Epicenter - Learn about Crypto, Blockchain, Ethereum, Bitcoin and Distributed Technologies - Guy Young: Ethena - USDe Synthetic Dollar via Delta-Neutral Staked Ethereum Hedging

Episode Date: October 20, 2023

Stablecoins represent a safe haven against crypto’s volatility, allowing participants to remain in the market, without off-ramping to fiat. While the major stablecoins are centrally issued (e.g. USDT, USDC, BUSD), there is a pressing need for an algorithmic variant or a synthetic dollar asset. (DAI is somewhere in between since approximately 50% of its collateral is USDC). After Luna’s collapse, many jumped to point out its design flaws, yet such a concept would be crucial for crypto’s decentralisation and self-sustainability. Arthur Hayes proposed an interesting concept of NAKA synthetic dollar (NUSD), backed by Bitcoin and its inverse perpetual swap short. However, yield generation on $BTC is significantly lower than that of $ETH. A sustainable yield would help balance the cases in which funding rates would be negative (a minority in crypto).We were joined by Guy Young, co-founder of Ethena, to discuss the stablecoin landscape and their synthetic eUSD, backed by staked ETH and its inverse perpetual swap.Topics covered in this episode:Guy’s background and founding EthenaThe history of stablecoinsThe stablecoin trilemma and transparencyeUSD synthetic dollar mechanicsEthena’s insurance fund and hedgingMinimising the impact of depeggingLeveraging centralised liquidity while maintaining self custody through MPCUser experienceRisk factorsCEX vs. DEX: liquidity, infrastructure, UXThe potential of fixed return ratesSupply & rates equilibriumEpisode links:Guy Young on TwitterEthena Labs on TwitterArthur Hayes' article proposing the Naka synthetic dollarThis episode is hosted by Felix Lutsch. Show notes and listening options: epicenter.tv/518

Welcome to Epicenter, the show which talks about the technologies, projects, and people driving decentralization and the blockchain revolution. I'm Felix, and today I'm speaking with Guy Young, who's the founder and CEO of Athena Labs. Athena is building EUSD, a delta neutral stable coin backed by stake ether. We're going to dive into what that all means. But yeah, first of all, welcome Guy, so glad to have you on. Yeah, thanks well. Awesome. Yeah, so generally we start more about your personal journey to crypto.

You know, Episandr is mostly about kind of how did people get into the space, what were either vision or why are they trying to work on future of decentralized finance. So yeah, when we start there, how do you like learn about crypto and get into the space and what did you do before? Yeah, sure. So before crypto, I was working at a hedge fund in the US, um, they, focused primarily on financial services. So we're investing into everything from banks, insurance company, especially finance. And that was across the capital structure as well.

I had a friend who was a defy founder back in 2019. He introduced me to Ethereum and to defy back then. And it was just something that was immediately interesting to me. And I was investing into it on the side of my day job all the way through until when Luna collapsed. and it was when Luna went down, Arthur came out with his sort of thought piece around how we might think about a more secure and scalable cryptonite of step-point and that's how we sort of landed with the idea of Athena. Yeah, that's a really interesting post there and like sort of origins with Arthur's pose and also like he's like a founding advisor to Ethanine.

And in the post, if you look through it in the edit, it actually says, right, If I find a team that does this, like, let's do it. So, really cool that you guys found each other there, I guess. We'd be curious to hear actually how that went down. Did you, like, reach out to, hey, if you started building it, and then later when it had like some shape, you were like, hey, Arthur, it's happening? Yeah, that's exactly right.

I thought we wanted to flesh out the idea a bit. So you actually had a bit of a game plan and architecture behind the whole thing. And, you know, a business plan that actually made sense that he could get behind rather than just sort of putting our hands up. So we'd bit that together, had this sort of design and everything ready to go, and then had a connection into Agchat, who is the guy who's running his family office, Mailstrom, had a few calls with them and then got introduced to Arthur directly that.

Awesome, yeah, that's pretty cool. So I also, like, I guess for our listeners, although I think many people know Arthur is, right? It's like a Bitmax founder and, like, a sort of person that's very famed in the crypto space around like macroeconomics and like trading in general. And he wrote this post called actually I don't know, dust on crust, I think. It starts about skiing and kind of draw parallels to, you know, the crypto space.

I think in general, Arthur is a pretty great writer. So I recommend like reading some of his blog post. And it goes into this idea of the Naka dollar, which is essentially like sort of a new type of stable coin that he was thinking about that is based around derivatives and sort of hedging, out the price risk of cryptocurrencies. But we'll get into that. I think I wanted to start because I guess stable coins, like huge topic,

you also already mentioned Luna there, right? In general, stable coins seem to be one of the kind of use cases of crypto that has gotten the most traction. So yeah, I wanted to kind of get your view of like the history of stable coins in the space. You know, what did you find interesting and what was lacking? like going from like die and and these sort of like more centrally backed stable coins. Can you like just kind of talk through a bit how you were thinking about stable coins before you started like if you know.

