Curve Armageddon, Base Chain BALD Memecoin, and Solana DEX Innovation | Eugene Chen (Ellipsis Labs)

Transcript

you know unisoft has been trying for five years to build a fully on-chain Decks that is sustainable and competitive and they couldn't figure it out so they're hoping someone else will figure it out through these through these apis in this lower cost of innovation what's up everyone welcome back to another episode of Xerox research brought to you by the atom accelerator if you're a developer looking for a home in the industry the atom economic zone welcomes you today is July 31st and as always we have a great interview lined up with Eugene the co-founder of Ellipsis Labs the development team that's bringing in next-gen central order book decks to Solana but before we do that we are joined today by Ren and

Westy to discuss the latest Market happenings Dan why don't I kick it over to you for the hot seat or cool Throne yeah we uh we definitely have some latest Market happenings this week unfortunately I'm finding myself of all people putting curve in the hot seat so curve on a couple days well I mean I'm totally like lost track of time over the past couple hours but uh within the last 24 hours and again recording this on on Monday the 31st uh curve suffered a re-entrancy exploit for around 70 million dollars and really quickly on re-engine exploits so this has been an attack that's kind of plague to Define smart contracts since the early days the actually the original Dow exploits that ended up forking ethereum between ethereum

classic and the ethereum we know today was actually due to a re-entrancy attack and what that really means is so when you call a smart contract let's say you're like withdrawing liquidity from maybe a specific pool you really want to go through like what's known as the checks effects interact interactions pattern and so basically when the way the code will be sequentially written in this function would be to like validate all the conditions that are necessary to execute properly so maybe like make sure in the case of withdrawing liquidity that the user is actually deposited funds into that step two would be then the effects piece of this so you're going to update the contract State and record any changes made to any of the data or any state changes and then

lastly you'd actually perform the action so you know perform any calls that need to be made and if like things aren't done in that specific order then you can like reduce this re introduce this idea of re-entering see where basically before the funds are actually moved the attacking contract can basically like call back into the the victim contract um and like essentially get stuck in like a loop until all the funds you know in the case of withdrawing liquidity have been withdrawn from the pool um and so specific to the case with curve uh it's a little bit more unique than just a vanilla re-entracy attack so the code itself was actually written correctly by the curve developers which is actually why this one stings a little more it was actually an issue with the

underlying programming language that caused this re-entrancy so basically they had um what's called like a reentrancy guard so a another function that'll like essentially protect reentrancy from occurring they had that in this contract um it actually was broken so they had like effectively a broken guard uh in in the case of these contracts and it was it had something to do with like the versioning of the compiler that was used to compile the code code into bytecode and so um really tough to see but one thing to note here is that curve uses this that development Community uses a specific smart Contracting language that's called Viper it's based on python the con uh the more popular coding language that

many people are probably familiar with at least the name if nothing else um and so they use this like smaller more alternative coding language that you know it's probably a net positive to have some smart Contracting language diversity within the ecosystem because you could really imagine like if everything was written in solidity and say solidity had a similar exploit to be honest I don't even know if that's possible um but let's just assume for a second it did then having some diversity in the languages used to write smart contracts would be a net positive so that's kind of the idea I think it's also just like in a more comfortable language to use as many people are familiar with how python is written and how the language operates but net net we landed with this issue creating about a 70 million dollar loss

in terms of the exploit it for a curve now it seems likely that some of this was actually collected by white hats instead of the exploiters themselves it looks like roughly 20 million dollars it's kind of just too early to call on what the specific amount will be and we saw a lot of interesting cases with front running so Mev Bots actually kind of came to Curves benefit one of the Mev thoughts is named coffee babe and she front ran one of the later exploits that actually was resulted in around like about eight or so million dollars and there's already been some Communications between the curve deployer address and coffee Babe saying like you know yes I'm willing to return these funds like let's

let's coordinate a recovery so there will be at least some funds recovered uh the the bigger takeaway I guess here is you know the grand scheme thing 70 million dollars isn't huge for a protocol that has about 1.7 billion dollars of TBL uh it's more so the loss of trust that we will have to like kind of overcome from this and again the fact that it was like with the smart Contracting programming language itself that one stings a little bit more that's like a little bit more nerve-wracking it's like how quite honestly I'm just like not technical enough to know like how like the root of that how like how how that gets drawn out how it gets fixed like what are the odds of it happening again like to me that's you know when you say an issue with the compiler like I know very high

surface level what a compiler does so um that's just like a bit harder for me to wrap my head around and so that's kind of the exploit itself and where it's at it's it's kind of a shame in terms of timing because curve has just launched its stable coining lending market and it's getting some pretty meaningful traction with around 100 million dollars of CRV USD debt outstanding um it seems like those contracts are totally fine at the moment at least um but again those are also written in Viper but it seems to be a version not impacted by the specific bug and so there's some interesting knock-on effects here we can dive into but I don't know I guess I should open the floor here and you know I'm kind of curious to get your thoughts like this is an OG D5 protocol that kind of was

viewed as you know one of these Safe Haven places um so like how do you guys feel about seeing you know one of these major players go through like a very unique bug I don't think we've seen a compiler bug like this before Andy's both for us and both for static more chat fire institutions if you're running like a large crypto hedge fund traditionally you and I all know like the D5 which is like synthetics maker uni curve and that entire curved ecosystem and those have always been the protocols that you don't really need to go look at their smart contract dollars to say okay I'm comfortable doing whatever I'm comfortable slopping through these photos or interacting with these protocols and we still know that D5 to

this day is run by a huge amount of whales right for example I think like more than eighty percent of liquidity in every single unit who is held by 10 people if I'm not wrong um and if something like curve is exploitable I think it makes you wonder right like most of us aren't like tech people most such fund managers probably also are technical but they're also wondering if courageous success is if curve is susceptible then who else is kind of open waiting to be exploited out there you know I just said lots of trust I think it's going to be pretty negative to the ecosystem and I wouldn't be surprised if you are like tracking some larger Waters right now and you're just

seeing them pull out of protocols other than curve too yeah I mean I guess we're recording this at like max pay in time it's not even like 24 hours since we really uh kind of like understand what's happening so it's still like very raw for all of us so keep that in mind if some of the facts changed before this episode releases uh but it is interesting to see the knock-on effects here so um her founder Michael eggerov is highly levered against the CRV token um for better for worse or for why that reason is is kind of speculation but the reality of it is is there's about 63 million dollars borrowed in usdt against Ave on about 17 million dollars of fracks through fraxland now obviously

when there's an exploit the token price tends to react negatively and given the risk around this we've also seen it's like a compounding effect right so we have the collateral asset CRV decreasing in value at the same time you know this has kind of been a hotly debated topic for a couple months now so people are aware of the issue of like there's a lack of liquidity around curve and the deepest on-chained liquidity pool was the CRV East curve pool that was one of the exploited pools so now we like lost the best place to liquidate any assets uh any CRV assets specifically through any on-chain pool of course there's still like some other fairly decent options but that was the the top dog for in terms of liquidity and like maybe two

percent depth so we have the collateral asset depreciation in value the loss of liquidity for the token uh and the both fraxland and Ave are like variable rate interest markets based on the amount of utilization for the borrowed asset which uh is the case of Ave is usdt and frax for fraxland both of the lenders have basically like you've seen large withdrawals people being like I don't want to be the last one holding the back here I'm pulling out some of the assets that I've deposited and because of that the utilization rate is running higher when utilization rate runs higher you see an increase in the interest rates so now there's like you have the interest rate rapidly increasing pushing up the

