Attempts to eliminate MEV failed? How developers deal with the MEV crisis
Original title: MEV resolution: Are we there yet?
Original author:CHRIS POWERS
Original compilation: Luccy, BlockBeats
A lot has changed in the past three years since Flashbots released MEV-Geth (with the goal of preemptively solving the MEV crisis). Ethereum has weathered market booms and busts and successfully moved to Proof of Stake, a major technical achievement. At the same time, competition for the value leaked by economic transfers on the blockchain intensifies. Attempts to eliminate MEV failed, but block proposers are always willing to exploit their privileged position.
Now, the lure of MEV profits threatens Ethereum’s sovereignty and censorship resistance. There remains a central issue, which is that the search for MEV will centralize stakes in Ethereum. If only those who perform the most complex verifications can earn rewards from MEV, then ETH holders will gravitate toward these validators. This is because they can offer higher MEV returns on top of those enforced by the protocol.
At least, this centralized future is temporarily avoided with MEV-Boost. MEV-Boost enables both large stake pools and independent validators to equally share the results of MEV extraction. However, MEV-Boost was always a stopgap solution. Although it has produced a strong MEV ecosystem of specialized players, the industry has become increasingly centralized and dependent on trusted entities and single software clients.
Last December, we argued: “The best solution to the MEV crisis will be an independent, decentralized network focused on ordering transactions.” However, this is still a slow process. Modifications to the core Ethereum protocol are needed to ensure Proposal Builder Separation (PBS), but this is not enough.
We still firmly believe that the solution to the MEV crisis does not lie entirely in Ethereum, but we are now convinced that there will not be a monolithic solution to all MEV leaks. Instead, we foresee the emergence of various decentralized and vertically integrated MEV supply chains that extract MEV for specific applications. In addition, we believe that too much attention has been paid to MEV from CEX-DEX arbitrage, which currently accounts for approximately 60-70% of MEV trading volume and profits. However, if DeFi truly becomes the infrastructure of global financial markets, this will no longer be the biggest MEV challenge we need to face.
MEV
PBS (Protocol-Level Random Selection) is an overall design concept aimed at ensuring that Ethereum remains decentralized and neutral, and is generally considered a basic condition for solving the MEV problem. Currently, Ethereum implements PBS through MEV-Boost, software provided by Flashbots run by validators, which allows randomly selected block proposers to auction the rights to create the most profitable blocks to the highest bidder. To date, PBS has popularized MEV rewards by only allowing stakers to participate, without the need to use complex MEV strategies. However, risks of centralization and censorship remain in other parts of the MEV supply chain.
This diagram shows the major components of the MEV supply chain, a term first coined by Frontier Tech founder and former Flashbots member Stephane Gosselin.
The architecture of the Ethereum protocol is meant to be much simpler. The design assumes that for each block, a randomly selected validator will build the block locally by sifting through the transactions in the public mempool, which were submitted by the users who provided the highest gas fees.
In the MEV supply chain, there are three other actors, namely searchers, builders and repeaters. They work closely with validators through MEV-Boost, and currently 93% of Ethereum validators are running MEV-Boost. Of the three, builders are the easiest to centralize, while relayers have the most limited rewards. Searchers were once so-called super coders, but have now transformed into teams of developers, some remaining anonymous while others have become large traders. The economic relationships between searchers, builders, and relayers are complex, and payment traffic, although it occurs, is difficult to observe due to its off-chain nature.
More and more builders are structurally taking advantage of vertical integration. This week alone, Blocknative, a major block builder, announced that it will no longer serve as a trusted relayer. It is reported that this is because the cost of running a repeater (about $500,000 per year) is completely disproportionate to revenue. While it might make financial sense if Blocknative had its own team of searchers, as a US company it would avoid engaging in any activity that might draw the ire of regulators.
Bloxroute, a major builder, has yet to shut down its relay business, but is also avoiding regulatory risks. It operates two relays, one called regulated, which is responsible for censoring blocks containing OFAC-sanctioned addresses, and the other, maximum profit, which is not censored. After Blocknatives exit, there are currently only four major relayers left: Bloxroute, Flashbots, Ultrasound (led by Ethereum researcher Justin Drake), and Agnostic (from the Gnosis team).
CEX-DEX Arbitrage: The Worst Thing About MEV?
After understanding the current supply chain and its challenges, lets review the mechanics of MEV extraction and why it has attracted so much attention. For many, MEV is simply about trading before and after trades in the trading pool to lock in profits, or worse, sandwiching them in the middle, which is certainly not good for the average user. The most productive form of MEV is on-chain arbitrage between different DEXs. Trading occurs through the Uni v2 pool, driving the token price up or down, with powerful MEV bots scrambling to rebalance other liquidity pools and incorporate the new token price. These are the easiest to understand examples of MEV, but they are not the most common.