Yeah, sure. So the history of stable coins goes all the way back to tether and it was originally called Real Coin back in 2014. And then they launched officially as tether on Bitfenax back in January to 2015. So that really the key innovation there was a lot of the, of the crypto businesses that existed were really struggling with banking relationships and being able to off-board and to interfere. And really the only innovation that they had there was if Tether could get the right banking rail set up, you only really needed them to be able to do that as

sort of an off-ramp to the real world and then everyone else within the crypto ecosystem could rely on Tether as transactional money within the space. And so that was really the first use case, which was just a trading pair on exchanges. And it's obviously evolved pretty, pretty significantly from then onwards. After that, I think actually the origins of Athena was probably the next example of how a stable coin-like acid surfaced. And so it was actually on Bitmex where everything was BTC denominated. So every single instrument that in order to margin the derivative required Bitcoin as the collateral acid. And so if you wanted to get into a flat position, you'd take out a short inverse perpetual against that BTC collateral. And that would in effect create a

synthetic doll position, none of those two things are netting off. And really, that's, that's kind of the origin of the Athena story, which is what we're trying to do is tokenize and export that as an actual product rather than just a trade that was sitting on BitMax. Since those days of Bitmax, I think you saw a few different evolutions going away from the centralized model of staple points. So you saw Maker Dow, and the originally was called Sai with single collateral being Eath, and then moved to die, which allowed different collateral types to come in there. at a very basic level what was happening is you're putting up collateral on one side and then you're able to borrow the Sturberkine into existence on the other. Really the key moment that before for MakerDAO was in the March 2020 crash where the system basically became insolvent with what happened with the price crash beneath.

And in order to actually remedy that situation, they needed to onboard USDC as a collateral asset debt. And so I think really that changed the course of history for MakerDAO where they sort of made the decision that they were going to take on centralized assets. and that fundamentally sort of changed the actual profile of dye as an asset going forward, and now you sort of look at it, and it's more than 50% centralized assets, whether it's RWA's or USDC sitting in there. And then we obviously get to Luna. I think everyone's pretty aware of what went on then, happy to sort of jump into what we think went right and wrong there,

but I think it's a pretty well-spoken story at the moment. Right, yeah, totally. I think very interesting to hear that, I guess, on BitMack. there was already like this product there, but I guess it was only on the exchange, right? It didn't make its way into on-chain or defy and that's sort of what you were your after since that also has like evolved a lot. So it goes back quite far, which I guess most people maybe don't know. I think you are also kind of mentioning in some of your materials the stable coin phrylamas.

Can you like sort of expand on what's the trade-offs that stable coins generally make? and then how we see enough hits in that? Yeah, well, I think Traynum is used quite a bit within pitch decks in crypto, I think to make up problems and pretend that we're solving them to get a capital raise. But I think actually the Tronomer, as it relates to stable coins, is actually a very real one. So the basic idea here is you've got decentralization, scalability, and stability, and you only can really have two of those three components.

We personally think that it's a pretty unhelpful sort of definition of the world, where if you're focusing on these words, which are just very broad, like decentralization and scalability, it's quite unhelpful to actually say, are you actually attaining those qualities that you're actually after? And so for us, we're trying to think a bit more deeply about what are the more narrow qualities of each of those three pieces that we actually want to try and retain within our product?

And what we arrived at is within this decentralization piece, what is actually the piece that you're trying to solve for here, and that is actually having censorship resistance? So for us, that meant you don't want bonds or real world assets sitting within SBBs or banking structures in the real world, which a regulator or some of the entity can come in and shut down in the space of the day. And so for us, that was really the piece that we were trying to unlock. And it's something that was echoed in the sentiment of Arthur's piece, which is you're not actually trying to create a purely decentralized stable coin because I don't actually think that they exist in reality. What you're trying to do is create a stable coin, which is independent of the Viet system. and that's actually the quality that you're going for.

So for us, yeah, like I said, it's a more narrow definition that we're going for. I don't think we've solved anything when it comes to the trimmer that's sort of presented as it is. But we are more narrowly defining what it is that we're trying to achieve. And in doing so, I hope that we're just presenting something that's honest to our users in terms of the type of trade-offs that we're making. Let's get into this. I think another one that's very common. like fear or like concern in the stable coin space, especially around Tether is sort of this

transparency angle. Is that like kind of related to the censorship resistance for you or how do you think about that? Since I guess in Tether's case, right, when you think is the backing actually there, there's not enough audits and yeah, is that something you're addressing as well? Yeah, 100%. So I think we felt like we weren't leveraging the openness and transparency to Defi within our product, we'd be doing our users a massive disservice with the way that we put it together. So it does look slightly different here because we are leveraging centralized liquidity to put on the hedges, and we can jump into the reasons for that later on. But what you do have are wallets where you can see where the collateral is actually sitting,

and then the corresponding hedges that offset that collateral. You can obviously just have a read from the exchange APIs to display that. So really what we're trying to do is elevate the level of transparency here well beyond anything that you see. even from like a USC where you're often waiting a month to get watered down order reports come back. And here you can see a real, real-time dashboard of every single cent of collateral and hedge that sort of corresponds to that. And we think that, you know, showing that to our users and being fully transparent about it is the way that we really want to present the product to our users. Right. Yeah, let's definitely get a bit more into this. I think maybe we take it back a bit

back a bit around, you know, this actual principle of how the stable coin is created. So I guess in Athena is called EUSD. And it mostly relies on this principle that you already said that you have the long asset, the core asset and short position against. Can you explain that maybe in a few sentences again for our listeners that maybe are not so familiar with like financial product. Yeah, for sure. And maybe just before we jump into that, I know we've mentioned like stable coins a few times

on this. I think it's a word that we as a team have been thinking about more deeply around in terms of how we're marketing these type of products to people outside the space. So I just wanted to be clear before we jump to the mechanics that really the risks around this product look very different to having a normal stable coin with bonds, basically sitting in a bank account in the real world. And so I think we're just thinking about repositioning the way that we're, we're we described the product as something that is closer to that synthetic dollar concept that I

described earlier on. And yeah, really what goes into like the synthetic dollar idea is you have, in Arthur's case, his idea was long BTC on one side and then short the inverse perpetual on the other. And the basic idea there is that for every percentage change in the underlying collateral, those two positions are essentially netting off so that you always have one dollar of collateral. of that's in about. One key change that we made to Arthur's post was thinking about swapping out Bitcoin for Stake Deuterium. There's a bunch of different reasons why we thought that that made sense.

