MEV-Boost: Ethereum Auction Dynamics

• MEV-Boost is an Ethereum protocol that auctions block construction rights via a just-in-time English auction, serving as the central mechanism for MEV extraction.
• The protocol differentiates between integrated and independent builders, with the former benefiting from a second-price auction mechanism while the latter face bid shading and the winner’s curse.
• Empirical findings reveal that multi-block MEV extraction is rare, with modest increases in payments for longer sequences, highlighting efficiency gains and centralization risks.

MEV-Boost is a protocol deployed on Ethereum following the merge that enables validators to auction their block-building rights to a pool of competing builders. This block-by-block open English auction serves as the principal infrastructure for extracting and distributing Miner Extractable Value (MEV) in the post-merge Ethereum ecosystem. Over 90% of Ethereum blocks are constructed via MEV-Boost, which has introduced distinct efficiency and fairness implications due to the structural differentiation between integrated and independent builders. Empirical analyses covering millions of slots since the merge have also investigated whether systematic multi-block MEV extraction exists and how payment and builder behaviors evolve under MEV-Boost dynamics.

1. Auction Architecture of MEV-Boost

In MEV-Boost, validators sell block construction rights in a per-slot auction. Builders submit bids to produce blocks, generally by aggregating profitable MEV transactions. Builders are stratified into two main types:

• Integrated Builders: Firms that directly combine the search for MEV opportunities with block construction, leveraging proprietary order flow and low-latency infrastructure.
• Independent Builders: Entities acting as neutral aggregators of transactions, typically lacking direct trading integration.

The protocol applies a just-in-time (JIT) auction design for each block, wherein the highest bidder wins the right to build and propagates the winning block to the validator. The observed MEV-Boost auction is implemented effectively as a “second price + 1 wei” mechanism for block payment.

2. Strategic Advantages for Integrated Builders

Integrated builders possess a critical structural advantage stemming from their capacity to delay surplus extraction until later phases of the PBS (Proposer-Builder Separation) auction after more information becomes available. For these builders, the auction functions analogously to a second-price auction—where the winner pays the second-highest bid—thus enabling truthful bidding. In contrast, independent builders must commit their bids upfront, absent exposure to competitors’ information, forcing them into a first-price regime and necessitating bid shading.

This dynamic is formalized via a hybrid auction model:

• All builders submit bids simultaneously.
• If an integrated builder prevails, payment equals the second-highest bid.
• If an independent builder prevails, payment equals their own bid.

Mathematically, the equilibrium strategy for a non-integrated builder with private value $v$ is presented as:

$$\sigma(v) = v - \frac{\int_0^{v} F_B^{n_B-1}(t) F_A^{n_A}(\sigma(t)) dt}{F_B^{n_B-1}(v) F_A^{n_A}(\sigma(v))}$$

where $F_A$ and $F_B$ denote value distributions for integrated and independent builders, respectively; $n_A$ and $n_B$ are their counts. For $n_B=1$ and uniform $F_A$ on $[0,1]$, the strategy simplifies to $\sigma(v) = \frac{n_A}{n_A + 1} v$. This formulation indicates surplus loss for independent builders relative to integrated builders.

3. Latency, Information Asymmetry, and the Winner’s Curse

MEV often comprises a common-value component, particularly in arbitrage scenarios (CEX/DEX) where all bidders pursue an identical opportunity. Latency becomes determinative; integrated builders (“fast” bidders) can update their bids after observing stochastic asset price evolution, modelled by:

$$v_t = v_0 \exp(m_t), \quad m_t \sim N(\mu t, \frac{\sigma^2}{2} t), \quad \mu = -\frac{\sigma^2}{2}$$

Slow bidders submit their bids at time $0$, while fast bidders adjust at $t = \Delta$. If the slow bidder bids $b > 0$, the fast bidder will always win if $v_\Delta > b$, leading to slow bidders being forced toward zero bid—the winner’s curse—where “winning” equates to overpaying relative to realized value.

For partial information, such as a candlestick auction where the fast bidder has revision access with probability $p$, the slow bidder’s equilibrium bid $b_0^s$ must satisfy:

$$(1-p)(v_0 - b_0^s) + p P(v_\Delta < b_0^s)[E[v_\Delta | v_\Delta < b_0^s] - b_0^s] = 0$$

This expression quantifies the direct cost imposed by latency disadvantage, encapsulating reduced expected surplus and heightened adverse selection.

