AMM Fee Optimization Study Finds Traditional Pools Capture Just 50% of Value

Every swap on a decentralized exchange leaves a small mark behind: a price that no longer matches the wider market. That gap, tiny as it usually is, quietly funds an entire industry of arbitrage bots and MEV searchers. New research from the MEV-X research team, working with academic collaborators from HSE University, digs into exactly how much value gets created by these micro-dislocations, and who actually ends up keeping it. The findings point toward a specific answer for AMM fee optimization: pools that charge nothing on arbitrage trades capture the most total value, even though that sounds backwards at first glance.

Key takeaways

• DeFi markets are efficient overall but constantly dislocated at the micro level, since almost every swap knocks a pool’s price away from the broader market.
• The combined value released by that mispricing — called dislocation value — is maximized when a pool charges a zero fee on the arbitrage that corrects it.
• Constant-product AMMs like the classic x·y=k model can theoretically capture only up to 50% of dislocation value through fees, and real markets often fall short of even that.
• Using AMM hooks, a pool can execute a mechanism that executes internal arbitrage at the atomic level following every swap, securing value that might otherwise remain untapped or be claimed by external searchers.
• Retail traders keep paying normal swap fees; only the pool’s own internal rebalancing trade runs fee-free.

DeFi’s Hidden Inefficiency: Where Arbitrage Value Comes From

Arbitrage in decentralized finance exists because pools react slowly to price changes elsewhere. Any swap large enough to move a pool’s price away from the market rate creates an opening, and someone, somewhere, is usually watching for it. According to the MEV-X and HSE University researchers, this isn’t an occasional glitch — it’s structural. Almost every trade produces a small dislocation, and each one becomes a potential profit opportunity for whoever arbitrages it fastest.

What Is Dislocation Value in AMMs

The researchers define dislocation value as the total prize released by a mispricing: the arbitrageur’s profit plus whatever fee the pool collects on that arbitrage trade. It’s a simple idea with a useful consequence — it lets you measure not just how much value a market inefficiency creates, but how that value gets split between the two parties involved.

Why Arbitrage Fees Erode Market Efficiency

Here’s the twist: charging a fee on arbitrage doesn’t just redistribute the dislocation value, it shrinks it. When a pool sets a positive fee on the corrective trade, part of the opportunity becomes uneconomical for the arbitrageur to bother collecting. That portion isn’t captured by the pool either — it simply evaporates, left unrealized because the arbitrage stops short of fully re-pricing the pool. The study’s simulations put a number on this: at the fee level that maximizes the pool’s own fee revenue, total dislocation value had already dropped to 74.62% of its maximum.

The Theoretical Limits of AMM Fee Structures

A zero fee on arbitrage produces the largest possible pool of value, but that value goes entirely to whoever executes the trade — usually an outside arbitrageur, not the liquidity provider footing the risk. This tension sits at the center of the paper’s argument for rethinking AMM fee optimization altogether.

Zero Fees Maximize Total Value

In the researchers’ numerical simulation, a zero fee on arbitrage yielded the maximum dislocation value of 24.75 units, with the entire amount going to the arbitrageur. Raise the fee even slightly, and the total starts falling immediately. Push it past roughly 0.99%, and the fee exceeds what an arbitrageur could ever recover — at that point, the arbitrage simply never happens, the pool earns nothing, and the mispricing just sits there uncorrected.

The team tested this across nine different market setups, varying the liquidity ratio between the volatile pool and the reference pool, and the size of the price dislocation. The shape of the result held steady every time — only the dollar amounts changed. They also ran the same test across six different AMM architectures, including Uniswap V2, Uniswap V3’s concentrated liquidity, Balancer, Curve, Trader Joe, and DODO. The zero-fee optimum showed up on all of them.

Constant-Product Pools Cap Out at 50%

For constant-product AMMs — the x·y=k formula behind pools like Uniswap V2 — the paper works out a hard ceiling. A liquidity provider tuning its fee purely to maximize its own fee revenue can capture, at best, about 50% of the maximum dislocation value when the price gap is small. In the numerical example, the optimal fee sat near 0.50%, extracting 12.36 units, or 49.94% of the maximum. At the more conventional 0.30% fee, the pool took just 10.42 units (42.08% of the maximum), while the arbitrageur still walked away with 12.08. Push the dislocation size up to 10%, and the fee-revenue peak drops slightly further, to somewhere between 47.1% and 49.1% of the maximum, edging back toward 50% only as dislocations get smaller.

Why does this matter beyond the math? Because it shows that under the current market structure — where the pool and the arbitrageur are separate, self-interested parties — liquidity providers are structurally capped in how much of this value they can ever realistically keep, no matter how carefully they tune their fees.

Merging Liquidity Providers and Arbitrageurs

If a positive fee shrinks the total pie, and a zero fee hands the whole pie to an outside searcher, the only way out is to remove the split entirely. The paper’s proposed fix is to merge the liquidity provider and the arbitrageur into a single actor, so the value never leaves the pool’s own hands in the first place.

How AMM Hooks Enable Internal Arbitrage

This merger became technically possible with AMM hooks — code that lets a pool react to its own state changes within the same transaction as a user’s trade. Using AMM hooks, a pool can execute an atomic, internal arbitrage the moment a retail swap creates a dislocation. Because the correction happens inside the same transaction, there’s no window of time for an external searcher to step in, no priority-fee auction, and no MEV leakage to a block builder. This is atomic arbitrage execution in practice: the pool effectively becomes its own fastest, most efficient arbitrageur.

What Changes for Retail Traders

Crucially, this internal, fee-free arbitrage applies only to the pool’s own rebalancing trade — not to ordinary users. Retail traders continue paying the normal swap fee they always would, and that fee is still what compensates liquidity providers for the risk of holding inventory. The zero-fee treatment is reserved strictly for the pool’s internal correction step.

There’s a secondary benefit worth flagging for anyone thinking about liquidity provider arbitrage strategies: internalizing the rebalance doesn’t just add fee revenue, it also partially undoes the price divergence the original swap created. How much depends on the depth of the reference pool and the fee charged — against a deep reference pool at zero fee, the pool snaps back almost exactly to where it started. A shallower reference pool, or any positive fee, leaves part of that divergence standing.

Why This Reshapes the Conversation Around AMM Design

The practical implication here goes beyond a single protocol. If the zero-fee joint optimum genuinely holds across AMM formulas — as the six-architecture test suggests — then any pool sticking with a conventional positive fee on arbitrage is, by definition, leaving dislocation value DeFi participants could otherwise capture on the table. That’s a meaningful competitive signal for protocol designers weighing whether to adopt hook-based internal arbitrage versus sticking with the traditional model of external searchers competing for MEV.

It also reframes what “efficient” fee-setting even means for a liquidity pool. Optimizing purely for fee revenue, as most AMMs do today, turns out to be a different — and lower-ceiling — goal than optimizing for total value capture. The distinction matters most for larger pools and protocols where even small percentage gains in captured dislocation value translate into real revenue at scale.

FAQ

What causes arbitrage opportunities in DeFi markets?

Almost every swap moves a pool away from the market price, creating micro-level inefficiencies that open arbitrage opportunities.

Why does charging a fee on arbitrage reduce the total dislocation value?

Positive fees on arbitrage lower the dislocation value because part of the arbitrage opportunity remains unrealized as the arbitrageur is disincentivized from executing it.

How can an AMM maximize liquidity provider revenue from arbitrage?

By charging zero fees on arbitrage and internalizing the arbitrage operation atomically, merging the liquidity provider and arbitrageur roles into one.

What is the role of AMM hooks in arbitrage optimization?

AMM hooks enable a pool to execute atomic internal arbitrage without MEV leakage, capturing arbitrage value for liquidity providers rather than losing it to outside searchers.

Article produced with the assistance of artificial intelligence and reviewed by the editorial team.