipeline of those. I think what we really need. If you’re listening to this and you think it’s, you please reach out is, just founders who can take a project from start to finish, right. Who are crypto native. We understand the problems, understand what the product needs to be like. Can interface with developers have built stuff before and can like take these projects from start to finish. We’re going to make a big push once we launch, these initial projects that we’re incubating, but we really want to, like now we’re going to make a push to attract more of those. We can do a few more things, but we’ll be growing very slowly. 

01:04:54 • Jose
Like we want to add projects like one by one and make sure that we’re nailing, like all of them and take like a concentrated approach, I’d say. Yeah. Super meticulous. Yeah. Yeah. Anything else? Anything else? We, we missed out on, 

01:05:09 • Yan
I dunno. I don’t think so. 

01:05:11 • Jose
I don’t think so either. Yeah. I think that’s pretty much it. Yeah, it was a labs kind of a mega session. It’s a shame, want to give a shout out to Luke, Sanders our CTO. Who’s, who’s a beast. Like, couldn’t definitely couldn’t do any of this without him and Gabe Shapiro who was coming and crushed it and all the labs team Sage and all the devs and stuff, because they’re not on camera Luke. Doesn’t like to it, doesn’t like to go on camera, but, he’s a beast and we couldn’t have done this without him. He, they were all super involved in all the stuff that we discussed today. Yeah, if you’ve got a cool idea, if you’re a, and we’re not, by the way, we’re not tearing Solana, Maxis, it’s kind of an important, point, we, a year ago, we just identified those as the highest opportunity places for us to build for the next few years. 

01:05:57 • Jose
Were looking very actively, especially at layer twos. We’re interested in stark where, if you have an idea and you’re interested in building it, we’re not going to like shoehorn you into terror or Solana. We want people to build wherever makes the most sense for their, for their project. So, 

01:06:13 • Yan
We’re interested in Starkware, if you have an idea and you’re interested in building it, we’re not going to like shoehorn you into terror or Solana. Now it is, we’re choosing a chain for a specific reason and building a product out that we like that needs to exist, and it can best executed on that chain. Every kind of decision ends up being, very particular, very specific in particular. It, yeah. The idea is, yeah, we’re happy to kind of build wherever, but those are some of the areas we’ve just spent the most time on recently. 

01:06:40 • Jose
Yeah, exactly. Yeah, we’re super excited about what we’re doing. Like, I think all of us have had, we had like chances to build in the space before, like from when we joined the space, a lot of opportunities like that, and this was the first time that we thought like, this makes sense, like now is the time I think crypto is at an inflection point, both with NFTs and we can now see the path to like mainstream adoption where in 2017 it was like, it was tough. Yeah. Yeah, yeah. 

01:07:08 • Yan
I guess the other, yeah. I don’t know if I want to go into that. It would be just the, kind of the, some of the NFT strategies and building that on top, but that could be an entirely different can of worms that we might not necessarily. 

01:07:17 • Jose
Community a different kind of work. Yeah. I mean, yeah. I think NFTs are like, I think all of us at Delphi, like super excited about NFTs personally, I’m as excited about them as I was kind of, all of crypto in 2017, but very skeptical about S most of the stuff that exists right now. I think we want to use them in new ways for these projects. In, in like interesting ways that are composable, we think like your on chain identity is going to be very much based on like, NFTs that you own, but for that to happen, they need to be earned for specific actions and they need to be composable in certain ways. Also I think they’re going to be part of like every project like loyalty strategy and retention strategy. 

01:07:55 • Yan
Yeah. Great. It’s a lot of stickiness, a lot of community, and just, a lot of clout they can be gained by achieving certain goals that give you certain access that, it doesn’t necessarily have to be like something that’s completely functional from anesthesia perspective. Cause sometimes those are tough where, owning a certain NFT, like there are ways to build it. But, the high level difficulty is that if owning this NFT gives you a discount, those are often zero sum. So, like that discount comes at the expense of either other individuals paying more or receiving less. Absolutely those are interesting, but are, can be a bit difficult to implement, whereas here, you’re getting it in response for achieving a certain task. You can kind of build on your identity where, a new protocol launches you’re the largest contributor or, you’re nearly, steak or something like that. As a result, you get some kind of, dynamic or some kind of component that you can add to your existing NFT and kind of keep upgrading. 

01:08:53 • Yan
So, yeah, that’s kind of the on chain identity where it doesn’t have to be, it’s not your name or your credit score. It’s what you’ve actually been able to achieve. 

01:08:59 • Jose
It’s kind of another tool in this token, economics like incentive, like toolbox, right. You can just use it to cause the other thing is I think, like game’s going to become your bank and also your bank is going to become a game, like everything’s going to be gamified. Right. I think NFTs were like a really important tool and we’re putting a lot of thought right now into how to use them in really cool ways to incentivize, the right, the behaviors that we want on these projects. So stay tuned for that. Yeah, I think that’s good. Yeah. Yeah. I really appreciate everyone tuning in. If you have any questions about labs, hit us up on Twitter and hope you enjoyed the podcast and, excuse the sweat and yawns irrational fear of bees, which caused a few interruptions. 

01:09:42 • Yan
And Josie’s flat hair. 

01:09:45 • Jose
I shampooed it this morning. If you look at other pictures, it doesn’t look. 

01:09:47 • Yan
As flat. It looks. 

01:09:48 • Jose
Just as flat and you should consider washing you as occasionally, Yeah. Is that. 

01:09:52 • Yan
It can’t get flatter than yours. It looks less flat on a relative basis. 

01:09:57 • Jose
It’s kind of rolling down your foreheads and just a sweat. It’s shiny, I think it’s the grease. 

01:10:03 • Yan

01:10:04 • Jose
On a relative basis. Better, better flat and clean. I say, cause then the bees don’t you didn’t see bees flying around me, but yeah. Thanks everyone for, 

01:10:11 • Yan
If you made it this far. Yeah. 

01:10:14 • Jose
We probably ruined the best hours of sun for the team and the pool, but yeah, let’s go. 

01:10:19 • Yan
Eat. Let’s go eat. Thank you.