---
title: Enshrined Proposer–Builder Separation (ePBS)
url: https://www.emergentmind.com/topics/enshrined-proposer-builder-separation-epbs-2b9c081a-47b2-4429-a22d-561e281a77cf
type: topic
---

# Enshrined Proposer–Builder Separation (ePBS)

Enshrined Proposer–Builder Separation (ePBS) is the protocol-level formalization of the Proposer–Builder Separation paradigm in Proof-of-Stake blockchains, exemplified by Ethereum’s adoption in the Glamsterdam upgrade (EIP-7732). ePBS integrates market-based block production directly into consensus, replacing reliance on off-chain relays with on-chain mechanisms for blockspace auction and execution payload delivery. ePBS aims to enhance decentralization, mitigate Miner/Maximal Extractable Value (MEV), and lower validator operational complexity, but introduces new challenges including risks of builder centralization, liveness failures, and auction-induced economic distortions.

## 1. Architectural Principles and Economic Motivation

Proposer–Builder Separation structurally divides block production between proposers (validators, who select and publish new blocks) and builders (specialized entities that aggregate, order, and optimize transactions for MEV extraction). In ePBS, this separation is enshrined at the protocol level: the proposer outsources payload construction to builders via a market auction; builders submit bids for their constructed payloads, and the highest bidder wins the right to publish its payload.

Formalizing this in the context of staking and block rewards, each block producer (BP) possesses a reward multiplier $μ_i$ reflecting technical skill, information advantage, or access to exclusive order flow [2401.12120]. The expected reward per block is:

$$
U_i(\pi_i; \pi_{-i}) = μ_i \cdot r \cdot x_i(\pi) - c \cdot \pi_i,
$$

where $π_i$ denotes the stake, $x_i(π)$ is the probability the BP is chosen, and $r$ is the base block reward. Competitive equilibrium is captured by:

$$
\max \sum_i μ_i \frac{r}{c} x_i \left(1 - \frac{1}{2} x_i\right),
\quad \text{subject to} \ \sum_i x_i = 1, \ x_i \geq 0.
$$

Without separation, heterogeneity in $μ_i$ provokes oligopolistic stake concentration, threatening decentralization and protocol security [2401.12120].

## 2. PBS Markets, Relays, and Builder Centralization

In the current market, MEV-Boost (Flashbots) connects proposers to builders via relays. This enables over 90% of proposers to participate, but creates vulnerability to relay centralization, collusion, and censorship [2305.19037, 2506.18189]. Builders optimize transaction order, often exploiting MEV through arbitrage and sandwich attacks. A heavy-tailed reward distribution emerges: ~20% of high-value MEV transactions contribute 72% of total revenue [2305.16468].

Relay dependence introduces implicit trust assumptions—proposers commit to block headers blind to contents, relying on relays for promised value delivery, and raising risks of faulty or censored payloads [2305.19037, 2305.16468]. The model is sensitive to vertical integration; combined builder/relay entities enhance revenue but can exacerbate market concentration and conflicts of interest.

## 3. Emergence of Private Order Flow Auctions and Monopoly Dynamics

Builders increasingly extract value from private order flows: select bundles dispatched directly (often through private RPCs) rather than public mempools. These flows, though comprising only ~12% of transactions, represent 54.59% of block value [2410.12352]. Builder valuation thus becomes:

$$
v_t^k = g(\Delta_t^k, r_t),
$$

where
$\Delta_t^k = N_t Z_t^k w_t$
($N_t$: number, $Z_t^k$: builder’s share, $w_t$: average profit per order).

Auction-theoretic analysis reveals that builders with richer private flow bid less aggressively yet win disproportionately, producing increasingly monopolistic outcomes. Lower bid ratios for dominant builders (26.87% below median) correspond to >95% auction win rates [2410.12352]. The positive feedback (winning attracts more private flow, further increasing valuation and win rate) drives market concentration and diminishes MEV revenue for proposers.

OFA (Order Flow Auction) modeling shows the Nash equilibrium bids by builders with asymmetric MEV capacities, solved via quartic equations [2502.12026]. The equilibrium exacerbates effective centralization—competitive advantages translate into disproportionately discounted bids and amplified profits for dominant builders.

## 4. On-chain Enshrinement, Economic Smoothing, and Burn Mechanisms

ePBS seeks to internalize builder-proposer auctions in consensus, eliminating off-chain intermediaries. Proposals include on-chain builder staking, MEV burn auctions, payload timeliness committees, and equitable reward distribution schemes [2506.18189]. The base reward formula (for proposers/validators) is

$$
\text{Effective Balance} \times (\text{Base Reward Factor}) / (\text{Base Rewards per Epoch} \times \sqrt{\sum \text{Active Balance}}).
$$

MEV burn proposals require builders to burn a fraction of their bid:

$$
\text{Base Burn} < \text{Effective Balance} - \text{Payload Tip}.
$$

MEV smoothing (redistributing extreme rewards to attestation committees in proportion to typical protocol reward shares, 22–41%) reduces reward variance and improves equity across validators.

