# Frequently Asked Questions about MEV Burn ## How can you burn MEV? The first step is quantifying the amount of MEV, and the conceptually simplest way to do this is with [Proposer/Builder Separation](https://notes.ethereum.org/@domothy/pbs_links) (PBS), where builders compete against each other and offer bribes to proposers for their block to be included, then the proposer simply has to pick the top bid. Here is a simple example to visualize this auction process: if there is 1 ETH of extractable value in a block, then a rational builder will not bid more than 1 ETH. With multiple builders competing over the same block, the top bid quickly approaches 1 ETH. For example, the winning bid might be 0.99 ETH: The proposer collects 0.99 ETH, and the builder collects this "delta" of 0.01 ETH as profits. Today, such auctions happen out of protocol via [MEV-Boost](https://boost.flashbots.net/). The first (and hardest) step in burning MEV is to **enshrine** Proposer/Builder Separation directly in the protocol. This has multiple benefits, but the most relevant one here is that the protocol itself will become aware of MEV bids. After this, actions can be taken in response to these bids broadcasted by builders, such as send the money nowhere – similar to how EIP-1559's base fees are subtracted from account balances with no equivalent balance increase anywhere else, a process colloquially understood as *burning* ETH. Another, less direct alternative to burn MEV would be to combine two distinct design proposals: first by [smoothing it across validators](https://ethresear.ch/t/committee-driven-mev-smoothing/10408) and then adjusting the yield curve so that [issuance becomes negative](https://notes.ethereum.org/@vbuterin/single_slot_finality#Economic-capping-of-total-deposits) past a certain number of validators. MEV would go to validators, but they would effectively be paying for their spot as a validator via a negative issuance. ## What's the incentive for builders to extract value that just gets burned? In practice, nothing actually changes for block builders and searchers: They're *already* doing the hard work of extracting value even though a big bulk of it makes its way to proposers. After MEV burn, the only thing that changes is that this payment to proposer instead gets burned. A simple mental model for MEV burn is to simply picture that the block builders would be bribing the blockchain itself rather than whichever proposer happens to be the lucky one for that slot. The builder will still collect the delta between the total value extracted and the bid that gets burned. ## What's the incentive for proposers to go along with their bribes getting burned? If the equation involves only proposers and builders, collusion is inevitable due to the incentives: both the builder and proposer would like to pretend there is no MEV in the block by bidding 0 ETH in-protocol, in order to keep it all to themselves. That is why mechanisms proposers involve a disinterested neutral third party, namely a committee of validators that all impose constraints on the validator in charge of proposing a block. By imposing their view of what the top bid is, the choice for each proposer is to either go along with it or not have their block included in the main chain at all and receive no rewards. Since validator committees are involved, being able to collude implies controlling a majority of validators, at which point much more important honest majority assumptions break and the colluding majority can do way worse things that just keep MEV for themselves. ## What does this change for people who use the blockchain to do stuff? Nothing: All this stuff happens in the background, the fee market will continue working the same way with more congestion resulting in higher fees. Unlike EIP-1559, MEV burn isn't a UX upgrade for users. Interestingly, submitting a transaction with a higher priority fee still results in prioritizing its inclusion even if most of that priority fee will be burned by virtue of becoming a block reward that builders can use to make higher bids. ## Does MEV burn solve toxic MEV? By toxic MEV, I mean the kind of MEV that directly extracts value from a specific user, the most common example being *transaction sandwiching*: when a user broadcasts a transaction for a sizable trade on a DEX with large slippage, MEV bots can place their own transaction right before and after the user's. This is in contrast to non-toxic MEV such as DEX arbitrage, CDP liquidations, priority fees, etc. which plays a positive role in the ecosystem and provides various benefits to users. To answer the question: **No**, MEV burn will *not* eliminate or disincentivize toxic MEV. The economic incentives to sandwich users will still be present at the builder and searcher level, and they will still outbid each other in a race to the bottom. These bids will simply get burned instead of passed along to the proposers. Tactics to mitigate toxic MEV are a seperate topic from MEV burn, and both can/should happen. At the protocol level, there are a few proposals, namely [Encrypted Mempools](https://www.youtube.com/watch?v=XRM0CpGY3sw) or [Shutterized Beacon Chain](https://ethresear.ch/t/shutterized-beacon-chain/12249). At the application layer, dapps and infrastructure can be built with MEV in mind to either capture it themselves or prevent it altogether (e.g. various Flashbots initiatives, CoW Swap, etc.). But even if toxic MEV gets completely eliminated, there will always be non-toxic MEV which should be burned. ## How does burning MEV impact transaction censorship? This is about "weak censorship", or the thing that's tracked over at [MEVWatch.info](https://www.mevwatch.info/). In a nutshell, this means specific censored transactions