The main one being that you now have a positive carry to being long Stake Ethereum, so you have paid some yield that's range between 3.5 to 6% this year. And what that does is not only enable you to create what we think are really interesting yield products around the synthetic dollar, but then also it gives you a really interesting margin of safety because these hedges often do pay you to be short. So over the last three years, you get paid roughly 7% to be short the ETH contract. But that's not always the case. And sometimes it is negative. So for periods during 2022, when the whole space is blowing up and centralized entities were getting liquidated, you did see that sort of briefly dip into negative funding territory.

and so having that stake to return on one side obviously helps for you to be able to cover that risk as well and just create a more secure product. Right. So this funding rate, just to go into this a bit more, right? If you're like shorting or using perpetual, you have to pay this funding rate to the other side, essentially, right? Usually you get paid for putting on the short. So you can sort of think about it as there's a natural,

long demand for leverage within the system and crypto is a long biased market where most people who are in the space think it's going up and so they're willing to pay for that leverage speed longer than one X their net worth and so in our case we're taking the opposite side of that position with a short and they're getting paid for

that. But essentially what's underlying to what if it's like negative or positive defunding rate is essentially like a market dynamic of like supply in demand for short longs right? Yeah, exactly right. Yeah. Okay. And you mentioned already, so they, that mostly it's like long biased. I guess we'll maybe get into a bit more like how this could turn out in the future and things. But I guess one other question that I had around, you know, like you already said, you're switching out Ether, Bitcoin for Ether. But is there also potential to, you know, have multiple assets as collateral in the Athena system like also Bitcoin maybe or even like other proof of stake assets?

Yeah, absolutely. For us, it's really just a question of sequencing. So Bitcoin's obviously got the deepest derivative market setting behind it, but it's obviously got that issue of not having as much interesting yield for you to be able to create a product. And so the way that we thought about this is in order to bootstrap something from zero and fund an insurance fund that can sit behind the whole thing and ultimately actually just drive demand for it, we know the people within crypto respond to yields. And that's how you can sort of address that cold start problem in the beginning by interesting people with something that has a bit more yield. But as we sort of grow in scale and you start to tap out the potential of the derivative to market and stake Ethereum,

you can obviously generalize that to look at Bitcoin and then also other proof of stake assets like Solana. The issue here is that on one hand, Bitcoin's more scalable due to the size of the derivative to market that exists around it. But you don't have the yield. But then on the other hand, you've got something like Solana, which, yes, it's got a proof of stake yield that's attached to it. but the route to market is like a 20th of the size of Ethereum. So really, it is a question of sequencing for us, but who knows where these markets grow up in the next three to five years. But we aren't sort of restricting ourselves to state Ethereum, aren't they?

Right. Yeah, makes sense. Makes all sense. I think what's also interesting, I guess, is this insurance one you just mentioned or like sinking fund, I think it's called in Arthur's Post. So like that's basically for these periods where the funding rate goes negative. can you like expand a bit how does it exist in itina how are you building this what like sort of i guess a portion of the yield is diverted to that how do you how do you how are you thinking about

the system to do that yeah you can you can think about the sources and uses for the insurance fund uh in in basically two ways so the first is just raising capital so when we're raising capital at the equity for the Athenalves business or within the if we eventually have a token and do treasury sales through there, you'll obviously have an ability to basically raise dollar capital in exchange for the equity or tokens. And I think that that's just a pretty clean and easy way to get an initial insurance fund set up. The other interesting piece about this is to the extent that you're generating yields on the product that are above market. So roughly this year, this products have been running at around 12% unleathered, and that's with crypto rates being pretty

close to the low of their cycles versus like the real rates in the real world at 5%. What's interesting there, I think, is yes, initially you want to pay out most of that yield, I think, to users in order to sort of bootstrap liquidity and supply in the beginning. But there does come a point where that interest rate differential to rates in the real world, I'm not convinced that you need to pay out the full amount to keep sort of users in the product. And so there might be an equilibrium where, you know, rates in the real world are at 5, 4, 3, and this product's at 10, 12, 15. That obviously gives you some scopes there to be able to capture some of that yield to the

insurance fund going forward. I think the interesting piece about that is obviously, as I mentioned, funding is typically skewed positive, where you get 7% over the last three years. And if you just look at the distribution of sort of the days of positive funding versus negative, when you include the stake debt yields, you get 89% of the days you're getting paid and you can add cash to that insurance fund, and only 11% where you're going to be drawn. from it. So really it is in your favor that you're going to be accumulating cash into the insurance fund through the life of the product. And we think it's quite interesting because

