MEV-Boost Block Auctions in Ethereum PBS

• MEV-Boost block auctions are off-chain, sealed-bid protocols under the Proposer-Builder Separation paradigm that separate transaction ordering from block proposal to democratize MEV access.
• They reveal market concentration as dominant builders leverage private order flows and bid shading, resulting in reduced proposer revenue and potential centralization.
• Proposed mitigations include auction redesign, randomized order flow allocation, and on-chain enforcement to improve auction efficiency and counteract monopolistic dynamics.

MEV-Boost block auctions are off-chain, sealed-bid auction protocols that determine the allocation of Ethereum block-building rights under the Proposer-Builder Separation (PBS) paradigm. By separating the role of transaction ordering (builder) from block proposal (validator), MEV-Boost aims to democratize access to Maximal Extractable Value (MEV) while maintaining high validator revenue, fast block times, and the overall security and censorship-resistance of the Ethereum protocol. In practice, the introduction of private order flows, builder integration, and feedback dynamics has led to new forms of centralization, auction inefficiency, and incentive misalignment between network actors.

1. Auction Model and Protocol Workflow

In each 12-second slot, a validator (proposer) is randomly selected to produce a block. Specialized entities known as builders aggregate public-mempool transactions and private order flows (exclusive bundles submitted directly by searchers or via private RPC) into candidate blocks and submit sealed bids representing the share of block value they are willing to pay to the proposer. Each bid $b_t^i$ is submitted to relays, trusted intermediaries that commit to paying the bid if their builder is selected. The proposer queries all relays for their highest bids, selects the maximum, and signs the associated block header. Once the relay receives this signature, the full block is transmitted for inclusion on-chain, and the proposer earns $b_t^{i^*}$ as direct compensation. This implementation realizes a sealed-bid, first-price auction with a reserve price equal to the expected public-mempool value $r_t$ (Wang et al., 16 Oct 2024, Koegler, 22 Jun 2025, Yang et al., 2022, Wahrstätter et al., 2023).

Formally, builder $i$’s slot-$t$ valuation is $v_t^i = g(\Delta_t^i, r_t)$, where $\Delta_t^i$ is private-flow profit and $r_t$ is public-mempool value. Only the builder with the highest eligible bid (above the reserve) wins; if all bids fall below $r_t$, the proposer produces a fallback public mempool block (Wang et al., 16 Oct 2024).

Mechanically, the protocol consists of:

Step	Action	Actor
1. Block Construction	Gather public and private transactions; assemble maximal-value block	Builder
2. Bid Submission	Compute valuation, shade bid as appropriate, submit to relay	Builder
3. Relay Aggregation	Validate block, commit to bid, forward top bid to proposer	Relay
4. Selection/Signing	Proposer queries relays, selects max bid, signs header	Proposer
5. Block Delivery	Relay sends block; proposer includes it on-chain	Relay/Proposer

The builder’s utility is quasi-linear: $u_i = (v_i - b_i)$ if selected, zero otherwise (Wang et al., 16 Oct 2024, Yang et al., 2022).

2. Valuation Asymmetry, Bid Shading, and Equilibrium Dynamics

MEV-Boost auctions are characterized by strong informational asymmetry: builders with higher private order-flow share ($Z_t^i$ in period $t$) have higher $v_t^i$ and thus systematically outbid competitors. In equilibrium, stronger builders (with more private flow) can shade bids more aggressively: for all $v$, $b_i(v) < b_j(v)$ (weaker builders), yet $i$’s win probability exceeds $j$'s for every $v$ (Wang et al., 16 Oct 2024).

The theoretical auction solution yields a pair of coupled, Maskin–Riley ODEs:

$$\frac{f_i(v_i)}{F_i(v_i)}\,\frac{d v_i}{d b_i} = \frac{1}{v_j(b_i) - b_i}$$

with analogous equations for competing builders.

Empirically, the top three builders commonly bid 26.9% below observed block value, while smaller builders bid at nearly 100% of value. Private flow’s win-probability effect is substantial: logistic regression of win events on $Z_t^i$ shows a coefficient $\beta \approx 4.2$ ($p < 10^{-5}$). Builders with more private flows retain up to 27.7% profit margins, while mid-tier builders capture only 8–13% (Wang et al., 16 Oct 2024).

3. Centralization, Market Concentration, and Monopoly Formation

Feedback between private order-flow advantage and block-winning probability creates a self-reinforcing loop: winning more auctions attracts further exclusivity agreements with searchers, yielding higher $Z_t^i$ and further increasing the odds of victory. Modeling $Z_n$ as a stochastic-approximation process, the mean-field has stable fixed points at $Z = 0$ and $Z = 1$, corresponding to total loss or capture of private order flow. Simulations confirm convergence to monopoly, even from modest initial advantage: on-chain Herfindahl–Hirschman Index (HHI) values drifted from ~0.2 in late 2023 to ~0.35 by May 2024, empirically confirming rapid builder centralization (Wang et al., 16 Oct 2024).

This monopolization dynamic is further compounded by differentiated order-flow sources and exclusive provider arrangements, reinforcing the “chicken-and-egg” cycle in which only high-market-share builders get access to exclusive/signaled flow and only those with such access can consistently win auctions (Öz et al., 18 Jul 2024).