4. Empirical Patterns of Multi-Block MEV Extraction

Analysis of approximately 4.3 million Ethereum slots post-merge reveals salient features regarding multi-block MEV extraction:

• The frequency of multi-slot sequences, where a builder produces consecutive blocks, is significantly lower than that predicted by Monte Carlo simulation based on daily market share.
• The longest observed uninterrupted sequence is 25 slots; for instance, an 11-slot sequence by BeaverBuild for Lido occurred on March 4, 2024.
• Most top builders, including ETH-Builder, f1b, and Blocknative, display fewer multi-slot sequences than expectations, and any surplus is limited to single-slot runs.

This indicates that builder strategies do not generally focus on securing extended multi-block runs, likely due to the risk-profile and auction structure of MEV-Boost.

5. Payment Dynamics and Weak Inter-Slot Correlations

MEV-Boost payments, mapped as a proxy for per-block MEV, exhibit the following empirical characteristics:

• Average payment per block increases with the length of consecutive sequences: ~0.05 ETH for single slots, rising to ~0.08 ETH for sequences of nine slots.
• There is a slight, incremental increase in payments for slots deeper into a sequence ($$\text{Payment} \approx 0.05\ \text{ETH} + \alpha \cdot (\text{sequence position})$$, with $\alpha > 0$).
• Auto-correlation analysis indicates only a weak positive correlation between consecutive payments; the Pearson coefficient attenuates quickly after 2–3 slots, and non-parametric measures show volatility dominates high MEV periods.

This suggests that, while longer sequences command mildly higher payments, builders do not systematically ramp up bids for extended runs, and MEV-Boost payment levels are largely unpredictable at the block scale.

6. Builder Specialization and Base Fee Volatility

Investigation of builder behaviors in relation to base fee volatility—a variable derived from gas fee fluctuations under EIP-1559—demonstrates negligible correlation:

• Classification by absolute fee change and Garman–Klass volatility both yield low Cramér’s V values (0.0664 and 0.0772 respectively), indicating minimal relationship between builder participation and fee environment volatility.
• No specialization or systematic adjustment in bidding strategies is observed among builders relative to base fee volatility.

A plausible implication is that, unlike token price volatility which can influence trading strategies, base fee volatility is not a determining factor for builder behavior in MEV-Boost auctions.

7. Efficiency, Fairness, and Centralization Implications

Structural features of MEV-Boost confer significant efficiency advantages to integrated builders by permitting truthful bidding in a regime approximating second-price auctions. Allocative efficiency is improved since bids better reflect actual MEV opportunity values, minimizing wasteful overbidding. However, the same mechanics amplify surplus extraction asymmetry: integrated builders with superior latency and informational access accrue greater market share, enhancing centralization risk. Independent builders face suppressed surplus and recurring “winner’s curse” scenarios, especially under rapid price evolution or common-value opportunities.

The rarity of multi-slot sequences and lack of strong bid correlation suggest that current MEV-Boost market architecture and block-by-block auction risk disincentivize deliberate multi-block MEV extraction strategies. Proposals for protocol evolutions (e.g., Attester-Proposer Separation) that mitigate inherent auction risks could theoretically enable systematic multi-block MEV extraction and reshape builder incentives.

Table: Key Empirical Observations in MEV-Boost Research

Metric	Value/Range	Context/Significance
Share of blocks via MEV-Boost	>90%	Dominant block-building method post-merge
Longest multi-slot builder sequence	25 slots	Rarity of extended builder runs
Avg payment—single-slot sequence	~0.05 ETH	Baseline MEV per block
Avg payment—9-slot sequence	~0.08 ETH	Modest value accumulation by sequence length
Cramér’s V (base fee volatility)	0.0664–0.0772	Minimal builder specialization by volatility

Conclusion

MEV-Boost has established itself as the central auction mechanism for block construction on Ethereum, enabling efficient surplus capture while structurally privileging integrated builders with low latency and access to fresh information. Empirical and mathematical analyses show that auction design and builder integration play pivotal roles in surplus distribution, market efficiency, and fairness. Multi-block MEV extraction remains infrequent and marginally incentivized under current market conditions, with payment dynamics and builder strategies displaying limited predictability and specialization. These findings inform protocol development considerations and the ongoing discourse on centralization and equitable MEV distribution within blockchains.