Curve Founder: Real Yield Is Coming (without impermanent loss)

Transcript

But if you pair stable coin with stable coin, you don't have to do anything tricky. You just do stable swap. We know, Michael, we know that you had the largest loan on chain known to mankind. Why is that the case that the price over is the square root of the the real Bitcoin price? Yeah, that's that would be quite an interesting story. So, welcome back to the rollup, your home for technical talks and open conversations that shape the modular future. Nothing on today's show is to be construed as financial advice. You can see our full disclosures list at the rollup.co/ventures. Sit back, relax, and enjoy today's episode after a quick word from some of our awesome partners. The intersection of AI and crypto is moving fast, and soon an agent will be managing your onchain transactions for you. But

without guardrails, it's a disaster waiting to happen. Newton locks agents to your intent with cryptographic proofs and real penalties if they slip. Learn more at magic Newton onx. What's going on, guys? Welcome back to the rollup. Today we have an episode about the DeFi Bitcoin black hole with our friend Michael here, the OG founder of Curve, one of the projects which I'm sure you guys have heard of. Uh I know Robbie and I were messing around on chain in 2020. Um Curve is known for stable coin swaps. Uh some of the best liquidity on the market today for Ethereum DeFi. We have Michael in the building. Good to have you, Michael, and thanks for coming on.

Yeah. Uh, hi everyone. Michael, I guess one thing that struck me right when I when I learned about curve and I learned why it was differentiated for stable coins was the idea that you could essentially uh change the xy equals k model that we had all came to know and love in unis swap v2 and I think you really set the foundation for you know the rest of the decentralized exchanges to move to a concentrated liquidity model. Could you kind of talk about like h what in your mind led you to that innovation and ultimately how did you how did you come to execute on that and figure out okay this is the right idea right well I mean it started from thinking about uh what

we know today as uh liquid staking right and um I was actually doing uh new cipher at that time and uh new cipher tokens. Well, they were stable and I thought, well, okay, what if someone wants to leave staking quickly? And of course, this also relates to Ethereum staking. And I uh thought that maybe something like concentrated liquidity is needed. Um but at the same time at the same time I was uh kind of a um well a big user of maker dow well not big by by

the size but I used it fairly often because I didn't want to sell my ETH so I borrowed against ETH borrowed die and exchanged to USDC and withdrew that on Coinbase to to my bank account. Oh, and we know, Michael, we know that you had the largest loan on chain known to mankind. That was what I was tracking. That That was not on Maker D. That wasn't on Maker D. No. No. Um, my uh loans on Maker D were relatively small, but the biggest bottleneck was that liquidity for D was not very good at all. And there was even offchain and there was no particular reason for that.

So I thought well what if the what if the algorithm which I'm thinking of for liquid staking because liquid staking is not was not the thing yet why not I don't I apply it to die and solve my own issue here I've uh well I I I wrote a simulator it it looks like it would work okay I estimated the market size market size looked Okay. Well, it actually proved to be exactly the market size which which curve captured eventually and um I started implementing that in what uh maybe September 2019 and by um by

January 2020 everything was fully ready uh for launch including the UI. So and uh I I launched it. So that's essentially how it happened. It appeared that already with half a million liquidity which the pool got in maybe a week, it outperformed all the existing um all the existing exchanges essentially in I think the um probably the best at the time for this particular purpose was some internal order book based exchange uh of uh running by by maker da I think um and well obviously order books on Coinbase. So the half a

million dollar pool outperformed all of those um and um users came um from using 1 in 1 in at the time was extremely fast in integrating anything. They integrated it in like half a day. So even before their smart contract code was verified, it was already integrated and it was amazing and uh yeah, so that's how it started. Yeah. So before we get too deep into kind of where we're going with yield basis and some of the technical DeFi concepts, which I'd love to touch on, I'd like to just uh kind of explain to the audience some of the lore here.