total amount of borrow as the collateral is falling and again we lost the liquidity to liquidate positions into so we have this like triple compounding effect of like oh [ __ ] there might actually be some major knock-on effects of this um the both of these do at the moment have pretty high Health factors around 1.5 ish a little higher than that for the Ave one um but we are seeing it looks like if one was going to get liquidated first it's likely to be the fraxland one because that position um fraction is a pretty interesting uh interest rate model where there really is no upper bound until about 10 000 APR and the way the model works is like it effectively doubles every four hours or

so when the utilization is at 100 I believe uh and so like this thing is utilization right now I think is like 96 so about like every six or eight hours you're seeing a double in the interest rate which is currently up to like 46 47 last I looked um so if that goes there'll be you know whatever's left at that point in time in terms of collateral that needs to get liquidated to keep that position solvent so if that one triggers I imagine you'd see the second one triggered through Ave and I as pretty good likelihood that Ave gets stuck holding the bag of bad debt here um so we have it like it's like just this chain daisy chain reaction of bad things right because if Ave gets stuck with bad debt then the users that are in

the staking module would be open to like sort of filling that back debt unless they use like protocol reserves and again it like very much so matters at the size of the bad debt that avi's left with um if anything could be done here so again still very too early to call it but there are some very important knock-on effects here to be monitoring yeah I've also noticed like basically as we're speaking igorov is selling his ldo tokens at like ten thousand dollars a clip which is probably him hoping to recover some funds to then pay back these loans before the interest rates get too high and he does get liquidated so yeah I I mean I look at these not second order effects these knock-on effects and it does look like if

nothing's done on liquidation is going to be likely and as a result like hopefully we get it resolved the situation soon because yeah this could have second order effects for for Ave for for excellent um and it's just something you don't want to see I think one interesting thing that I'm thinking about is if you're faxed then or your facts uh and your Ave you are like the governance token holders what do you want to do because say say you kind of acknowledge that it's over like this guy's gonna get liquidated somewhere and there's only so much liquidity between these two lending and bottling protocols or whatever like lending involving protocols that Mitch has a position on you don't want to be the last person liquidating his position

if anything you want to be the first person uh but none of his positions on any of these protocols have crossed the sort of liquidation threshold yet so do you kind of uh liquidation just avoid protocol bad debt or you just let the markets do a thing and kind of see how it plays out I think it's an interesting thing that governance token holders are considering right now yeah no it's uh I think I actually misspoke it I said it doubled every four hours I think it doubles every 12 hours nonetheless it is increasing at a rapid Pace uh specifically as it relates to the fraxland pool um so yeah like uh it'd be a real shame if this exploit

like Daisy chained into defy imploding um but again maybe that's just the the harsh reaction given that it's it's less than 24 hours old in my mind hate to see it though like her was on such a great pace and they're doing such great things I really like that's why it sucks that we're seeing like this huge leverage I think play out because you know largely I think now like okay the exploit has passed it wasn't great but I don't think it's the end of the world but now like you're gonna see people hunt this position that's like the downside of having everything fully transparent you can quickly do some math and figure out what the liquidation value is or the liquidation point for this fraction alone is and like it's

going to get chased which is like just how these things work which I think it's about 33 34 cents for CRV um given the current market conditions but again interest rates climbing which pushes that liquidation Point higher which makes it happen sooner so I don't know man it uh it doesn't feel good right now that is for sure yeah I saw a tweet from roon uh created maker who said this might be sort of like the black Thursday moment of this cycle and that D5 will bounce back as a result and I'm hoping that's the case I'm feeling that's the case that like after sort of this washes washes away that you know there really is no negative uh occurrences on on the horizon so D5 can sort of flourish from there does anyone

have any possible guesses as to why egrov is so levered on his protocol's native token like that's the one thing that I just cannot for the life of me trying to understand why why would he do that like if if that if those positions aren't open this isn't even a thing like we're not talking about this for 20 minutes on the intro like that yes it would still be bad but now it's kind of got the potential to turn into something so much worse so I don't know if anyone has any theories there but I'd love to hear them I've heard that um Australia's weather is quite nice and the real estate market keeps on going up so um but on other more serious note I think I think if

you're a founder and you're holding your protocols native token or if in the case of sort of slightly more web to chat Fire Company you're holding your company's Equity you don't really want to sell that because like that's just a bad luck in general and one thing that rich people do really really well is make their assets efficient you know they probably can they access financing opportunities that normally it's like good me can't access and that's one way they compound the wealth so rather than starting his curve you know he's bought usdt and whatever he's thinking of his usat is kind of like none of our business but as an example right the founder of Peloton at one point like this guy was Giga loaded he

borrowed against his stock to go by his private jet to go buy his houses you know if it keeps going up and you don't get liquidated fine but obviously pet isn't it's not doing so well these days so one day if I'm not wrong Goldman Sachs comes knocking on it so I say hey we're gonna have to liquidate you're gonna have to call back your private jet your house I mean when it goes up it's fine um but when it goes down I think that's kind of bad and I think most people when they borrow against their assets I don't want to say they get greedy or they're probably like a bit optimistic and they borrow against it into like a liquid assets like purchasing real estate or whatever and that means you can't just pay back your loan like

within a day or two yeah I'd agree with that rent on your last half there um you know like look if he if we make it out of this and like whatever reason curve price recovers and life is good and you know bear Market kicks in in a couple quarters and resume up but only then yeah he made Max efficiency with the assets he had and like you know it's is it if he just dump the same amount of CRV tokens like you'd be like oh the Founder's quitting on the protocol like that wouldn't really work so it's like almost not an option um obviously being in the tokenomic style that curve has the CRV price is fairly important to its future incentives and the attractiveness of the protocol so again dumping the CRV tokens really wasn't an option there either

um so like this was basically his his exit I don't think that in a malicious way okay I think like that was just a way to get liquidity and you know it's like you gotta anytime you open leverage like you're writing a fine line and so that's just kind of uh where we're at right now yeah that's a good take I guess too like the curved devs like out of any team I feel like they value decentralization and like really sticking to the ethos of crypto more than most groups out there so I guess kind of reaping what you sow and taking out loans in Ave kind of supporting I guess just the broader ethereum defy ecosystem might have been a an angle there as well but anyways on to more uh Happy topics let's who's got a cool throne for today who wants to go next

a few months back voted on their die savings rate they increased it from some negative number to 3.49 or your daily earlier this month they toned it down a little they reduced it to 2.19 but a recent governance vote just went through that basically proposes a new thing called the effective die savings rate and that's basically when the amount of dye being deposited into that the die savings rate vault is low they'll sort of use all of their income and sort of concentrate it into Whoever has deposited into the die savings thought the kind of like the fractive model that kind of

um but basically when the vault is at a zero percent utilization you get an eight percent interest rate um just some extra numbers when the utilization is 50 you get a 4.16 and right now the utilization of of the vault is around 10 there's about 360 million in the fall and 4.5 billion in outstanding die Supply so I mean this industry could be pretty good if you deposit it right now I want to say like six seven percent compared to the current Reserve rate of 5.25 to 5.5 and I think that's a growing Trend that we've seen lady in crypto right like I want to make every token Capital