are delayed by a few seconds due to block builders and/or proposing refusing to include them in their block. The most common example of external incentives playing a role in weak censorship is OFAC compliance. The impact of MEV burn on this form of censorship is to disincentivize proposer-side censorship: Today, a proposer running MEV-boost can decide to connect exclusively to censoring relays and ignore all bids from non-censoring block builders. After enshrined PBS and MEV burn, the proposer would be forced to either pick the top bid — censoring or not — or bear the opportunity cost of the censored transaction themselves. e.g. if the top censoring bid is 0.04 ETH but there is a higher bid of 0.05 ETH from a non-censoring builder, then the proposer would still be forced to burn 0.05 ETH even if they insist on picking the censoring bid – incurring a personal net loss of 0.01 ETH. The same dynamic happens on the block builder side: adding an extra fee-paying transaction always results in more block rewards available for a higher bid compared to the same block that doesn't include the transaction. Thus a censoring builder won't be as competitive as a non-censoring builder. Although it's not always that simple, a censoring builder who has more optimized MEV algorithms and/or private access to transactions can potentially outbid a non-censoring builder simply by virtue of being more profitable. This is why ePBS designs must come with [various mechanisms](https://notes.ethereum.org/@fradamt/H1TsYRfJc) to enforce censorship resistance, such a inclusion lists who force builders to include transactions seen by validators. ## Won't it reduce the staking yield? In aggregate yes, less ETH is going to the class of ETH holders that are stakers. But the long term result is that this "class of ETH holders that are stakers" would shrink, such that each staker would individually make the same yield as before. The argument is that the staking yield is a self-balancing mechanism: if the yield is too high, more stakers come in and dilute it, and when the yield is too low, stakers exit and use their capital elsewhere. Whatever constitutes the "correct amount of yield" is left to the market to decide. For example, if the market prices the proper staking yield at 3%, then that's what it will converge to, with or without MEV rewards being part of the equation. With MEV burn, the only thing that changes is that this 3% comes from the beacon chain issuance, rather than some combination, e.g. 2% from issuance and 1% from MEV. ## What are the benefits of MEV burn? Burning MEV solves a few problems present in today's landscape with the "MEV lottery" giving a few lucky proposers a lot of extra reward: ### Smoothing the staking yield A design goal of proof of stake is that owning x% of the stake should result in getting x% of rewards. With MEV rewards being what they are today, this is not the case except on very long time frames: Large staking pools who propose blocks more often are much more likely to benefit from rare big spikes in MEV, while your average solo staker will only propose 4-6 blocks a year with a very small chance of getting a high-value MEV payout. Worse still, staking pools who get a large MEV reward can immediately compound it into more validators, increasing their chance (and decreasing yours) to get more MEV the next time. When the MEV instead gets burned, the staking yield becomes much smoother across the board: proposing blocks more often simply means burning MEV more often, otherwise everyone gets the same proportional staking yield that comes from issuance. Additionally, this effect of smoothing the staking yield drastically simplifies the design of decentralized staking pools, since they no longer have to worry about node operators potentially running off with MEV rewards that belongs to the pool when the benefit outweighs the cost. With MEV burn, the smoothing happens automatically for everyone. ### More stable validator set By burning the MEV, we decouple the beacon chain's security from the economic activity happening on-chain. This way, we can preserve a stable number of validators across the board regardless of whether we're in a bull markets with lots of MEV or a bear market with nothing happening. When the staking yield comes from protocol-controlled issuance, the size of the validator set can be controlled more effectively, which is handy when it comes to things like [Single Slot Finality](https://notes.ethereum.org/@vbuterin/single_slot_finality). ### Reducing incentive to reorg the chain If there is a lot of MEV rewards in the previous block, it can become an incentive for validators to ignore that block in order to steal the MEV opportunities for themselves in the next block. Even with slashing penalties, there are scenarios where the value extracted from reorganizing the chain is higher than the penalties from slashing. Instead, when it's burned, the dominant strategy is the one we want: every validator continues the main chain, since there is no point in stealing MEV that would just get burned in your block rather than the previous block. This results in a more secure and stable chain for everyone. ### Enhancing economic qualities of Ether as an asset Burning MEV is a way for the blockchain itself to capture value from the economic activity happening on top of it, and redistribute it proportionally to all ETH holders. This has some benefits, mainly from enhancing the [ultra sound money](https://ultrasound.money) qualities that were brought by EIP-1559: having a scarcer asset is good when this asset is predominantly used as *collateral* (as opposed to *debt* assets like fiat where inflation is better). Since Ether is the collateral asset used to secure Ethereum, it