So we we joined uh kind of doing this podcast and whatnot in like early 2020. Rob and I found DeFi in late 20 2019. You know, we we were some of the first yield farmers on Compound, on year um and these places and of course Curve, um and and you know, we're we're active in in just vividly remember the first Curve interface feeling like a a 2000s um Dell computer vibe. Um, and this kind of initial evolution of DeFi was what really brought in a lot of the brains and minds that are in the space still today like Michael, like ourselves, where it really truly felt like a zero

to one moment in in terms of new capabilities on chain. And so that kind of played out uh through 2020 2021. Um 2022 came, a lot of the yields compressed um and you know kind of DeFi went through its like kind of growing up moment where a a lot of what was built was kind of exposed for basically the lack of sustainability uh core found uh foundational problems such as impermanent loss uh causing LPS to have a lot of uh permanent losses which I'm sure we'll touch on and much more. So, the part of the lore that that I I I want to ask you about though, Michael, is what I found to be quite funny, which was all of crypto Twitter was watching

your liquidation price on your Curve loan. Um, and this was going around in circles. You were getting me'd for having a mansion. They were trying to liquidate you and pump Curve and dump Curve and and do all these things. So, maybe you could just walk us through your thinking of that position. um and just kind of like what what happened in that in that like couple week time period. Um I I think the audience would appreciate the Well, I don't know if I can uh really um well I I don't really like talking about that too much I would say but um I think in brief I would say this that I don't like to uh to to sell crypto especially crypto

of the project which I created right so I I really don't like that and uh this is exactly why I used maker dow and that's uh because I didn't want to sell e but uh selling CRV is also something I really did not plan to do at all and this is why this uh essentially loan was created but uh yeah it's it was I would say not so easy to avoid um to avoid uh liquidation And um it's um how to say um the value of the token

can quite significantly deviate from uh what you can say the fair value should be. So you probably if you hold any crypto you can um quite say that the true the val market value can be deviating for from the real value by factor of 10 in either direction. Right. So I think they call that they say that the market can stay irrational longer than we can stay solvent. Right. Right.

So I think this is exactly that and really the the way to do it uh properly is probably not by I don't know not by selling not by borrowing but by building something which is um which is u I would say sustainable well curve obviously is sustainable right but the real sustainable yield which can be earned not for just LPS but also So for the system and eventually uh benefit everyone should be coming from from somewhere from from from the activity which is happening on the platform on curve. Well, what's activity happening

on curve today? It's obviously stable swap uh swapping between uh currencies of the same denomination mostly and uh curve stable coin which is well minting which is borrowing the stable coin. So, and um well, I would say that it's well stable swaps at least they have some limits of how big this market is and they they reached that limit many times already. Uh like so stable swap reached its limits. So, uh definitely we need to go somewhere beyond. Yeah. And um well somewhere beyond I guess should

be some different source of revenue which uh I think should be uh volatile swaps. Um but yeah solving the problem of volatile swaps without having impermanent loss was uh always an issue. Yeah, it always has been. And I mean, I remember LPing for the first couple times thinking that it was going to be easy yield and um over time you kind of learned that it was a much more actively managed strategy than you thought it was going to have to be due to the you know price discrepancy between two different assets. Um now just one more one more thing on the lore and then let's talk about yield basis. Did you ever have any any uh any interactions with like uh Andre Cronier when when you were

building Curve and like uh did you guys build anything to together there? Oh yeah, that's a yeah that that would be quite an interesting story. So I think Andre started interacting with me in late January 2020. He immediately recognized the potential. at the time he was building this well uh not yur but I earn so with letter I and the essence of the project was essentially um well um kind of um yield aggregator for for lending platforms that's what it really was it was tokenized as a sort of vault and

basically uh he told me well my even the first curve pool was using compound uh compound tokens so compound deposits as stable coins so why not use his thing and I said yes this is a great idea so let's let's do it and on well in the beginning of February uh we uh uh we launched uh a pool which used his uh his vault tokens essentially sorry ypool and well I think I think something was uh before before the ypool I think first it was USDT pool and then then y poolool something like that anyway uh one of the

first pools and it worked uh pretty well um uh later he uh thought well we thought of something even crazier so he contributed a code with for for the first SUSD pool. And that was crazy because it was um how to say first metapool. Well, I figured out how to do meta poolool and uh uh explained it to Andreas. So he uh plugged in um uh he plugged in the existing pool and uh it was uh swapping between SUSD and LP token of Y poolool and everything was uh

kind of assembled together in that pool. Well, unfortunately that pool is dead by now because a month after I figured that there was a vulnerability. So we asked uh uh so Andra made a small mistake and we asked all the uh LPS to migrate to a um uh to a new plane pool SUSD pool which is still up today. And um another thing with that pool was that because it was using a pool which swaps between uh between yield aggregator tokens. uh it was using essentially 2 million gas for every swap. So which appeared to be pretty painful as uh Ethereum gas prices