you're capturing some value within the token within the product and actually helping to create a more safe and secure product going forward by capitalizing the insurance fund. Right. Right. That makes sense. And then I guess if it turns negative, the insurance fund would like sort of subsidize that to get the yield back for a while. The stats are sort of the mechanic that it would... Yeah, well, it's self-importance is zero in that case. So we don't want to be in a position like with what you saw with Anchor and Luno, where I think actually creating that inflexible,

endogenous interest right there, I think, was one of the biggest mistakes that they made because you really need external market forces, I think, to be calibrating the supply of the whole system. And whenever you're stepping in and trying to pay more yield than it's naturally producing, I think it's just a bit of a slippery slope where, it eventually leads to breakages in the system at some of the point. And so really what the insurance fund is doing there is when it would be negative, you're just getting it to zero so that, you know,

users aren't losing principle in the stable coin. And then naturally what's going to happen is when you're getting paid 0% to hold the stable coin, we expect that some users would step out of the product. And in the process of doing so, we'd have to lift the shorts on the exchanges, which causes funding to mean revert again above zero. So just to be totally clear on this one. the funding rate is not something that we're scared of.

It really is like a call to the whole design here, which is we want users to be responding to positive and negative interest rates. And when it does get too low, that means the supply of the stable coin is too high relative to dynamics in the rest of the market. And it needs to shrink. And we're not going to stand in the way of that. Right.

I read this like tweet where you answered basically like some concerns of like Degent Spartan there. And I found it very interesting that essentially, the stable coin can adjust by like just people withdrawing and then you're closing all the shorts like you just said but there's also the the just like general mechanics are not as many people might think I think in the that are like scarred from from Luna that it would be like totally collapsing at once or it's very very hard to do that can you can you explain maybe why the like a deep pegging wouldn't be like you know from one dollar to zero versus

instead maybe going like more gradually losing value in this sort of case? Yeah, for sure. So I think the key difference here is that you've actually got the collateral sitting behind the standpoint. So Luna was obviously backed by basically just the faith that the Luna token had some sort of value and then market maker's ability to stand in and save the system a bit like a central backward in the real world. And the key difference here is, yes, even if you do have some of these issues and there are other risks outside of the negative funding, but we can just focus on this one. Even if it does go negative, if you look at sort of the bottom 25 centile of funding, are you like the worst

quarter of where funding gets to, it's typically around sort of the negative 5% range. If you think about that on an annualized basis, if you're bringing that down to a daily attrition that you'd see within the staple coin, it's literally one basis point per day, right? So it's like not even comparable to a swap-bomb curve or a little bit of slippage that you'd be paying like anywhere within this entire system. That's how much you're losing every day if this thing is at like the bottom quartile of the funding. And so really the risk are very different here where if something is going wrong, it's a very slow, gradual attrition of the principle of the stable coin rather than something that sort of collapses to zero in the space of a day like you saw with Luna.

Right. Totally. Okay. I think that's very interesting. I think maybe we can get a bit into like technically how it works. I think the interesting piece in Ecinas case, I guess it has like this sort of interaction between CFI and DFI. So like there's elements on chain and there is like interaction with centralized parties. Can you, yeah, you know, just explain how that works and why you need that maybe? Yeah, for sure. So this is something that Arthur was pretty interesting on his piece,

which is the demand for this type of product we think is here right now, and it's a product that's sort of the need is immediate and urgent, and we don't really have the time to be able to wait for perpetual dexes on chain to grow to a size that would be able to accommodate for it. So centralised liquidity, rough numbers, is between 25 to 30 times larger than what you see on chain. And part of the issue is that even on-chain dexes, they're not a single unified source of liquidity.

So you have synthetic sitting on top. optimism, you got GMX and Arbitrum, you got D-YDX doing their app chain on Cosmos. These are all disparate areas and pools of liquidity, which you can't even sort of aggregate into one place to try and build this product. And so for us, it was really accepting the fact that you needed centralised equity in order to get this to scale to achieve the goals that we have as a team and deliver a product that we think is useful for millions of people rather than thousands. And given that fact that you need to make those trade-offs, we also ourselves the question

of can you do that in a trust-minimized way while you still retain the core pieces of decentralization that we care about? And so just going back to that original point that you're making around the Tramma, for us, it was really asking a more narrow question of what is it that you actually care about when you have your assets on an exchange and what is it that you're trying to protect yourself from? And for us, it's really just you don't want your assets sitting on a centralized server if you have FTX 2.8 out and again, where you want to be able to hold your assets and make sure that centralized exchange is not going to basically withhold you being able to take them out. And so we've seen a really interesting unlocks when it comes to custodian setups that have

been put in place since FDX, where you have an ability now to hold your assets outside of the exchange, but then still use them as a margining instrument for the derivative on the other side. And so what that unlocks is our ability to be able to, A, provide the transparency that I was describing earlier, so you can actually, you know, see those wallets and be able to read into them as a user on the other side. and we'll be providing that dashboards on the app. And then B, it reduces that counter-party risk to the exchangeers is obviously enormously by being able to disaggregate the assets from sitting on that service.

