What has changed in the MEV supply chain on the first anniversary of the Ethereum merger?

To celebrate the first anniversary of the Ethereum merger and the launch of MEV-Boost 5, we are bringing you insights and trends from the first year of PBS outside the Ethereum protocol. We will first look at MEV-Boost adoption by proposers, and then summarize the behavior of relays, block builders, and searchers.

1. Rapid adoption of MEV-Boost

The Ethereum merge brought significant changes to Ethereum, as the responsibility for block production shifted from a handful of mining pools to thousands of stakers around the world. This change also resulted in a “hard reset” of Ethereum’s MEV infrastructure. With this new group of proposers taking control, all previous relationships between searchers and miners disappeared.

A year before the merger, Vitalik introduced the Proposer/Builder Separation (PBS) concept as a direct response to the risks that MEV could pose to PoS Ethereum. If left unattended, MEV will be accumulated by proposers who have the most complex infrastructure and are associated with searchers. This will discourage individual staking, promote economies of scale, and threaten the network's reliable neutrality.

Flashbots, in collaboration with the Ethereum Foundation and the Eth2 Working Group, developed MEV-Boost as a temporary PBS solution that does not require any changes to the core protocol. MEV-Boost, as a secondary tool for the grand merge of all CL clients, implements out-of-protocol PBS and establishes an open, competitive market for block builders. Through this separation, MEV-Boost democratizes access to MEV, allowing all validators to participate equally. This is especially important for individual validators, who would otherwise be unable to participate and be blocked from receiving any MEV benefits.

After the Ethereum merger, MEV-Boost was adopted quickly, with 50% of validators running it in the first month, and by the end of the second month, this number reached 90%. Proposers who outsourced block construction through MEV-Boost received the same median block reward, thus reducing the centralization pressure caused by MEV on the Ethereum validator set.

Proposer benefits and major events

In the first year after Ethereum PBS, validators received more than 300,000 ETH from the blocks they proposed, including gas priority fees and MEV. Although gas priority fees usually account for the majority of proposers' earnings, in periods of high market volatility, MEV fees may significantly exceed gas fees.

There were three periods last year where large MEV fees occurred. We will briefly touch on each of them here.

• November 9, 2022: FTX crashed, followed by a five-day market liquidation that resulted in a total of 14,585 ETH in gas priority fees and MEV being paid to proposers.

• March 11, 2023: The SVB bank run and USDC decoupling triggered a surge in on-chain activity, resulting in the proposer receiving a total of 7,694 ETH.

• July 30, 2023: Following the Vyper hack, arbitrageurs raced to eliminate the price difference between depegged assets in various Curve pools. A total of 7,187 ETH was paid to proposers. During the turmoil, one transaction paid 570 ETH to ensure inclusion, eliminating the arbitrage opportunity between Alchemix ETH and Frax ETH, the second largest MEV fee in Ethereum history to date.

The transaction with the highest single MEV fee ever was 678 ETH paid by whitehat c0ffeebabe.eth to Beaverbuild, which received this huge reward because Beaverbuild was the first to discover a vulnerability in the contract and avoided the funds being stolen by hackers. Beaverbuild transferred this MEV to the block proposer to win the bid.

2. Relay Ecosystem

MEV-Boost introduces the role of a trusted relay between builders and proposers, responsible for verifying and hosting blocks. Anyone can run a relay node, and proposers can register as many relay nodes they trust as possible. In order to help Ethereum merge to launch a diverse and healthy relay market, Flashbots opened our relay under the LGPL copy-left license a month in advance. Blocknative followed closely and also made an open source deployment.

With Ethereum’s transition to Proof of Stake on September 15th, validators can connect to 7 different relays, which in turn receive bids from 27 unique block builders. To avoid enshrine flashbots relays, we intentionally do not set it as the default in MEV-Boost.

