The LP Profitability Problem: LVR, Dynamic Fees and Auctions | Season 6 Episode 2

Transcript

all right everyone welcome back to another episode of bell curve before we jump in quick disclaimer the views expressed by my co-host today are their personal views and they do not represent the views of any organization with which the co-hosts are associated with uh nothing in the episode is construed or relied upon as Financial technical tax legal or other advice you know the deal now let's jump into the episode welcome back to another episode of bell curve today Dan and I are going to be joined by Alex NIS Lin and Jason mil owners to be diving deep into the world of loss vers rebalancing also known as lever or lvr before we get into the nitty-gritty of what exactly lever is which I'm sure many of you will

already be familiar with is the why we decided to do this so amms have been a remarkable innovation in the world of defi but one of the big problems that they've suffered from is how to make LPS profitable how to calculate that profitability but also how to make especially retail passive LPS profitable the first the very first idea of how to quantify that was this idea of impermanent loss which many of you will be familiar with but that is evolved to lever so we're going to get into how exactly we should Define lever and what are some of the solutions for Lever today namely Dynamic fees and auctions this was an extremely Illuminating and fun episode I hope you all enjoy all right everyone welcome back to another episode of bell curve uh today Dan and I are joined by Alex as Lin and Jason

Milanos guys welcome to the show thank you for having us we're excited guys and just to set the scene for the audience and listeners we are going to be diving into one of the thorniest and most complicated but simultaneously most interesting and I think critical areas of crypto today which is lvr and LP profitability so I want to I want to do our best to deconstruct in as simple language as we can for for listeners what lvr is and why it's important and then I want to spend the meat uh of the conversation getting into some of the proposed solutions for lvr which I think are super interesting interesting and um and worth spending time on so before we get into it I'll just open up the floor

uh Alex and Jason could you guys give us a definition of lvr yeah of course um so I guess um you know there's from what I see there's many angles to it there's many interpretations where it is that um correspond to like several equivalent definitions I think our paper so actually gives like all five equivalent definitions but you know if you want to analyze all of them it could take a whole episode of the series to do that so I think the intuitive idea behind lever is that essentially it's the price of knowledge so whenever the price of an asset Moves In in an external Market say if trades on on coinbase uh the amm does not know if it's um the updated price like it's behind so essentially what the

amm does is that offers liquidity at stale prices and the thing is that somebody else and these people are called arbitrageur they do know the latest price and the amm will trade with them at worse than market prices so basically the LPS of the amm lose exactly because of the informational disadvantage that um the prices of the amm puts them into so I think that's my my intuition around it but you know more more formally I guess as a computer scientist the formulation that sticks to me the most is is the comparison with alternative strategies that the lp could have done instead of um lping at the amm and so I think that's also The View with impermanent loss but I guess we can

discuss that a bit later I love the first definition I think from my perspective lvr is essentially a transaction cost yet another transaction cost that is imposed uh on LPS um if you are a passive lpce and you want to make money by making the market then um you can um contribute some amounts to the protocol of one token of another token and then there is a wide range of outcomes what can happen in the future future the price can go up in which case the permanent loss is probably going to be high and um this may not be the best outcome the price can go down in which case um the outcome is also probably not going to be very

good for LP so maybe the price will fluctuate around the same value in which case it's great for LPS the lp will be making money um one issue however is that when you contribute money you should be thinking about what is going to happen in expectation and if you take a step back and think about a world without any transaction cost then there is uh a theum in finance there is a theum in statistics as well saying that if you believe that prices follow something like a random walk then you cannot R the time it so basically all these strategies I will buy now I will sell later if the price goes to some uh higher uh number they in expectation don't really generate value

unless you have some what people call Alpha about the price so if you just think that the price follows a random walk then there's a wide range of outcomes where you make money in some you lose money in some others but on average that's a fair game and what makes this game a little bit unfair is the lvr because every time the mm quotes a price the price is a little bit stale and so the amm end UPS trading at a price end UPS buying at a price that is a little too high end up selling at a price that is a little too low and over time this little transaction cost accumulate and so that makes the game unfair to the

LPS um so that's my idea of um lvr I I really like both of those definitions and maybe to sum it up and pull out some of the key Concepts there's I love this idea Jason of the price of knowledge and then there's also an idea of information asymmetry so if you imagine a world where price was a Martin Gale right which means prices are basically a random walk and there is no directional Alpha and everyone is trading on one venue there would be no lvr however we're not living in that world we're living in a world where price is discovered on one venue and then it adjusts to other venues and that discrepancy in terms of timing and knowledge generally leads to the concept

of lvr and this idea of expected idea of profit as opposed to realized uh profit is another very important one we could get into the the weeds of how this works in the world of traditional Market making which is probably the best analog for talking about LP profitability but it's also one that's a relevant problem for businesses as well right so folks who'll be familiar with the concept of a p&l balance sheet cash flow is it tells you different parts about the profitability and viability of a business right so to ask a very simple question for a regular business you know you might say how profitable is this business well how much money did you start with and how much did you end with but actually once you get into the weeds

it's not really that simple because let's say you're building a factory you might actually have a huge outlay of cash flow in the you know at one period of time but what you're trying to do with the p&l is get an expectation of profitability over a certain period And there's a similar concept I think to be applied to lvr and what makes it so difficult I think I actually heard Doug kit who we're going to be hearing from on our next episode he he mentioned that probably this this isn't even a lot of people aren't even aware of this issue because you might be if you're not a particularly sophisticated LP lping along you might be making money but you might be losing money in expectation from lvr which just makes it a very devious problem so that's that's the way

that I would sort of sum it up and and you guys give really great helpful definitions there and I want I want to turn it over to Dan in a minute to get into some of the solutions but Alex I I heard you lay out a great um in terms of solving the information asymmetry part of this you said that information sort of flows from three separate sources you have to consider the source of the flow you have to consider information that's going on in the world you have to consider order flow so maybe if you could lay out those three sources of information and then we can transition into the categories of solutions for mitigating lvr which would be auctions and dynamic fees but turn it over to you to talk about sources of information for pricing I didn't mean the least the list

to be exhaustive so I think that there may be some other sources that uh I might be we might be overlooking um in general as a know mechanism designers might be overlooking but I think that these three sources are sort of the most natural ones so basically when you think about pricing your liquidity where should it be positioned where should the best B bid be where should the best be how much liquidity do I want to provide at this best bid and dusk prices you should uh take into account pretty much all the information that you have but primarily if you talk to traditional market makers it will come from several