Can you explain a bit more how that works, like these off-exchange custody accounts, who are actually the holder, like what's technically happening there? Yeah, so it's brand names that you'd be familiar with, guys like copper, buy blocks, with Bitcoin and Anchorage coming out with their products in Q1 next year. But essentially all that's happening is you have usually, and this is obviously different between different custodians, but you have a key share that's split using MPC into usually three different keys. There's usually one pair that's sort of sitting with the custodian,

one with the trust of third party, and then one with the client on the other side. And the reason that the exchange can allow you to have your assets in there without actually sitting on the centralized servers, that they know that we on the other side or a user on the other side does not have unilateral ability to be able to pull out those assets at any point. So as long as the rules of that wallet are being adhered to, i.e. you have enough margin in there for the position that you have in that exchange. They have comfort that the custodian is obviously undertaking their job correctly

and have the sort of correct checks and balances that sit around it. So I know it sounds like strange within crypto, this type of sense. setup, but it's actually kind of how the real world works where you don't have, uh, like assets sitting on the place that you're actually trading. Uh, those, those functions are kind of different in the real world and, uh, you know, purposefully separated. And I think, uh, this is just the natural evolution where I think a lot of the times in crypto, we kind of have to learn the lessons, uh, uh, ourselves and in a painful way to kind of see that some of the pieces of traditional market structure actually

make a bit of sense. And for me, actually having custody of your assets outside of an exchange is, uh, It makes a lot of sense, and I hope that the industry sort of may have stored that over the next top chronicle months. Yeah, it's definitely a really, like, interesting direction, probably also useful for other use cases or things outside of Athena's use case specifically. So just to clarify, the trust third party here would be then like something Athena Labs does. And the actual user, so these shorts are also something that doesn't really involve the user, but more like, now having accounts on these different centralized exchanges and kind of placing the hedging sort of trades on the exchanges, correct? Yeah, so the user is actually basically sending their assets to this custodian.

So the user does not hold the keys to their assets anymore, and that's a key trust assumption that's sort of built into the design here. So really what Athena is doing is providing a front end where a user is depositing assets. it's being sent to one of these institutional great custodians where we no longer have control and title over those assets. And, you know, it's a bit like a user base of just sending their assets to a five blocks or a coin base where they no longer are holding basically the keys to those assets. But it's programmatic in the way that they can come in and out where if they choose to redeem,

we don't have an ability to sort of stop them from doing that. Right. Okay. And the, again, on, on the trade and how, like, if you have to take these short positions, how are you like basically bringing back this information to the, to the people holding the coin or like, you know, what kind of trades were made? Is this also like something, yeah, you're exposing somehow or like, how can you make this like sort of trust minimized, I guess? Yeah, so the way it's actually working on chain is, um, it was probably worthwhile. I may actually just step into the user work for it, but you arrive at the front end and you put in

an EIP 712 signature request. So that's you basically requesting or signaling an intent that you want this collateral to be hedged. Our off-chain servers are basically reading those messages coming through, reading exchange APIs to understand where can we actually efficiently hedge that collateral on the other side. And then we're providing a quote back to the user, which says you can hedge this at 99.99D1 to accept. And if it is within sort of that the range that the user was expecting when they sign that request, then we can actually execute. And the key piece here is that we cannot issue the USDA to the user on the other side

until we have confirmation that the hedge is actually being executed. So really the sequencing here is user comes in, request a hedge, if it's in line with what they're expecting, and where we can execute, we execute, and then issue the USDE on the other side to the user. And why does the user do that? He wants the USDA to like earn a yield on it. or what stairs? Yeah, that's correct.

I have sort of, there is two different types of users within the system. There's like what we can call retail users who are just wanting to swap in from dollars to dollars. And so the front end that we're providing there is basically just a front end that rewrites people into liquidity pulls on chain. So you'll arrive with USDC, USTC, whatever stable coin.

And all you're doing is actually swapping it in a curve pool into our stable coin. But then in order to achieve what I've just described there, we need to abstract away all of the complexities that I've just described to previously. And the sort of mechanism that allows for that is we have market makers who are doing the mint and redeem arbitrage against our curve pool. So you can think of the flow as a user comes in with $100,000 to swap into USCE. That slightly imbalances the price of USDA versus USDT in the curveball. And when that becomes imbalanced, a market maker now has an arbitrage

opportunity to pick up that asset and come to our mint and redeem competition. and add new assets into our system. And it's actually the market makers who are doing that request for quote that I was just describing where they're performing the arbitrage and keeping things in line with the dollar over there. So really that complicated system that I was trying to describe that, that's only really into how a market maker is interacting with us. And retail flow basically just looks like a uniswob transaction that just hits an AMM

on the side. Right. It makes sense. And then maybe I guess. in in terms of like risk we talked a bit about the funding rate but now you're like interacting with all these centralized exchanges and like sort of other players can you expand what sort of additional risks exist there and how they might impact like

uh... tapecoin holders yeah for sure um so yeah as you said i think we've we covered the negative funding one a bit uh the other one was just around credit risk to the exchanges right um we thought we made a pretty thoughtful decision in terms of how we disaggregated the assets from the exchange and tried to help alleviate some of that concern. But it doesn't take the risk down to zero. The risk that still exists there is you still have the derivative instrument sitting on the exchange. And so if you just play through that scenario of an exchange going down on the other side, let's say you've spread the portfolio. There's five exchanges. You've spread it 20% across

each exchange and one of them goes down. Let's assume that on the day that that happens, ETH is probably going to be falling, you know, 10, 20% on that day. So you're going to have a hedge that is massively in profit on that day where you might have lost the P&L to that hedge that corresponds to the collateral that's still sitting outside of there. So you still retain some of the credit risk on the hedge side of the whole thing, but you have an ability to basically move that down to basically daily settlements of the PNL. So you can just, you can keep that risk to just one day of PNL that you'd have on that exchange. And so it really was you'd be losing there is 20% of 20% in terms of the whole portfolio. And so, yes, we haven't