## 5. Future-block Auctions and Game-theoretic Mechanisms

The “Flashback” model extends ePBS by including future block proposers in auction design [2405.09465]. Builders may “reserve” high-value bundles for future proposers determined by epoch scheduling. Mathematical analysis defines the reserve threshold $ρ$ and rate $r_1$ for reserved bids:

$$
r_1 = \frac{R[Q_t^{1:k}] - (1+\epsilon)(1-r_2)\cdot \bar{E}[V[t+1]]}{R[Q_t^{1:k}]},
$$

where $R[Q_t^{1:k}]$ is the sum of bid fees, $r_2$ is default fee share, and $\bar{E}[V[t+1]]$ is expected proposer return.

Experimental evaluation shows this mechanism increases primary builder rewards by ≈20% and block award rate to 55%, with beneficial impacts on user experience and proposer revenue [2405.09465].

## 6. Centralization Risks in Advanced Auction Designs

Execution Tickets and ahead-of-time Execution Auctions present liveness and fairness hazards [2408.03116]. The ex-ante strongest builder, by winning future block rights, consistently sets reserve prices above realized values, always winning the primary auction and further dominating in JIT resale auctions:

$$
v_2^* - \frac{1 - F_2(v_2^*)}{f_2(v_2^*)} = v_1,
$$

where $F_2$, $f_2$ are CDF and PDF of the competitor’s value. This advantage trivially extends to imperfect resale markets, magnifying market concentration even absent multi-block MEV extraction.

## 7. Liveness, Free Option Problem, and Economic Trade-offs

Enshrined separation (as in EIP-7732/Glamsterdam) grants builders a short-dated “free option” to drop payloads at no cost, by withholding execution during the option window [2509.24849]. With theoretical model

$$
\Pi_{\tau}(y) = \mu + (1 + r_{\tau})y - P_{DEX}\left(\frac{y}{P_0}\right),
$$

the builder maximizes

$$
V^* = \max_{y} E[\max\{0, \Pi_{\tau}(y)\}],
$$

with exercise probability

$$
P^* = Pr[\Pi_{\tau}(y) < 0].
$$

Builders whose block value is highly sensitive to DEX-CEX arbitrage are more likely to exercise the free option, especially in volatile periods (average exercise: 0.82% of blocks, spikes up to 6% in high volatility; individual builders see rates up to 23%). This creates liveness risk—empty blocks, delayed execution, and DEX mispricing—all magnified by market volatility or elongated option windows.

Mitigation is possible by penalizing exercised options or shortening the option window. Static penalties as small as 0.075–0.15 ETH suppress exercise rates by 75–83%, while dynamic approaches employ online optimization to balance builder participation and liveness risk.

## 8. Protocol Design Implications and Directions

ePBS yields robust decentralization in the proposer/validator layer (stake shares evolve as a martingale, preserving initial distribution [2502.12026]), but remains fragile to builder centralization due to auction dynamics, information asymmetry, and order flow concentration [2410.12352, 2305.19150]. The true decentralizing effect relies critically on competitive builder markets; otherwise, risk simply migrates from proposers to builders [2401.12120].

Protocol designs must intertwine auction engineering, penalty mechanisms, and dynamic fee models (both per-transaction and non-distortionary per-block) [2505.19708]. Active regulation of private order flows, auction unbundling (separating top-of-block and block body), and committee-driven reward smoothing [2506.18189] are central to sustainable decentralization and MEV mitigation.

## 9. Summary Table: Core Attributes and Results

| Protocol Element           | Centralization Effect                       | Mitigation/Design Lever                            |
|---------------------------|---------------------------------------------|----------------------------------------------------|
| Auction Market (Builders) | Positive feedback, monopoly risk            | Auction unbundling, fairness enshrinement          |
| Private Order Flows       | Dominant builder wins, monopoly formation   | Limit info asymmetry, promote open access          |
| MEV Extraction            | Amplifies reward disparity                  | MEV burn, committee-driven smoothing               |
| ePBS On-chain Auction     | Reduces relay colln, introduces liveness risk| Builder penalties, option window tuning            |
| Validator Layer           | Decentralized via martingale stake shares   | --                                                 |

## 10. Conclusion

Enshrined Proposer–Builder Separation introduces a protocol-integrated market for block production, eliminating relay trust assumptions and enabling enhanced MEV mitigation, fee equity, and operational simplicity for validators. However, builder centralization persists as a latent risk through auction and order flow mechanisms, particularly under private information and advanced auction designs. Liveness and empty block risk via builder free options further motivates careful incentive design. Advanced mechanism engineering—auction unbundling, dynamic penalties, reward smoothing, and competitive builder regulation—is essential to realize the promises of ePBS: sustainable decentralization, fair MEV extraction, and resilient protocol equilibrium.

Source: https://www.emergentmind.com/topics/enshrined-proposer-builder-separation-epbs-2b9c081a-47b2-4429-a22d-561e281a77cf