sources right so they will be monitoring other exchanges very closely what is happening elsewhere the information about the outside world they will be monitoring the news so this is what this is one very important piece of information obviously the second source of information is just the flow that they observe and the finance literature the academic literature has done a lot of work on this and of course in practice people um uh use that a lot too so if you see that for some reason you are putting your limit orders on the book and they're getting uh and somebody is taking them way too eagerly then there is a chance that you they might know something that

you don't in which case you should start taking some defensive action so even without any outside information about the outside world you should be learning from the order flow That You observe and finally if you have any information about the particular incoming order that you can use then of course that should be used as well um almost any Market maker would price an order coming from Robin Hood differently from an order coming from a traditional exchange right so that's those are two very different sources of uh order flow for them um and uh

basically in uh in the context of defi if we had more information about the transaction where it's coming from was it observed in the public M pool or is it coming only through the the private man pool um everything that you can imagine you could learn about the transaction might behave helpful in pricing liquidity for it so that those are the three sources but maybe there is something else hey everyone wanted to take a quick second to shout out this season's partner Maverick protocol now many of you probably know Maverick is an Innovative amm which they are but in reality they're a lot more than that as well Maverick is a suite of tools for defi users and builders that allows them to put liquidity where it will get the

most work done since Maverick launched in March theyve been gobbling up market share and at the time of this recording which is the end of September on a trailing 7-Day volume basis Maverick is now a top three decks by volume and they support over 50% of the volume on the L2 ZK sync a chain Maverick enables LPS and token pairs to process higher volume with limited tvl which allows them to support some of the highest levels of capital efficiency for lsts like WAP Ste another very cool feature is something called Maverick boosted positions so that allows protocols looking to bootstrap their token liquidity to Target the shape of liquidity of any token pair with surgical Precision Maverick is backed by some of the leading institutions in crypto Founders

fund panta coinbase Ventures binance lab they are all backing Maverick in their Vision to revolutionize the next generation of defi daps and helping them build their liquidity in all market conditions click the link at the bottom of this episode let them know that I sent you thanks guys the example of Robin is is a really good one and the reason why makers would actually be able to provide better uh quote tighter spreads is because they're not worried about getting run over right so this is a concept they'll be very familiar with sort of trafi market makers which is if you're making a market for Goldman Sachs and Goldman Sachs places a huge sell order there's a good chance that Goldman Sachs knows something that you don't and you might lose money on making that making markets for that particular trade

whereas Robin Hood the reason that order flow is so valuable right is because it's what you would call dumb dumb order flow I wonder how long that's going to be a thing after the movie Dumb money is out now everyone knows that it's called that but it's dumb order flow and as a maker you're not particularly worried about getting run over so you can actually quote better tighter spreads so maybe Dan I could turn it over to you here and start to ask these guys about um what what are some of the solutions for for lvr here I know in the past uh you've been more on the the sort of auction side of things now you've you've lost a lot of that religion and are looking at kind of dnamic fees and and oracles as well but be curious like kind of introduce this concept of these two

different um or the different categories that people have for for uh solving or or helping to mitigate lbr yeah and yes and I think what what we just talked about is a great lead into that and just before I do I would say I I consider myself mostly agnostic on the on the concept of how to reduce um loss versus rebalancing it is a project that I think makes sense as a project and I think we talked about this a little in our last episode but one of the things I loved about Jason's lever papers that it really crystallized for me this as a as a meaningful research Direction um and when you like zoom in into the micro of basically what's happening in every block it makes a lot of sense whereas up until the lever paper came out a lot of people I think had been framing this as

how do we reduce impermanent loss which to me is it's almost an incoherent concept for a market maker to to the idea that a market maker could make markets without having some risk of the market going going away from them and being and being run over um but I think with lever it's it's specifically talking about the part of the risk that can't that can't be hedged um and that generally comes from these from these sort of short-term generally short-term uh uh informational disadvantages um and one more thing for I think when talking about these I do think lever is very different in or the the concept applies very differently if we were talking about say like a longer tail um token that maybe one that's only traded on this particular um exchange I do think it's still a meaningful concept

potentially there but I think when people talk about solutions to lever we're often talking about it in the in the context of um of Highly liquid tokens with with a lot of price Discovery especially a lot of offchain price Discovery where it becomes it's slightly easier to measure and I think possibly a the concept means I think a slightly different thing in that context um I don't know would Alex and and Jason would you agree with that as as maybe the the main um or maybe that's that's sort of a more meaningful thing to limit our discussion to is about the the shorter tale of tokens yeah I think that sounds just about right it's just that um I think that when you have price Discovery also happen on chain it's more an unknown as to how this price Discovery is going to reverse affect

these um external markets where you U previously assumed that you have price Discovery so for something like let's say eth we can safely assume that you know since um I don't know 78% of the order flow is essentially on centralized exchanges um basically the price Discovery happens on these exchanges uh once on something else you have um amm be the dominant part of the order flow then you can't essentially um perfectly uh distinguish as to what is the the quantity that you're losing due to adverse selection which is basically what that lever represents so I think that poses um you know a difficulty that I think it's it's a very interesting

research Direction and definitely one that I like thinking about but I don't think we've um certainly crystallized to this point as to um what's the um right way to think about price discovery on chain I agree I think that there are certainly different market segments um in uh in this debate and I think that they should be uh addressed uh separately uh deep liquidity major tokens would be one and of course we probably have the expectation that price Discovery is happening elsewhere then uh stable coins uh would be a different example and we may not worry about lvr for stable coins too much but there is still uh a related

problem how do you make um lping in this particular Market even more profitable can we can we make it even more profitable uh then there are assets uh that people call longtail assets uh the market there is obviously very different so some of the ideas may carry over but not necessarily everything and uh then there are other some there are some other interesting markets as well but I think the big one should probably be uh the major tokens for now and then one more question on this is do you feel like markouts might actually be a better uh uh metric for measuring this with longer taale tokens like does does

law ver rebalancing maybe mean a little less as a concept in the in the longer with a longer tail and we didn't really talk about markouts as maybe the an alternative way to talk about this but is that is that maybe a weakness in the in the lever frame if you're actually talking about a token where where price discover is happening on the same exchange so I don't personally I don't think it kind of is because if you there's this basically um theorem that if you take markouts and um you take the time interval um go to zero as to how you measure your markout then that essentially is going to be exactly lever so you know what what markouts are essentially doing and it might not be obvious at first glance but I think that's um that's the case is essentially that um they try to interpolate between

lever and impermanent loss so if you have a market with like um a huge time Horizon then that's exactly equal to impermanent loss whereas if you have a markut with a time interval that goes to zero then that goes exactly to um to lever basically so I'm not sure as to um what the benefit is if you have an intermediate time interval uh maybe Alex can speak more to that but I think that in the two edge cases it's kind of relatively clear what what markout corresponds to so I don't think there's a meaningful difference in that direction yes I think there is this traditional debate on how long the markout period should be if you take it all the way to zero if you somehow know the fair value of the token at the time when the transaction is made then to me

