How Shared Sequencers and Proposer/Builder Separation are Changing the Game in a Modular World

Introduction

When Ethereum was just starting out, Maximal Extractable Value was a bit of a fringe concern. It was a technical term meant to define the different ways that miners could reorder, add, or remove transactions in order to generate a profit in their mining operations. Today, MEV has become one of the foremost challenges in blockchain economics. But the tides are shifting again: modular blockchains are gaining traction, and MEV extraction and mitigation are unfolding in ways that simply aren’t predictable.

This article aims to present the ways that shared sequencers and Proposer/Builder Separation are changing the risk, benefits, and incentives of MEV. It is intended for crypto developers, protocol designers, and researchers who care about how changes in modular architecture change what we can consider fair, decentralized, and economically aligned in Web3 infrastructure.

Basics of MEV

At a fundamental level, max extractable value (MEV) means the maximum extractable value based on the ordering and inclusion of transactions into a block. For instance, in the case of monolithic blockchain systems like Ethereum pre-merge (before they were split into layers), miners were left to sequence all the transactions into blocks, which led to the rise of techniques such as frontrunning, sandwich attacks and arbitrage. In general, these techniques would lead to worse and worse user experience, while creating "dark forest" conditions where only the more sophisticated bots could make any gains.

Flashbots came along in 2020 and created off-chain MEV relays and auction mechanisms that sought to bring some level of openness and reduce risk. For the first time, builders would be competing against each other to bundle transactions in the best possible way, while validators could freely choose the most profitable bundles independent of builders, as builders could not even modify the bundles. But this MEV auction paradigm, all working in the context of Ethereum, required the existence of a single consensus layer that was responsible for contents of execution and block proposals.

With the emergence of modular blockchains, which split those duties across multiple layers and entities, the question becomes: Who controls MEV?

A New Economic Map with Modular Blockchains

In several architectures, consensus, execution, and availability are split across modules run by different people groups. As an example, rollups would use data availability layers such as Avail or Celestia; shared sequencers or validators would perform transaction ordering for rollups (and for many other things).

This fragmentation allows greater specialization, but also complication. Sequencers, builders, and proposers are all separate entities with some share of agency in the transaction ordering process. This means the MEV (I.e. maximum extractable value) occurs in a much more dispersed way, yet becomes much more complicated.

In a monolithic chain, MEV was whoever ordered the transactions swapping their own preference/s (potentially) at the expense of others. In a modular system, it is instead, who makes sure ordering occurs across domains. That is a far harder problem, and it is already causing new infrastructure and novel threats.

Shared Sequencers

A shared sequencer acts as a neutral ordering layer between multiple rollups with regard to transaction sequence, even ensuring atomicity between these rollups. Espresso, Astria, and Radius are leveraging this approach to reduce fragmentation, and mitigate front-running across rollups.

However, shared sequencers do not mitigate MEV altogether, they just distribute it. Being in the middle of different execution layers, shared sequencers can see cross-chain opportunities (e.g. arbitrages between rollups) that individual sequencers for each rollup would not be able to pick up. Hence shared sequencers can be quite potent, and in some cases, too potent.

Practical challenge: When a shared sequencer becomes the primary way to order dozens of rollups, it becomes an MEV clearinghouse. Centralization also opens up potential for attackers to gain additional ways in: e.g. collusion by validators, or prioritizing builders who pay more.

From the user's perspective, shared sequencer congestion can look like unsettled confirmation times or that cross-rollup swaps only complete as expected when the other rollups are not also competing for confirmation/ordering. This bears resemblance to the "gas wars" during the early days of Ethereum, except now we see them in modular ecosystems.

PBS: Separation of Proposer and Builder

The Proposer/Builder Separation (PBS) was a new concept in Ethereum to separate the block building process from the block proposing process; builders group transactions together and attempt to build the block that they believe will be the most valuable; proposers (validators) simply pick the highest price.

New applications of PBS have already begun to be invented in modular settings; for example, builders are now working across multiple rollups or domains, creating bundles that span multiple sequencers, or even multiple data availability layers. Protocols like Flashbots SUAVE (Single Unified Auction for Value Expression) are perfect examples of getting started with modular PBS through cross-domain MEV auctions, which will allow builders to simultaneously bid on opportunities across multiple networks.

