Ethereum MEV Dark Forest: From Gas War to PBS

Domothy and I co-authored this article. PBS (Block Proposer-Builder Separation) is still an active area of ​​research, but this comprehensive article aims to summarize the research progress so far and where research is going.

In the existing architecture, block proposers (formerly miners) select transactions from the storage pool (mempool) to build blocks. On Ethereum, these consist of validators, who run both the consensus client and the execution client. In theory, block builders will select transactions that pay the highest fees - but this is not a strict requirement. Because a block builder has full control over tx ordering and inclusion, block builders can run complex MEV strategies. MEV is extremely common wherever there is activity, and as a result, Ethereum is the most profitable MEV battlefield, with over $675 million extracted to date.

Data source: Token Terminal

01. M E V   Quick Start

Technically, an Ethereum transaction is just a series of bytes. Therefore, the costs involved in creating and sending it are negligible. A good example of this is on-chain arbitrage. When the same asset is priced differently between exchanges, it creates opportunities for value extraction. Any money made from this transaction, then minus the gas fee, is essentially profit extracted from economic activity on the blockchain.

However, there is a problem: due to the transparent nature of the blockchain, anyone with a node can identify and submit the same transaction. No matter how many people send their transactions, there will only be one winner: the person whose submitted transaction is confirmed. As you might imagine, this greatly benefits the person who decides which transactions go into the next block, and in what order, and they can simply take advantage of them!

This example illustrates what is known as a “MEV opportunity.” Given the list of pending transactions waiting to be confirmed, and the current global state of Ethereum, there is a specific order of transactions that allows block builders to extract the most value from the ecosystem. Most of the time, MEV opportunities are more complex and require multiple transactions in the correct order, but the principle is the same: block builders have a huge advantage.

02. MEV development path

A few years ago, when Ethereum was still protected by Proof of Work (PoW), the concept of MEV was mostly theoretical. Instead, transactions were simple: “Here’s what I want to do, and here’s how much I’m willing to pay the miners to include it in the next block.”

Miners will simply look through the list of pending transactions, sort by highest fee, and fit as many of them as possible into a block. The complexity of the transactions themselves does not matter. This satisfies some notion of fairness: every miner will run the same code, follow the same rules, and make the same amount of money (proportional to hashing power).

But in short, this can’t last. Financial incentives breed competition, and the MEV transaction that provides the most inclusion wins, because miners are also financially incentivized to choose the most profitable transactions. In these cases, miners are the ones who profit the most, and they don’t even have to play any games. If there is enough competition, profits will always tend to miners (or validators, in the case of PoS Ethereum).

Short term: Gas war

The first consequence became apparent quickly. Every MEV transaction, whether successful or not, was still included in the attempt on the chain. This wasted a lot of Ethereum’s already limited block space and led to incredibly high fees.

Data source: Etherscan

Medium term: Centralization of consensus

A new centralizing force emerged when miners realized the power they had as block builders, who were guaranteed to include their MEV transactions regardless of the bid.

Note: After EIP-1559, block builders still need to pay the base fee.

The new strategy is for block builders to actively monitor the chain for MEV opportunities to directly exploit their power. It’s easy to see why this is a powerful centralization vector, independent of the consensus mechanism: more frequent block creation will lead to more MEV extraction. But unfortunately, due to economies of scale, MEV breaks the guiding ideal of fairness by strengthening extraction over time.

Long term: Misaligned incentives

This centralization leads to a second-order consequence: in addition to ordering transactions in a block, block builders can also choose to add their blocks as expected or roll back the blockchain by a few blocks to effectively "steal" all MEV opportunities from recent transactions. In some extreme cases, stealing the MEV of the previous block may be a more profitable option, which can easily lead to an unstable blockchain that constantly splits into smaller forks, resulting in longer settlement times.

Thankfully, this last issue has never materialized at the base layer. However, relying on a small number of prominent actors to act in good faith against strong economic incentives is not the level of resilience that Ethereum is striving for. Over a long enough timeframe, and if left completely unchecked, MEV will pose an existential threat to Ethereum.

03. Potential Solutions

Regarding MEV, there have been many proposals to mitigate the above consequences, and it remains an open area of ​​research. Some potential solutions include:

1. Force blocks to be sorted by fee

result:

(1) Back to the gas war;

(2) Due to the lack of storage pool consensus, validators can ignore inconvenient transactions;

2. Enforce random order of included transactions

Open question: Who decides what is random?

result:

(1) Return to the gas war;

(2) Increased delay in transaction consensus;

(3) Validators can artificially keep blocks empty except for their MEV transactions, which results in a bad experience.