started growing up. Um so but anyway uh the idea of metapools uh was implemented later in a more proper way and worked uh fairly well over the years. If I know Andre, he is not happy that his mistake is being exposed. Well, I mean, uh, it's it didn't lead to any, um, any losses and, um, I mean, yeah, it it was a kind of a wild time. So, abs, absolutely. I mean, this is this is DeFi composability at its finest, right? We have year operating as intended and and we have curve also in parallel operating. and the two founders Michael and and Andre

get together and say, "Hey, wouldn't it be great if we composed our two products together?" And we were able to hard was a great example of composibility and later farming of Ypool to get Y worked uh worked really really well. Yeah. Yeah. One last thing on the lore here as well is I remember quite vividly in the days leading up to the US collapse uh the USDC USDT uh curve three pool which is like this metapool idea that that you guys introduced and then started introducing these three pools. You could slowly see the ratio start to get filled up much more with US versus USDT and USDC. And

this was this was like the leading indicator for something fishy was about to happen. Um you know what was going on in your mind at that point and uh Yeah. Yeah. Yeah. Well, it was uh definitely a feeling of increased risk. So the I was I would say I was initially very uncomfortable with US even when everyone was super excited about it and I was deeply uncomfortable about it because it was a stable coin collateralized by its own protocol token. So it felt not uh not not correct to me that to do that

really. So like too like too dangerous. And I did talk to to docu and uh at some point well there was also llama risk assessment of US and it was uh it did look risky really uh so what uh docu proposed to do is to do um something which was called LFG um so some reserve where it's not protocol token which collateralizes USD but uh Ethereum and Bitcoin essentially And um I mean it it looked kind of better to me at that point. So I thought okay maybe maybe if they diversify

collateral towards that maybe it will be better. Uh well it appeared that uh well later much later I found that it's not necessarily they used LFG to support the PEG as they should have and really it was probably too small of a reserve which is not reliant on the uh protocol token. And another thing, you probably still can make a stable coin b u backed by its own protocol token if this protocol does something else than the mere stable coin with yield. Uh for Terra, it was kind of that there was mirror uh protocol which allows to swap

between uh between kind of synthetic versions of real assets. But activity from that from that was much much smaller than the yield they've been given away. So really it was uh kind of stable coin backed by its protocol token where the whole uh task of the protocol is to I don't know to to to print yield right so yep interesting times this is the snake eating its own tail. Yeah.

Yeah. So it's you you really need some externalities here and it's not necessarily even the token which are tokens which are externalities. It would be um I don't know activity which is giving the value to collateral should be mostly not your stable coin. Yeah. So, I'm I'm curious how this kind of relates to to the Curve tokconomics and the Curve protocol as a whole because one thing that took DeFi summer and turned it from farming sushi coins and yam coins and food coins with, you know, thousand% APYs and made it something that felt more real. That was the curve wars and and that was like this ecosystem around curve with convex and

vote escrow and like all of this very exotic and very interesting behavior that was happening as you as you say the externality to the protocol itself. So could you could you describe why it's important to have the surrounding ecosystem that is outside of the protocol itself? How does that help bolster the value and make make a project more sustainable? Yeah. Yeah. Well, that that was not something I thought about initially. I mostly thought about something a little different. I thought about aligning protocol participants uh with the long-term success of the protocol.

That's why I introduced uh voting escro thing where you get governance power and power to earn uh proportional to not just your token amount but also to the lock time and that um that appeared to have not really only governance consequences but also economic consequences. it appeared to be a much much better supply sync than I don't know buy and burn model for example. Um but also something I did not expect. Now I did not expect uh v people selling votes. That was not

planned. And this was again first introduced as a sort of bribing platform uh launched as a simple um simple smart contract by Andra Kronier. And by the way it had a bug as it appears. Um but anyway it was not a critical bug. Um so but yeah anyway so um you guys are so funny man. You guys are so funny. Yeah. Well I mean it's uh you you don't deposit funds there right?

So bugs cannot be critical there. Um but it you know the all the information which allows you to compute what the incentives to what are the incentives to distribute that all that information is already on chain. uh historic balances of VECRV are recorded onchain and accessible from any smart contract which makes it very uh convenient to to write these platforms which measure who had who voted for what and and for how long and whatever. Uh so that's um that allowed for to create uh so-called bribes and they they were integrated with all the secondary platforms like we

had convex appearing um and why didn't really play the bribes game but you know stake da uh so I think uh yeah that's that was something which gave apparently extra value to to vecv Yep. Okay. So, you know, we we've had lots of, you know, several years that the Curve project, Curve Dow, has been around. It survived. Um, there's been lots of learnings that you've taken from Curve. Um, and now you have a new project. It's in stealth, but part of this podcast is bringing this out of stealth, yield basis. So, let's talk about yield basis. Let's talk about what