The clear benefits of PBS are:

• Less manipulation by validators: Validators cannot reorder transactions to get richer.
• More efficiency: Specialized builders compete on their capabilities and optimizations.
• Transparency: MEV flows can be seen, and bid on, by anyone.

But, it's not without new risks. Builders may coordinate across chains to form cartels, which could capture the MEV market. Cross-domain PBS may also tilt the scales toward large firms with more data and better infrastructure creating a different sort of MEV centralization from miners to large institutional builders.

The New MEV Meta: Changing the Game

The combination of shared sequencers and PBS alters the structure of MEV. While sequencers provide the structure for the ordering, the builders are the ones that truly improve the economy. Together, they create a cross-ecosystem MEV coordinating layer.

Let's illustrate the scenario:

The builder discovers there is an arbitrage opportunity between two rollups, say a DEX on Arbitrum and one on another DEX on Optimism. They create a cross-rollup bundle to send to the sequencer, to send the transaction, which anyone can use. But the sequencer is managing multiple builders and simultaneous competing agendas. Latency or congestion may cause the arbitrage to fail, or worse, another builder with better access to the sequencing could extract from that opportunity instead.

This interaction can reveal a new kind of MEV, which we might call cross-domain latency arbitrage. The builder who is able to coordinate the fastest, by making decisions across layers (modular), will win, but that isn't necessarily the builder that provides the most utility to the users.

As PBS and shared sequencers gain traction, scaling, we should expect to see the functional emergence of meta-builders or firms acting as coordinators to bundle the best transaction across multiple PBS and sequencing markets. This could make modular MEV as complex as global high-frequency trading, with the same types of risks to liquidity, latency, and collusion that impact the entire ecosystem.

Problems to Face

Modern protocol designers are trying to achieve three often competing goals. The first goal is fairness, which looks like preventing MEV cartels and providing a fair shot to everyone to place an order. It also means that we should not concentrate too much power on sequencers or builders. The second goal is efficiency, which means keeping the throughput and latency low across all chains.

It can be difficult to strike a balance between all three goals. Shared sequencers can help maintain consistency, but when too much power is located in a central authority it risks centralization. PBS may constrain manipulation by validators, but it doesn't stay the power of builder monopolies. Some projects are experimenting with encrypted mempools, e.g. SUAVE's 'intent-based' architecture, and zero-knowledge order flow, which delivers secrecy of transaction information until the order is set. Others propose that multi-party sequencing or threshold encryption would help spread the decision making to mitigate MEV.

The questions will remain about DEXs or lending applications on the shared sequencing delay if latency is an important application. How do builders and sequencers avoid becoming economic stakeholders and establishing secret cartels? What are the necessary governance structures to want representative councils to ensure that modular MEV auctions are fair? These answers would be critical to the future of MEV and potentially modular blockchain economics as a whole.

Conclusions

MEV will be a thing of the past in modular blockchains instead it will change and evolve. Because of shared sequencers or PBS the way value moves across decentralized systems is changing. Miners no longer have the ability to "be ahead" of consumers. We now have builders, sequencers and data layers negotiating value in a modular ecosystem.

The changes will make us more efficient and more open to new ideas, but also burdens we had not pictured coming. Essentially, we sit today at a crossroads for what the industry will look like: Will the modular MEV infrastructure serve to decentralize new ideas and leverage open and competitive coordination or simply create central throttles that operate as neutral?

R&D shouldn’t solely think about how people get an MEV fast but how they build and develop the market itself: a fair coordination game, aligned incentives to problem solving, and verification of trust.

MEV is not going away and is not a bug or a side effect in the modular era it will become the economic heart of the system. How we build and develop the framework around it will be the next generation of decentralized networks.

This article is contributed by an external writer: Razel Jade Hijastro.

Disclaimer: The content created by LBank Creators represents their personal perspectives. LBank does not endorse any content on this page. Readers should do their own research before taking any actions related to the company and carry full responsibility for their decisions, nor can this article be considered as investment advice.