sort of reduced that risk entirely to zero, but like with an entire exchange going down, we think we sort of box up that risk reasonably well for, yeah, what we just described that. And then I think one of the pieces just to mention here as well is there is like a liquidity element to this as well. So you're obviously tapping into perpetual markets that you obviously need to be liquid as people come in and out of the product. This is something that you can just think about conceptually very similar to slip edge of people coming into Unisorp transaction. When you want to come in and out, there's obviously a cost to doing so. It's the same way that there's a cost to open and close the edge. Perp markets are considerably deeper than spot markets. And so this is

like low single digit type basis point of impact. But if like everyone wants out at the exact same time and you have thin liquidity at that moment, there might be, you know, higher slippage for people to get in and out, but really it's a cost of liquidity and it's something that we can't try and disguise in any way. I think we just need to be clear with our users that it does exist both on the way in and the way out. Right. And I guess, yeah, just to clarify, if there is like the scenario that one exchange kind of can honor the short or like whatever happens to the funds on this exchange, this impact is kind of socialized across all the, say, going on and not like by the specific user that did the trade, right?

Like it's kind of... Yeah, that's correct. Yeah. Yeah. Okay, cool. So, yeah, really it's a lot about working with the exchanges that are like trustworthy also and managing how much goes where. Is there like some sort of, yeah, like, are you just using like the best rates or like,

how are you managing what liquidity goes where in terms of the exchanges? Is it also like kind of risk adjusted? in some way based on your judgment of the exchanges or something like that. I don't know what one could use, but have you thought about that? Yeah, so in the beginning, I think you can do a bit more solving for the best return because you're small and sort of not, you know, impacting the funding rates and the liquidity on single venues. But I think in an end state, really, this whole thing is going to be determined by, you know,

global risk parameters that sit around how much of each exchange you're willing to be a part of. And so in the end, it's just going to look like the market share of the entire market, if that makes sense. So if we significantly diverge from that market share, I think that there's something going wrong in terms of our risk controls. And so, yeah, you can do a bit of solving for yield in the beginning, but then be on that point when you get too big, you can't even lose the ability to do that. And then just in terms of the exchanges itself, yeah, we are only working with exchanges

that we as a team feel comfortable. There is sort of different lighting systems and stuff that are at times. to the custodial services we use as well, where they're obviously flagging for any type of activity that looks slightly suspicious or on-chain wallet movements, that type of stuff. And so we do lean on some of the third-party infrastructure providers there as well, who can help with that. I would just sort of circle back to the same point, though, which is if an exchange is going down, you have sort of isolated to just one day of hedged risk, not even collateral. And so that number on the example we're giving,

if it's an exchange between 10 to 20% of the entire portfolio, only a single day and there's a 10 to 20% in ETH price, it's only somewhere between 5% to 1% of the total portfolio that you'd be losing. So again, it's not something that sort of takes the whole thing to zero. And a 5 to 1% is a type of risk that you can actually cover with the insurance fund as well. Yeah, thanks for clarifying that. I guess the other, this is like specific to the exchanges, but I mean, custodian, I guess there is a little bit of a risk maybe with these MPC wallets.

That would be the only way how collateral could be lost in case there is like something wrong with this MPC setup in some ways, right? Yeah, that's correct. I would also just say that the custodians that we're using have not actually lost a dollar to date. So we compare that 7 billion was in DFI in the last three years, north of 15 billion in CFI. if you just like empirically look at the data and say where has it been safer to the legal assets in a smart contract or with the custodians that we're working with, the data says it is a custodian. So yes, it's not, it's not perfect in terms of reducing risk to zero, which I think is impossible to do with any of these things. It's really just looking at how risky is it relative to other things. And I think the data gives us a bit of conflict around that.

Right, yeah. I do like that approach a lot to actually look at the state and the current numbers. I think also like in terms of the open interest and the choices you have made there and like the analysis. So I'm also recommending we will probably put it in the show notes like some of the kind of data that you have around that, which I also found very interesting to see just like also the differences between open interest and centralized exchanges versus like the dexes. So I guess, yeah, maybe also like a question around that. Do you think, because right now, I guess it's like 20X centralized exchanges. Do you think at some point more and more will move to decentralized, like sort of perpetual protocols?

And, you know, what is needed for that? Or like kind of why are people choosing centralized exchanges still over Dex? It's still like user experience type of questions or, yeah, liquidity type questions that if at some point they flip, that would change or, you know, is there anything else there? Yeah, sure. So we are working with synthetics. So we put out a bit of a post where we described how we're thinking about integrating with them. So that is absolutely the intention that we sort of start where liquidity exists today and then walk back to decentralization after that. And I am bullish on the on-chain growth of derivatives. I think it was something that a lot of investors and a lot of people were quite excited about last cycle. But actually, I think the infrastructure wasn't really there to enable to grow to the size of people. people were expecting. And so now really like derives and being able to run off chain order books, it really does take, not even an L2 can really handle it. Like if you speak to DYDX and their decisions to go and build on an app trade in Cosmos, they weren't even happy with the way that an L2 or even Solana could cope with the capacity that they needed. And so I do think we should have

got there now and already the cycle in terms of infrastructure to allow those things to grow. really it does become a question of liquidity and it is quite difficult, I think, to unseat these big pools of liquidity on sexes where ultimately market makers and big traders have shown that they're comfortable to trade and it does take quite a big step for them to move onto new chains and application sitting within Defi.