(1) Review issues

The rapid adoption of MEV-Boost continues to outpace our efforts to guide a diverse relay market. A month before the merger, OFAC added the Tornado Cash contract address to the sanctions list. This caused many relayers, including Flashbots, to choose not to receive blocks containing these transactions. The way MEV-Boost is architected, there is a complete separation between block builders and proposers. Validators must blindly sign the blocks they propose without being able to add any further transactions and bypass this censorship.

The Flashbots relay is the only permissionless relay in the Ethereum merge, which allows any builder to submit a bid. The second permissionless relay, and the first censorship-free relay, began running on October 26 of last year. Even so, Flashbots remained the most popular relay in the first few months, reaching a peak on November 11, forwarding 69% of all new Ethereum blocks. On November 21, the proportion of blocks from censorship relays reached an all-time high, accounting for 79% of all new Ethereum blocks.

On November 22, we published the article "The Cost of Resilience". The article encourages: If the bid for MEV-Boost is lower than this minimum bid value, validators should use the newly added -min-bid flag to propose locally built blocks. By using this feature, validators will no longer risk missing high-value MEV-Boost blocks, while still being able to maintain Ethereum's network resilience by proposing locally built blocks that are not related to content (containing transactions) when MEV is low.

On November 30th, two new permissionless, content-agnostic relays were announced - Agnostic Relay and ultra sound relay. To support their initial growth, Flashbots developers began sending blocks to them the following week. In addition, we released a technical guide and knowledge base on how to run MEV-Boost-Relay at scale to help everyone effectively operate their own relay infrastructure. Aestus began operations in December last year, becoming the fourth content-agnostic, permissionless relay, bringing the total number of relays to 11.

(2) Diversification and stability of the relay market

While Flashbots remains by far the most popular relay for validators, the slot sharing of this relay has diversified considerably over the past 6 months. In April, the slot sharing levels of Agnostic relay and ultra sound relay grew to the same level as Flashbots relay. As of now, ultra sound relay and BloXroute are in the lead, and while the top 6 relays have a combined 90% market share, no single relay has a market share of more than 25%.

The number of blocks coming from censored relays hit a low of 17% in April. Since then, it has shown an upward trend and is currently close to 30%. It is worth noting that after the Ethereum merge, less than 2% of blocks contain one or more transactions that interact with sanctioned addresses. Due to the scarcity of these transactions, the vast majority of blocks coming from censored entities do not actually delay any unconfirmed transactions from being included in the chain. Furthermore, since these transactions do not rely on timely inclusion, they will not fail due to added delays.

(3) Use optimistic relay to reduce latency

To reduce delays in block production, Ultra Sound Relay introduces the concept of optimistic realy. In optimistic relay mode, blocks are forwarded directly to proposers without going through relay simulation to ensure their validity. This relay process saves about 100 milliseconds and allows builders to create higher-yielding blocks, giving relays an advantage in the competition. To ensure that builders operate in good faith, they must provide collateral for the relay.

This allows builders more flexibility to adjust at the end of the slot. The chart below shows that bids submitted in optimistic mode by Beaverbuild and Titan Builder are nearly three times more likely to win than normal bids. Ultra Sound Relay enabled optimistic relaying on March 17 and remains the only relay that effectively utilizes this feature.

(4) Brief introduction to the relay attack on April 2

On April 2, a malicious proposer exploited a vulnerability in MEV-Boost-Relay and manipulated the ultra sound relay to steal approximately $20 million from multiple sandwich bots. The malicious proposer received the block body from the relay, unbound the sandwich they induced the seeker to create, and sandwiched their transactions together. The vulnerability allowed the proposer to send an invalid block header to the relay to ensure that they would win the block fuzzing competition. The vulnerability was patched that day, but in the following days, the number of forks on the network increased.

