
Conservation of MEV: Why Ordering Value Cannot Be Eliminated, Only Changes Its Hiding Place?
Author: Jascha Samadi
Compiled by: TechFlow
TechFlow Insights: Jascha Samadi uses a physics-based premise to debunk the industry's biggest illusion: MEV simply cannot be "designed away." Whether through Ethereum's open auctions, Canton's permissioned ledger, or TradFi's "payment for order flow" and "last look," ordering value has never disappeared; it has merely changed hiding places. For investors and practitioners, the real question has never been "is there MEV?" but rather who are you handing over the pricing power of ordering to?

In physics, energy cannot be destroyed: it only changes form. Transaction ordering shares the same property. Wherever a ledger executes transactions against valuable shared state, execution order carries economic value, and this value is conserved across every architecture we know. It can be auctioned, hidden, burned, or renamed, but it does not disappear. Ethereum turned it into an open market. Permissioned networks like Canton shifted it behind contracts and non-disclosure agreements. TradFi spent a century repackaging it under names like "payment for order flow" and "last look." Hyperliquid recently started burning it.
Over the past few years, we have spent considerable time working inside and around the trading supply chain—the path a trade takes from a click in a wallet to its final placement in a block, and the market structures that organically assembled around that path. This article is the structural perspective left to us by that work. It traces the conservation of ordering value across four regimes: open markets, private ledgers, regulated intermediaries, and public auctions. The truly interesting question has never been "is there MEV or not," but whether the price of ordering is open or hidden. TradFi’s enforcement record meticulously demonstrates what happens when this value is denied rather than priced.
This article continues our Block Building series, which currently covers an introduction to the block space market, as well as the demand and supply sides of Ethereum block building since 2024.
Open Markets: Ethereum Prices Ordering Value
An open, permissionless blockchain is built on a handful of foundational design choices:
- Anyone can run nodes and independently verify the network's history.
- Anyone can submit transactions.
- Participating in consensus (i.e., the right to produce blocks and extend the chain) is open, earned through capital or work, and not granted by a gatekeeper.
- Underpinning all of this is the most critical choice: every participant replicates the same ledger. There exists a single global state visible to all participants, where each node re-executes every transaction to confirm rules were followed.
The globally shared state is what elevates these systems beyond ordinary distributed databases. Because every application resides in the same state machine, any contract can read and compositely compose with any other. A lending protocol can consume a price oracle it never coordinated with, accept collateral tokens minted by a third party, and be wrapped by an aggregator approved by no one: all of this within a single atomic transaction. Integration is an `import` statement, not a commercial development negotiation. This feature (permissionless composability atop communally verifiable state) is the source of most genuinely novel aspects in DeFi: open verifiability, trust neutrality, and markets assembled at deployment speed by unaffiliated parties.
The same choices carry a second, less heralded implication. If everyone shares the same state, transactions must be applied to it in some sequence, and because many transactions touch the same state, order can change outcomes. Two transactions targeting the same liquidity pool are non-commutative: the one sequenced first gets the better price. A liquidation is available to only one caller: the one who arrives first. An arbitrage opportunity exists in only one transaction: the one placed ahead of the others. Position in the sequence is access to economic opportunity, and scarce, valuable access does not stay free for long.
Ethereum provides the clearest example of what happens once this value becomes recognizable. An industrialized supply chain (searchers, builders, relays, validators (see our Introduction to the Ethereum Block Space Market)) organically assembled around the path between user wallets and the eventual block. Searchers scan the public mempool and state for extractable opportunities: cross-pool arbitrage, liquidations, and on the predatory end, sandwich trades around visible user orders. Builders assemble candidate blocks. They pack and sequence transactions and bundles to maximize total block value. Under Proposer-Builder Separation (PBS), validators no longer build blocks themselves; they can auction off block-building rights via relays to the highest bidder. Further upstream, order flow itself has become an auctionable asset, with wallets and apps selling the right to interact with user trades.
This mechanism is often described as parasitic, and parts of it may indeed be. More precisely, PBS is what a market for a scarce resource looks like after that resource is acknowledged.
Ordering has always been valuable; Ethereum’s architecture made this value visible, contestable, and measurable.
Subsequent research agendas—order flow auctions, MEV rebates that return extraction to users creating that transaction value, encrypted mempools, batch auctions—are not attempting to deny this phenomenon, but to govern it: consciously deciding who captures ordering value and how much flows back to users.
Another approach is to outright remove the prerequisites for an open sequencing market. This permissioned approach now backs monthly trading volumes in the tens of trillions of dollars.