Yeah, that's super interesting. I think it will be interesting to see how it plays out in the next cycle and after that if actually liquidity kind of moves more on chain and I guess also how you're designed and then adapt because I guess for now you are, are you supporting any texts? I think synthetics, right? You already announced, but it's essentially, again, like a pragmatic sort of thing

where you're like thinking about the sequencing of how to roll things out. Yeah, that's exactly right. So I think also from an engineering perspective, it's a lot easier to integrate with CFI. It's sort of one type of model and you can just plug in different exchanges and then actually picking different chains, different DIPI applications to get into, just takes quite a bit more engineering lift on our side. But we are really excited about what we see with a bunch of the different Daxes on chain. It just sort of takes time, A, for them to grow to a size when they're more comparable to sexes, but then B, also for us to just do the integrations on our side.

I do think it's pretty interesting, though, if you start to think about the different ways that Athena can plug into different DEPI applications, it's not just perpetual Daxes. So if you think about the core product itself where you're capturing a yield and tokenizing it within a product where you've got censorship-resistant collateral with an embedded crypto-native yield coming from the two sources that we described earlier. What's really interesting there is that you're actually importing a new source of yield on chain for defy, where at the moment all of defy is basically just running on stake to eth, right, and there's no way of extracting any more return or bringing any more yield to

defy. Your ability to import that 7% return that I described from shorting Heath perps from C-Fi into D-Fi, then unlocks a whole sort of a new suite of things that you can start doing when you compose with that asset into different applications. So just one example I could give you here is, if you look at them, make a Dow or a Frax, if they're filling their entire balance sheet and the backing for their own stable coins with RWAs, what happens when rates start to go from 5 to 4 to 3, and this product is sort of sitting there under 12, does it become quite interesting to think about, you know, diversifying away from pure centralized assets, which is something that's a bit more pertinentative as.

as what we're calling an internet bond. And so, yeah, it's really beyond just trading on Dexas, but it's actually, we hope, delivering a really interesting bond-like asset for the rest of Defi. Right. Yeah, that's super interesting. I think you also, yeah, I guess maybe let's go a bit into that as well. Right?

Like now it's very based on perpetuals, but there is also, I think, some ideas around, like, fixed rate and, you know, like kind of using future. or options maybe to also hedge. Can you explain how that would work and what it would unlock for or this sort of concept of the internet bond or for anything I guess at large or end the space? Yeah, for sure.

So you can think of the hedge that we described earlier where you're long spots take deep from one side and then short a perp. There's both variable rates of return. So the perp is resetting every eight hours and the interest rate is changing. You do have traditional features contracts within crypto as well.

They're quite a bit less liquid. So roughly the size of the market is a tenth of the size of the perps market. But what you can do there is actually hedges the collateral out for a predefined maturity date. So you can say, I want to look out from now until June of next year and lock in what that, how much you can get paid for that hedge on the other side. And so what you're able to construct there is actually fixed return dollar denominated instruments where you're putting down the stake deeth collateral and saying, I know I'm going to get paid 4%. and then all of that conversation that we just described around negative funding and all those other elements,

it completely drops out of the equation because you know you've locked in that collateral as a dollar format with the return. And so again, think about that asset. If you look at some of the treasuries within DFI, that's a pretty interesting proposition right, where you can go to these treasuries, which are sitting on hundreds of millions of dollars of either USC or ETH and say, do you want to buy into something that has, you know, a fixed rate of return of whatever it is, 6, 8% of the next six months. And you can actually take that yield and use it for sort of working capital and paying employees within the down and stuff like that.

So that's, we think, a really interesting asset to be bringing to defy and the concept of like fixed rate is really started to, you can really unlock it when you're using an asset of the size of state death and then also the futures market across the entire sex space. Yeah, that's one piece there. And then one other item that you sort of called out there, which I can just jump into a bit more detail is thinking about this idea of repo financing. So if you have the assets that we're describing now with the returns that we described are all obviously on an unleathered basis, but to the extent that you have an interest rate differential between that asset and your

ability to borrow dollars in the rest of defy at a different rate. So, you know, if Ave is 3 to 4 percent to borrow USDC or USDT, if you can put up this asset and start to lever it up by borrowing USDC against our USDA asset with the yield and you gain paid 12 on the asset side and borrowing a three or four, that post-leverage return starts to become really interesting when you start to loop it up to five or ten times over. So again, it's a bit of sort of financial engineering that I think is unlocked by having an interesting base layer of yield that were obviously importing from CFI and to defy. Right. If you were to do that, does it, I guess it would impact maybe. like more people do it

and might push down the yield of EUSD also like to like go closer to the borrowing rates of these money markets elsewhere that's exactly right and there's like a big conceptual question around what we're doing and what is the equilibrium interest rate of this entire thing right? Because right now you're getting paid too much to put on this long spot and short prep position

eventually it has to sort of come down to some equilibrium level what does that equilibrium level look like? It's probably where borrow dollar costs and the rest of the market start to converge as that number. So if you can borrow dollars at 3 or 4, put away the complication of everything else with all of the custodial risk, all of the credit risk, all that type of stuff,

and there's just a very clean way to borrow dollars and put it into what we're doing. Those two interest rates should sort of like come together. But really the outcome of those two things coming together, what needs to happen is a supply of our stable coin needs to go through the roof, right? And so really that repo thing that I described there is just a really clean mechanism within Defi to allow people to basically force that interest rate differential to go voucher.