it's almost the same thing Mark out or lvr there's almost no difference between the two it's literally the same thing isn't it it's literally the same thing yeah so I I I wanted to hedge my bets a little bit because Jason has a special interpretation of fees but we we may not want to go there right now so let's let's call it the same thing but then uh the issue is that uh uh liquidity providers they like to provide liquidity to uninformed flow and when uh there is somebody buying in size then there is a chance that uh they're buying on multiple venues at the same time so there is always a question that uh when you use the price from Central itic

changes or from some other venue or some weight at average price or maybe the global price uh the mid of the best bid and offer everywhere is that price biased relative to the fair value because of the temporary pressure on the price that the uninformed flow creates everywhere and so the traditional thinking is that we should allow for some period of time to make sure that uh there is some uninformed flow going on right now that it dissipates we need to give it some time to dissipate there will be some other uninformed flow later of course but we don't know in which direction it will be so the tradeoff in choosing the markout period is that the

longer the markout period the uh less bias you should expect to see in your mark out but this lesser bias comes at a cost of higher noise because you're just giving some more time for the price to walk away in each Direction you don't know in which direction it will go so there is no bias but there will be noise so basically there is this tradeoff which uh brings me back to Dan's question uh would that be an appropriate metric for uh longtail assets for longtail assets you don't necessarily have a good price from centralized exchanges even 15 seconds later because there's simply no good

price and so the question is at what time period can you expect the price on the deck on the same deck that we are talking about today to be unbiased and uh for the all the uninformed flow to dissipate and so I don't want to go into uh much detail on the mechanics uh of uh how Unis swap pools work but basically since Unis swap pools can only be moved by takers of liquidity it takes a little bit longer for the mid price on Unis swap to sort of reflect the average true price so basically takes a little bit longer for uninformed

flows to dissipate so basically you do have to take longer markouts I think that that the approach still works you just might need to give it a little bit more time and use logo markouts if you are talking about longtail de longtail assets well so I I did want to ask um get into some discussion about uh Solutions and possible families of solutions to to last res I think maybe before we go there can I interject a little bit um I think what Alex said um gives me a kickback to discuss something else that also came back a bit um earlier and that's like I think we should make clear that like the term like noise Traders or uninformed order flow it's not a dimunitive term I don't think it's meant to say that these people are are dumb or being taken

advantage of and actually think on Top This is actually wrong and that's because um what it solely means is that uh informed order flow does not have a short-term Edge so immediate Market information longterm I think that uninformed order flow is actually exactly what um what moves an efficient market uh it's basically the purpose of any efficient market to um uncover and aggregate this long-term information so I think that the distinguishing bit between Arbitrage flow and uninformed flow is essentially whether you do have or you don't have shortterm uh Market information for the next let's say second or millisecond and I think that's actually what um what constitutes the um

impetus for essentially having markouts with different time periods and actually just just one interjection there for folks who aren't aware of a markout a markout is again one of these ways that more traditional market makers or highfrequency Traders use to measure expected profitability right so if you're a highfrequency Trader you're placing many even sometimes hundreds of thousands of Trades per day you want to get a sense of hey am I placing am I placing these trades in a way that I'm getting good execution and is my strategy profitable and sometimes the period that you actually hold your trades doesn't measure up to the sort of expected Alpha timeline that you have so again you'll sort of take a a period of time you'll place the trade and say after five seconds I'm gonna measure whether or not

this was a good trade so just to I just realized we haven't defined what a markout is so just to set the scene it's it's another one of these sort of expected measures of profitability that high frequency Traders or market makers will use and I think Jason I think a great point because I do think when you hear people talking about uninformed Traders um you might think right that the market makers are scheming about basically how to take advantage of them but in fact the brutal competition generally among professional market makers is people competing to give the best possible execution to uninformed Traders and just try trying to avoid trying to take advantage of or avoid being taken advantage of by the by informed Traders but generally knowing that Flo is uninformed is a reason to give it better execution uh traditionally

um and that's that's again I think an interesting an interesting property of the how these markets work there's there's a great there's a great I'm gonna blank on his name I will find it in link it in the show notes for people who want to take a look at there's but this guy named Patrick something I will I will link this who wrote a great uh sort of treatise on payment for order flow and how it works in traditional markets and why people have it exactly backwards and actually most of the time retail gets better pricing than institutions um but we we'll link that in shown us if some nerds want to go down the rabbit hole of payment for order flow I think that does that leads pretty nicely into our um next topic talking about solutions to um POS possible mitigations for loss res

rebalancing and I think if we start with thinking about Dynamic fees um what what they've been called um Alex you want to describe this general family of solutions I I think that LPS talk a lot about fees uh when I think about fees I mostly think about the uh Bas spread uh because this is what fees do in an in an amm essentially you have the mid price which is the mm price and since you have to pay the fee when you buy and when you sell then the best bid and dusk prices that the amm essentially offers are different and for a market maker determining how to best position the spread is a uh probably one of the most important exercises you do have to make

other decisions as well you will probably want to be present at many LEL lels but the best levels the the best bid and dusk play the most important role in trading if you look at traditional markets most trading happens exclusively at the best bid and dusk ticks so essentially not much is happening um deeper inside the book deeper inside the book you're just competing for priority you don't have to compete for priority in amm because it's um prata essentially uh so there is no competition really but what the amm can do you can try to come up with a better ways to position your bed ask spread and so when you start thinking about this better ways to position the bask spread

you have to ask the following question what if I had perfect information how would I want to position my bites spread what if somebody told me that the fair value of the asset is X and you can immediately say that well if the fair value is X then you probably want to Center your spread around that X and if you look a little bit more closely at what happens in the actual amms uh it's not necessarily what's going to happen so one observation for example is that since uh the spread is only moved by takers of liquidity in a traditional mm is that uh if you have an Arbitrage transaction it will move the

best ask price to the fair value of the asset but then for the next block the spread will stay it will not be centered relative to the fair value of the asset but the fair value will be at one side of the spread which is not great because you have liquidity that is Overexposed on one side and you have liquidity that is pretty far away on the other side you always want to be somewhere in the middle this is just one example but I think that there are many ways that you can think about how to position the spread and actually you vi4 gives quite a bit of flexibility because you can