efficient in the sense that if I can make an interesting version of it I'm going to make an interest-bearing version of it and I'm going to make this interest rate as high as possible um but yeah basically the whole goal of this is to grow the demand for die and the die savings rate to ensure a growing user base that's ready when maker dial sort of launches out of their sub-dials and participate in the whole end game plan my my only concern with this is that within crypto just becomes part of a larger movement to chase yield and historically that hasn't ended well for Chad fight when everyone starts chasing yield and starts taking on more risk so we'll see but so far it seems like the reward asset portfolio is decently stable in the grand scheme of things and

it should be pretty stable for us the time being now I like the move here I think the the DSR as a whole is a really exciting development I think like we need to get off chain yields on chain that's that seems straightforward to me I definitely probably flip my position from maybe 12 months ago where I uh I was I was like no we're gonna build our own on-chain ecosystem and we don't need to rely on off training yields and well that's probably true in the long term I think there's no reason not to try to pull uh these long existing institutional yields on chain I think that's really exciting um I have a lot of opinions about how that's going to happen but nonetheless it's exciting to see maker kind of using this interesting proxy with with the DSR rate so

um we talked to Sam McPherson a couple months ago and he like talked about how the dsrl change defy one thing that we haven't seen happen is we had like crypto's not efficient that's the tldr here but we haven't seen like for example you can borrow uh die on Ave and just stake it into the die savings rate and earn that yield um and based on the current bar rate you get like a one 1.2 percent spread which obviously is not huge but you would think that would have closed by now you know it's been a couple a couple of months live so I think like the goal here uh is really just to like kind of put the DSR and the like the ability to earn that like put that into the

Forefront here um they took a very interesting approach with how they're going to get this in so it's kind of like this One Direction one uh this yeah like unidirectional uh process where once the utilization of the savings rate like eclipses a certain uh percentage then they die savings will like decrease it's essentially getting back down to that three-ish percent range um but stablecoin yield they're like eight or so percent when this thing first launches is pretty attractive so I do think we're going to see a meaningful move into it a second order effect of this that I'm like curious to see play out I I can't verify the accuracy of this so please like I saw a tweet about it five minutes

ago before we were recording this so definitely go check this out make sure it's right but um I guess I saw Richard Hart from the founder of hex and pulse chain and all that he has like 700 million to die or something which is like a meaningful percentage uh Supply a percent of the total Supply um so that's pretty interesting like why does he have so much dye I guess you know he probably wants something that can't be blacklisted like usdc or usdt um I don't know man that that's just a cool random thing I saw with the ice I definitely need to verify that yeah another thing with the DSR is it's almost like like you said bringing off-chain yields on chain and there are other ways the protocols are doing this but it's usually through like some kyc

system like ondo Finance like bringing the treasury yields on chain but you have to kyc give your information whereas I think having a proxy via dye and Via its lending protocol I think this is just a better way to have like more permissionless access to to these off-chain yields so I I also agree that this is a great way to grow the dot Supply and I think it's going to be very attractive I could take it over for my hot seat for the week I've got a Lido dual token governance in there so I actually said before the merge that I thought or sorry yeah before the merge that uh I thought the the Lido dominance rate amongst staked eth would go down just because it's hard to keep up the

percentage dominance in absolute terms when you're already absolutely huge but we're still chilling at 32 percent it's been what 10 months nine months something like that since the merge so there's kind of resurfacing conversations that got brought up in June of last year of giving steath holders the LST of Lido itself the ability to veto governance uh proposals in Lido governance so I don't really know how I feel about this to be honest because without um Steve holders you really don't have any value for ldo so like yeah I don't really know and then on top of it like values only value proposition light is only value proposition sorry is pretty much to be active in governance today

um you know you think maybe at some point in the future they turn on some type of fee switch but even their revenues aren't massive today by looking at the in relation of the the multiples that ldo trades at based on current Revenue because they only get five percent of the um six percent staking yield on all the AUM so it's actually a lot smaller number than a lot of people think um and I also think like in general LST providers like yes it's not great to have one have 32 Mark and share but at the same time they help client diversity it's better than having large centralized exchanges hold all the youth Supply um so I I don't know if it's really all that necessary I personally believe that the the Lido dominance is a little bit overblown um so yeah I don't know what do you guys

think of it yeah this has been in the works for a long time now I think you said June of last year and something that I've always been a huge fan of I mean there's a big contention within a portion of the youth community that like Lido having 32 market share and all East staked is a problem and it's continuing to grow and that they should self-limit in some way and a lot of these concerns come around obviously one entity having control of that much stake obviously people on the other side say well if you look at how their stake is structured there's many different entities and they're adding DVT and ways to basically decentralize their or operator base in a way that

it's really not one entity but at the same time if you have one token that governs all of the state eth you really do have centralizing power in that way and so having the ability to have steath holders who are more so aligned with the ethereum protocol than they are a lot of the Dow that allowing them tell power that if there's a governance proposal that obviously negatively affects ethereum as a protocol and therefore it's teeth holders that they're going to do the job and veto that proposal like I think having them have this power or maybe they can't vote on proposals but at the very least they can provide a backstop I think is a really good idea and helps Lido to like sort of self-limit in a way that's

allows them to maintain large market share while remaining decentralized and good for ethereum I think one interesting point about D5 is everyone kind of say Defy is very composable right and in Ideal World like just crypto as a whole is decentralized but part of me thinks that because Diva itself is so composable and that means kind of network effects are Amplified in a permissionless in the overt manner right that kind of ends up leading to sort of more monopolizing forces then perhaps like it's the same thing were to happen in like web 2 or like outside of crypto I don't know if that's a slightly

controversial take but it's just like a random thought I had yeah it'll definitely be interesting to watch this play out I mean Lido will be will be with us throughout the thick and thin of D5 so definitely rooting for him I just don't really understand that like like if a leo tokenholder is ultimately vote on something that's negative for ethereum then the value of the ldo token goes down already so the incentives are already relatively aligned in my opinion but I guess yeah if you have the the Steve veto power like like whatever like they're supposed to be aligned anyways so what's it hurt it's not like they're gonna just veto things Willy willy-nilly so I guess I'm with USDA I like your take who you got in the hot Cedar cool Throne though yeah absolutely can finish this off here

it wouldn't be a week of Xerox research without coinbase and in the cool throne but this time it's their new L2 base so they just launched this week for anyone to bridge funds over and they had over 200 million dollars in Daily Dex volume over the past 24 hours which is higher than any chain besides ethereum and binance smart chain although Arboretum is pretty close and of course since the Chain's only been live for a week uh this is mostly the result of copy and paste Dex protocols and meme coin speculation most of all the most popular of these meme coins was of course called bald which was up over 8 000 in just 24 hours uh before being deployed but of

course um right as sort of the bridge it was a one-way Bridge up until early this morning and as soon as the two-way Bridge was enacted uh the deployer of bald basically he bridged over a bunch of eth to make it look like he was about to buy more balls and put up the price and basically bait people into buying more bald before rugging the liquidity um but yeah it was definitely a fun weekend if you're paying attention it feels like in crypto if you blink you're gonna miss another 100x opportunity even in this bear market and this is just another one of them um yes I'm wondering what do you guys think about all the base speculation that was going on did you guys bridge

over I didn't reach over myself but I think it's just kind of a funny story in how crypto is so unique like for example if you're bringing a consulting company right and you tell them okay I have this platform I want to sort of like grow its adoption and get meaningful use of metrics like what should I do they're gonna come back it's like oh you know like social media ads like have people building on it and then someone launches a meme concave fault and it's probably 10 times more successful than whatever plan they may have come up with or even probably like the build on basis team on Sanic I'm sure they had like a nice like fancy like oh today Face is Going live on Main net for our users here's the 10 list of protocols that we'll be

supporting on day one and chances are that's probably not going to be successful as whatever happened over the weekend so I just think that's very funny and unique to crypto I think the second thing is I'm not really very sure whether the bridge was like shooting one way I think there was like a way to cut a smart contract and you get your funds with Drew after seven days but most people didn't know that I think um like the whole of yesterday it was just like ten thousand threads of my timeline saying oh you know here's how you reach your base it's a one-way Bridge you send your funds like zero x49 blah blah blah whatever and I think the takeaway I had from that was that basically like no one cares about decentralization or like sequence your life you have like these like VCS these like info guys saying

like oh I have to cut out uh but big boss Mike here and he's saying like oh you know like ethereum is like not decentralized because uh the sequencers are centralized blah blah blah like there's no secrets to likeness but like 99 of you would have Vision faces so that I guarantee you did not care that you could have told them that the bridge was locked for say like two weeks or one month and I guarantee you still would have seen like the 50 million dollars of info and I just think that's hilarious I think the the counter argument therapy the type of user that bridged yesterday is incredibly specific that is like someone who's comfortable being on chain and is going with one purpose and that is to chase meme coins so like that user

I couldn't agree more does not care um but I do think like you know if you're going over there to use like a lending protocol or you know perform some Financial action like yeah and then like that's the kind of user that would probably care a little more rather than you know just a degenerate Gambler um not there's anything wrong with that but that's just two different types of users that uh have different preferences so I I tend to think that the at least in the medium to long term like the person that does care is probably a larger group uh I don't know that's actually a great question like do more or less people carry in some of people who like use a blockchain I actually don't know I gotta try and pull the receipts but I said on Xerox research

like two or three months ago I was like I bet All Season I'm basically gonna be insane because there's literally nothing else to speculate on and I love that it actually panned out I bet it's gonna get even crazier but uh yeah just the fact that that many that amount of money flowed over there without incentivization is is pretty insane and I'm really really excited for like the official mainnet to open up because I I'm just really intrigued to see how much money goes over there just to try it out because we really haven't seen like any type of organic non-incentivized behavior in crypto and I think everyone's super super excited about bass um but yeah I mean just wild stuff I can't like it's part of me is like man like we deserved this curve exploit like look at what we're

doing right now we're just trading bald on a place that's basically a black box where we can't get out of like it's just crazy sometimes when I step back for a minute and really look at the state of things I'm coming home from a nice relaxing Beach vacation and I checked slack and it's like the curve exploit like at Dan what's going on like is this curve Armageddon I'm like dude come on I'm literally driving on the highway right now this is a terrible time for this and then I like pull over you know get gas or whatever figure out curves getting [ __ ] and I'm like this is terrible and then I like scroll through Twitter for five seconds and it's like you know people making millions of dollars off of a 500 purchase of bald and I'm like man it is two ends of the spectrum one one side

you just have pure degenerate gamblers absolutely crushing and on the other side you have like Giga brain devs trying to you know front run the black hats before they can continue their exploit like it's heren's Point like this is just such a unique industry man it's For Better or For Worse it's it's fun to be here there's no doubt about it and I'm not [ __ ] leaving love to hear that love the enthusiasm Dan you want to tell people what's good with the atom accelerator yeah for sure and you know as always love to give a shout outs to the great atom accelerator for sponsoring our show if you're a developer looking for a home in the industry come build in the atom economic zone interchange security gives you the ability to launch with the economic security of the cosmos Hub rather than trying to bootstrap your own Economic

Security IBC gives you the ability to interrupt with other chains already building in the ecosystem and we've seen like Duality and stride really take advantage of this and I'm kind of like helping push forward development within the atom economic zone Noble is going to bring native usdc into the ecosystem and really Kickstart D5 by having that native Union of account in that ability to be have a stable coin right people love stability especially in the heat of a bear market so if you think that you have an idea that could help the out of economic zone it would like to get looking uh look for funding to help kick start that idea the atom accelerator is doing grants on a rolling monthly basis ranging from ten thousand to one million dollars be sure to check out the link in

the show notes for more information on how to get Inc in touch with that team but now on to the awesome interview with Eugene from Ellipsis Labs all right we're joined by Eugene from Ellipsis Labs the development team behind a solana-based order book decks called Phoenix thanks for coming on Eugene thanks for having me yeah so I was hoping we could just kind of get right into it and get an overview from your perspective of you know what Phoenix is and and why you decided to build it yeah absolutely Phoenix is a spot limit order book decks on Solana it's fully on chain so there's no off-chain component whatsoever the matching engine all the limit orders the order book um all lives fully on chain um

you know we we started working on Phoenix as a result of a couple of things first of all I think uh you know serum originally was this limit order book Dex on the Lana it was actually one of the first applications put on salon and it really showed why having a high throughput low fee short block uh blockchain enabled much more in D5 much more efficient protocol designs than what we see on other blockchains that are much more constrained on the performance side um but Serum is sort of not under active development for like two and a half years there's some fundamental architectural things that are uh you

know wrong with it where if you're building on Solana in 2020 I think there's just a lot of uh knowledge that's kind of common today that wasn't really understood back then I think uh Sarah was always sort of supposed to be this V zero kind of proof of concept type thing and we're really productionizing that today so we have gone and taken uh well we've written Rewritten the entire order book from from scratch we've really focused on the market maker developer experience so making sure that their interfaces are as as clean as possible uh fixing some of the core issues around settlement so

settlement is now fully Atomic and a bunch of other quality of life improvements okay I see and why did you choose Solana over maybe like an Ethel 2 or a Cosmos app specific chain like what is it that's unique about Solana in particular that you're like all right this is the best spot to build a limit order book on the is unique about Solana today is really just the high throughput and the low fees when we started the company so we're not religious about infrastructure we're not religious about what blockchain we build on on top of and when we started the company all we really knew is that we wanted to build Financial Primitives that are fully on

chain that don't currently exist and we did a pretty thorough survey of the infrastructure that was available at the time looking at the l2s I think they have a pretty promising road map but there's still a lot to execute on including decentralizing sequencer implementing fraud proofs or permissionless fraud proofs and the throughput is still just not high enough and the fees are still too high I think after EIP 4844 they'll probably still be maybe an order of magnitude maybe a little bit worse than that um too high to support active liquidity on an order book and then the app chain question is interesting one as well we really wanted

to build first spot liquidity as opposed to derivatives liquidity and I think for Druid is liquidity um or for a derivatives exchange it makes a lot of sense to be on an app chain you sort of bridge one time you only have one asset you really care about which is whatever base asset you denominate all your perpetuals in uh and I think it makes a lot less sense for a spot where kind of the idea is you'd have to be bridging in all these different types of assets because they're not going to be native to your app chain you're going to sacrifice Atomic composability which you know when we ask when we actually get to the question of why does liquidity need to be on chain I think a big part of the answer is

um this demand for Atomic composability which we haven't seen yet today but I think in a world where crypto and D5 flourish we should see a lot of demand for atomically composable liquidity and so that all points to being on a monolithic blockchain as opposed to an app chain I want to dive a little deeper into that right because there's kind of been this narrative coming about that's like uh you know Atomic composability is pretty overrated and you know that all gets Dad arbed away anyways but that's like really coming at it from the Mev lens but what you're talking about seems to be more like okay well why are we coming at it from that lens like there might actually be new use cases that can be derived from from having that composability between applications and