Right. Yeah. I think I think that's a very interesting case for like the supply growing actually. And when there is like so much demand, let's say like scenario that you, it does succeed and there's like huge supply. now it does impact the funding rate. Do you would, yeah, I guess that's also like an equilibrium where one is not clear where would that lie then if there's like just much more short demand from your end through just such a product, right? Yeah, and I just want to be clear on it, which is if we do reach this theoretical equilibrium and this like we know in crypto things never like actually land on their theoretical like

equilibrium, it's always just like a mess between here and that. but really if you are getting to that stage that is just an outcome of us doing our job correctly right which is we've produced a crypto native stable coin that is large and it sort of breached the size that the market can actually handle and so really it's not something that I'm like

awake and worrying about if you sort of get to that size it's actually something of you've sort of succeeded in what you've set out to do and if you do get to the stage where it's zero it'll start to push down the funding as I said it's part of the design where if that is starting to happen it means that the supply needs to shrink and we need to get to a new york equilibrium of supply after that.

So it's something, the things I stress about are have I put the appropriate measures in place to cover for the tail events, i.e. the insurance fund is that, you know, appropriately sized relative to the whole thing. I'm not like awake worrying about if this thing caps out at three, four, five or ten. It's going to cap out where it caps out. And that's just sort of like the market dynamics determining where that is. All we need to do is put in the infrastructure for us to gap from zero to that point and then make sure that we're covering the tail risk appropriately to create a product that's as safe as possible for our users. So I think it's interesting to discuss these theoretical outcomes. I don't think we ever landed them because it is crypto and then

I would just to just comment. It's not something that I'm spending a lot of time thinking about. I just need to think about sort of the tail risks. Right. Yeah, that makes a lot of sense. I think why I guess it's interesting to discuss is that if we think about, yeah, I guess, coins like really being used in the actual economy and other things, then you wanted to be like quite a big supply. I guess that's what generally the scalability or like the lack of in maybe like things like die come to mind or people are like maybe thinking about. And that's what's probably like the promise of Luna for a second there. And it's like very scalable. And I guess you know, you can't have free lunch. And that definitely showed. So yeah.

I think that's a very pragmatic sort of approach and makes sense, right? First look out for the, that it actually works on the smaller scales too. I do think, yeah, it's also interesting that, I guess, just to know that there's like, obviously, it's kind of obvious that there's like long demand for crypto bias. And maybe that in the future will also change. But I guess, you know, also Athena will evolve. And who knows what else, what our assets you will be able to collateralize in the future or other pool of stake assets or who knows what this will turn into.

Do you have any thoughts about that? Like, you know, could you do like a similar mechanism with like non-crypto assets? Or is that like not on the table for now or something like that? Yeah, you've got like, you obviously got features that exist on equities in the real world. They don't have perps in the same way that we see it in crypto. There's perhaps are very much crypto specific phenomenon. But what you could do, and this is just conceptually speaking, it's not something that I put like a ton of thought into is

if there is a positive carry or like a contango within the epities market as it relates to a single equity, and so on sort of allows you to collateralize that with an apple stock or whatever. Yeah, really, the whole concept here is you've just got one asset as a spot thing on one side and then a derivative on the other side that offsets it. So as long as there's sort of contango within the derivative instrument on the other side you're going paid to put in that position, you can do that. I think you're thinking further ahead than I have with this thing.

I think we're just pretty focused on getting something that sort of works and the useful for the crypto space to start with. Awesome. Yeah, I think, yeah, maybe we went a bit very far there. I thank you so much for coming on and explaining, you know, how it works and bringing like all the pieces together. I hope this was a, episode for our listeners.

Yeah, and maybe as a final sort of question, like where can people learn more about Athena, how, what's the road web like, things like that? Yeah, sure. So the website Athena dofai, the Twitter is Athena underscore Labs. I'm on Leptocritic underscore, which is a bit cryptic, so hopefully you can thread in the show note. And then, yeah, in terms of the roadmap, launching, basically. basically in a couple of weeks, slow sort of roll out of the product to internal parties

where we're just going to be doing some testing with internal capital. And then the idea is that we slowly start to roll it out the month after that to the public to be able to come in and test the product. So yeah, excited to get it out and hope that everyone who likes the product as much as we do. Awesome. Yeah, very excited to see how the user experience goes and how a launch goes and like, yeah, congrats. And yeah, thanks for coming on and see you soon. Well, see you not. Thanks for the time.

Thank you for joining us on this week's episode. We release new episodes every week. You can find and subscribe to the show on iTunes, Spotify, YouTube, SoundCloud, or wherever you listen to podcasts. And if you have a Google Home or Alexa device, you can tell it to listen to the latest episode of the Epicenter podcast. Go to epicenter.tv slash subscribe for a full list of places where you can watch and listen. And while you're there, be sure to sign up for the newsletter, so you get new episodes in your inbox as they're released.

If you want to interact with us, guests, or other podcast listeners, you can follow us on Twitter. And please leave us a review on iTunes. It helps people find the show, and we're always happy to read them. So thanks so much, and we look forward to being back next week.