come up with your custom fee mechanisms um for um different pools you can write a hook that uh basically um governs the fee for a particular trade and uh with by changing the fee you can basically adjust the bask spread for each individual trade in theory and now basically you just want to use as much information as you can gather to position the spread optimally yeah so I think an important thing here is that the the fee happens to be the mechanic when we're talking about amm that you're using to reposition the spread but really what you're doing is you're changing you're changing the spread and you're changing potentially the price the midpoint price that you're

using and when you're thinking about how to change that price you could be using any of the signals that Alex was talking about earlier you could be basing it on on recent activity you could be basing it on offchain information like the prices on other exchanges and you could even be basing on the characteristics and the source of that of that particular volume um and so that's where that's where some of the Discrimination um uh order discrimination might uh might come in now why when we talk about amms are we always or at least you and I tend to be talking about adjusting the fees rather than adjusting just the price I think Alex and I are very are are have have strong feelings on this but I've seen I've seen others try to build amm that that try to adjust the price based on based on offchain

information so I think the nice thing about uh the underlying algorithm why you may not want to change the price too quickly is that it does provide certain guarantees to liquidity providers in particular the passive liquidity providers that basically I know that if I contributed my liquidity in a certain range and I know that the price moved within that range or maybe even outside that range but for every price move I can bound my losses so my losses are essentially bounded depending on how far the price moves relative to where it started and uh that bound on the losses the upper bound on the losses comes

specifically from the amm mechanism if you start changing the price all together changing the mid price without any constraint so basically instead of just changing fees you say all right I'm going to move the price and you can also start doing other things you can concentrate liquidity more and do all those things it's great if it works so if you have a very good solution for lvr it's great but the issue is that if you are leaking value to arbitragers then you can significantly increase the losses and there will be simply no uper bound on your losses you can imagine a situation where the price moved somewhere you reposition all of your liquidity there you sold completely all

of your inventory of one token and then you turn out that the price was wrong uh and uh your losses are basically your losses can be everything so um I think that uh having some constraints meaning that you you you can adjust the spread around the price but you actually have to an arbitrager has to do some real work to move the price create some uh guarantees on the maximum amount of losses that lpce can uh encounter yeah and I I tend to think this is particularly uh uh nice combination with how unisoft V4 works because um if you are just adjusting the fees you can't make fees negative on on unisoft V4

which is which is nice which means that you know you can do whatever you want to fees during the Swap and liquid but liquidity providers will still ultimately who are providing a hook they'll know that they can still get uh that they bound their losses will be bounded assuming you don't have any any hooks that can that can that operate on on liquidity when it's being burned um and so I think there's there's a nice property there where we could have more experimentation with these in unisoft V4 um and have liquidity be able to provide to that while knowing that their losses would still be bounded which I think is is an interesting combination there now the what limitation does that does that come from like what's sort of the downside of doing things trying to do things entirely with Dynamic fees so I think if if um if I can speak

to that a bit I think what I consider the most significant drawback of many approaches that are related to dynamically setting fees um is that in general um you could have let's say an amm with a fe of 100% of the trade right like that would be perfect it has no arbitr no losses to lpce but what do we achieve with that we achieved absolutely nothing because now nobody no real person is going to trade with this amm so there's actually a trade-off between minimizing the losses 12ps um and actually having a meaningful order flow that we need to think carefully about and so similarly um Dynamic fees are also a friction for the uninformed flow and so this will kind of lead the amm to have um lower income besides just

lowering the losses from Arbitrage so what you're basically trading off here is how much the LPS are losing versus how up to- dat the price of the amm are with respect to to the market and subsequently uh the gain in fees from from the uninformed order flow so I think that's that's one of the considerations that we need to uh think about when we consider um Dynamic fees as a as a solution to lever maybe you know minimizing lever is not your your only objetive it's actually maximizing the income to the liquidity providers that is your objective Alex you agree with that wholeheartedly I agree with this 100% uh there are easy solutions to lvr

that uh people we didn't we don't mention because we know about them right but uh there are solutions you can charge a very high fee to everybody so essentially you are not making markets then uh there is no lvr you can reduce block times if you could um this reduces lvr as well you could reduce fixed transaction cost which is a very very important component of that and that reduces lvr so that all these Solutions when you think about charging different fees you have to you absolutely have to find a mechanism that helps discriminate between Arbitrage and uninformed flow at first so the idea is that on average an

uninformed Trader should face the same spread whereas the arbitrager should on average pay more and so only when you have some good ideas on how to discriminate the two flows only then you should be raising the fee otherwise you can still raise the fee there is no problem with that but uh you are reducing lvr at the cost of also having a lower uninformed flow and so that's that's that may be a solution but uh it's it's not necessarily something that is super exciting we would like to lower the lvr holding the spread faced by uninformed Traders unchanged I have a I have a quick question guys is is are these are

Dynamic fees one thing that I'm thinking about listening to this is all right this makes really good sense right one of the things that Dan and I are thinking about is like when you scale a deck you can't just totally favor LPS you can't totally favor swappers you kind of have to gradually sort of in a step step kind of way make the experience better for both of them and there's a positive flywheel that starts to kick in so one of the things I'm finding myself wondering here is are Dynamic fees something that only a market leader like Unis swap could really Implement because I'm imagining like a deck you know like kind of the nth deck after something like Unis swap where there's not a whole lot of liquidity saying hey you know with very well intentions you know my LPS are an

important stakeholder within my ecosystem and I'm going to start charging toxic flow for like more discrimin I'm discriminating basically doing price discrimination against toxic flow but in that world if you make up a very small percentage of volume in liquidity and Unis swap for instance is not doing that then I would imagine all of that liquidity or all of that trading just flows to Unis swap do you see do you see what I'm saying like I'm does it do you have to kind of implement this from the position of a market leader like would Unis swap have to move first here and sort of set the precedent or would this be available for anyone to just Implement right now and you think they would see positive results are you talking about all of the toxic flow will flow to Unis swap if this is a working

discrimination order discrimination method right then like ideally it should only be punishing people who are informed Traders and generally as as a rule you don't want informed Trader basic your your liquidity fiters don't want informed volume on your um on your decks right I agree I guess maybe what I didn't articulate there is I wonder what percentage of volume today is actually toxic versus retail order flow I I don't all of the people on this call would probably have a much better idea than than me I would guess it's majority toxic probably these days um so on the one hand it's like we don't want that volume but then I could imagine implementing some sort of discriminatory pricing scheme against the toxic volume

and then your volume Falls by 80% you're like wait wait wait wait wait wait come come back uh I'm not going to discriminate that much so do do you see what I'm saying like I could see we talked about this a little bit a little bit I think in our last episode um but I do think I I'll I'll put it slightly differently in that I think this this is something that maybe only Unis swap only a market leader could have the courage to do which is to basically sacrifice what a lot of what looks like volume but is in fact something that you don't want um and say like we're going to we're going to give up we're g to give up this this vanity stat um in favor of actually better returns and and and a better offering for liquidity providers I think