3. Evolution of the Builder Market

After the Ethereum merger, Flashbots builders initially won 95% of the blocks from MEV-Boost. But the diversification of block building happened faster than in the relay market. Just one week later, with the addition of BloXroute, the number of Flashbots builders dropped below 60%, taking 25% of the share, while other early adopters such as Blocknative and builder0x69 also began to win blocks. In November, we open-sourced our block builder and saw more builders join.

Beaverbuild entered the market in January, further decentralizing the builder landscape, with no single entity currently producing more than 30% of blocks. However, Beaverbuild's unique position as a seeker-builder (with access to exclusive high MEV arbitrage streams) gives them a significant competitive advantage. Vertical integration allows Beaverbuild to produce 40-50% of blocks in a short period of time when the market is volatile, far exceeding other builders and temporarily centralizing the market.

More competing builders have entered the market and gained significant market share, with rsync joining in March and Titan Builder joining in May. Unlike the relay market, the builder market is driven by economic incentives. The competition landscape remains very fierce, with new players joining and strategies improving in order to win the competition. As of the time of this article, the top 4 block builders account for 90% of the total market share.

The blockchain builder ecosystem has evolved into a complex game, with builders adopting a variety of strategic approaches to gain competitive advantage, such as:

• Vertical Integration of Seeker-Builders: Given Ethereum’s 12-second block time, seekers in off-chain high-frequency markets such as Binance benefit from more control over transaction finality and inclusion. Therefore, from an economic perspective, seekers can also become builders. Notably, CEX-DEX arbitrage constitutes the bulk of seeker-builder strategies and relies heavily on access to off-chain capital. On-chain data suggests that seekers generally view seeker-builders as less trustworthy than neutral builders. Some seekers strategically avoid sending data to these entities to mitigate the risk of data leakage or censorship, especially when there may be competition for the same data stream. With the emergence of multiple seeker-builders, seekers need to strategically judge which builder to send their transaction packages to.

• Strategic bidding: builder0x69 is the first notable builder who doesn’t just bid the total value of the block. They adjust their bids by subsidizing blocks with their own funds, sometimes bidding higher than the block value, and sometimes lowering their bids to make a profit. By subsidizing blocks, they have a higher chance of winning the slot, increasing their market share, and thus attracting more private order flow. This strategy is increasingly being adopted by new builders.

• Order Flow Sharing: With the rise of order flow auctions such as MEV-share and MEVBlocker, and the emergence of Telegram Bots requesting private memory pool settlements, builders are under increasing pressure to quickly include transactions and optimize the user experience. Due to the decentralized nature of the builder market, sharing order flow among builders has become a solution to speed up transaction inclusion time. This practice complicates the tracking of order flow until a more standardized solution is developed.

• Private Order Flow Access: Access to private order flow containing MEV value is very important for builders to compete for blocks. Most prominent builders have already adopted private order flow, either through their RPC endpoint products or through exclusive transactions purchased from the app.

Thomas Thiery's "Empirical Analysis of Builder Behavior Profiles (BBPs)" provides a deeper analysis of the strategies used by builders. The research analysis introduces a series of indicators to describe how builders build blocks and how they act in MEV-Boost auctions, providing insights into their strategies and optimizations. The research data shows that between June and mid-July, exclusive order flows accounted for approximately 25-35% of the total number of transactions for all builders, but accounted for 80% of the total value.

A major problem with the current PBS architecture remains builder centralization and censorship resistance. MEV-Boost does not create these centralizing forces, but MEV does, and MEV-Boost simply transfers the risk from validators to builders, where it is easier to resolve these risks.

4. Seeker Profits after Ethereum Merger

(1) Atomic MEV: Most of the profits are paid to the proposer

Using mev-inspect-py to identify Atomic MEV, such as arbitrage and liquidation, we were able to quantify that Atomic MEV’s monthly profit hovers around $2 million, surging to $4 million during periods of high market volatility.

We estimate that the total proposal revenue that benefited from atomic MEV transactions was between $1-4 million. This represents only 5-10% of the overall proposal revenue in the same time period. Please note that this is only a conservative lower bound estimate due to the limitations of mev-inspect-py's scope.

