← All articles

Ethereum's Blockspace Market Under the Microscope

@0x3hu79
@0x3hu79
April 4, 2026 · Market Structure
#PBS#block-building#MEV#relays#order-flow#preconfirmations

In December 2025, a group of Ethereum’s core infrastructure operators published a joint assessment of the blockspace market: Kubi M. from Titan Builder, Alex T. from Ultrasound Relay, Kevin L. from ETHGas, and Justin D. from the Ethereum Foundation. A builder, a relay, a pricing researcher, and a protocol researcher - four perspectives on the same pipeline they all operate in daily.

The paper isn’t an EIP or a specification. It’s a problem statement: PBS solved the validator centralization problem, but the out-of-protocol market that emerged has structural issues across economics, robustness, performance, and services. The authors map these problems and propose principles for addressing them.

This is a significant paper. The problems it identifies and the principles it proposes will likely shape the blockspace market discourse for the next 1-2 years. Here’s what it says, and what it means.

The Core Framing: Blockspace Is the Product

The paper opens with a powerful lens: Ethereum’s core business is the production and sale of blockspace - verifiable compute and data storage. Everything else - DeFi, rollups, stablecoins, NFTs - is consumption of that blockspace. The protocol produces it through validators; the out-of-protocol market determines how it gets priced, allocated, exchanged, and consumed.

Through MEV-Boost and Proposer-Builder Separation (PBS), over 90% of Ethereum’s blockspace flows through this out-of-protocol pipeline. PBS preserved validator neutrality - proposers stay lightweight and don’t need to be sophisticated block constructors. But the pipeline itself has problems.

1. Economics: Price Discovery Is Broken

The Exclusive Order Flow Trap

This is the paper’s most important economic observation. It describes a game theory trap that most users and even many protocol researchers don’t fully appreciate.

Here’s how it works:

A user wants a transaction included. They have two options: send it to the public mempool (all builders can see it) or send it privately to one specific builder.

If they send to multiple builders, the priority fees get competed away in the relay auction - builders bid higher against each other, passing those fees to the validator. The user doesn’t benefit from the competition.

If they send exclusively to one builder, that builder has a monopoly on the transaction. The builder doesn’t need to bid as aggressively (competitors don’t have the transaction), retains some of the priority fees, and can refund a portion to the user as a kickback.

The result: users are economically incentivized to send exclusively to a single builder. The paper cites research suggesting that up to 84% of total fees in winning blocks come from exclusive transactions.

This creates a cascading problem. Each builder only sees its own exclusive flow plus the public mempool. The winning builder constructs blocks from a partial view of all available transactions. High-value transactions that went to a different builder go unincluded, spilling into later slots. Users experience worse inclusion times. Validators earn less than the theoretical maximum. The system as a whole produces suboptimal blocks.

It’s a classic collective action problem: each builder individually benefits from exclusive flow, but the market collectively suffers from the fragmentation it causes.

Volatility Risk Falls on the Wrong People

The paper argues that end-users currently absorb all blockspace volatility risk. Gas prices spike, users overpay or get delayed, and there are no instruments to hedge against this uncertainty.

The proposed direction: blockspace futures - financial instruments where specialized actors (blockspace wholesalers, market makers) absorb volatility risk, and users can hedge. A rollup could buy guaranteed blob slots 24 hours ahead at a fixed price. A DeFi protocol could hedge against gas spikes during volatile market conditions.

This is how most commodity markets work. Oil, electricity, bandwidth - they all have futures markets that transfer risk from consumers to specialized intermediaries. Blockspace doesn’t have this yet, and ETHGas (one of the paper’s co-authors) is actively researching it.

For block builders, this is a potential new business line. Builders are already the actors most capable of making credible forward inclusion commitments. Offering rollups and institutional users guaranteed future blockspace at a fixed price is a natural extension of what builders already do - it just adds a time dimension.