it is, how it works, and we'll talk about the learnings that you learned from curve and how you applied those to yield basis. But first, let's start with the basics. What is yield basis? Right? So basically it's a project where I where I try to solve impermanent loss and um well what I mean by that um so well you if you deposit something in a any flavor of volatile swap be it on curve be it on unis swap two or three or four you always have the same problem. Uh the problem is if you uh deposit essentially two currencies there

um I don't know Bitcoin and USD and price of wrapped BTC goes up or maybe it goes down doesn't matter the value or in your pool appears to be almost always smaller than the value of these coins if you if they don't do anything. So you could just uh put them on a shelf and they collect dust, but actually they would be more expensive than uh your your the value of your pool if the price of your crypto is significantly different from where you started. Um of course you earn fees some fees on top of that, but they usually don't close this gap and that's

kind of a kind of an issue. Um, I would say this this is really well the the fees produced in in volatile swaps are really large. They are I think they are much larger than the fees produced in stable coin swaps. But this impermanent loss issue really prevented people from uh from keeping liquidity in there. So that's it's just uh well imagine that you you deposit uh funds in in in a crypto swap because it showed 100% APR or whatever and uh you you held funds there and then you figured out

that you you have actually less than deposited and uh then you you hear that this is not a loss because it's impermanent and uh you know it's you you just didn't didn't lose anything. you just have less money that's all. uh it it kind of sounds wrong and um yeah so that's why uh that's why TVL in in um I don't know in unis swap for example is not necessarily super huge especially if you look at u like democratized liquidity not coming from any uh any large players and um that kind of uh that problem uh well I I mean I know that

it's impermanent loss is not actually a loss. It's a uh not something you lose to someone. It's just a mathematical property of of the of liquidity in these sorts of pools. But for user, it doesn't matter like it's if you withdraw less money than you deposit, it's a loss, right? Um so anyway, I figured how to actually get rid of this um u like ma mathematical artifact so to say. and have it uh interfaced as um as like staked bitcoin for example or maybe staked ethereum if it's for ether why not so that's uh so user basically

deposits crypto and has yield bearing crypto and that's it right so sounds very easy for a user to understand and all the this management is abstracted away in inside the algorithm itself. But I mean it's easy for user but not not easy how how to create this algorithm. So I figured that the it actually builds very well on top of curve crypto swap which by the way existed since 2021.

um well it just needs few modifications but uh most of the logic uh should leave outside uh outside of the crypto swap and I thought that maybe it's a great opportunity to to make um also a new token when where I would apply all the all the learnings I've got from observing uh CRV and maybe uh to do to do few modifications there. Yeah. Okay. So, um maybe we can talk about where this uh new protocol design works the best in and uh where it might struggle to earn yield and kind of like how exactly those dynamics affect the

protocol design. Um I'd like to talk about that a bit and then we can get into kind of some of the token design uh as well. Yeah. Yeah. Well, I mean I probably makes sense for me to mention how approximately it works and where impermanent loss at all comes from. So, uh, as I said, it's built on top of curve crypto pools which do have a little bit of concentrated liquidity which is automatically managed and it was working like that since 2021. So, um I would say if you optimize it for for APR, it would probably be higher APR than XYK pools, but just a little bit higher, not like orders of magnitude higher. Um and curve crypto pools by themselves would have had the same

exactly the same impermanent loss as unis swap 2. So, this is impermanent loss we are trying to trying to get rid of. Um and well this is impermanent loss is just um a property of rebalancing strategy right so if you uh have any kind of rebalanced portfolio you will have that problem so it's it's just inherent to that um and basically another way to to to look at it is this if you deposit something in curve crypto swap or unis swap 2 let's say in wtc USDT be pulled. For example, if you deposit funds over there, then your

liquidity uh will have base price not equal to bitcoin price but equal to square root of bitcoin price. Right? So you have this could you break that down? Could you break that down for people like why why is that the case that the price over is the square root of the well it's uh it's just a mathematical property of how this this rebalancing strategy works. So um well how to say imagine that you have um I don't know 100k worth of USD and 100k worth of BTC uh and let's say BTC went up by I don't know 2% right and then it appeared that

you have 2% more BTC than USD but the pool it actually should strive to to make itself balanced. So it kind of sells to traders uh 1% of BTC and then you have um 101k worth of USD 101k worth of BTC right but if you are doing that all the time then it appears that later if it's not changed by 1% but changed by 10% or 100% you're uh well the um as you you starts having less BTC. It appears that price of all the portfolio is uh

not like not equal to to sum of prices of these assets but it's actually a little bit smaller and to be precisely yeah it appears to be uh the price of LP appears to be proportional to square root of bitcoin price. Um so and and this is a really annoying dependency. Uh so that because square root appears to be always smaller than a straight line a straight line which goes through some point on this graph. So anyway um and uh so thinking was this what if I try to construct some position which inverts this square root dependency. So to invert square root I need to um to do

what I need to do something which squares it. So if you have square root squared you you have Bitcoin price back and what sort of position should that be? It appears that this position is leverage but this leverage should be not like not typical leverage like where if thing is going up then you're actually having less leverage. If thing is going down, you have you liquidated. That's not the leverage which we want. The leverage which you want is leverage where you keep the leverage constant. So if I don't know, Bitcoin goes down and your liquidity price goes down then you decrease the loan size and sell a little bit a little bit of liquidity to do that. Then your leverage will be always

constant. And this is the thing which apparently inverts the square root dependency. So which does the square which you apply to square root. So, it's a leverage token on uh on liquidity token and that's uh that's what appears to get rid of this impermanent loss. But it appears that it's not as simple because if you get rid of impermanent loss, you actually realize some permanent loss which uh if if you applied that to XYK pools, pools like in unis swap 2, it appears that uh these losses would eat away entirely all your profits which you you would have made.

But for with curve pools, it appears to be not like that. But with curve pool, crypto pools, you actually do have quite a bit of profit left on the table after you do that. And this is what uh yield basis kind of is uh feeding on. Got it. Okay. So in this in in the instance that that Bitcoin becomes more and more of a stable asset, would that be a would that be a beneficial outcome for yield basis versus extended periods of volatility? Yeah, that's a good question. So it should be somewhat volatile but its volatility should be not crazy. So what it appears if uh well this this method works well if concentrated liquidity

works well for the uh for the asset and concentrated liquidity works well. Apparently when has some volatility but tends to return back to some trend line. major crypto assets do seem to return to to to the trend line. So it's not entirely random walks. It's uh they they are kind of gravitating to some trend line of the asset price. So if they fluctuate around this trend line but return to it then they are good assets.

So Bitcoin appears to be a very good asset for that. Ethereum seems to be okay. Major uh major other tokens probably seem to be workable as well. Meme coins don't appear to be okay. If something works as crazy as memecoins, it's too volatile. Uh concentrated liquidity doesn't seem to work well. So the yield you would you probably wouldn't be able to um to make I don't know to um to to make staked uh a staked meme coin unless it's uh well unless it's a kind of well established long living uh staked memecoin right so

um anyway staked me coins are probably not possible but stake um bitcoin staked staked staked Ether um I don't know something with I don't know BNB or whatever that that works that should work fairly well I think and and how about something like stable coins that do have a very flat trend line they tend to always return to that trend line they don't have too much volatility but you know they tend to oscillate around that stable yeah which uh stable coins do you mean yeah well not US and some of the recent ones right but you know USDC USD PT some of the more more common pairs.

Yeah. Yeah. But if you pair stable coin with stable coin, you don't have to do anything tricky. You just do stable swap because the trend line is is just constant. Or maybe if you have yield varying token uh like uh I don't know SUSD from from Athena then the tread trend line is also having a well it's it's going up only. You don't have to do anything tricky. You you do that on a curve already. So it's like nothing special you need to do. And by the way, it's not like really huge yield in in those swaps, but there is some of course it but anyway uh maybe more interesting would be can you do it with foreign foreign pairs or can you do it with

gold, right? And with that one is a much more tricky question because it appears that maybe if you take Euro dollar pair it's uh all the profit is pretty much taken away by traders. So it's uh price is very very close to its trend line at all the times and Euro dollar pair probably doesn't really have too much of a profit sitting in it. Maybe if euro stable coins start getting accepted uh on chain and there is some volatility above the international markets then it probably would work otherwise euro is

probably not the best one. uh so stake euro I I don't really know but also maybe uh maybe for these very stable assets you don't really have to make it a yield basis way you can still have a pool which still has some impermanent loss but prices of your assets don't change too much so maybe this impermanent loss will be really small uh staked gold would be probably more interesting I didn't really simulate that but maybe maybe it's possible to Perhaps staked gold. I don't know.

