MEV burn

Transcript

[Music] foreign [Music] which is a piece of research that I'm very excited about because it's a potential upgrade to ethereum which I think of as being the equivalent of eip1559 but for Mev okay so there's this mental model that I want you to keep in your mind which is that there's two use cases for Block space on the one hand you can use blogspace as a means of including your transactions of getting confirmation of

your transactions but then another use case of the blockchains is specifically to get an ordering so for example you might want to be at the very top of the block and when you're competing for inclusion you get congestion just like you know traffic on the road where lots of cars are competing to be on the road but then when you're competing to be at the very top of the block when you're competing for ordering you have contention because the various people are contending to be in that first place um and the block value really is the sum of these two like the congestion value and the contention value and so we're all familiar with the the eip1559 transaction basically kind of going up and down with time

and we're also familiar with Mev which also kind of spikes a little bit more randomly uh uh with with time and so the the contention fee I guess is is another word for for Mev and really what we've done with eip1559 is we found a way to burn the congestion fee and Mev brand is nothing more than the other side of the coin where now we can burn the contentions be and there's two kind of classes of of benefits for Mev burn on the one hand we have benefits from smoothing and we have benefits from redistribution now it's moving I mean specifically

rewards moving so as a validator right now the Mev spikes kind of bring a lot of volatility a lot of variance and when you burn the Mev you kind of bring this this clean um issuance based rewards and rewards moving comes with security benefits as well as we'll see and then there's this this cash flow distribution redistribution aspect of Mev brand where basically right now the recipients are I mean of Mev burn as as a group are the stakers and we kind of redirecting this cash flow towards the holders and there's some economic benefits that come with that um as we'll talk about

but before talking about the benefits let me try and talk about the construction itself just to give you a little bit of intuition as to how it works so the context for Mev burn is proposal Builder separation so let me try and recap that real quick so you have builders that are doing the heavy lifting of building blocks and they submit bids uh in in a bit pool right now the bid pool is maintained by a so-called relays that are kind of these semi-trusted operators that kind of tells the proposers when you have a a valid bid The Proposal generally speaking will pick the the highest paying bid so the top bid from the bid pool

and they will commit to it by signing it with their with their secret key and they're going to broadcast this this sign bit and as soon as they've done that they've basically guaranteed some sort of payment to themselves of the bid value basically from the Builder to to The Proposal and then once there's been a commitment to the bid the builders are now um you know invited to reveal the the payloads corresponding to the bid so that's proposed to build a separation and there's a very specific thing called enshrines proposal Builder separation where we put this logic at layer one within consensus and the way it works is is as follows so you have this this this

one slot and before the slot even starts you have this bidding period where the builders send their bids to to a bid pool and at the very beginning of the slot The Proposal is meant to commit to the bid as we've discussed and then there's the reveal phase by the Builder kind of somewhere later in the in the slot so this is all um you know was all covered in the in the previous slide and the thing I wanted to show you here is that we have a testers kind of at layer one whose role is to enforce that proposal Builder separation is working smoothly and really what they're doing is that they're attesting to timeliness so they're making sure that the signed bid was broadcast on time and same for

the payload that the payload was revealed on time you actually don't need to know these details but it's mostly if you're curious um now what is the how do you get from here to Mev burn it's actually very very simple and this is what I wanted to illustrate is that MEP burn is like technically speaking a very small diff on top of Enterprise PBS and this is it so basically what we're going to do is we're going to ask the attesters to observe the bid pool specifically the bitpool two seconds before the start of the slot and we want them to calculate the so-called burn flaw which is going to be the top bit the maximum bid that was observed two

seconds before the start of the slot and we also want them to enforce this burn flaw at the proposal level so basically we're saying that if the proposer picked a bid which didn't respect this burn flow so it was not a bid which kind of maximized the value then that bid is just not going to get included on chain okay and this is the formula for the burn flow basically we want to take the maximum of all the the bids that came two seconds early meaning that they were two seconds before the start of the slot and we want to maximize the sum of the transaction base fees so that's eip1559 and this new thing which you know we