3. Encrypted storage pool

result:

(1) Reducing toxic MEVs;

(2) It is necessary to increase the honesty assumption of the verifier;

(3) Increased delay;

(4) Increased complexity;

(5) Increased censorship resistance;

(6) The validator’s self-contained MEV is not affected

Researchers are exploring more potential solutions, both cryptographic and cryptoeconomic. “Solving MEV” as a whole is impossible. Any obstacle will lead to more esoteric MEV extraction, exacerbating centralization. Surprisingly, the next best thing is to accept that MEV is inevitable and will always be a centralization vector whether we like it or not.

04. Blockchain construction market

The solution is by separating the process of proposing a block from the process of building it, hence the name Proposer-Builder Separation. The idea of ​​a "block proposer" already existed in Ethereum's PoS design. Every 12 seconds, a validator is elected as a block proposer. It doesn't matter how the block is built. The validator's expected job is to propose it to the rest of the network, and the job of every other validator is to validate it to make sure it doesn't break any rules.

The general idea of ​​PBS is that block builders will compete and submit bids to any validator responsible for proposing blocks. The only thing this validator needs to do is to propose the block with the highest bid.

With PBS, the highly adversarial MEV battlefield is carefully isolated to where block construction occurs. Additionally, the auction system allows all validators to easily benefit from MEV opportunities discovered by block builders, regardless of size.

Giving up and turning the whole thing into one big auction without actually trying to fix MEV itself doesn’t seem like something we want to see. However, MEV is simply too profitable to treat it as some rare edge case. If we don’t mandate such an auction directly in the protocol, it will still exist in some form, except it will be out-of-band, opaque, and subject to the centralized forces mentioned earlier. If left unchecked, it’s easy to see what private block construction auctions would look like, and all the damage they could do to the decentralization and trusted neutrality of a blockchain.

With PBS, we benefit from this inevitable economic reality instead — we can design the rules around what such a market should look like, and eliminate the parts we don’t like. We can make it fair, transparent, and easily accessible to every validator.

The ideal blockchain market

What should an ideal market look like?

1. Neutral

(1) Block proposers should not have an incentive to favor certain block builders;

(2) Block builders should have no incentive to favor certain block proposers;

2. Minimum block proposer overhead

(1) Low hardware requirements allow for more block proposers in the market;

3. Bundled security

(1) Block proposers should not be able to intercept a constructed block and replace their own transactions with MEV transactions;

4. Simplicity of consensus

(1) Blocks built by the block construction market do not require major changes to the consensus layer.

Existing Blockchain Market

Implementing PBS directly in the protocol will be complex and will take some time. In the meantime, some out-of-protocol solutions have emerged to achieve the same goals, but they have some subtle trade-offs. While not formally written into the protocol, out-of-protocol implementations allow for a higher degree of experimentation and provide an excellent testing platform for eventual implementations. Flashbots created a block construction market that includes all-or-nothing bundles and conditional payments.

Their tools include:

1. MEV-Geth

(1) Not conducive to client diversity

(2) Work only before the merge

2. MEV-Boost

(1) It is client-independent and solves the problems of previous implementations.

(2) Validators cannot manipulate blocks (they only receive the block body after signing the block header)

But the off-protocol market still has some disadvantages:

(1) Relays still need to be trusted

(2) Review issues

Image credit: Jon Charbonneau and Delphi Digital

05. Remaining Questions

Appropriate PBS design remains an active area of ​​research. We have guiding principles, but ultimately implementation is not easy.

1. Centralization of Block Builders

Block builder centralization is less harmful than block proposer centralization, but still not ideal. Therefore, we are conducting research to eventually create a decentralized network of block builders, which we hope can outperform centralized entities.

2. Review issues

Implementing crLists still allows censorship, but in a competitive market, long-term transaction censorship becomes too expensive and requires block builders to fill blocks. If block builders do not fill blocks, space is filled with transactions selected by block proposers. This added defense greatly reduces the existential threat posed by extreme block builder centralization. Encrypted storage pools can also solve the censorship problem and toxic MEV at the same time, but require upgrades to the underlying protocol.

3. High MEV   difference

Even if the returns are democratized through PBS, the luck of the block proposer is still an incentive to pool staked ETH. The more validators an entity has, the more likely it is to propose blocks and receive MEV rewards. Implementing MEV smoothing in the core protocol allows us to reach an ideal fairness scenario where staking rewards are proportional to the amount staked, as MEV rewards will be aggregated and paid out equally to validators. Additionally, if in-protocol MEV smoothing is implemented in an optimal way, it will reduce the incentive to circumvent the in-protocol PBS design.

4. Formal Specifications

There are many important research initiatives in the PBS space, but a notable lack of formal analysis of incentives or economics. Furthermore, the research space is changing rapidly, and there will inevitably be more developments before we reach the analysis and implementation stage.

06. Conclusion

A lot has happened in MEV and protocol research over the past few years, and the more we learn about it, the broader the topic becomes. Although we still have unanswered questions, we are on the path to a decentralized, highly resilient, neutral global network. If you want to go deeper, Domothy has compiled a list of further reading here.