Private Ledgers: Canton Removes the Market, Not the Value
The most prominent permissioned network in production today has inverted (intentionally and coherently) each of those design choices above, yet what it eliminates is the sequencing market, not discretion over sequencing. Canton lacks global state and global transparency. Its architecture is a network of networks: each participant runs a validator, holding only the contracts in which they are a party, with data propagated strictly on-demand. This privacy model extends to individual transactions. Sub-transaction privacy means that even counterparties in different branches of an atomic transaction cannot see one another: a bank settling the bond leg of a DVP trade won’t learn the commission terms of another branch in the same atomic submission. There is no public mempool: pending transactions are never broadcast, only delivered to their stakeholders. The sequencing layer is blind: synchronizer operators sort cryptographic proofs without seeing transaction content.
Canton’s stance on MEV is almost mechanically derived from these choices, built on four mechanisms:
- No mempool to front-run: transactions are never publicly broadcast, so the raw material for front-running (visible pending order flow) does not exist at the network layer.
- Sequencing is blind: operators cannot front-run what they cannot read.
- No sequencing market: no builder or searcher auctions, no priority fee bids for position, no PBS supply chain: sequencing is a utility function executed by identified, contracted, legally liable institutions, not anonymous actors.
- Sequencing rules live at the application layer: they are embedded directly into smart contract logic, allowing exchange-style fairness rules (e.g., price-time priority) to be enforced constructively, independent of validator behavior.
The adoption scale behind this design is substantial by any metric. Broadridge’s distributed ledger repo platform settles approximately $8–9 trillion in monthly repo volume on Canton infrastructure (~$354 billion daily), making it the largest live tokenized RWA settlement platform today. Public tracking data shows roughly $345 billion in representative asset value on the network, with claimed monthly throughput nearing $9 trillion. The alignment with flagship use cases is genuine: interbank repos and collateral markets are precisely where privacy, permissioning, and actionable finality matter more than censorship resistance. No bank will broadcast its financing positions to a public mempool (real-time exposure of who is borrowing overnight cash isn't transparency; it's a bank run accelerator), and no regulator would license key market infrastructure where a software failure equates to irreversible loss.
We consider this divergence highly durable: these permissioned networks will likely continue succeeding in institutional back-office workflows (settlements, collateral mobility, post-trade processing), where participants are known, trades are pre-negotiated, and confidentiality and recourse are requirements, not preferences. That market is huge, and Canton is winning it through architectural advantage. But the immediate question is narrower: does this architecture actually eliminate MEV, or is it doing something far more familiar to it?
The structural observation is simple: someone is still sequencing.
On Canton, sequencing discretion did not disappear; it was relocated: to synchronizer operators at the public layer, and to single companies on private synchronizers, where the vast majority of today's volume actually settles.
Discretion remains; it is uncontestable, unobservable, bound by NDAs and terms of service, not eliminated by cryptography. Therefore, what a permissioned network can strictly claim under the banner of "zero MEV" is far narrower than the phrase implies: there is no permissionless MEV market, nor public evidence of anyone extracting MEV. These are two distinct statements, and the distance between them encompasses most of market structure history.
A second explanation for observed MEV approaching zero on these networks, alongside design, is use case. Canton’s flagship workflows (bilateral repos, pre-negotiated DVP settlements) contain virtually no contested shared state. In a repo agreement where price, size, and counterparty are fixed before the trade hits the ledger, there is nothing to sandwich; sequencing a pre-negotiated settlement queue generates no economic value on any chain or under any design. Thus, the current record is better described as unfalsified rather than proven.
The falsifiability test will come when such a network hosts genuinely competitive market structures: shared central limit order books, clearing engines, oracle-triggered margin calls, scaled-tokenized Treasury markets. By then, arrival sequence to the synchronizer will once again carry economic determinism, and those capturing it will be those with latency advantages, physical proximity, or operator relationships. Rent does not vanish. It migrates into privileged, opaque channels: priority connections that never appear in fee schedules, information proximity that never appears anywhere. This architecture reintroduces trusted intermediaries with sequencing discretion onto a ledger and reconstructs TradFi’s trust model.
Whether this is a stable endpoint depends on that trust model’s track record on this exact issue. And that record is quite instructive.
Regulated Intermediaries: How TradFi Has Hidden the Same Rent for a Century
TradFi has operated the "trusted intermediary with sequencing discretion" model for decades; its enforcement record shows that opacity did not prevent ordering value capture—it enabled it to be profitable.
Traditional market structures typically view what we now call MEV as a system design problem: an issue already solved through correct rules, accountable intermediaries, and enforcement. The structural perspective argues otherwise.
MEV is valuable shared state paired with sequential execution, and every trading venue possesses both. From this viewpoint, TradFi never eliminated the problem; it banned it via rules, imperfectly regulated it, and pushed extraction into channels invisible to victims. Enforcement records align perfectly with this structural view.