Ladies, gentlemen, anons, and dens, it is stable coin season brought to you by your favorite, the best in stable coins, fracks USD dollar stable coin, backed by the best in the business, the biddock and superstate. This is stable coin season. It is the yieldbearing stable coin by Fracks, coming soon to a chain near you. One thing that I think our audience really, you know, appreciates about our podcast is that, you know, we we try to ask questions that help extract like the edge in each of these protocols. Um, it's not something that, you know, oh, if some if a token is volatile, right, then this

makes it great or if it's not volatile, it doesn't make it great. And one thing about the unis swap v2 and even curve to a certain extent was look, it's actively you just put it in, it's passive, right? Right. and and yeah, you experience some impermanent loss, but it's easy. So, if you're looking at yield basis and you're thinking, okay, what is the best way to play this game to to get the most edge and use the protocol to to it its fullest capabilities? How would you suggest users use yield bases? Well, I think there are two ways. It's it's very simple from the user perspective. you e either earn real yield so you deposit I don't know crypto some sort of bitcoin

wrapper and then you you earn yield on that so you have um you you get more bitcoins over time that's that's one way but there is another way uh because there is a protocol token you can earn protocol token so what I uh wanted to to make it easier for users I just I'm just splitting between users who can earn who want to earn protocol token and users who want to earn real yield. So if you want to deposit Bitcoin and earn just more bitcoins over time, you you you just deposit and and that's it. You don't do anything. Um if you want to earn governance token, you um you

explicitly choose to not earn real yield. So you can you want you can earn either one or the other. I think the market will equalize both. So so you will probably have a similar APR and it it's really if you want to hold this token and uh if you want to to earn it, you you probably deposit this way, right? Got it. Um and yeah, you you have both options when depositing. So so you can choose the denomination of your yield. This is quite a bit different from curve because you didn't give users the option. Yeah. And different from just traditional yield farming, right?

Where it was all just pool one, the whales would get into pool one, then they would farm the pool two effectively, and then the pool two had zero value theoretically over time and and it was basically a race to see who could get the most size in to pool ones to be able to to basically unload on the pool twos. So, your new token design is basically Yeah. is is a a result of this reality effectively. Yeah. Yeah. So, um well, it's it's just making uh it's it's kind of the same in terms of numbers eventually. But if you have a traditional uh traditional design with pool two, you kind of have to have everyone going to to pool two other

otherwise they are missing out. here it doesn't matter if you go to pool two or not you are not missing out right so maybe there will be uh some discrepancy between APR so maybe some APR will be 1% more or whatever but it's not like not heaps right so maybe some um traders will uh will kind of exploit this discrepancy and bring the APRs closer together by doing that uh but really the idea is that users who don't want this complexity of uh why should I stake and or like oh I forgot to stake so I didn't earn any yield for for several months this did happen in the past and uh

really we pro probably better to not have that so probably better to have users earning uh regardless and if they want to uh earn real yield they they just earn real yield and then they have this uh yieldbearing uh crypto and this yield bearing crypto would be integrable with all the lending platforms for example uh probably much easier for them to integrate because like if there is something which farms governance token integrating that with a would be harder right and so uh probably much easier to have something in the in the token itself and then all sorts of

um you know looping plays appear over there but this would be external to the protocol but also kind of spreading integrations around which is good. So uh um so yeah in brief it would be easier for users and more integration friendly. Okay. So then what was the exact breakthrough around the token model then that that supports yield basis? It's around the optionality for the users with regards to their yield as as we explained. I think yes I think that's probably uh the well I mean the biggest breakthrough here is not in the token model obviously but in the token model I do want to to to do something new and this is exactly that it's uh the

splitting between users who want to earn real yield and users who want to earn uh who want to earn tokens and um I think another another thing to explain here imagine that imagine a situation If everyone wants to earn tokens and no one wants to earn real yield, right? That's theoretically possible. Where would real yield go in this case? So, it should go somewhere, but nobody wants it. Where does it go? In this case, it would go all to the system to those people who locked YB token. So actually admin fee fee of the system it

actually grows up as people stake uh as people try to earn governance token and if 100% of people try to earn governance tokens then 100% of uh yields go to admin fees for the system but of course if there is uh um well the market will find some equilibrium so there will be some split between people who want to earn governance token and people who want to earn real yield. So there will be well some admin fee which probably will depend on the state of the market.