call Mev but in this context we're calling the Block Base B so there's two base fees one per transaction and also one per block and we want to maximize the value of the block and we kind of want to maximize the burn as it were and the bid value instead of going to the proposal it just gets destroyed just like erp1559 okay so what are the the benefits of of of this and they they kind of fall into three categories so first uh we're kind of enshrining infrastructure and so we're kind of pushing the markets in a certain direction which we feel is specifically healthy in terms of decentralization in terms of censorship and things like that

so I'll talk about that we also have benefits from smoothing that are security benefits as I mentioned and we have benefits from redistribution which are economic and so I'll go through these three types of of benefits one by one starting with enshrining so one of the the first nice things that we get from Mev brand is we get an Mev Oracle what does it mean it means that the ethereum blockchain now becomes aware of how much muv there is on its own chain and this is very similar to the congestion um basically well now we have this this you know oracle that can be used in small contracts for example if you want to do block space futures or anything like that

and it's it's and I guess the other thing the other advantage of having a a base fee is that we we're kind of enshrining if right so if you want to pay for the congestion fee you have to use if and you will get the same benefit with Mev brand so with Mev burn we have this new base fee at the at the Block Level that has to be paid in if so if it's kind of a privileged unit of account in the context of ethereum and we get this very nice Oracle that gives ethereum some level of introspection okay another um advantage of Mev burn is that it forces the market towards a more trustless equilibrium

uh and the reason is that enshrined PBS gives us this new bit pool but it's kind of optional to use the builders and the validators can still if they want to go through the relays and there's actually reasons to use the relays whenever the bid value is greater than the collateral posted by the Builder because now the proposal wants to have a guaranteed payment from the Builder but the build the collateral is not sufficient and so you need to use the The Trusted relays anyway long story short is muv burn kind of removes the option for the proposer to to go through the um through the trusted relays and it kind of forces the market through this more enshrined more

decentralized equilibrium another nice advantage of Mev burn is around censorship so right now we have some basically we have the bit pool is split into two sections on the one hand we have bids from censoring builders for example builders that sensor ofac transactions and then we have bids that from builders that don't censor and so some some of the proposes might only listen to the sense of bitpool but not to the uncensored one like they're kind of pretending to as if this uncensored bid pool didn't exist and here are some of the the the the the operators doing that for example Celsius and jump capital but with Mev burn we're kind of forcing

the these these operators that want to censor to listen to both mempools and if it so happens that the top bid is in the uncensored bit pool then they have to include it and so we're kind of forcing the hands against censorship another uh kind of secondary uh censorship benefit is around the cost of censorship so before eip1559 if you didn't include a specific transaction which you didn't know which you wanted to censor then you kind of need to pay the full you know transaction fee for for that transaction but with eip1559 you only have to forego the tip and generally speaking the tip is you

know at least in order of magnitude smaller if it might be one one way per gas relative to the base fee which might be I know 30 gray per gas so it's 30 times smaller and so it's 30 times cheaper to censor specific transactions on the theorem but with Mev burn with this burn maximization kind of idea we go back to um you know the cost the full cost of censorship being the full uh transaction fee okay now let's talk about uh um benefits from from smoothing and this is what got me interested in the first place in in in in in Mev band because there's some various attacks where Mev can

affect the stability of the layer one itself and then maybe burn addresses those so one of the the things uh to to be cognizant of is that the median Mev return is significantly smaller than the average Mev return so the the average is about three times larger and there's some really extreme blocks out there so for example there was this one block which had 691 if of rewards in a single slot which is about 12 000 times um the the median so you have this this distribution which kind of skews very much on these very very large Mev Rewards and this is the exact same

situation as a Bitcoin proof of work right with Bitcoin proof of work if you buy a mining rig and you run it for five years more likely than not you're never going to produce a block and so what you need to do to remain you know competitive especially as a small consensus participant is you need to pull and so pooling is basically a way to start with this Lottery and then kind of smooth it out and have fairness where you know every validator can get the average as opposed to small validators only getting the medium but burning gives us another option which doesn't have the centralization pressures that that pooling has so that's nice Mev burn removes

centralization pressures another aspect is basically proposes are incentivized to fight each other for the Mev so imagine that you have three proposals in a row and you know they each should be having this the the share of Mev but then the very last proposal decides to be evil and to DDOS the the previous two proposals so what's going to happen is that proposal n minus 2 and N minus 1 is not going to receive their share of Mev and instead what's going to happen is that the Mev is going to go to the malicious proposer and so Tada just by dosing a few IP addresses you've tripled your Mev

and Mev burn addresses that because Mev doesn't flow to the proposals it just goes it flows to the ifolders and so you remove these these incentives to basically destabilize the the layer one chain and dosing is only one such example of an attack there's kind of a whole set of attacks you can do to steal Mev from other participants another problem with Mev today is what I call rug pooling is basically when operators of a staking pool are going to rug the the stakers meaning those who provide the eve stake and if you think of Rocket pool or if you think of Lido there's operators that

are collateralized so for example with Lido you have reputation as your collateral in the case of Rocket pool you have if and RPL tokens as collateral and most of the time the Mev that you get in any given slot is going to be less than the collateral of the operator so everything's incentive aligned there but you know some of the time what's going to happen is that there's going to be a huge Mev Spike which is going to be greater than the collateral of the operator and here the operator is incentivized to just run away um take the Mev reward for themselves and forego the collateral and what we're seeing is that these decentralized pools you know are

basically competing to lower and lower and lower the amount of collateral so you have more collateral efficiency but this opens up this attack more and more and more and so Mev burn is a way to basically make decentralized staking pools much more stable and robust another aspect of smoothing is toxic Mev let's say there's a user who's been who's a victim of a sandwich attack or maybe the victim of some sort of hack and then that leads to Mev going away from the user to the Staker toxic muv is defined as Mev where there is a clear victim um that's that's that's that's losing value

and so here what we have is this kind of conundrum this dilemma should we return the Mev um and you know there's kind of an ethical a moral dilemma but there's you know maybe a legal dilemma going on and one of the the bad things here is that if you're really good-hearted and you want to return the Mev then suddenly you become uncompetitive because some operators for example the Lido operators kind of as a policy will say well we won't return the the the Mev and so you would have an APR which is an expectation lower if you're good-hearted than than those who don't return and so Mev band just removes this this whole

question um but then there's a there's a kind of another similar scenario which is what I call systemic Mev where you have huge amounts of toxic mud not you know if you if but a few millions of Eve so imagine that one of the Roll-Ups gets hacked it has 10 million users and 10 million if gets hacked now a single validator just received 10 million if out of nowhere you know it's almost like the Dao hacker just stole a huge amount of if and that could you know be systemically dangerous for the blockchain because okay we need to ask ourselves okay do we want to bail out this roll up do we want to or you know maybe the the Staker might decide to use

this if to to perform some sort of of attack or just be misaligned with the community and we don't want to have one single actor having 10 percent of the each Supply that would that would not be uh palatable another aspect um of smoothing is that it unlocks so-called stake capping so it's healthy to have a relatively small number of of uh a small amount of stake we want we don't want to be the position where you know there's 80 of each state maybe you know 25 is is more than enough and if we do want to have the state capping what we can do is have an issuance curve which goes down as soon as you get close to your target

so for example let's say we want to cap the total amount of Eve to 30 million if stake teeth to 30 million as soon as we get close to the 30 million the issuance curve starts going negative and so the closer you are to 30 million the more and more negative you you get but in order to do that safely we need Mev burn because otherwise you kind of really accelerating the pooling effect like the the variance when you remove the issuance and you only have Mev then the the the variance is even is even larger now State capping gives us validator capping which means that we have a limited number of validators we like this and the reason we like this is because it means we can have smaller

validators so right now we need 32 if as a minimum to start the validators but maybe in the future we might only need eight if and capping the the worst case number of validators is is a way to help us get smaller validators but we can do other things as well we can have shorter slots if we have fewer validators and we might also be able to unlock big upgrades like single slots finality when we have fewer validators and then kind of another interesting aspect of State capping is that we're burning restaking proceeds so very soon you know with projects like eigenlayer we're going to have three components to the to to to staking we're going to have the issuance as a reward we're going to

have the Mev as a reward and then there's gonna be this third bucket of yield from restaking and is State capping is basically a way to do restaking but meaning to burn the yield from from restaking so that's a you know a potential future upgrade as well that's unlocked by MAV bands okay final section in terms of benefits of Med burn is the the redistributive aspect of instead of giving the um the Mev to the um to the stakers we give them to to the etholders now today this is the amount of take teeth that we have and it can be

you know classified in two buckets on the one hand we have staked if that's there you know from the issuance rewards and then we have some sort of additional state if that is there just because there's just more um more rewards to to to to be had and so if we were to do Mev burn we'd only get the the state Eve from from from issuance now the good news is that we've designed ethereum in such a way that even if there's zero Mev the state E from issuance to be sufficient to secure a theorem and so arguably anything on top of that anything above any basically non-zero amount of Mev is wasteful because we're wastefully paying for security which we don't need because

we have Security even if there's zero Mev now why is this why is this a good thing so one of the reasons is that we have if which is this really unique asset on ethereum it's the only pristine collateral asset is this collateral money and there's two big case use cases of collateral money on the one hand you can stick it into staking and we call that Economic Security because you're you're securing ethereum or you can stick it into D5 and we call that economic bandwidth because it unlocks applications like stable coins that requires a lot of collateral and with Mev we're potentially going to be in a really lopsided situation where

you know the the the beacon chain is gonna absorb like a sponge all the all the liquidity and there's going to be very little economic bandwidth left and so Med burn is kind of a way to have more predictable and more balance between if that's deployed for Economic Security and if that's deployed for economic bandwidth another aspect is reduced issuance so the total amount of issuance is roughly speaking well it's exactly I guess the the square root of the total amount of Eve State and so if there's less Eve staked with Mev burn then there's going to be less issuance and so the cost of proof of stake is going to go down we're saving this little orange area

and so this brings me to the economic efficiency that goes up on the one hand Mev burn reduces the amount of issuance removes wasteful spending for security which we don't need and there's this burn aspect um which you know further increases the economic efficiency now I try to put some numbers concrete numbers on on what this were and I try to be you know roughly conservative and what I got was about a quarter of a million uh per year according each per year and so Mev burn very roughly speaking is a half a million each per year optimization and what that means is that you know the the if Supply would

decrease by an additional half a million per year versus what it's currently doing and so that's you know a very ultrasound uh uh for for ethereum and that gives us more economic bandwidth and more Economic Security from a US dollar denominator denominated standpoint because we have more scarcity of E30 asset and then this is my last slide around tax efficiency which is maybe like a really weird slide uh to put to put here but um I just want to you know end with this analogy with companies let's say as a company you have profits and you want to distribute those profits to shareholders well you have two options option number one is that you pay dividends to your shareholders

or option number two is use share buy packs like apple does they reduce the total supply of Apple shares um and then that kind of inflates the price of each Apple share and the clever thing is that the BuyBacks are taxed as capital gains whereas the dividend attacks as income tax and so for me and most jurisdiction as I understand income tax is significantly higher than capital gains so in the UK income tax is 50 capital gains is 20 and so you know we're paying two and a half times more taxes than we should be paying and so Mev burn is kind of a way to reduce the cell pressure uh on each from paying tax bills to nation states

around the world and that's pretty much it that's all the advantages of Mev burn thank you [Applause] I have 30 seconds so maybe just one question thank you very interesting um I wanted to ask about the two seconds was that just an arbitrary length and also do you think this will affect the effectiveness of lmd score boosting and would just incentivize validators to wait longer into a slot and wait for more bids to pile up so the the

band was chosen as a conservative synchrony assumption basically we want the property that all the bids that are sent from the builders will reach the proposes and the testers within two seconds so they basically get gossiped out within two seconds so that the proposal at the very start of the slot has received whatever bid there was two seconds ago and so they act they have the option of you know being compliant with the bank flow um in terms of might this change the the game Dynamics where Builders might be incentivized to post their bids after the two seconds

um it turns out that from the perspective of the builders the incentives don't change at all builders make the exact same profit um the only thing that changes is the profit uh is like the the the Mev going to the e-folders as opposed to going to the proposal so proposals don't really have an incentive to change that game and the nice thing about Mev burn is that if there's even a single Builder who is willing to to you know to play this game of submitting bids in real time before the two second threshold the whole thing works um and yeah that's pretty much it happy to talk about it more offline thank you [Applause] [Music]

foreign [Music]