Payment for Order Flow (PFOF) turned retail orders themselves into monetizable assets. In 2021 alone, US brokers collected approximately $3.8 billion in PFOF: capturing the value of seeing and internalizing order flow through bilateral agreements between brokers and wholesalers, entirely invisible to customers viewing the order as a product. Robinhood paid a $65 million SEC settlement in 2020 for misrepresenting execution quality related to this arrangement. Dark pools pitched protection from predatory order flow but generated a series of opposite settlements: Barclays paid $70 million in 2016, Credit Suisse $84.5 million, for misrepresenting how their dark pools operated, including high-frequency order flow that customers were told had been excluded actually participating in trades; and ITG paid $20.3 million in 2015 for operating an internal book within its own dark pool that traded on confidential client order information. In FX, Barclays paid $150 million in 2015 for using "last look" (a dealer's option to decline a trade after observing price movement direction) and coordinating shared client order information prior to benchmark fixings, with cumulative fines for such industry practices exceeding $10 billion.
The pattern across these cases is consistent. Ordering value is real. Extraction is carried out by trusted intermediaries positioned in the sequence. Affected parties cannot observe it in real-time. Facts only emerge years later through whistleblowers and enforcement actions. This is the operational model imported onto ledgers under the guise of "design-zero-MEV": ordering rent is adjudicated by compliance, uncovered by regulators, rather than priced by an open market.
Public Auctions: Hyperliquid Burns It
The transparent alternative is not hypothetical: Hyperliquid took that same structural rent, funneled it into a public auction, and directed revenue to a public treasury.
Hyperliquid’s architecture (no public mempool, sequencing handled at the consensus layer) has long been described as MEV-resistant, and by mempool-extraction standards, it indeed is. Yet execution advantages persist, because structurally, they always do: priority belongs to those who engineer the lowest latency. Sophisticated market makers win queue positions through infrastructure optimization and network proximity, functionally equivalent to co-location. This advantage has existed for a long time, carries significant economic weight, and remains accessible only to participants with the capital and pipelines to build it. Rent is being captured; it just hasn’t appeared on anyone’s dashboard.
This April, Hyperliquid formalized this dynamic. It introduced priority fees: public, continuously repeating Dutch auctions priced in $HYPE, auctioning exactly what low latency buys: earlier visibility into incoming transaction data, and earlier positioning in the execution queue. Its impact can be measured with exceptional precision: paying one basis point in priority fees improves end-to-end latency by roughly 45 milliseconds. An advantage that was previously implicit, private, and constrained by engineering capability has transformed into an explicit, public, continuously repricing, open-to-all market.
This mechanism has two attributes worth discussing separately. First, it formalizes a paid priority tier, which can genuinely be criticized: a visible queuing tax where none existed before. But the relevant comparator isn't a market without priority tiers; it's the same tier previously allocated invisibly through capex and connectivity relationships at zero price.
An open, priced advantage is more honest and easier to correct than an implicit, free one.
A visible market can be measured, debated, capped, and redesigned; an invisible one can only be discovered retroactively.
The second attribute is where the revenue goes, which is where we believe true novelty lies in market structure. These fees are burned. The clearing price for execution priority is not paid to venue operators, not shared with favored counterparties, nor embedded in any bilateral agreement; it is removed from supply (economically equivalent to proportionally allocating it to every holder of the network asset $HYPE). The contrast with how equivalent value flows in traditional markets is direct: colocation fees go to exchange shareholders; PFOF is distributed among wholesalers and brokers via contracts terminal clients never see; information rent from dark pools and last look rights goes to privileged seat holders until enforcement recovers a portion. Hyperliquid puts the exact same structural rent (the value of position in the sequence) in the open, allows a permissionless auction to discover its price, and directs revenue to a public treasury.
The queue remains; what changed is that its price is open, its economic yield shared with every holder, rather than settled in closed bilateral structures.
This is not claiming ordering value has been designed away. It acknowledges it cannot be eliminated, then decides to price it in an open market and redistribute proceeds.

Conclusion: The Difference Lies in Whether the Price Is Open
Meaningful comparisons between chain designs have never been "MEV vs. no MEV"; every venue that sequences valuable state transitions will structurally generate ordering value.
The questions that truly differentiate market designs are narrower and easier to answer: who holds sequencing power; is it contestable or delegated; is it observable or hidden; is its rent priced and redistributed in an open market, or quietly captured by whoever is closest to the sequencer?
Permissioned networks answer these questions through governance: clear operators, legal accountability, and institutional guarantees. For pre-negotiated settlement flows, where sequencing carries no value from the outset, this answer suffices, and adoption data reflects it. For genuinely competitive markets, evidence jointly supplied by the brief history of crypto and the long history of TradFi points in the same direction: denied ordering value tends to be captured in the shadows, while acknowledged ordering value can be priced, audited, and returned to the markets that create it. In this sense, ordering value is conserved across different designs. The difference lies in whether its price is open.