Got it. Okay. So that my next question was where does the yield come from? I think this is one of the more fundamental questions and and it seems that you answered this it comes from the admin fee. Could you just and maybe there's some other fees. Could you explain just a little bit more in detail what is the admin fee and is there any any other fees that you know? So admin fee is fee on top of the profit which the system is making. So imagine that you deposited one WBTC right and uh one WBTC turned into 1.1 WBTC's right and uh admin fee will be a fraction of this 0.1 uh WBTC increase.

Okay, got it. Right. But but since we have only those people who want real yield uh having real yield right so we take this 0.1 WBTC and this goes not to all the one WBTC but maybe maybe half of WBTC sits earning YB token and half sits earning WBTC. So this 0.1 uh WBTC goes to this half of WBTC, not to the whole one of WBTC, that kind of stuff. And and where does the 0.1 WBTC come from? This comes essentially from uh the the actual source of it is exchange fees in curve

crypto swap pool. Okay. Right. Okay. The thing is that this pool spends a whole bunch of these fees on rebalances, right? And also yield basis uh will experience some value decrease like real value decrease due to real leverage keeping the leverage constant and what is left will be will actually be the the profit. So the more vol the more swaps in and out of WBTC on the crypto swap pool on curve will generate more fees and theoretically in a way yes but also volatility of WBTC will generate

uh losses right because you you keep the leverage constant and and there is this uh uh this difference between profits and losses which you are speaking. Right. Right. So, usually more volatility in the pool creates more more swaps as the pool rebalances. But that's right. That's balances out with the leverage rebalancing. Yeah. Yeah. Yeah. But also, I would say that um more volatility at the same time means more losses, right? More losses on uh on keeping the leverage constant on rebalances of the concentrated liquidity, right? So the so and the trick is to have this balance positive

not negative right so as I was alluding to earlier effectively I suppose the longer term thesis for yield basis is is based on the idea that bitcoin has gone through a logarithmic growth curveesque uh phase right with the majority of the day-to-day volatility behind us still one of the most volatile assets assets when you're looking at it from a traditional portfolio perspective. But the thought is for users, if you believe that that BTC is is a core holding that you want to have, which I think all of crypto has uh you know has gone through a whole cycle of like loving Bitcoin, then not wanting Bitcoin, then usually coming back to Bitcoin. um they if you

think that this kind of logarithmic growth curve is basically coming to a more narrow uh uh chart formation where there's going to be uh less volatility on both the upside and downside then yield basis is a product to to take that bet and earn outsiz yield as a result of holding Bitcoin uh itself. Well, yeah. Yeah, you kind of have it interfaced as yield bearing bitcoin. So yeah, you uh yield base itself doesn't make any assumption but uh users make an assumption that uh they want to hold bitcoin. Wait, but doesn't the protocol design make the assumption that this will work better in if bitcoin becomes less volatile like in those economic models that we're working through? I

would say that it's more like a safety assumption which is made and safety assumption is that over time cycle to cycle bitcoin will probably be less volatile and which uh when algorithm is optimized uh when algorithm is optimized then we kind of assume that volatility will be not higher than the highest volatility over the over the data which we used for optimization and the data which we used is data over let's say last cycle right so in this cycle bitcoin volatility will be smaller than in the last cycle then

we are uh totally safe and uh value uh of holdings will in in bitcoin terms will never go down right yes um and which is not the case if bitcoin becomes more volatile than in the past cycle. Uh so then you you could lose like I don't know 5% for example if Bitcoin becomes more volatile than than it ever was over the last cycle. And and just to to to clarify that when you say lose 5% you know hypothetically you're not saying that the number of Bitcoin if I have one Bitcoin I still have one Bitcoin. It's just that the value of my Bitcoin might drop. No no no it's uh I mean it's it's

value in bitcoins actually. So let's say uh if everything works normally like if volatility is not higher than what you observed then you have one bitcoin 1.1 bitcoin 1.2 bitcoin so on but let's say if volatility of bitcoin becomes higher than what you ever observed uh then maybe you know it suddenly becomes from 1.2 two bitcoins to like 1.15 bitcoins and then continues growing after that when volatility is come down. So I tried to simulate that um around different times where volatility was high like I don't know FTX collapse for example. Uh

it was more affecting Ethereum than Bitcoin by the way. Uh but yeah, it's uh something where you can uh you know you you can induce value loss by deliberately tweaking parameters to be too aggressive. So when you make parameters more aggressive, they actually work better at low volatility times, but at high volatility times they can have a kind of roll back of value, not unlike what you have regularly in I don't know hyperlquid vaults. But I think that's not a good model. We probably don't want to have roll backs of value ever. Even if the total APR is higher with roll backs, it's probably

better to not have roll backs. And this is all uh kind of um driven by uh by how the pool is optimized and all depends on volatility and on liquidity concentration. So basically the less the volatility is the higher liquidity concentration you can afford. Right? And if you have something like a memecoin then uh liquidity concentration just doesn't work. It just doesn't work with concentrated liquidity. So for me coins the best is to have just simple XYK pools.

Got it. Okay. So in the in the model then when when we're thinking about applying capital to earn yield in this model um with especially with the new token uh design um are are we thinking about a specific strategy or ecosystem around this yield basis uh product? Well, I think I think our ecosystem some ecosystem around will definitely grow. Okay. But to actually earn yield, you don't really have to use uh the ecosystem, right? You can do it directly.

But maybe what would be really great is integrations with all sorts of lending platforms. for example, you well um I can imagine that uh um looping strategies will be will be very interesting. Okay, got it. And I know this may be kind of a difficult question to answer with regards to the yields, but do you have any expectations or forecasting as to what to expect from the different yields on this product in comparison to other uh Bitcoin yield type products?

Yeah, it's a really difficult uh difficult question because uh well uh hard to predict but at least uh based on the past data I would say that in uh in bare markets like when let's say when the bare market is at the bottom of it you maybe will have a little bit sub 10% I don't know 9% or something right uh if you look at the peak of the bull market let's say if you take the APR over average APR over year 2021 it would have been 60% APR so really large uh if you take let's say the start of the bare

market 2022 um it shows around 20% APR in that time because volatility was reasonably high but actually volatility around the bull market and uh crash after the peak is still higher. So over there you have higher APR than in the start of the bare market and uh beginning of the bull market is also kind of uh probably higher than 10%. So right now um right now we have quite increased volatility thanks to all the all sorts of news positive and negative. They shake Bitcoin in both directions, but it's all good. Uh it's all good stuff. Uh so um I

probably would expect something uh something higher than 10% APR here. Cool. Maybe maybe 20%, but it's a guess, right? Uh I probably should look in practice even not necessarily just simulations. Got it. Michael, really appreciate you coming on. I I do have one more question and and it kind of points to these shakes. You know, we're going back and forth, market, but also like, you know, the sentiment kind of gets shaked left and right. Yep. For someone who's been around since 2019 and built multiple pro, you know, projects that have succeeded and sustained through multiple years and multiple cycles, what kind of advice would you

have for someone that is, you know, maybe a little bit jaded, maybe just getting a little disillusioned with the whole crypto thing? it's maybe not really what they thought it would be when they got into it. What what kind of advice would you give to someone that is kind of feeling this way? Um to to just kind of you know like you know what what has kept you going over these years? Well, I think uh c the ideas of crypto seemed very right by design and also I used crypto fairly often and over time I use it even more often and uh over the last year it's not easy to find a day when I

did not use crypto uh I mean crypto to interact with real world not to just earn not to to yield farm or whatever. So, it's uh it seems like it's well also the the world is kind of kind of more divided today than it used to be. And when the world is more divided, it's um well, traditional finance is a little bit less connected and crypto doesn't care. crypto just moves around and it's the only global thing which which is left and you cannot really make it less global. So I think uh the situation in

the world is such that crypto is becoming more and more useful. Maybe uh maybe some people use crypto only because they used it for the sake of going up. All right. uh but uh really the true purpose of crypto is not just is not an investment vehicle. It's uh it's useful beyond that and uh this is exactly what makes it uh more and more valuable. So I think uh so I think all crypto yeah all crypto is which is fundamentally useful like any any deeply good fundamentals is is something to stick to. Agreed. Last thing on mind is

uh what is the timeline here to getting um yield bases out into the wild? Right. Well, I think uh we are around uh final stages of audits of the core of yield basis. It will take a little bit more time but not too long to get uh uh to get the token stuff out. Uh and after that it can be fully launched. Before it fully launched though, I probably will run some uh little testing prod instance where uh where I want to to look at yield uh being real yield being earned in in in production before actually

spinning the system to be too big via the token because you don't really want to uh uh to scale something before before you try to optimize it. And uh yes, so I just wanted to optimize it a little bit not in uh not just in calculations but also to check in productions in production again that yes indeed it uh coincides with calculations. So this is something we want to scale. Amazing. Thank you Michael for joining us today. It's been an awesome conversation. Looking forward to seeing yield bases out there in in in the near future. Excellent. Thank you, Michael.