4. Auction Efficiency, Proposer Revenue, and Fairness

In symmetric (independent-private-values) first-price auctions, proposers attain maximal expected revenue. However, with value asymmetry and bid shading by dominant builders, proposer (validator) revenue is strictly reduced:

• As HHI increased from 0.23 to 0.35, top builder profit margin increased from 5.4% to 27.7%, and the proposer’s share of MEV declined.
• Over 1 million slots, top three builders win >95% of auctions. Private order flows, though only 12% of transactions, contribute 54.59% of block value. Large builders capture 60–70% of private value; small builders, near zero (Wang et al., 16 Oct 2024).

Auction inefficiency is endemic under real-world conditions. For example, only 79.74% of auctions are fully efficient (winner had highest true value), and 0.98% of potential proposer revenue is lost in uncompetitive (CI < 0) slots due to bid-shading or block subsidization. Competitiveness and efficiency metrics degrade further at higher MEV levels, reflecting strategic behavior by dominant builders (Yang et al., 2 May 2024).

5. Latency, Integration, and Builder Type Advantages

Beyond private order-flow, integrated builders possess two powerful structural advantages (Pai et al., 2023):

1. Truthful Bidding via Deferred Profit Extraction: Integrated builders running both search and build infrastructure perform internal bundle merges and submit bids close to their full block value. For them, the auction outcome approaches a second-price mechanism, whereas neutral builders face first-price auctions and shade their bids, depressing equilibrium payoffs for non-integrated participants.
2. Latency and Winner’s Curse: Under stochastic price evolution (e.g., real-time CEX/DEX arbitrages), faster builders can wait for fresher value signals, while slower builders bid on stale information and are subject to “winner’s curse,” forcing them to shade bids toward zero as fast revision probability increases.

Simulations and empirical data corroborate that integration grows builder surplus multiplicatively in the number of integrated participants, and correlates with block-winning rates especially in periods of high volatility (Pai et al., 2023).

Latency confers only a marginal edge compared to order-flow exclusivity. Simulated win-rate gain from a 10ms latency advantage is ~0.26 percentage points, whereas EOF access in the range 0.8→0.98 yields exponential win-rate and profit escalation (Wu et al., 2023, Wu et al., 24 Dec 2024).

6. Market Outcomes, Empirics, and Quantitative Benchmarks

Empirical studies of block auctions from January 2023–May 2024 confirm persistent concentration and inefficiency:

• Three builders account for >90% of blocks; the top-2 control >85% on many days (Yang et al., 2 May 2024).
• Private providers (MEV-Share, MEV Blocker, major searchers) can sway >50% of auctions (Yang et al., 2 May 2024), and high builder profit margins correlate strongly with exclusive-signal share (Spearman ρ ≈ 0.62, $p < 0.01$) (Öz et al., 18 Jul 2024).
• Entry requires block subsidization: to clear access thresholds for a major private order-flow channel, new builders must pre-fund dozens of unprofitable blocks per day for a week, with subsidies increasing over time (Yang et al., 2 May 2024).
• In slots with high MEV, auction efficiency and competitiveness degrade. At high order-flow asymmetry, the equilibrium HHI exceeds 2000, indicating oligopoly, and aggressive bidding rates drop in the dominant group (Wu et al., 24 Dec 2024).

7. Mitigations, Protocol Evolution, and Research Directions

Proposed mitigations span order-flow redistribution, auction redesign, protocol randomization, and transparency enforcement (Wang et al., 16 Oct 2024):

• Redistribution: Mechanisms to rotate or randomize private order-flow providers across builders.
• Auction Format Changes: Hybrid first-/second-price or VCG-side-payment schemes, multi-winner auctions, and periodic “public slots” to force baseline competition.
• Protocol Modifications: Random assignment of bundles, on-chain bid distribution audits, anti-collusion rules, and minimum-fare enforcement.
• Builder Market Reforms: Developments such as MEV-Share, MEVBlocker, or committee-based builder selection to democratize order-flow access (Wu et al., 24 Dec 2024).
• Enshrined PBS (ePBS): On-chain bid/payment settlement, builder stake requirements, and committee-based timeliness/reveal enforcement to formally penalize misbehavior and limit market power (Koegler, 22 Jun 2025).
• Order-Flow Confidentiality and Fair Settlement: For robust decentralization, auctions must guarantee trust-minimizing confidentiality (e.g., cryptographic commitment schemes, TEE-based enclave execution) and verifiable leakage detection with on-chain enforcement, to remove the need for permissioned, reputation-gated private channels (Yang et al., 2 May 2024).

Only by directly addressing both private order-flow asymmetry and the synergistic feedback loops between builder market share and exclusivity can future block auction protocols avoid monopolization, proposer revenue attrition, and systemic censorship risk (Wang et al., 16 Oct 2024, Öz et al., 18 Jul 2024).

---

Key papers referenced:

• "Private Order Flows and Builder Bidding Dynamics: The Road to Monopoly in Ethereum's Block Building Market" (Wang et al., 16 Oct 2024)
• "Structural Advantages for Integrated Builders in MEV-Boost" (Pai et al., 2023)
• "From Competition to Centralization: The Oligopoly in Ethereum Block Building Auctions" (Wu et al., 24 Dec 2024)
• "Who Wins Ethereum Block Building Auctions and Why?" (Öz et al., 18 Jul 2024)
• "Decentralization of Ethereum's Builder Market" (Yang et al., 2 May 2024)
• "SoK: Current State of Ethereum's Enshrined Proposer Builder Separation" (Koegler, 22 Jun 2025)
• "Strategic Bidding Wars in On-chain Auctions" (Wu et al., 2023)