Before the Ethereum merge, searchers had to raise gas prices to prioritize transactions within a block. After the merge, with the adoption of MEV-Boost, in addition to the standard gas fee, searchers can now receive additional tips through MEV fees to prioritize their transaction packages. The percentage of MEV paid to proposers shows an interesting paradigm shift before and after the merge:

• Before the merger, the proposer’s MEV revenue share was on average less than 50%.

• Just after MEV-Boost was launched, when competition among builders had barely formed, the share of profits forwarded by searchers to proposers first fell and then rose.

• Starting in 2023, as more builders enter the field, the share forwarded to proposers stabilizes at over 50%. Proposers increasingly capture the majority of AtomMEV’s profits, sometimes as high as 80-95%.

(2) Non-atomic MEV: Higher revenue, higher profit margin

Seekers can engage in cross-market arbitrage, involving both DeFi and CeFi platforms…Tracking these profits becomes particularly challenging because transactions are partially off-chain. In the article “A Tale of Two Arbitrages,” the authors conclude:

“…Comparing Atomic Arbitrage and CeFi-DeFi Arbitrage in Q1 2023, CeFi-DeFi Arbitrage generated $37.8M in revenue in Q1 2023, while Atomic Arbitrage generated $25M in revenue. 91-99% of Atomic Arbitrage revenue was paid to validators, while only 37-77% of CeFi-DeFi Arbitrage revenue was paid to validators.”

This helps us conclude that non-atomic MEV flows ultimately contribute similarly to proposer revenue compared to atomic MEVs, while seekers can earn a higher profit margin on the total revenue.

5. Conclusion

(1) The Future of PBS

MEV-Boost was developed as a temporary solution for Ethereum PoS until in-protocol research is conducted and enshrined PBS matures and becomes part of the Ethereum roadmap. Relays were initially seen as a temporary component in the current out-of-protocol PBS structure, but will likely continue to play a role even in enshrined PBS. There is still much work to be done in relay diversification, and we are committed to encouraging the development of an open, transparent, permissionless MEV market by open sourcing our work, sharing our experiences, and actively participating in discussions.

To help support the current and future research, development, and operations of PBS on Ethereum, the MEV-Boost community jointly proposed the PBS Guild. The PBS Guild is a non-commercial ecosystem R&D funding organization that supports the R&D of the PBS ecosystem. Its goal is to leverage decentralized features to address short-, medium-, and long-term research, development, and operational challenges of PBS. The goal of the initial phase is to raise $1 million for independent relays, R&D, data transparency, and PBS education.

(2) Illuminating the future with data transparency

As the MEV supply chain evolves into a complex supply network, it becomes challenging to track the order flow lifecycle across the stack. While MEV-Boost’s open source relay data API helps understand the final steps before on-chain settlement, the initial entry point for tracking orders remains obscure.

Flashbots has always been committed to providing transparency to the MEV ecosystem and providing research data to contributors and collaborators. The recently released Mempool Dumpster provides historical memory pool transaction data from five different node providers. It aims to provide easy access to free data for any community research or construction development.

To ensure good data quality and increase protocol coverage, we invite contributors to maintain and improve mev-inspect-py through new data licenses.

(3) Looking to the future

Ethereum’s transition to proof-of-stake, coupled with the rise of MEV-Boost, has fundamentally reshaped the transaction supply chain landscape. While this achievement is huge and worth celebrating, the centralized power of MEV will continue to pose inherent challenges to the neutrality and decentralization of blockchains. The road ahead will be filled with analytical research, vigorous debate, and innovation to ensure that these centralized effects are carefully addressed. Ethereum’s development journey has always been community-driven. As we move into the next chapter, let us direct our collective energy to protect the spirit that brought us here in the first place. The future is exciting, let us rise to the challenge together.