In fact, most MEV withdrawals are CEX-DEX arbitrage. This MEV mainly harms DeFi liquidity providers, not traders. The deepest market for ETH is not on Ethereum, but on Binance. When the price of ETH changes on Binance, there is a frantic trend of placing trades with Uniswap LP that has not yet incorporated the new price. The winning MEV must be the first transaction in the new block after the price change, which is the top of the block.
After a major price movement on Binance, MEV bots must bribe lucky validators who are randomly selected to propose the next block. In the world of PBS and MEV-Boost, the MEV bot first sends its transactions to a block generator that is populated with other transactions, and the block generator then pays some of the validators who proposed the block. All of this happens within 12 seconds, the time between blocks on Ethereum. Max Resnick of Special Mechanism Group (SMG) explained CEX-DEX arbitrage in detail at Flashbots MEV Salon Paris and in a full technical paper.
What’s disturbing is that when there is volatility on Binance, say when the price changes by 1% or more, the most sophisticated generators always win the bid for the next block because they are willing to pay more to get the most profitable blocks. The top position of the graph block. This creates a cycle that goes back and forth. Complex generators pay more when there is a MEV chance at the top of a block, meaning they win more of these plump blocks. This makes them more likely to get private order flow from seekers, and with more private order flow, they can bid higher to get their blocks included. Or, as SMGs Max puts it, Winners are better at winning when theyre motivated.
It’s a delayed war that does little good for society. Unfortunately, this is almost impossible to prevent. The early bird catches the worm, and the most sophisticated players will always arrive first. There has been a lot of research devoted to solving this problem, and the current consensus solution is to allow auctions of partial blocks. We believe these research efforts are noble but misguided. The ultimate goal of DeFi is not to act as an appendage to TradFi. It’s hard to imagine now, but the whole point is that price discovery shouldn’t happen on centralized exchanges.
Of course, there will always be market-impacting information from the off-chain world, and being the first trader in an on-chain transaction brings certain advantages. The most obvious example is when the Federal Reserve announces interest rate changes. It is never on-chain first. Going forward, MEV in the future will not look like MEV extracted today. MEV solutions should be universal, not tailored to the current problem. After all, if Ethereum succeeds, wouldnt Ethereum become the place for ETH price discovery?
MEV rewards are like “whack-a-mole”
It is very difficult to understand the full picture of MEV because it can be called a dark forest. However, EigenPhi has some of the best MEV visualizations when it comes to data. The figure below shows the profits earned by each participant in the supply chain. You can see that the relayers profit is $0.
The most interesting finding is that EigenPhi noticed that a significant portion of validators’ MEV rewards are generated outside of MEV-Boost. This means that some validators are already developing MEV supply chains separate from MEV-Boost. This would be fine if these validators were all independent stake holders, however, it is more likely that large stake pools have started running their own internal MEV strategies. If true, this could trigger an upward spiral, with higher profits leading to greater returns, attracting more and more investors, thereby enabling stake pools to afford higher private order flow.
ePBS still has shortcomings
Just as MEV can extract opportunities from transactions sent to the public mempool, MEV opportunities can also be preempted by different entities in the MEV supply chain. For example, a searcher discovers an MEV opportunity and sends it to the builder in exchange for a portion of the proceeds. The builder then packages this valuable transaction into the entire block, paying the winning validator to propose the block. However, in the current design, there is nothing to prevent validators from directly copying the searchers transaction submission, replacing it with their own, thereby excluding both the searcher and the builder. Repeaters are the safeguard against this alpha preemption.
Since PBS is implemented outside of the protocol via MEV-Boost, there are no cryptoeconomic guarantees. There has been a lot of research into how to incorporate PBS (ePBS) into the Ethereum protocol so that third-party relayers do not need to be trusted to facilitate builder payments to block proposers. This is technically challenging and would require changes to the Ethereum protocol (but may not require consensus). Protocol Enforced Proposer Commitment (PEPC), developed by the Ethereum Foundations Barnabe, is the most mature example, but is at least 18 months away from being possible.
This design does eliminate the need for relays by creating a neutral way for builders and block proposers to transact, but it does not create an economically viable, trustworthy competitor to relayers. As Mike Neuder explained in an ETH Research Forum post, for high-value blocks such as CEX-DEX arbitrage, relays that also support cancellation functionality are still preferred.
One potential advantage of ePBS is increased censorship resistance. Many designs have an inclusion list where transactions must be part of the next block. Therefore, even if the largest builders are highly regulated entities, they cannot collude to exclude sanctioned addresses, such as OFAC.
The best solution will depend on the source of the MEV
When it comes to dealing with the MEV crisis, we have many solutions. However, these solutions are often presented as a panacea that will instantly solve all problems, such as fair sorting. We believe that any solution to the MEV extraction problem must start at the application design stage. Many outstanding application developers are already working hard to reduce the impact of MEV, but they should also be aware that there will always be some degree of value leakage. MEV cannot be completely eliminated, so developers should be more proactive in interacting directly with participants in the MEV supply chain. Paying for order flow is not inherently evil, it only causes problems when it is hidden.