Relays Have No Sustainable Economics

Relays perform critical functions - they solve the fair exchange problem between builders and proposers, host the whole-block auction, and provide latency optimization - but the protocol doesn’t compensate them. They operate as pseudo-public goods funded by donations or cross-subsidized by other business lines.

The consequences are predictable:

The paper makes an interesting observation about ePBS: it removes the trust function of relays, but trust isn’t what drives relay adoption. Relays are used for their services - bid delaying, cancellation, price discovery. Even with ePBS, relay-like service providers will be needed. The economic model question remains.

Block builders should be paying close attention to relay economics. Any change in how relays monetize - whether through fees, bid adjustments, or new models under ePBS - directly affects builder cost structures and which relays are worth routing through.

2. Robustness: The Pipeline Is More Fragile Than It Looks

Concentration Creates Correlated Failure Risk

The PBS pipeline has winner-take-most economics. A small number of builders win most blocks. A small number of relays handle most auctions. This concentration means a single event can cascade:

When a dominant builder loses the relay auction, its exclusive order flow faces delayed inclusion. There’s no graceful failover mechanism - the transactions simply wait for the next slot where that builder might win.

Censorship Resistance Has a Blind Spot

The paper acknowledges that FOCIL (Fork-Choice Enforced Inclusion Lists) addresses censorship for public mempool transactions. But it identifies a critical gap: private transactions can’t benefit from inclusion lists without being revealed.

If a DeFi user sends a private transaction for legitimate privacy reasons, FOCIL can’t force its inclusion - the FOCIL committee can’t include what it can’t see. The paper hints at block merging as a potential solution: different parties each construct partial blocks (FOCIL committees handle public flow, builders handle private flow) and these get merged into a final block. This is still an active research area.

After the Merge, up to 78% of blocks were OFAC-compliant censored for a period. The paper treats this history as a warning: liveness alone isn’t sufficient if access is conditional based on transaction content.

For block builders, the censorship question is becoming a competitive factor. As the ecosystem moves toward stronger censorship resistance guarantees, builders who can credibly commit to non-censorship - whether through technical mechanisms like TEEs or through transparent, verifiable policies - will be more attractive to order flow providers who care about neutrality.

Proposer Agency Is Eroding

Validators are becoming increasingly dependent on specific builders and relays. Switching costs are growing. The paper argues proposers must retain effective control: the ability to select and rotate providers, enforce allocation policies, monitor for censorship, and recover safely when delegates fail.

Without this control, out-of-protocol infrastructure can diverge from the protocol’s neutrality, embedding dependencies that increase switching costs over time.

3. Performance: The Pipeline Has No Measuring Stick

No One Publishes SLOs

The paper maps the critical latency path - the “hot path” from transaction origination to finalized block:

Transaction originator
  -> Builder (blockspace allocation)
    -> Relay (whole-block auction)
      -> Validator (signature/commitment)
        -> Network (propagation to attesters)

Every hop in this path is ad-hoc. There are no published cut-off times. No queue depth metrics. No formal Service-Level Objectives. Late transactions are unpredictably dropped or pushed to the next slot. The result is unstable time-to-inclusion that users defensively price in with higher fees.

The paper calls for explicit SLOs:

This is table stakes in any mature market infrastructure. Traditional exchanges publish SLOs. Cloud providers publish SLOs. The blockspace pipeline doesn’t.

In my opinion, block builders should start thinking about publishing their own SLOs. Today, no builder publishes performance metrics. The first to do so - with credible, measured numbers - gains an immediate trust advantage with order flow providers and sets the bar for the rest of the market.

Burst Demand Causes Multi-Slot Cascades

When demand spikes - a popular NFT mint, a market crash triggering liquidations, multiple rollups posting blobs simultaneously - there are no explicit admission policies. The burst propagates through the entire pipeline: builder backlogs overflow, the relay auction runs hot, and what could have been a single-slot spike turns into multi-slot delays.

The paper proposes explicit slot budgets: stop accepting new bundles after a defined cut-off time, and publish queue depth and drain rate so wallets and rollups can pace their submissions.