makes sense from a sustainability perspective to have it capture value that is otherwise extracted. Burning MEV is also in line with the Ethereum community's approach of aiming for "[minimal viable issuance to secure the blockchain](https://www.reddit.com/r/ethfinance/comments/izhkvk/minimum_viable_issuance_why_ethereums_lack_of_a/)" More MEV attracts more validators, which arguably leads in the chain "overpaying" on security by issuing more. So on top of burning more ETH, MEV burn would also result in less issuance by virtue of having fewer validators. For example, let's assume that 3% APR is staking yield's equilibrium, and that there are about 500,000 ETH of MEV distributed to validator a year: we can [calculate](https://docs.google.com/spreadsheets/d/1V8xxJc7r9p6Q5435I3osP9BiWzrv49w1vevnjS9B4WU/view) that MEV burn would not only result in those 500k ETH getting burned, but there is a difference of about 360,000 ETH that would never get issued compared to the same 3% APR coming from issuance+MEV. Finally, a less significant factor but still noteworthy is that MEV burn would enshrine ETH as the only possible currency used in the MEV market, in the same way EIP-1559 enshrined ETH as the only currency used to pay gas fees. Any value extractable on-chain, no matter how it is denominated (NFTs, stablecoins, some other tokens, a CBDC, you name it) would result in an equivalent amount of value *denominated in ETH* getting burned. ## What if the MEV ends up getting captured by rollups? Some (possibly a lot) of the MEV will indeed be captured by Layer 2 sequencers: Many rollup designs involve centralized sequencers, or a rotating cycle of various sequencers each taking turn batching L2 transactions to L1. In those designs, the L2 sequencer in charge effectively has a monopoly on transaction inclusion and ordering, which means they can capture MEV for themselves from the economic activity happening on their chain. However there are other things to note in the relationship between Layer 2 scaling vs. Layer 1 MEV: 1. There are alternative designs for rollups, such as [based rollups](https://ethresear.ch/t/based-rollups-superpowers-from-l1-sequencing/15016), that allow anyone to be a sequencer for any block. This means the L2 MEV becomes a free-for-all at the stack below, on L1. This sort of design is arguably simpler and safer, as there are fewer moving parts and inherits liveness guarantees from Ethereum's base layer. 2. Cross-domain MEV will always exist, e.g. arbitraging between two DEXes on two different rollups means carefully placing the two batches together in the same L1 block to extract value, which ultimately results in MEV being captured by layer 1. 3. Ethereum's Layer 1 will still be used for high-value transactions for the foreseeable future, even as rollups scale. This will always results in *some* base amount of MEV that can be captured. 4. The end goal plan to scale Ethereum's Layer 1 execution involves enshrining a zkEVM directly in the base layer itself, so any work done to capture MEV today will still bear fruit in the long run. ## How much more ETH will get burned? It is impossible to predict how much economic activity is going to happen in the future, so the next best thing is looking at [MEV-boost dashboards](https://github.com/superphiz/dashboards#mev-dashboards) like [this one](https://mevboost.pics/) – the amount of MEV going to proposers today is just about the best proxy to see how much ETH would get burned. As of early May 2023, [it is estimated](https://youtu.be/nb7x7n8Ga3U?t=4570) that burning the MEV since the merge would have resulted in the supply declining roughly three times faster. ## Wen MEV burn? At this point in time, we're mostly in the research phase of enshrined PBS, trying to form consensus on the best design before eventually specifying changes to the beacon chain specs, before even more eventually getting those new specs implemented and deployed on mainnet. So the unsatisfying answer is that there's no telling exactly when it's gonna happen other than that we still have to wait before the timeline even upgrades to "Soon™". Stay tuned for development! ## What are the similarities between EIP-1559 and MEV burn? It is natural to compare EIP-1559 with MEV burn (as has been done many times throughout this FAQ), but the similarities go beyond the burn itself: * Both mechanisms make the protocol aware of externalities it cannot control or predict: EIP-1559 deals with *congestion*, whereas MEV burn deals with *contention* * Both mechanisms capture this externality as a form "revenue stream" used for the chain's sustainability via burning * Both mechanisms rely on an honest majority of validators: * 51% of validators could collude and prevent blocks from ever going above 15M gas used, in order to nullify 1559 and return to the first priced auction where all gas fees go to validators * 51% of validators could collude and pretend that the highest bid they hear is 0 ETH, so that no MEV gets burned so that it flows back to them * In both cases, any validator part of that 51% personally benefits by defecting * Both mechanisms are partial burns – priority fees for 1559 and the delta between value extracted and top bid for MEV burn ## What are some good links about all this? * [Proposer/Builder Separation link hub](https://notes.ethereum.org/@domothy/pbs_links) * [Simple MEV burn design](https://ethresear.ch/t/mev-burn-a-simple-design/15590), May 2023 * [Committee-driven MEV smoothing](https://ethresear.ch/t/committee-driven-mev-smoothing/10408), August 2021 * [MEV burn via proposer auction](https://ethresear.ch/t/burning-mev-through-block-proposer-auctions/14029), October 2022 * [MEV burn on Bankless](https://www.youtube.com/watch?v=nb7x7n8Ga3U)