that obviously ends up ending benefiting the end user is that kind of how you're thinking about atomic composability yeah exactly um I agree completely on the med side I think when this cross-chain Mev narrative started to pick up about a year ago I have some tweet from probably last August that said yeah cross chain atomicity or crosshand Meb is overrated it's all it's all basically going to be stat ARB in the same way that on ethereum mainnet today most of the Mev is into this uniswap Sushi swap Atomic Arbitrage actually unisot versus binance and that's done in this non-atomic way obviously and I think the same will be

true for cross chain Mev you're not going to see too much optimism Arbitron decks Mev you're going to see optimism binance and binance arbitrum um but when we talk about use cases first of all this is all highly speculative because I think like no one can really say with too much confidence what the dominant use cases are going to be and I think that really does end up informing the infrastructure choices that we make so one of like this big one of these big bets we've made with this company or at least developing this particular protocol is like yeah spot liquidity is going to be important and it has to be on a monolithic blockchain because there's going to be demand for Atomic composability and I think the shape of the the shape of this demand can

you know look pretty different in in different Futures uh one example is a lot of let's say a lot of Finance moves on chain um and you have a lot of financial products that tap into a bunch of different protocols on chain maybe that's exposed to the user maybe it's not uh you can imagine that some of these Atomic actions end up uh hitting many different protocols perhaps in like a single transaction or a single larger larger action and then you can also Imagine a bunch of other types of applications that make it fully on chain that are somewhat financialized that end up with some sort of trading component where you do need to have this

um liquidity that is that is fully on chain this is pretty similar to this like OG D5 maybe not OG but it's like 2020 Vision that was articulated for D5 which is you're going to have these on-chain fully on chain right transparent uh composable D5 Legos D5 building blocks we still believe in this Vision I think a lot of others in crypto have kind of kind of given up on it uh I think especially when you're so limited by the infrastructure there's just not that much you can do and so coming over to the Solana side which is not where my you know crypto background is from is sort of this breath of fresh air where all of a sudden all these like super

heavy constraints that are on you as a protocol developer are just gone and you know they're replaced with a different set of constraints of course and there's all sorts of other problems we are still quite early on on the infrastructure side but as a D5 protocol designer there's just like so much more you can do when you don't have to deal with like 10 20 TPS you don't have to deal with every transaction is going to cost you at least five bucks but usually significantly more yeah I I wanna I feel there's two different directions you can take this now but you know you mentioned the technical constraints of building on maybe like ethereum for example and uh I want to get your thoughts on how that influenced your decision to go Club versus amm because on ethereum obviously we see some increasingly complex amm

designs but you know we haven't really found like the Silver Bullet that could probably like outperform a club so uh can you dive into like you know a why where you think the amm versus Club debate is at and you know be what what exactly what constraints are maybe lifted when you're building in in Solana that it kind of enables more Innovation yeah let me start with the second question how does the high throughput and low fees enable Innovation it sort of allows this active participation in the market that is still going to be somewhat cost effective so as an example today on Phoenix we will see on a single Market anywhere from like five to ten uh updates on a market which is like placing a new limit

order or canceling an order or some trade goes through we'll see like five to ten of those every second and we're still in the very early stages here so we expect that number to go up pretty significantly uh as the markets become more competitive and as more users come on this is just not possible on a chain that does 10 or 20 TPS you just are not going to have 10 TPS going through a single decks or if you are the the cost is going to be prohibitive so our market makers who are updating the orders I believe every half second sometimes every second they're paying on the order like twenty dollars per day uh for gas which is like the cost of one add liquidity or remove liquidity on unisot

V3 so when we think about the design space uh it just is so much more open because we're not constrained by these by these heavy costs and I think a lot of the like when you think about like why did am s get developed in the first place they were born out of this constraint right no one in 2017 or 2018 when AMS were coming around I thought that oh yeah you know all else equal to amm is a superior design to what exists on the centralized exchanges but this is kind of an approximation and there are kind of some interesting properties uh of this XY equals k thing let me put it on chain but most importantly you can support

this passive liquidity versus active liquidity and so you don't need your market makers to be updating their quotes super frequently it just you said it and forget it and I think that really was quite innovative and it does make a lot of sense for this 10 20 TPS train and uni B3 is another pretty significant Improvement over the the XY equals Kamm but you end up in the situation where the liquidity is also just not that competitive with limit order books that exist off chain and so this is why for pretty much all of the trading pairs that binance has listed or coinbase has

listed or any of the other centralized exchanges the price Discovery is not happening on chain the price Discovery is happening on these venues that don't have 12 second block times where market makers can put the liquidity exactly where they want to rather than being so constrained by the the preference space that you're allowed to choose from when you're providing liquidity on an A M and so I mean I think like the limit order book looks very different when it's on chain there are like a ton of concerns with Mev and discrete block times in the same way that these are pretty big challenges for amm to overcome we certainly or I I

don't want to speak for the company here I would say I personally don't think a vanilla limit order book like the one we've implemented today is good enough to be like the final form of D5 liquidity but I think it's a much better starting point than the amm I think the amm design space is fairly restricted especially when you're on ethereum mainnet I do expect to see a lot more innovation in Dex design as more and more uh defy activity on the ethereum side moves to l2s and I'm pretty excited to see that so you mentioned Mev there I was actually hoping to peel that back an additional layer there's a there's a few differences between me be on salon and

Meb on ethereum uh first blocks are usually built continuously on Solana as opposed to built discreetly and so concretely what that means is as transactions are coming into the leader of or the proposer of the block they're sort of building the block on the Fly whereas on the ethereum side say the previous block was committed at time T equals zero all these transactions all this private order flow builds up until say like T equals 11.9 seconds and then the next block comes out at 12 at the 12 second so this block is built in this very discrete way now I'm not sure if this is always going to be the case for Solana

um but it does really reduce the amount of Med Dot is exposed so when we come to the Dex design in particular there's a there's like a few forms of Mev that really harm the amm and when I say harm I'm not talking about like necessarily to Makers or to takers but sort of you have this Dex ecosystem which is comprised of makers and takers and say the the protocol itself and then you have this base layer the so where the med is accruing to and that's sort of extracted from this from this isolated system and

so on ethereum today you know uniswap loses hundreds of millions of dollars perhaps billions of dollars per year to the base layer which is partially Mev that's that is toxic to the makers partially Mev that is talking to the takers and obviously a big portion of it is just gas fees as well uh and so kind of the goal for a protocol designer is to make that number as small as possible so the faster block times help a lot having active liquidity helps a lot because the makers can move their quotes themselves as opposed to having the uh having like a toxic take or come and fix your quotes for you where the Surplus value doesn't actually end up with the taker because it's

competitive it actually ends up going to the base layer uh and then some of the like you know sandwich attacks are not really a thing in in Trad by front running uh if not regulated away can still be pretty bad for takers but not nearly to the extent that it is with sandwich attacks sandwich attacks are uniquely enabled by this two-sided liquidity that we see with a M's um and so we sort of expect there to be far less Mev on limit order books now that being said we're certainly not like declaring Victory on on this question I think there are a ton of open questions

related to Solana and maybe some of which we've written up and posed on the uh Umbra research site the the the competition today on ethereum and maybe is much more sophisticated than it is on the Solana side and so I don't think we can say too concretely with a ton of evidence that uh you know the externalities of Meb when D5 becomes much more competitive on Solana those could still be like crippling to the chain we just don't have the evidence to to show that it's that it is or it's not right now and so there's a lot of research work that still needs to be done to to to make the chain robust to to more sophisticated greedy actors showing up yeah that's

interesting and uh one is I definitely recommend umber research to the to the listeners here that you guys have done some incredible work there we'll put a link to this in there in the show notes to that at least to the Twitter page if nothing else but um I want to see earlier you mentioned that there's uh like sex to Dex Arbitrage and instead of going like uh you know say ethereum to arbitral like decks to decks between ethereum and arbitrome it goes like between binance and like the market maker kind of becomes this bridge like why is that and yeah I just like why is it not really dexodex but more like using binance as a central point and bouncing between yeah so this is one of many areas where I think a lot of the discourse has been

informed by this path dependency in the past so Mev originally really was thought of as atomic thing when you go to like the very early days 2019 2020 sort of uh coinciding with this original booming of D5 when you think about the type of person that was a Searcher or you know really likely to be playing on chain in 2019 and early 2020 they were far more likely to be like engineers and hackers and much less likely to be you know Finance guys and so here are the the atomicity guarantees actually make a lot of a lot of sense where yeah your trade might be a little bit worse but your worst case

is really really bounded you don't need to understand how to hedge out your risk on binance you don't need to do inventory management um there's like a lot that goes into sex to Dex uh stat arm and yeah I think there was a lot of like leaning into this you know Atomic narrative where actually the stat are sex next side was has always been growing uh we did a piece on this where we looked at some of the numbers on Frontier research where uh yeah we looked at the market share of sexadex Arbitrage versus Atomic Arbitrage on chain on ethereum mainnet and for the liquid pairs so for like tokens

that are in the top hundred by market cap almost all of the Mev that's extracted is done in the sex to Dex way and the fundamental reason for that is because there's far better liquidity on binance than there is on a uniswap or a sushi swap so usually what is happening in one of the there's two ways these sexadex moves can occur the first one is someone makes a big trade on chain so they move the price of a pool and so now that the price of that pool is no longer in sync with the binance price so someone is going to come in and fix that price and then they can either hedge atomically on chain or they can hold a little bit of risk and hedge it out later off chain

with the understanding that yeah you're going to hold on to some risk but maybe the total cost to hedge is lower so you can bid a little bit more in these bundle auctions that are happening off chain and so this is sort of the back running case and then the other the other way one of these sex detects opportunities can open up obviously is if the price on binance moves so the price on binance moves even if all the prices on chain are still in line with each other there's not going to be an atomic ARB that's opened up but all of the prices sort of need to be fixed to go to the price on the the new quote unquote correct price on binance and there's a lot of nuance here like

what do we even mean by Price Discovery what do we mean by correct price um but really what the trader cares about is where they can hedge out and how much how much size they can hedge out and this is a pretty sophisticated operation where you care about okay so I I tried to do this trade I have some probability that I that I'm going to get filled at some point I know I got filled then I might have to move inventory around if I don't have a ton between the chain and and a binance or whatever centralized venues or other venues I'm hedging out on maybe I'm even hedging via OTC uh maybe I'm hedging via Futures so this is just like this really um intensive operation and yeah a lot of it is figuring out how to play these

on-chain games but I think in general especially after uh the merge when we went from proof of work to proof of stake and then we entered this PBS world there's a lot less like pure on-chain Alpha and more of this like uh Meb supply chain as I like to say is doing this stuff that is off-chin like how do we figure out what is the fair price for ethusdc or how much you know how much am I willing to take um uh some amount of risk and there are already a bunch of types of companies that are really good at this and it's not necessarily going to be these crypto

native Mev shops it's actually going to be these training firms that I've been doing this for 20 30 years yeah and that like kind of coincides with the launch of UE swap X and their like RFQ style um release and so we're kind of seeing this mini narrative play out there it's like yeah so we need to do we need like sophisticated actors to perform you know this action off chain and then let's see if we can like bring that back in create an Efficient Market on chain do you think we can like get to a point where we can have price discovery on chain with like the liquidity and like the Dex construction like is that is that a possible end State here I think it's possible that's certainly our goal is to build a venue where yeah at least to some degree price Discovery happens uh

right now first of all there is already a lot of price discovery on chain it's just in these all these tokens that um binance hasn't listed which is the vast majority of tokens but not the majority of tokens by market cap or or trading volume or anything like that and we would love to be in a world eventually where uh for all the major tokens and major assets not just crypto native tokens right uh as much of that price Discovery is happening on chain as possible and what you need for that is a couple things first of all like as much of the um the non-toxic flow is is occurring um on chain so this comes back to like providing competitive products that are better than the centralized alternatives

um with regards to uniswap I think there's a couple ways to look at it right so they've had a they've had these two big product announcements one of them is uniswap V4 one of them is unisot X um unified B4 is introducing these hooks on top of the V3 model but you get a lot of the Primitives that uniswot V3 had for free so I think it really does like uh reduce the barrier to Innovation on the ethereum deck side I'm not sure how much room there really is there especially on ethereum Mana in this capital or sorry this compute constrained world

I think there might be some really interesting stuff that happens say on uh on arbitrum uniswap V4 uh and another way you could look at it is you know unisoft has been trying for five years to build a fully on-chain Decks that is sustainable and competitive and they couldn't figure it out so they're hoping someone else will figure it out through these through these apis and this lower cost of innovation you know uniswap X introduces this off-chan component uh they they call it the the relayer so essentially the user will submit an intent uh which is effectively a signature that allows someone else to

take the signature put it on chain under some constraints uh with these limit orders where the price sort of decays over time and I think from an architectural perspective this makes some sense but when you really look into the details it starts to make less sense with these like 12 second block times your Dutch auction is just not going to be that efficient I don't think there's anything uniswap can do if they're sticking with ethereum mainnet I think this might look a lot better uh again on an L2 um and then they also have this permission like totally permissioned RFQ system that lives on top of it that's actually

so if you go to app.unisoft.org you turn on unisob x the thing you're hitting first is not unisop X you actually end up hitting their permissioned RFQ first and this is like a set of five or six professional market makers that are giving you a quote and then whoever wins who whoever gives you the best quote if you accept it then it's their responsibility to land your trade on chain and this really feels like giving up on price discovery on chain where you're saying all right we can't do it so winter mute is going to find the best price on binance or any other you know jump trading or whoever else is going to find the best price on binance and they're going to return that to the user

and sort of be this bridge between the the chain or unisoft X and and binance and I mean I don't know what else they could have done and it also does very neatly solve the routing problem for them but I think this isn't gonna net be like quite good for the takers because yeah the system is just probably going to be far more efficient than swapping in these unisot V3 or Unison V4 pools and it's going to be very likely bad for the LPS who have already been in this terrible situation for years they're going to get much less than non-toxic flow and essentially you're only gonna if you're an LP you're

only gonna get the flow that winner mute doesn't want uh that's not really a great place to be in what does that look like for the protocol like did it like you know obviously when trades get routed through their pools their LPS accrue fees but in the world where like every trade is hitting the the first pass through the RFQ system and winter mutes or whoever's these actors are are uh you know filling these trades probably through Sexes then like you know does do you think we see a significant decrease in uni swap volume because all that stuff is being all that volume is being moved off chain yep so you will lose all the non-toxic flow or a decent chunk of the non-toxic flow you will lose a bunch of the back running flow which one could argue maybe shouldn't really exist anyway uh that's

