Backrun Arbitrage Isn't the Problem. Misallocated Capture Rights Are.

On the distinction between adversarial extraction and post-trade value, and why protocol-level execution changes the allocation.

The MEV discourse has a framing problem.

By collapsing sandwich attacks and backrun arbitrage into a single "MEV" category, we obscure a fundamental economic distinction: one extracts value at the user's expense, the other captures value that exists independent of and subsequent to user execution.

This isn't semantic. It determines whether your MEV strategy should focus on prevention (impossible for backruns, the opportunity is intrinsic to CFMMs) or capture (achievable, with correct execution architecture).

The LVR Decomposition

Loss-versus-rebalancing quantifies what LPs pay for price discovery. But LVR itself decomposes into components worth distinguishing:

Adversarial LVR: Value extracted through information asymmetry exploited during execution. Sandwich attacks, toxic flow from informed traders, front-running oracle updates. The user or LP is directly harmed, their execution occurs at a suboptimal price because of the extraction.

Structural LVR: Value created by the mechanical reality that CFMM prices update discretely while reference markets update continuously. Post-swap, the pool quotes a price that differs from external markets. This differential is arbitrage profit for whoever corrects it.

Adversarial LVR is a tax on users. Structural LVR is a byproduct of AMM architecture.

The common prescription: "reduce LVR" conflates these. Reducing adversarial LVR through private mempools or MEV-aware routing protects users. But structural LVR isn't reducible in the same sense. The post-trade price dislocation exists. Someone captures it. The question is who.

Capture Rights as a Design Variable

Under default configurations, protocols implicitly assign backrun capture rights to external searchers. This isn't a conscious decision, it's an architectural absence. No mechanism exists to internalize the capture, so it externalizes by default.

The "Flow Owner" framing reframes this as a design variable rather than an environmental constant. Protocols generating meaningful swap volume are producing valuable order flow, specifically, information about post-trade price dislocations that have economic value.

Current state: protocols generate this information, external parties monetize it.

Alternative state: protocols capture this value and redistribute to LPs, treasuries, or fee reduction.

The transition requires solving an execution problem, not a pricing problem.

Why Pricing Mechanisms Alone Don't Solve Contested Arbitrage

PFDA and similar auction mechanisms efficiently price marginal arbitrage, small dislocations where gas costs approach profit and competition is minimal. By creating price discovery for these opportunities, protocols can extract value that would otherwise go uncaptured or leak to searchers with gas advantages.

But arbitrage opportunities aren't uniformly distributed. They follow a power law. The largest dislocations—from significant swaps, volatile markets, cross-DEX inefficiencies—concentrate most of the value. And these fall into what we'd call the contested regime: multiple sophisticated searchers competing for the same opportunity.

In contested regimes, pricing is necessary but insufficient. You can perfectly price a $50K arbitrage opportunity, but if you can't execute before competing searchers, the pricing is academic. The value flows to whoever lands the transaction.

This is the gap that execution mechanisms fill. Not replacing pricing, but completing the capture stack for contested opportunities where execution determines allocation.

Same-Transaction Execution as Race Elimination

The traditional backrun capture problem is a race condition. Opportunity emerges (swap executes), information propagates (mempool, block), searchers compete (gas auctions, builder relationships), one wins. Protocol has no privileged position in this race, external searchers with latency optimization and builder integration typically dominate.

Same-transaction execution eliminates the race by capturing within the originating transaction. Using hook architectures (Uniswap v4's afterSwap, Algebra Integral's callbacks), the sequence becomes:

1. Swap executes
2. Hook triggers
3. Backrun computes and executes
4. Settlement, atomic, single transaction

No mempool propagation. No block delay. No external competition for this specific opportunity. The protocol's execution mechanism has a first-mover advantage by architectural construction, not infrastructure investment.

This changes the game theory. External searchers can't compete for opportunities they never see. The capture right is assigned by transaction structure, not by race outcome.

Implications for Protocol Architecture

For protocols evaluating MEV strategy, the backrun/sandwich distinction suggests a bifurcated approach:

For adversarial extraction (sandwiches, toxic flow): Prevention mechanisms remain appropriate. Private order flow, MEV-aware routing, dynamic fees that penalize informed trading. The goal is protecting user execution quality.

For structural arbitrage (backruns): Prevention is the wrong frame, the opportunity is intrinsic to CFMM mechanics. Capture is the objective. This requires execution infrastructure: hook integrations, same-transaction settlement, atomic guarantees.

The error is treating all MEV as something to minimize. Adversarial MEV should be minimized. Structural MEV should be captured and internalized. Conflating them produces strategies that are defensive where they should be offensive.

The Incomplete State of Current Solutions

Most MEV "solutions" address half the problem:

• Private mempools / MEV protection: Reduces adversarial extraction, doesn't capture structural value
• Batch auctions: Reduces some adversarial flow, doesn't address post-batch arbitrage
• Dynamic fees: Prices some MEV, doesn't execute capture
• Searcher integration / OFAs: Captures MEV but externalizes it—protocol shares revenue rather than owning execution

Full-stack MEV strategy requires both protection (for adversarial) and capture (for structural), with capture meaning protocol-owned execution rather than revenue-sharing with external parties.

The protocols that will define the next era of DeFi economics are those that recognize order flow ownership implies arbitrage capture rights—and build the execution infrastructure to exercise them.