What is Proposer / Builder Separation (PBS)?

Written by Frederik B

Reviewed by Brady Werkheiser

Published on August 4, 20226 min read

---

The blockchain emerged as a solution to decentralize the trust required to run a digital economy, which traditionally has been managed by central authorities. In a blockchain, a network of validators work together to process transactions and add new blocks to the chain. 

Block proposers effectively have control over which transactions get included in a block and can use this power to censor or otherwise manipulate transaction flow. Researchers have proposed various solutions to this censorship problem, one of which is known as Proposer/Builder Separation (PBS).

What is in-protocol Proposer/Builder Separation (PBS)?

Proposer-Builder Separation (PBS) is a potential solution to the blockchain censorship and MEV attack problem where block construction (i.e. block building) and block proposing are assigned to different roles in the network.

Block builders create "exec block bodies," which are ordered lists of transactions, and submit bids for these blocks. The proposer's job is simply to accept the exec block body with the highest bid. Proposer/builder separation isolates building, which is prone to centralization, from transaction validation which should be highly decentralized.

Unlike the traditional approach, the proposer (and everyone else) cannot see the contents of an exec block body until after they select the header that wins the auction. This "pre-confirmation privacy" is necessary to prevent the stealing of Maximal Extractable Value (MEV) by parties that could otherwise control the transaction ordering in a block.

The long-term plan for PBS is to be protocol-native in Ethereum. Flashbots, an R&D organization mitigating the negative externalities of MEV attacks, is leading the protocol standard today.

What is hybrid PBS?

In hybrid PBS, each transaction comes with a witness to prove the balance and nonce of the transaction sender. These witnesses can be added by intermediary nodes, such as builders. This extension allows for validators to be stateless, as they would no longer need to keep track of the entire transaction history. 

What is a block proposer?

In today’s Ethereum, a block proposer is an entity which creates a block of transactions, and propagates it to the network for inclusion in the blockchain. A block proposer can choose to include or exclude transactions from a block, as they look at which transactions in the Ethereum mempool pay the highest priority fee. The design of post-merge Ethereum is the same.

In PBS, the proposer instead relies on a market of outside actors (the builders) who produce bundles with complete block contents, and the fee for the proposer, with the proposer choosing the bundle with the highest fee. 

Who can be a block proposer?

In the traditional approach, any node can be a block proposer.

With the Ethereum Merge, where Ethereum became Proof-of-Stake, validators are randomly selected to be block proposers in every slot. This validator creates the new block and sends it to other nodes in the network. At the same time, a committee of validators is randomly selected, who will vote to determine the validity of the proposed block.

Under the PBS proposal, there's not yet a concrete answer as to who would become block proposers, but since the only thing the proposer needs to do is select the highest-fee bundle, it could be done inside a simple MPC (multiparty computation) to prevent cheating.

What is a block builder?

A block builder is an entity which creates an "exec block body" – a list of transactions – and submits it, along with a bid, to the Ethereum network. The block proposer then chooses the exec block body with the highest bid to include in the new block. 

In one proposed PBS solution, there are two types of builders:

1. Primary builders are block builders that can build primary exec blocks and can see the pre-state they are building on top of

2. Auxiliary builders are block builders that can build auxiliary exec blocks and don’t know what state they are building on top of

Who can be a block builder?

Since committees currently exist in Ethereum to help finalize blocks, it's likely that these same committees could act to vote for the selection of builders in each block period. It would be possible to have a group of semi-trusted builders that are voted on by the network. 

What is required to be a block builder?

Vitalik Buterin explains that proposers "are totally able to be their own block-builders," but they wouldn't have to be. Any node that can pay the fee for the proposer and run the algorithm to create an exec block body could be a builder. 

How does Two Slot PBS work?

In two-slot PBS, a beacon block is needed in every slot to include the winning exec header, which is attested by committees. If a beacon block is missing, then the next slot is switched to be a beacon block instead of an intermediate block.

How is proposer/builder separation related to MEV?

Under the current design, miners are able to capture a large portion of MEV by choosing which transactions to include in a block. This allows them to front-run other users or conduct other MEV attacks.

PBS would mitigate this problem by separating the block construction process from the block proposal process. Builders would be selected by the highest fee they're willing to pay, and proposers would simply select the exec header with the highest fee.

At this time, PBS is simply a research idea. It's not being used in Ethereum, or any other blockchain.

Alternative Solutions to MEV

There are other alternative solutions to MEV costs, including request-for-quote (RFQ) exchanges and TWAMM (Time Weighted AMM) designs.

Request-for-quote exchanges use off-chain quoting to avoid the problem of sandwich attacks where an entity will include two transactions around the victim's transaction, profiting from their slippage tolerance. With RFQ exchanges the client does not reveal their order to the market until it is ready to be executed. This way, there is no window for another entity to place their transactions around the victim's.

TWAMM also prevents sandwich attacks by executing transactions lazily. In TWAMM, orders are executed as if they were placed in between blocks. This way, an attacker would need to straddle blocks in order to take advantage of the victim's slippage tolerance, which is riskier than simply including transactions around the victim's. 

How does PBS work in relation to danksharding?

The PBS design makes it easy to implement danksharding on top of it. Danksharding is a new sharding design, which has a merged fee market, where one proposer chooses all transactions and data that goes into a slot. This forces high system requirements on validators. PBS could be used such that builders bid on the right to choose the slot's content, and the proposer simply selects the valid header with the highest bid.

How does separating builders and proposers help scalability?

PBS improves scalability by allowing for stateless validators. If all builders include a witness for each transaction, then the proposer can just select the header with the highest fee, without having to process any data. This means that validators wouldn't need to keep track of the entire blockchain history.

How does separating builders and proposers protect security and decentralization?

PBS protects security and decentralization by making it more difficult for block proposers to censor transactions or manipulate the transaction order. By isolating the block construction process from the block proposal process, PBS makes it more difficult for block proposers to control which transactions get included in a block.

Why is it likely that block proposers become centralized?

There is extractable economic value in the selection of the next block. Mining pools or other centralized entities are likely to emerge and control the block proposer role due to this economic incentive.

What are the censorship challenges of PBS?

Front-running, or when a party submits a transaction to take advantage of another party's transaction, is a major issue with the current approach to transaction ordering.

If a large enough actor is trying to censor a particular transaction, they can raise the priority fee required for that transaction to such a high amount that it becomes uneconomical for anyone else to include it in a block. The actor effectively controls which transactions get included simply by paying higher fees.

PBS needs to mitigate that risk, alongside the following censorship-resistance requirements:

Denial of service (DoS) protection

Data that no one pays for, and thus bloats up the chain, must not be included in a block. For example, if a transaction is found to be invalid due to insufficient balance or other problems, the protocol must still charge the proposer for including it in the block.

Minimal bandwidth

‍The solution should not require excessive data to be exchanged between network participants. For instance, having many redundant exec block bodies floating around the network would use up unnecessary bandwidth.

No reintroduction of centralization

‍The proposal should not incentivize proposers to become more sophisticated and thus enter into relationships with other entities or join pools.

Allowing stateless proposers‍

It would be ideal if validators could remain stateless, as this would further decentralize and scale the network. 

Proposer/Builder Separation Takeaways

PBS is a potential solution to the blockchain censorship and MEV attack problem that is currently being researched. By separating the roles of block construction and proposal, PBS could create a more decentralized, secure network. 

However, there are a number of challenges that need to be addressed before it can be implemented, such as the risk of builder centralization and the use of bandwidth. Additionally, there are other potential solutions, such as request-for-quote exchanges and TWAMM designs.

To start building on Ethereum, create a free Ethereum developer account today.