also an artifact of poor protocol Design This double-sided liquidity um but you will still have all of the toxic flow you'll have all of the Arbitrage flow that is still going to go straight to the chain so the situation yeah it's going to be a lot worse for for LPS no doubt about it um because yeah the takers are going to get a better deal because a more competitive liquidity provider is stepping in through this centralized system so again when we think about um you know Solana versus ethereum I have a lot of Sympathy for the uniswap team because they're developing in this world that is so constrained uh where if you want efficiency you have to introduce these off chain systems and

unisop is definitely not the only one who who's doing something like this uh and on the Solana side you know we want to enable active liquidity because we fundamentally believe that yeah we want winter mute and jump to be competing to give the best prices to the Tigers that's sort of how competition and liquidity Works in traditional finance and it works quite well so we just think all of that should happen on chain as opposed to on some uniswap centralized server somewhere else that doesn't really give you any guarantees on on execution there's a lot that is going to be very difficult to audit like uh like fill rates or what the what the auctions tend to look like

or if there's any collusion between the RFQ providers you just won't be able to tell and yeah I feel like it's just reverting to a completely centralized world and if you're going to do that you might as well just put an order book off chain I don't see why you need to go through this you know jump through all these hoops to to make this um RFQ system happen yeah that's actually a good point that I hadn't considered it's kind of like the same same sacrifices but getting degraded performance going the other route but um you tweeted a while ago uh agreeing with Zaki like when he tweeted like oh like the time is now for the cosmos to really you know start innovating and building the apps that people want to use to keep up with ethereum and it's kind of L2 Centric roadmap

um is that the reason that you kind of started building Phoenix you're kind of like the time is now let's build a really good spot decks like there needs to be a really good one that everyone goes to on Solana well I think this realization came to me later at least as it relates to Solana but I do agree with the general point that a lot of these roadmaps are converging you see ethereum moving towards this this like combination of of app chains but also with with shared sequencers and whatnot which if you can solve some of these fairly difficult problems around interoperability you might get a lot of the like the benefits of both the app chain

as well as the benefits of composability on the monolithic chain and you know I think Salon is engineering is is very very strong relative to the the competition out there but it doesn't matter how good your infrastructure is if you don't have users if you don't have applications that are differentiated right because the ethereum world has pretty much all of the Mind share it has pretty much all of the capital uh both like dollar Capital as well as social capital and the only way that you can drive a wedge into that is by having products that are so differentiated from their counterparts

and so this is why I fundamentally don't think like a new Ave clone in 2023 or a new uni swap clone in 2023 yeah the fees are going to be lower yeah it's going to be a little bit more efficient because the block times are shorter uh it's probably not going to be a not to bring Solana over the line I think you need a lot of products or maybe just one like super super good product that is actually only possible to build on Solana because of the high throughput because of the low fees and and yeah I think if if I could like wave the hand of God and allocate the resources among the Solana protocol developer Community I think yeah we probably want

way more like wacky experiments in the protocol design space and fewer clones of things that were designed in 2018 we're designed to be super safe on ethereum mainnet were designed uh again in this heavily constrained world that where all these constrains just lifted on on Solana the performance is going to get better on the ethereum side with these l2s as the engineering gets better with uh you know ZK EVN and the proving and whatnot and if the actual like protocol Innovations are always coming from the ethereum side it doesn't matter if the engineering is is stronger and so on it doesn't matter if the fees are better on

Solana because if uh yeah Solana is always just if the Solana Protocols are always just copying the ethereum protocols there's never going to be a real differentiator what are some of those like applications that you would love to see get built like what is it about like what is it about you know the Solana architecture that specifically lets you build something else that you'd like to see built on Solana if that makes sense yeah I think uh I can like vaguely point to a couple of things um you know if I if I knew like exactly what the killer app was going to be for crypto um you know we would definitely be all in on that uh I think on the defy side there's uh so like having active markets for

everything I think is a clear Improvement now you need to find a way to make this translate into an improvement for the end user as well who is probably the the swapper but we have seen a lot of this to Solana's credit already in the past with uh like derivatives daxes that are order book based and fully on chain like mango markets and a few others uh spot dexes like serum like Phoenix open book um I think we can see a lot more interesting stuff happen on the margining side where yeah most of the most of the margin models are still essentially clones of Abe you can probably get away from this interest rate as a piecewise linear

function of pool utilization you can probably get to a much more efficient model that is just Market based because you have the bandwidth to implement markets Market mechanisms so you should be able to end up with much more efficient borrowing and lending so that means more Capital efficiency throughout the ecosystem um I think on the nft side this is not something that I've explored too much but if you think about what an nft really is it's sort of this ownable piece of code right it's like a piece of code that lives on chain it's ownable it's transferable maybe it's modifiable and today it's pretty much the design space for nfts has been restricted to we've

had this Turing complete piece of code but all we ever do with it is have a URL that points to some JPEG that lives on Amazon S3 it seems like there's a lot more you can do with that especially when you're not to constrained by by computational limits so I'd like to see a little bit more like experiments there at least what can you do with these ownable pieces of code can you like have them interact with each other interact with the chain in some way um that is not just you know 10 000 pixelated pictures seems like we could probably do better yeah no that's I'm always yelling about the same thing it's like you know I love the idea of uh The Interchange scheduler

from the Cosmos 2 atom 2.0 white paper they just like we're like all right let's create essentially a block space Futures market and this is how we'll use nfts in a very interesting way right because the right to uh order the transaction in a given block is of course non-fudgable with every block being unique so I would love to see uh just the whole industry from like in developers on every chain be like all right well maybe that's a more use case a better more interesting use case for for nfts um so hopefully we can continue trending trending in that direction but I kind of want to steer things back towards Mev and I I know actually one of the pieces you wrote um if not two actually for umber research or kind of like on this topic and I want to get your take so

Mev in ethereum obviously gets redistributed to the each stakers and this is all through like the Mev Mev boost supply chain that's you know flashbot so it's kind of you know really implemented into the system now um do you think that's like a net positive for the chain is that the best way to be extracting and redistributing value uh or do you think like Solana kind of has the opportunity to recreate and reconstruct a different mechanism that's maybe more beneficial for the trade yes I think there are a lot of areas for improvement on the med policy side uh someone on the uh ethereum side describe this to me

really really clearly which was ethereum has sort of psyops everyone on the Chain into agreeing with this vision of this hyper capitalist base layer and then almost this communist protocol layer on top where the base layer is like okay for it to be maximally extracted and at the protocol layer you know if you even want to take like a small fee people will get really really pissed at you and with like the unisop example we see you know a lot of people opposed to unusual turning the fee switch on but you know unisaw today is returning hundreds of millions of dollars if not billions of dollars per year to ethereum holders and ethereum stakers and yeah

the problem here is this is just not going to be sustainable so you need to find a way to like really reduce the amount of enemy which you know V4 and Unison x uh are are doing while trading off maybe uh the level of decentralization that unicep is always had uh I think there's a lot of interesting ways you can think about the redistribution the challenge is making sure that everything is still incentive compatible and like one really nice property of all the Meb returning to the base layer uh and you know burning the base fees um and you know Meb smoothing if that if that ever goes through and PBS is you

end up having very very little skill running a validator and you can run a competitive validator with low skill which means you don't have the centralizing Mev Force at the validator layer now in practice how centralizing is it I'm not sure we actually know uh I think there are there's like a lot of evidence that when a player ends up looking like they're going to become really big they're going to self-limit I think we've seen this before uh with Bitcoin mining when uh you know mining pools really really don't want to get to like 40 of the hash rate because then Bitcoin becomes very very fuddable you don't want today we see like block builders in the PBS block markets uh there's some

evidence that some of the larger Builders are self-limiting the their market share because yeah it kind of looks bad if one block Builder has 70 of the share so how how important is really making sure validators are both skill I'm not sure but I think that's a question that's worth exploring that's sort of been accepted as almost dogmatic in the ethereum world that you just have to have to turn this economic incentive to as close to zero as possible in a way that very conveniently also benefits people's eat bags uh but the thing is like we don't have a great solution we don't know exactly uh what to do I think if you just allow

the validators to extract all the nav keep it for themselves and it's a very high skill game yeah you probably do end up with just like the biggest trading firms in the world running all of the validators or at least having the yeah the most sophisticated validators I'm not sure if that's a fantastic end result either so I mean there's a lot of trade-offs here uh I think there's a lot of design decisions to Think Through but we almost certainly can improve at least on the sustainability side um the current situation that we see on ethereum which is almost like dystopian in my opinion in terms of uh fire dancer

is it gonna have a an impact on uh Phoenix directly and also like how significant are you expecting the performance enhancements to be so I haven't been following along the fire dancer developments um too closely my senses they're really optimizing every part of the validator from uh you know even from the networking layer to all the way down to like execution and and consensus and signature verification and whatnot they're like really looking at every component and trying to optimize it all the way down as far as impact I think you should end up with even higher throughput and even lower fees in the steady state um now it kind of depends how many like

what percent of the stake weight moves over to firedancer because if less than two-thirds does then uh you know everyone else on the chain is still going to have to be able to to keep up so maybe you can't just crank the throughput up a factor of 10. um but if everyone is running fight answer presumably you could yeah just crank everything up by a factor of 10 by a factor of 100 what that means practically speaking is uh yeah if you're if you're still in this world where you have continuous block building and I imagine fire dancers is you know still going to be implementing their own default scheduler there's no way that yeah it seems very likely to still involve continuous block building

we should we should just see like a lot more efficiency on the order books themselves uh and I think uh probably some of these latency games that are already getting played on the Meb side will become more important which is definitely not ideal but I think it's also this Natural Evolution of how financial markets go so as it goes to like the ethereum narratives right they've spent a lot of effort really focusing on like etha's money and because of this it makes a supreme type of collateral and now you see like you know obviously stake staking derivatives have taken off in Lido State deep it's like this new Prime collateral that's kind of you know if you look at uh like Ave backing or Ave

collateral curve USD you know down the list rapstick teeth is like by far the largest asset and when you look at the Solana side of things there's been like almost the opposite idea of you know like Anatoly being like hey Salon is not money it's you simply need a small amount to play transaction fees and that's how you participate in the network but like the asset's not money do you think that like that's been a part of the reason why D5 maybe has been a little bit slower to take off on Solana just because you don't have that like one asset that is prominent to be used in defy and like people can rely on to be this like form of you know digital money no I don't agree with that I think the the money narrative is is really overblown on the ethereum side I think

it's uh just a pretty cynical way to capture some of this monetary premium uh and really the the main argument for oh ethereum is money is just that it's an asset that has any sort of business model whatsoever which to be fair is very rare in crypto it might be like the only one that has a sustainable somewhat sustainable business model but that certainly is not enough to make an asset money I think it's yeah really just this narrative push I also do think on the Solana side there is um you know the the problem on Salon G5 is not there's like too many assets and we don't know what the dominant asset is the problem is there's not enough assets there's not enough stuff to do I think Seoul and uscc are totally dominant in the Solana D5 in the

same way that ether and uscc are dominant on ethereum D5 so today we have such a small market share relative to traditional markets and you know that's probably very okay right we're still moving fast we're still breaking things it's not really ready to hit scale um on any major chain today so part of that's very okay but if you fast forward it out it's like okay well what moves the needle and you know it's probably the applications themselves improving their Tech the base layer is improving their Tech but like what I don't know like do we need tokenized t-bills or something of this nature to like be like okay we have this perfect design for a lending Market but you still need that like something to be used as collateral on chain yeah it's a I think Ruble gases are this deal with the devil you already

see this with circle and tether which basically are you know this real world asset has been bridged on chain and the problem is the same as all the other Bridge assets we already see in crypto the problem is this dependency on the bridge so with the real world asset uh at least like a you know a tokenized wrapped asset like uh I don't know you want to bring some real estate on Chain by tokenizing it and splitting it up and sending some of the revenue to the the token holders the the problem is this bridge is this really really important Central piece of infrastructure and its importance only gets bigger and bigger for example today like effectively anytime ethereum wants

to do a hard Fork Circle and tether have a veto over it I don't think they've ever exercised this and if they did they would certainly do it in private they're probably like really not incentivized to do this ever but if push comes to shove and circle says Hey like I'm not going along with this upgrade uh there's not too much ethereum can do right because the like circle gets to decide which version of usdc which Fork it is accepting for creations and redemptions so if all of your usdc on your quote unquote canonical chain goes to zero that just like totally destroys the ecosystem so having more of a Reliance on real world assets I think is this scary

proposition do you sort of just end up with a bunch of other actors having a lot of uh yeah a bunch of centralized potentially censorable very likely censorable actors with a lot of influence over the chain in this way that's maybe not as easy to quantify as oh someone controlling 33 of the stake or 66 of the stake it's really this sort of political power that they that they have over the chain do you have opinions on this is kind of right field too but do you have opinions on decentralized stable coins because you know you say it's kind of unfortunate tether Circle Etc have like kind of a Stranglehold over choosing

ethereum upgrades so curious if you have any strong opinions there uh uh I think it would be great if we could make one that worked I haven't seen a design that I think works yet and we've seen a lot of designs that don't work uh uh it seems really difficult for stable decentralized stable coin to scale I'll say that I think that's the the biggest issue we see today uh and then you sort of get back to this question of can we have another asset as the money asset and this is where the actual value of that comes in where we say like hey if we can sort of socially meme everybody into not caring about dollars and into caring about some other

asset then we can make that asset crypto native whether it's BTC or ether or something else um but that also requires consolidation and you know if if there's no consolidation the very natural selling point is US Dollars which is the currency around the world that the vast majority of people prefer prefer to hold so coming up with some sort of decentralized version of it that like you know follows it reasonably I think is is a very Noble Mission I just don't know if it's if it's possible with the tools we have today yeah I completely agree and share your views on that topic but uh Eugene this has been a fantastic conversation thanks so much for coming on you're like

clearly like just you've got you've got research on very wide array of subjects that will also Link in the the show notes so you kind of got your hands in a lot of different pots it's incredibly impressive and it was fun to have a chat but do you want to share with the audience where uh they can find you learn more more about uh just what you're up to uh yeah you can follow me on Twitter I'm at Xerox [ __ ] Trader you can follow uh our company Ellipsis labs Ellipsis underscore labs and if you are more interested in the research side we post some stuff at Umbra research awesome thanks Anton Eugene it's been a great call I appreciate it man cheers [Music]