I think that I think that is potentially true and that in that if you want to actually make a mark as a new as a new decentralized Exchange I've seen a lot of decentralized exchanges that get very get a lot of attention because they have a ton of volume this happens a lot with like themm perss these types of of uh dexes that are designed like terribly um and that lose money on every tra every trade and that don't have no bounds on the amount of losses Etc and they get insane amounts of volume when they launch they get so much volume that like they have to shut down um because they've lost all their money and I think that can be a good marketing um strategy for a new deck to kind of to burst onto the scene whereas right one I think it

might be harder to get attention as like we're getting no volume but it's you know we're we're it's all bad volume that we that we got rid of I think that might be a harder sell whereas I think again someone someone like un might have a little bit more um be in a more of a position to to make that long-term choice that you actually don't want this volume I don't know if Alex and Jay have thoughts on that I don't know I think I think to what Mike said I I wouldn't say so um because essentially by discriminating for um toxic order flow and by um kind of having a better fee income stream from uninformed low and maybe providing better execution to them as well essentially what arises in this nth deck as you said um is essentially that you have competitive pressure for

liquidity providers so essentially you have liquidity providers being better off in this nth deck than in the um dominant essentially um decks so I think gradually because of this competitive pressure for LPS what you'll see is that you'll see that if if this end deck achieves um better um better price discrimination both Arbiters and um essentially uninformed flow you'll see the endex will become in minus one and then it will become in minus two and then it'll start dominating because exactly the LPS will will start leaking from uh from the leading decks to this uh this n deex exactly because um competitively it's better for them uh to

be uh there than the first one Jason I'm actually GNA push back on your push back there um and I think the thing that might be debating here is the idea that all LPS are behaving perfectly rationally and I think there's pretty demonstrable evidence that that's not necessarily the case or there are broader strategic implications that LPS are thinking about Beyond just the profitability that I'm making today so the big question that a lot of people are curious about um in LP land is who are these big univ V3 LPS Dan I've heard you Muse in question about that and we've talked a little bit about who who are these people who have been pretty empirically demonstrated to be losing money and it's it's another uncertain pools uncertain pools right uncertain pools

but another but but an additional layer of paradox or irony is that actually more active LPS are are these these larger LPS tend to be more active and that makes them lose even more money gas cost things like that but I I like your explanation Dan which is that basically like there are a couple big pools on Unis swap if you're not an LP or active in these pools and you're basically just out of the game right so even you might lose money for a specific period of time and if if you're a very large sophisticated well-funded entity that's taking a concentrated bet you might be okay losing money for years as an upstart decks you don't have years to demonstrate traction that so that that would be the only push back that would be the only push back that I would say yeah I think another way to put is you

do need to bootstrap retail volume and um in order to you know to get to have any returns above um uh any returns basically at all in expectation and so that happens right now to be on uniso and so in order to if LPS moveed to somewhere that just like happens to perfectly set their fees correctly but they don't get any retail volume um then then I think it won't really work I do think you I think the more efficient routing gets the hopefully the the better um the more competitive that gets and again I think part of the idea of we talked about this last time about of Unis swap X and Unis B4 in combination is to try to make that make it easy to Route volume to new uh to new more competitive uh models for for decentral exchange provision um Alex I yeah I

think that this is a very this is a great question uh people talk a lot about discriminating different groups of Traders informed versus uninformed we also need to think about the different interests of LPS as we use the same word for them we just call them LPS and we don't really look at the incentives why they're doing what they're doing um and I think that this is absolutely a critical component and all of this so let's say that I gave you an optimal Oracle price you can rely on it 100% then you can build a great decks around it you can position all of your liquidity right around that price offer it without slippage and maybe just charge the lowest fee possible on both sides then you will get uh only the uninformed flow there will be zero

Arbitrage flow because you're always pricing around the correct price and it looks like you're doing great you may not be showing as much volume as some other exchanges but all of your volume is great are we done is this mission accomplished and I don't think that the answer to this is yes because uh the RPS who actually willing to lose a little bit of money for some useful service that they get namely directional execution with a algorithm like the one that I described before where you position everything around the optimal price it's very difficult to direct to execute at trate directionally by being a liquidity

maker and uh in traditional markets I have not seen a great estimate of the number all of the estimates seem to be a little bit handwavy but uh people say that 60% of liquidity comes from about 60% of liquidity can be coming from people who trade directionally if you lose let's say three basis points on each trade as NLP it's horrible if you want to be in the market making business because you lose essentially three basis points in expectation on every trade but if you want to trade directionally three basis points is great execution uh this is why univ V3 is actually a big improvement over univ 2

because people could start trading directionally with short range liquidity position and they are willing to position their liquidity very close to the me to the true price or maybe even cross the spread a little bit and uh position it sort of on the other side uh to get that directional execution so I think that uh strategically protocol should start thinking about uh having these different groups of LPS some are interesed and passive Market making some actually want to get something out of it and those who want to get directional execution out of it they should be willing to pay a little bit more and they should be fine with some losses I looked at some of these large trades that happened recently vitalic had a famous one not not a long time ago yeah

well it was it wasn't a very huge one but it was 600k something so he was selling maker I'm not sure if he was aiming for high price impact in that particular case because he was not selling uh uh out of uh you know there were different speculations for why he's doing that but uh I looked at the execution cost and it's pretty high according to my back of the envelope calculation it was maybe something to the tune of 30 40 basis points uh he could have executed better I think with short range LP positions potentially um so that's why I think that it's not necessarily uh uh it's not necessarily just about offering liquidity to uninformed flow it's also about offering

a way for people to trade directionally with liquidity positions and just just to highlight for listeners who this might not be you know as common knowledge I think zero FBI femboy was the first one who wrote a post about this where basically people were actually providing extremely uh on on like univ V3 where you can do concentrated liquidity across an extremely tight range that was very close to the current price in hopes that essentially the price would cross the spread and that they would basically be getting paid a very small amount to make the trade so as your point there Alex basically that those people you should be able to charge them less or even uh even even a small fee for doing that and that would still be cheaper than the execution they would get on another

venue and that's something that would make sense yes I think that it people should be they they're probably fine with small losses right so if you maybe lost 5 10 basis points on this but you managed to trade directionally in the direction that you like and you didn't have to pay 60 basis points in an RFQ then that's a good outcome still again in traditional markets most people try to most largest institutions try to uh execute directionally as makers not as takers of liquidity they end up being takers sometimes because they throw in a limit order and turns out that it's a marketable limit order so it actually is uh it would it falls in the other side of the book when it lands uh so that