Here’s what block builders can anticipate going forward: as the discourse shifts toward measurable performance, builders will be expected to demonstrate graceful degradation under burst load. Internal capacity management - knowing when to stop accepting bundles, how to triage during spikes, and how to communicate queue state to order flow providers - becomes a differentiator.

Proximity Should Not Be the Primary Competitive Edge

The paper argues that persistent geographic proximity advantage (colocation with builders, relays, and validators) should be bounded. Competitive edge should come from measurable service quality - latency, reliability, block value - not from having a rack in the same data center.

Proposed mechanisms include timing windows that cap the proximity advantage, portable routing across venues, and substitutable operators without bespoke integrations.

4. Services: Innovation Is Bottlenecked by Gatekeepers

Builders and Relays Control Access

Anyone wanting to offer a new blockspace service - preconfirmations, inclusion guarantees, custom execution policies, priority lanes - must convince builders and relays to integrate it. This concentrates decision power, raises integration costs, and lengthens feedback cycles.

The paper envisions an open services layer with standard, neutral interfaces where services can be expressed directly:

Instead of bilateral deals with individual builders, originators and applications would request specific behaviors through standardized interfaces, and providers would compete to fulfill them.

L2s Are Underserved

When PBS first emerged, the L2 ecosystem barely existed. Now rollups are major blockspace consumers with specific needs that the current pipeline wasn’t designed for:

The paper suggests these services should be formalized and standardized, not negotiated bilaterally.

Block builders should be watching the L2 services space closely. Rollups need guaranteed blob inclusion, predictable pricing, and composability with L1 state. These are concrete service offerings that builders are uniquely positioned to provide - and it’s a less competitive space than L1 MEV extraction.

Guarantees Need Economic Backing

Currently, commitments from builders and relays are enforced by reputation alone. If a builder breaks a preconfirmation promise, the penalty is trust erosion - which in practice favors established incumbents (they have more trust to spend) and entrenches the existing market structure.

The paper calls for economically-backed guarantees: bonds, slashing conditions, refund policies. If you promise inclusion and fail to deliver, you lose money. This makes commitments credible regardless of who you are, and it levels the playing field for new entrants.

The 10 Principles

The paper concludes with a set of principles for evaluating any proposed improvement to the blockspace market:

  1. Transparent price discovery - pricing maps to state access, overpayment is minimized
  2. End-to-end robustness - consumption paths meet production-level guarantees
  3. Fair access & censorship resistance - non-discriminatory across public and private flow
  4. Sustainable incentives - key infrastructure providers have durable economics
  5. Proposer agency - validators stay lightweight while retaining policy control
  6. Hot-path latency discipline - explicit queuing, deadlines, and capacity signals
  7. No proximity lock-in - compete on quality, not geography
  8. Open services with verifiable guarantees - standard interfaces, measurable outcomes
  9. Observability & accountability - published metrics, staged changes, clear fault attribution
  10. Governance of critical infrastructure - explicit, accountable, capture-resistant

These aren’t just aspirational statements. They’re a framework that will likely be used to evaluate every blockspace market proposal over the coming years.

Reading Between the Lines

It’s worth considering what this paper is beyond a technical analysis.

The authors are the incumbents. Titan is one of the largest builders. Ultrasound is a major relay. They benefit from the current market structure even as they identify its flaws. The paper is careful to say it’s a “starting point” and a “call for engagement” - it frames the problems and sets the principles, but deliberately avoids proposing specific solutions.

Whoever frames the problems influences the solutions. By publishing this paper, the authors are defining the terms of the industry debate. The principles they propose will be the measuring stick. That’s a form of influence worth recognizing.

The paper also doesn’t address one fundamental tension: the call for “open services” and “portable guarantees” could disrupt the exclusive order flow relationships that builders currently depend on. Whether the authors believe they can navigate to a new equilibrium that still favors their position, or whether they’re genuinely willing to accept a more competitive market - that’s the open question readers should hold while evaluating the proposals that follow.

Sources

Original paper:

Referenced projects and initiatives:

Referenced research: