Maximal extractable value (MEV)
Maximal extractable value (MEV) refers to the maximum value that can be extracted from block production in excess of the standard block reward and gas fees by including, excluding, and changing the order of transactions in a block.
Maximal extractable value was first applied in the context of proof-of-work, and initially referred to as "miner extractable value". This is because in proof-of-work, miners control transaction inclusion, exclusion, and ordering. However, since the transition to proof-of-stake via The Merge validators have been responsible for these roles, and mining is no longer part of the Ethereum protocol. The value extraction methods still exist, though, so the term "Maximal extractable value" is now used instead.
In theory MEV accrues entirely to validators because they are the only party that can guarantee the execution of a profitable MEV opportunity. In practice, however, a large portion of MEV is extracted by independent network participants referred to as "searchers." Searchers run complex algorithms on blockchain data to detect profitable MEV opportunities and have bots to automatically submit those profitable transactions to the network.
Validators do get a portion of the full MEV amount anyway because searchers are willing to pay high gas fees (which go to the validator) in exchange for higher likelihood of inclusion of their profitable transactions in a block. Assuming searchers are economically rational, the gas fee that a searcher is willing to pay will be an amount up to 100% of the searcher's MEV (because if the gas fee was higher, the searcher would lose money).
With that, for some highly competitive MEV opportunities, such as DEX arbitrage, searchers may have to pay 90% or even more of their total MEV revenue in gas fees to the validator because so many people want to run the same profitable arbitrage trade. This is because the only way to guarantee that their arbitrage transaction runs is if they submit the transaction with the highest gas price.
This dynamic has made being good at "gas golfing" — programming transactions so that they use the least amount of gas — a competitive advantage, because it allows searchers to set a higher gas price while keeping their total gas fees constant (since gas fees = gas price * gas used).
A few well-known gas golf techniques include: using addresses that start with a long string of zeroes (e.g. 0x0000000000C521824EaFf97Eac7B73B084ef9306) since they take less space (and hence gas) to store; and leaving small ERC-20 token balances in contracts, since it costs more gas to initialize a storage slot (the case if the balance is 0) than to update a storage slot. Finding more techniques to reduce gas usage is an active area of research among searchers.
Rather than programming complex algorithms to detect profitable MEV opportunities, some searchers run generalized frontrunners. Generalized frontrunners are bots that watch the mempool to detect profitable transactions. The frontrunner will copy the potentially profitable transaction's code, replace addresses with the frontrunner's address, and run the transaction locally to double-check that the modified transaction results in a profit to the frontrunner's address. If the transaction is indeed profitable, the frontrunner will submit the modified transaction with the replaced address and a higher gas price, "frontrunning" the original transaction and getting the original searcher's MEV.
Flashbots is an independent project which extends execution clients with a service that allows searchers to submit MEV transactions to validators without revealing them to the public mempool. This prevents transactions from being frontrun by generalized frontrunners.
MEV emerges on the blockchain in a few ways.
Decentralized exchange (DEX) arbitrage is the simplest and most well-known MEV opportunity. As a result, it is also the most competitive.
It works like this: if two DEXes are offering a token at two different prices, someone can buy the token on the lower-priced DEX and sell it on the higher-priced DEX in a single, atomic transaction. Thanks to the mechanics of the blockchain, this is true, riskless arbitrage.
Here's an example of a profitable arbitrage transaction where a searcher turned 1,000 ETH into 1,045 ETH by taking advantage of different pricing of the ETH/DAI pair on Uniswap vs. Sushiswap.
Lending protocol liquidations present another well-known MEV opportunity.
Lending protocols like Maker and Aave require users to deposit some collateral (e.g. ETH). This deposited collateral is then used to then lend out to other users.
Users can then borrow assets and tokens from others depending on what they need (e.g. you might borrow MKR if you want to vote in a MakerDAO governance proposal) up to a certain percentage of their deposited collateral. For example, if the borrowing amount is a maximum of 30%, a user who deposits 100 DAI into the protocol can borrow up to 30 DAI worth of another asset. The protocol determines the exact borrowing power percentage.
As the value of a borrower's collateral fluctuates, so too does their borrowing power. If, due to market fluctuations, the value of borrowed assets exceeds say, 30% of the value of their collateral (again, the exact percentage is determined by the protocol), the protocol typically allows anyone to liquidate the collateral, instantly paying off the lenders (this is similar to how margin calls work in traditional finance). If liquidated, the borrower usually has to pay a hefty liquidation fee, some of which goes to the liquidator — which is where the MEV opportunity comes in.
Searchers compete to parse blockchain data as fast as possible to determine which borrowers can be liquidated and be the first to submit a liquidation transaction and collect the liquidation fee for themselves.
Sandwich trading is another common method of MEV extraction.
To sandwich, a searcher will watch the mempool for large DEX trades. For instance, suppose someone wants to buy 10,000 UNI with DAI on Uniswap. A trade of this magnitude will have a meaningful effect on the UNI/DAI pair, potentially significantly raising the price of UNI relative to DAI.
A searcher can calculate the approximate price effect of this large trade on the UNI/DAI pair and execute an optimal buy order immediately before the large trade, buying UNI cheaply, then execute a sell order immediately after the large trade, selling it for the higher price caused by the large order.
Sandwiching, however, is riskier as it isn't atomic (unlike DEX arbitrage, as described above) and is prone to a salmonella attack.
MEV in the NFT space is an emergent phenomenon, and isn't necessarily profitable.
However, since NFT transactions happen on the same blockchain shared by all other Ethereum transactions, searchers can use similar techniques as those used in traditional MEV opportunities in the NFT market too.
For example, if there's a popular NFT drop and a searcher wants a certain NFT or set of NFTs, they can program a transaction such that they are the first in line to buy the NFT, or they can buy the entire set of NFTs in a single transaction. Or if an NFT is mistakenly listed at a low price, a searcher can frontrun other purchasers and snap it up for cheap.
One prominent example of NFT MEV occurred when a searcher spent $7 million to buy every single Cryptopunk at the price floor. A blockchain researcher explained on Twitter how the buyer worked with an MEV provider to keep their purchase secret.
DEX arbitrage, liquidations, and sandwich trading are all very well-known MEV opportunities and are unlikely to be profitable for new searchers. However, there is a long tail of lesser known MEV opportunities (NFT MEV is arguably one such opportunity).
Searchers who are just getting started may be able to find more success by searching for MEV in this longer tail. Flashbot's MEV job board lists some emerging opportunities.
MEV is not all bad — there are both positive and negative consequences to MEV on Ethereum.
Many DeFi projects rely on economically rational actors to ensure the usefulness and stability of their protocols. For instance, DEX arbitrage ensures that users get the best, most correct prices for their tokens, and lending protocols rely on speedy liquidations when borrowers fall below collateralization ratios to ensure lenders get paid back.
Without rational searchers seeking and fixing economic inefficiencies and taking advantage of protocols' economic incentives, DeFi protocols and dapps in general may not be as robust as they are today.
At the application layer, some forms of MEV, like sandwich trading, result in an unequivocally worse experience for users. Users who are sandwiched face increased slippage and worse execution on their trades.
At the network layer, generalized frontrunners and the gas-price auctions they often engage in (when two or more frontrunners compete for their transaction to be included in the next block by progressively raising their own transactions' gas price) result in network congestion and high gas prices for everyone else trying to run regular transactions.
Beyond what's happening within blocks, MEV can have deleterious effects between blocks. If the MEV available in a block significantly exceeds the standard block reward, validators may be incentivized to reorg blocks and capture the MEV for themselves, causing blockchain re-organization and consensus instability.
This possibility of blockchain re-organization has been previously explored on the Bitcoin blockchain. As Bitcoin's block reward halves and transaction fees make up a greater and greater portion of the block reward, situations arise where it becomes economically rational for miners to give up the next block's reward and instead remine past blocks with higher fees. With the growth of MEV, the same sort of situation could occur in Ethereum, threatening the integrity of the blockchain.
MEV extraction ballooned in early 2021, resulting in extremely high gas prices in the first few months of the year. The emergence of Flashbots's MEV relay has reduced the effectiveness of generalized frontrunners and has taken gas price auctions off-chain, lowering gas prices for ordinary users.
While many searchers are still making good money from MEV, as opportunities become more well-known and more and more searchers compete for the same opportunity, validators will capture more and more total MEV revenue (because the same sort of gas auctions as originally described above also occur in Flashbots, albeit privately, and validators will capture the resulting gas revenue). MEV is also not unique to Ethereum, and as opportunities become more competitive on Ethereum, searchers are moving to alternate blockchains like Binance Smart Chain, where similar MEV opportunities as those on Ethereum exist with less competition.
On the other hand, the transition from proof-of-work to proof-of-stake and the ongoing effort to scale Ethereum using rollups and sharding all change the MEV landscape in ways that are still somewhat unclear. It is not yet well known how having guaranteed block-proposers known slightly in advance changes the dynamics of MEV extraction compared to the probabilistic model in proof-of-work or how this will be disrupted when single secret leader election and distributed validator technology get implemented. Similarly, it remains to be seen what MEV opportunities exist when most user activity is ported away from Ethereum and onto its layer 2 rollups and shards.
- Flashbots docs
- Flashbots GitHub
- MEV-Explore Dashboard and live transaction explorer for MEV transactions