happens but they generally try not to liquidity when they can avoid it and I think that this is uh this is something that uh the dexes might want to work so of Might focus on as well uh to offer better execution this way I I know we're we're winding down here or getting to the um sort of tail end of the new here and I I want to bring up the so we've been talking a lot about Dynamic pricing as one potential solution for lvr so how do we identify uh unempl uninformed versus informed order flow and maybe have different a differential pricing strategy for each of those two different groups but there's another maybe family of solutions which would sort of broadly

fall within the category of of auctions as a way for mitigating lvr so uh Alex Alex Dan or Jason whoever wants to take this can you just kind of introduce the idea of auctions and how that might complement Dynamic fees as a as a solution for mitigating or reducing lvr I think my idea around auctions is that essentially their purpose is to um try to take back lever from from the arbs before or after they trade with the amm so basically the idea is that there is um this um total amount on the table that is um lvr and you might not be able to reduce it straightforwardly but you might be able to take it back from um from the arbs and so auctions try to do

exactly that and so how do they try to do that you try to organize a competitive market between all of the arbs um to do exactly that and you have two um ways of doing that you could either um do that um just in time basically as the trade occurs or you could try to do that before so auction out the right essentially to be um the arbiture um and so you pay some some amount of money in this auction in order to have some exclusive right that grants you are privileges let's say yeah and I think one advantage of auctions potentially over over Dynamic fees one is that you you may reduce the need for an offchain some kind of possibly

trusted Oracle offchain because you have if you have this competitive process um to uh for where where basically anyone you know there's incentive compatible mechanism um in theory here for people to be competing to move the price to the to the uh to the right price the other potential Advantage you you maybe lift some of the constraints you have on Dynamic fees right with Dynamic fees you can only move so far off of the um uh off of the current price um and uh you can't have negative you can't have negative fees on uh you can't set like fees on on your am to to a negative value so in in some cases you may not be able to adjust the price to the proper price um I think with auctions it's it's cleaner potentially to to actually have

that rebalance happen at the at the new price because you just like Au Expos auctioned off the right for anyone to trade against it um so in theory you should actually get uh get that to clear um Alex you want to talk a little about some of the disadvantages of auctions I think that auctions can be a great idea but there are a lot of details that would need to be very carefully implemented and I I think that some of these issues seem to be pretty tricky to me but there is hope I guess that's my overall assessment but so so one of the first things is that when you when you running an auction I think it's a good idea to actually collect uh uh money regardless of whether transaction lands on chain or if

it or or it doesn't and the reason why it's a good idea because if you don't then I can beat whatever I want and then at some point I can just decide hey I don't want the transaction right so you need some mechanism to make sure that big bids are actually respected and so one of the ways to do it is of course to collect some amount of money up front but then there is a difficult issue with this upfront collections how do you actually allocate them to LPS you don't know exactly who is at risk are those just the LPS in the current te or are those the LPS in the SS is it just this particular pair or was the arbitrager trying to use this pair to Arbitrage some other pair so the allocation of this Surplus becomes a pretty difficult problem I know that there was a great

proposal by um by the cwap team that attracted a lot of attention called uh MCM ma ma mm and uh in that particular proposal I think that that problem turned out to be pretty difficult so I think that that was one of the reason reasons maybe it was not implemented there are other issues as well so how do you make sure that when you sell the right to Arbitrage to somebody how do you make sure that they don't slow down things intentionally we know from Jason's paper that lvr is high when blocks are slower if I have the exclusive right to open the exchange to everybody then I can artificially slow down the process I can artificially increase the block time and so on and so forth and then finally there is a

question of how competitive this auctions really are because right now we see people competing in auctions but when volatility is high there is a grand total of one or two participants who can actually Arbitrage a pool like e usdc on uni V3 so it may not be as competitive uh as it seems at the end again I think that uh there is hope here but there are also quite a few problems to address one thing that I think bear is mentioning here and the reason I mention it is because it ties into something Dan and I are looking at this whole season which is how much value is leaking out of call it the Dex ecosystem and how can we stop some of that leakage at least when I think about auctions it's it's kind of a

transfer of wealth away from ethereum stakers actually and towards LPS because you know when you when there are Arbitrage yours that are um you know we've been talking about arbitrageur as if they're extracting all of the r but they're not really because they ultimately have to participate in an auction which is r um at the Block space layer which is meev Boost today and they end up surrendering a whole bunch of those profits back to the the stakers of ethereum who are who are the real winners right now right so when I think about auctions it's actually a way for hey how can we transfer through another auction the the the excess profits here right to to LPS of the of the decks

instead of uh basically paying the platform tax of ethereum to St Akers so I don't know if you guys see that in a similar way or just wanted to make that observation yeah I think it's a good question actually to whom this value should um should go like um there's many people that could go it could go first of course um to the arbitragers as it is it could go back to the LPS the the argument could be made that since they're the ones that provide liquidity it should be returned to them it could also be that it should be returned to the users because these are um the swappers because these are the people who are using the amm and so you know these people should get better um execution price or like um higher

quality of um execution and it could also um you know there could also be the argument that could go to the block proposer as it uh defined because um they have essentially the um the right to um do whatever they want in the block and so they could just as easily exclude your uh transactions From the Block and provide you with no execution whatsoever so I think there's um a mix of um situations that um can occur and actually that's why you know that's why I like the um lever as a as a whole as the whole idea of um capturing the entire value that could be distributed anywhere and then where you want to distribute it might be something that

could be more uh fine you I view this trade-off between uh LPS and um stakers as a bare Market uh zero Sun game mentality I don't really think that uh we should be thinking in those terms uh everyone is going to be better off if we can come up with more efficient mechanisms to trade and that's because when you start implementing more efficient mechanisms people will be providing more liquidity there will be more trading volume and this is how you really build systems that uh can compete with centralized finance and can just attract more volume there is always this thing hey you know Brokers they used to

charge five that was great uh for them right $5 a long time ago for transaction I'm old enough to remember that uh now they charge uh maybe 10 of a basis point on average relative to the true price but I think they're happy now and the reason they're happy now is because more people are willing to trade at that transaction cost there is high volume for everybody um and so I think that in in in my world there is really this tradeoff is sort of secondary or maybe even you know third thing to think about after you address the big ones and uh if you come up with a really efficient mechanism to trade

then this will be much more efficient for stakers as well just because people will be doing more of that trading uh more of all of those activities that we love ethereum for so I think I know we're maybe just to to close here one thing that we haven't T we've sort of danced around this but not directly addressed it which is why is this topic of loss first rebalancing so important because I think the the end State um Alex I know you just sort of indirectly alluded to this but at least my goal would be I I want decentralized Finance to not only compete with but eventually optimistic enough to say I wanted to surpass surpass what gets done on traditional finance and there's

obviously a lot of room for improvement there and lvr is a very important component at least from from where I'm sitting so for folks who are out there because whenever you get into something as as technical as lvr there's sort of a oh like it's these people arguing about very Arcane stuff but I viewed is very very fundamental to one of the core Primitives of defi so can we just you know just as like a high Lev takeaway for the audience Alex and Jason you guys have both dedicated an enormous amount of your your personal time and professional career towards exploring this issue isue like why do you view it as something critical and worth spending time on lever is essentially extremely important for deck design basically like you like you said Mike so I just want to um Echo this basically um I think uh the

upset is that no one is going to actually provide liquidity if um if they don't find it profitable to do so and so this will eventually lead to their not being any dexes any longer so basically um that's the um the stakes of the game basically with um with lever and in order to essentially avoid this this um outcome that can be alleged to be bad you need to build the incentive structure in such a way that you could um avoid it so I think you know building an incentive structure such that you can avoid this situation of um dexes being inferior to um centralized Finance is exactly why um the subject matter of

lever is is important to me people sometimes think of trading as uh speculation and uh just games but you can use a different term for it right it's resource allocation and uh all of a sudden it sounds much nice and I think that this is why uh trading is really one of the most important uh activities in traditional finance and it's going to be one of the most important activities in uh defi as well and uh if you cannot in my if you cannot solve some of these issues with inefficient trading then everything else becomes a really sort of remote question right we can do

other things on chain of course but uh I think making trading more efficient and making coming up with mechanism of more efficient resource allocation is probably uh the number one most important point in all of it so coming up with more efficient uh mechanisms on this particular topic I think is very very important and lvr right now is the biggest um inefficiency that's out there I believe so it's a very natural thing to study I I whole agree with both of those and then I think you know for me uh getting as much crypto trading at least as possible non-custodial is is like my mission in life but I think even more than that I

do think trading is essential for the security of the having reducing lever is essential for the security of the chain just because right now a lot of the um uh of the Mev that is leaked to current block proposers today comes from Lever and these and this latency ARB and it's a potential centralization Vector for the for the whole chain and I think we trying to plug this hole and capture it for users it's for Dex users it's good for users it's but it's also really good as Alex said for the long-term future of the protocol because I think actually it's not in anyone's interest for um for this to destabilize the chain and so I think it is really important to to to try to capture this and and make it as efficient as possible so that you don't end up with these unhealthy incentives

uh kind of leaking into the core of the protocol really well said guys I think that that's a great place to end it and Alex and Jason thank you both so much for coming on and helping Dan and I kick this season off if if folks want to find out more about you guys follow you or the work you do what's the best way to do that people can essentially follow my Twitter account it's um Jason uncore ofor CS so Jason of Cs with with underscores and they can of course follow my my archive for the academic papers that um we're going to release with my collaborators my academic field is quite far from uh defi so I I will not advertise that but I do have a Twitter account where I post some stuff

it's 0x 94305 so that would be the way thank you for the invitation Michael oh of course and we will link in case uh in case anyone didn't get those right off the cuff we're going to link those in the show notes and guys I highly recommend that you follow both Jason and Alex there's a lot of Alpha and very very um very thoughtful content that comes out of both of your your accounts if I can plug you there so guys thank you so much again for just taking the time to do this this was a ton of ton of fun all right Dan what an episode there that that what a treat to get those two real Heavy Hitters on there yeah yeah that was great ton of fun and there's a lot to dig into I actually want to go in reverse order in terms of takeaway here

and that last question that we asked of why is lvr important thought we got a lot of good responses there and one of the points that I I just want to underscore and highlight because I like to push back on this when whenever I possibly can but this this idea that speculation and more trading volume is is a bad thing and that's not really real I would just wholeheartedly push back on that because there I think one of the takeaways that I've actually had of previous seasons of bell curve is that speculation actually provides sort of the building blocks of financial alchemy that enable very useful products so for instance if there wasn't a deep and liquid market for interest rate derivatives you couldn't have something like the 30-year fixed rate mortgage in

the US so I understand the temptation to say oh this is just trading this is just speculation there's a lot of that that happens in traditional Capital markets and Wall Street and things like that but it actually feeds into better bigger products and I think that mission that you outlined if getting bringing trading volume on chain is a very Noble one and people shouldn't just thumb their nose at it so I just want to highlight that at the end absolutely yeah I think on a on a micro level this is something we've talked about I think um this is true of for example minor extractable value which in the you know and the existence of arbit treasurers who take advantage of it which in the short run means you know that that maybe uh again people people making money off of off of these like Risk free trades and I think you

you want there to be basically efficient Arbitrage and in equilibrium ultimately their their profit is zero if actually the system is working correctly um but you actually need you do need some some initial profit to basically kick that off and then on a macro level I I think this is important for crypto in general I think uh my my colleague Matt hang U just published a piece called the casino on Mars about the general among other things the dynamic where um speculation often does drive a lot of infrastructure buildout and and growth into into new areas that can ultimately bootstrap a new better system and I think that that is important for for crypto when thinking about uh this on on the macro scale for what it means for crypto's evolution it seems like a very old type of business that Warren Buffett would get involved in and famously has gotten

involved in but there's the old analogy of the the railroad lines right that there was a huge boom in speculation around railroads people came in and funded huge amounts of infrastructure which was these Rail lines that uh that led cross country there was a lot of money lost but it it paved the way for a booming railroad industry eventually one point that I wanted to underscore because we initially almost made it a theme of the season but you you brought up a really interesting point that I hadn't thought here which was how can we move trading activity on chain and what does how does that impact lvr like I I had one of these naive thoughts um or I can never really tell how naive it is when I was just learning about lvr which is hey if we didn't have this alternative trading venue in the form of

centralized exchanges but really binance a lot of this lvr problem would just go away so I thought the point that you brought up about lvr on very popular trading pairs versus very unpopular trading pairs where price Discovery is actually happening on chain that's a very different conversation so I just want to highlight that again for people to go back and listen to and the the question I have for you ultimately I'm Tre P lot of this how can we bring price Discovery even for like the usdc Bitcoin pair or something like that on chain instead of on a centralized exchange like what's the the through line to that future yeah it's it's a deep question and I think there's two ways that we're that that it's is sort of happening one of it is just by by reducing lever on

these on these uh uh you know short tail pairs like like eusd and I think the more we can reduce lever the more volume you end up actually having um the more liquidity you can get on on chain and therefore the more volume you ultimately get on chain and then we try to shift the primary Market um happening on CH I think you've done some work that uh it's it seems like um uh already at at least at certain times liquidity on on chain has been greater than on binance at least liquidity visible on the book um and I think the uh you that doesn't because of latency because of other of uh other factors that doesn't necessarily mean that price Discovery is happening um on chain and I think the more we can actually move it for those

the more these dexes improve um the more we will actually potentially be able to move it more on chain um and then I think uh on the flip side we we do see a lot of innovation in terms of uh this of this longtail where price discover is already happening on chain but I think there you still do have Mev you still do have I think it's it's a deep and subtle question but you do have some kinds of of lever it's maybe just harder to measure and so I think figuring out how to reduce the um the lever and the uh or the the the meev on those uh on those and that may involve like getting smarter about how to measure it I think you know you still have you still have the possibility of losing money to an informed Trader and it just becomes a

little bit maybe harder again to measure in this in the shortterm way um but I think if we improve both of those then hopefully we can start to see some of those longer tale um assets becoming a little more having Better Price Discovery um you know things that are happening with is still happening on chain for them and then conversely where we can have the assets is happening off chain see if we can move it on yeah all right let me ask you some questions that are going to start to lead into next week's episode I want to underscore again and it's a Pity we had limited time with with Alex and Jason but this is something that we talked about offline as well I want to underscore the importance of latency in price Discovery and I want to ask you just kind of a candid question about

ethereum's block time so latency is an extremely important variable when it comes to price Discovery and I guess my question for you is can we ever get price Discovery with a 12 second block time and like how do you think about the role of latency within price Discovery so I think we we do see price Discovery um on chain and longer taale tokens with the 12C block time ultim you know if those if those tokens end up getting a lot of volume and becoming major tokens you will start to see people um basically you know creating doing more offchain volume doing more RFQ doing maybe listening it on centralized exchange is getting more volume there and so I do think that's a that's a potential uh a threat essentially to the stability of that of

that as a market structure so I do think in some ways trying to um uh you know in the in the naive way now like uh just having or you know ordinary onchain you know amm um markets you might yeah you you may it may be hard to to prevent price Discovery from happening off chain I think there are interesting mechanisms possibly for saying we actually don't need price Discovery faster than that and I think batch auctions are one are one such where it actually it's not entirely clear to me that there's a lot of social value in having price Discovery in a lot of these assets that is shorter than 12 seconds um and I think there's also possibly a lot of um uh costs to having really fast price Discovery and this high frequency

trading arms race um that we see I'm not sure 12 seconds is necessarily the right period but it seems clear that you know nobody really naturally needs price Discovery within a second it just kind of uh uh happens or within 100 milliseconds you know down to the nanc no one really needs that um for you know um any like practical real world thing it's just it's a consequence of just these these um uh Market forces that causes it to happen and so trying to figure out is there a way to do something more like a batch auction where we can have slower price Discovery um you know I think I think without without losing to competitors who it faster I think it's a really important open problem man I want to believe that so badly because I think anyone who's spent you know even any amount of time

looking at high frequency trading has kind of zoned out for a second and ask themselves what are we really doing here like what social good is this providing and I think that just like anything else there's sort of a curve right where you do want markets to be deep and liquid and actually there are a lot of good things that come from that but what's the difference between price discovery on a onec time frame versus 500 milliseconds versus 10 milliseconds I it's it's I think it hard it gets I think you probably see less and less and one thing that we didn't discuss actually was something cool to check out but the the third bucket that we talked about with Alec from Alex about just different buckets of information to be aware of where your flow is coming from so one thing that I think is starting to

happen on balancer is that that the flow that's coming from cow swap is actually getting better prices because in general right if you're a Trader that's placing trades on cow swap you're a comfortable waiting for a little bit which means you're probably not one of these high frequency Trader uh toxic orderflow people and balcer has essentially priced that and said okay there's a high likelihood that people who are trading here are retail and we're actually going to give them a better price so something kind of cool to look out at one thing that I thought about with lvr and I know we're in our final minutes here is this is a very sophisticated approach to profitability for an LP so I I would just go out on the limb and say lvr is only relevant really for professional or institutional market makers I just can't see a retail

person first of all even grocking this concept but second of all there's no way that they could really measure this in any sort of consistent way so does the prevalence of lbr presuppose a more sophisticated professional class of active LP or what do you think about that idea I don't think so in that I think ultimately what it reflects is something is something very real which is um which is a it's a real opportunity cost and a possible loss and expectation for for retail liquidity providers um and I think it it does so it measures it in a more reliable way um and in a clearer way a more a more in some sense a subjective way than um uh than impermanent loss or or someone just

looking at their realized p&l over a period it's very hard you know when you think about for for a um a retail liquidity provider for example if they're looking at just whether they made made or lost money it's sort of like a like a poker player um you know looking at just oh did I play good well tonight and they look at just whether they won or lost money on their um uh on the hands they played um or or for the for the whole night and they said oh like I lost money so I must have made some mistake right and that would be a mistake if you if you've got like a shortterm um uh a short time Horizon because you just there may there's a lot of luck or a lot of just you know how things worked out that could have been

diff you could have played perfect poker but still lost lost some money and I think similarly this concept for even amateur poker players I think it's important to figure out like how well did you play the cards that you were dealt and I think that's roughly what what lever is trying to measure I I agree with you I think the question I was really trying to get at is will they do that I I know that is the right thing to do but will they in the long run I think I think so and in the long run I think there's a lot of competitive reasons why um yeah why why I think like they should care about it and they should and they will they will tend towards solutions that do it and this is this is separate from again whether ultimately like I think liquidity

Fighters are people who are who are comfortable with basically the the Delta risks that they're taking there right that's why they're doing the rebalancing portfolio I think they they sort of want that that's more in their preference set but then everything else is kind of the um it's like vers you know how well are you doing vers compared to that and I think it's I think it's something that in theory could actually be very legible to them or made legible yeah all right now uh guys just to give you a little bit of a preview of next week's episode The the reason we talked about designing sort of a gas constrained environment is we're going to have a central limit order book Builder from salana and someone who's a little bit more in the amm side from ethereum so we're gonna get have a very fun discussion about

what the differences are of Designing kind of with gas cost and sort of a gas constraint environment on eth eth Main Net versus an environment like salana where that's not so much of a cost and then we're going to talk about sort of the spectrum of amm and Central limit order books and how there's a little bit maybe of a convergence that's happening there as well but Dan this was a ton of fun and I will see you next [Music] week