While a trader secures a superior Solana (SOL) swap rate, a passive pool depositor stays vulnerable to traders exploiting outdated quotes, according to a preprint released on Sept. 29.
For quiet-market SOL/USDC transactions, propAMMs—pools managed by professional operators—achieved a reference-relative execution cost proxy of 0.26 basis points, compared to 2.59 for standard public automated market makers (AMMs).
The research spans Sept. 1, 2025, through Aug. 31, 2026, alongside shorter observational windows for Base and Monad. It weighs execution amounts against Bybit’s size-weighted top-of-book USDT microprice, converted using its USDC/USDT midpoint. The study’s authors are affiliated with ETH Zurich and Category Labs.
While swappers desire more tokens for a given input, depositors supply the underlying inventory for those trades and require compensation for associated inventory risks. Consequently, low execution expenses may entice the former participant without necessarily presenting a viable investment rationale for the latter.
Swap prices and depositor returns on Solana
Across the Solana dataset, the paper notes two-second gross maker markouts of +0.37 basis points for propAMMs and −0.22 for public AMMs. A markout evaluates a fill against a subsequent reference price, where a positive value benefits the market maker.
Quiet-flow execution measures the concession a trader makes against a relatively stable reference. This proxy demands less than 1 basis point of reference movement spanning from five seconds prior to one second following a fill.
Maker markouts track the trajectory of a trade’s value after a pool accepts it. Combining these two metrics would mistakenly transform findings regarding pricing dynamics and adverse selection into a profitability assertion unsupported by the figures.
If an external market shifts first, a pool maintaining a stale price risks selling too cheaply or acquiring assets too dearly. Although an arbitrageur eventually realigns the prices, this correction happens via a transaction executed against the liquidity already residing in the pool.
Loss-versus-rebalancing studies view such arbitrage expenses as a single element of liquidity provider (LP) economics. Because total returns also encompass asset exposure and collected fees, an evaluation of investment performance requires a specific position, holding period, and its attributable revenues and costs.
Accurate accounting requires proper allocation of trading fees, inventory adjustments, hedging strategies, overhead expenses, and transaction costs. The brief observation window leaves these accounting factors unresolved, meaning venue averages cannot definitively prove that professional pools generated overall passive-LP losses.
Depositors require a comprehensive balance sheet review to assess returns accurately, whereas swappers benefit from liquidity whose operator actively manages pricing risks.
An April publication from Jump Crypto outlines how propAMMs—including its proprietary BisonFi platform—adjust pricing and available liquidity based on inventory levels, quote freshness, and incoming order flow quality. Because Jump is an active market operator, implementations naturally vary.
A market maker holding an excess of a specific asset can deter trades that add more of it, whereas an out-of-date price might justify shrinking depth or widening spreads. Furthermore, a routing path linked to adverse selection may face different terms than order flow deemed less risky by the maker.
From an economic standpoint, these mechanisms allow a maker to tighten quotes when anticipating lower risk. Enforcing identical terms across all counterparties would strip away a vital tool for managing that exposure, meaning the price ultimately secured by an everyday swapper still requires direct measurement.
Documentation regarding Jupiter’s AMM integration highlights a dedicated signer that flags trades originating from its frontend, categorizing that volume as retail and non-toxic. Nevertheless, identifying the origin of an order is distinct from independently verifying that every trade is entirely harmless to the market maker.
Behind public settlement layers, private market-making logic often operates. While the capacity to defend a price enables a firm to provide cheaper liquidity, accessing that price remains contingent on the specific route and counterparty involved.
Quote reliability is a separate test
Regarding Tessera on Base, execution lagged behind reconstructed previous-block-end quotes by an average of 1.08 basis points per trade and 0.56 by volume—a block-timed fee phenomenon researchers label as “spoofing.”
By comparing reconstructed pool output against actual execution while excluding individual screen quotes from the calculation, the observed pattern offers no conclusive proof of operator intent. Consequently, better execution relative to a market reference can coexist with worse execution relative to an earlier quote.
In a March 20 report, routing provider 0x detailed how Base prices degraded between quote selection and final settlement due to block timing shifts and changing spreads. Because specific operators went unnamed in the report, Tessera cannot be explicitly identified as the subject. Additionally, 0x noted an enforcement policy of cutting off data sources until execution issues are resolved.
If an advertised output lures an order only for a different output to be delivered, competition centered on the advertised figure risks rewarding the wrong trading venue. The core question becomes whether routing systems properly evaluate what a trader actually receives under the specific conditions of that transaction.
Jump asserted that routers selecting executable prices dynamically during transaction execution can significantly narrow the display-to-fill gap. The valuable design takeaway is that a market maker can preserve inventory, freshness, and counterparty protections, provided the router contrasts outputs that already incorporate those safeguards.
Jupiter’s active Swap API documentation outlines competition among routing engines alongside a mechanism that sidelines underperforming liquidity sources. Its integration guide likewise mandates quote-and-execution parity checks evaluated against identical pool snapshots.
Although a parity check confirms agreement on a single snapshot, maintaining consistency through subsequent updates remains an open question. Engine competition also leaves it unproven whether individual venues are continually reassessed during an executing transaction.
While comparing executable outputs provides a clear design direction, its practical effectiveness still needs to be measured.
For a valid comparison, the executable output must account for identical trade sizes, callers, prevailing pool states, and applicable fees. Otherwise, a favorable price restricted to one routing path could easily be mistaken for a price available across the board.
Jupiter’s documentation notes a platform swap fee on its Meta-Aggregator pathway and none on its Router pathway, highlighting how integrator fees and settlement arrangements vary. A protocol-level spread cannot substitute for the exact amount ultimately received after all deductions.
Future empirical evidence should directly compare quoted versus delivered outputs across matched transactions, clarify which costs are factored in, and demonstrate how routing mechanisms handle persistently lagging liquidity sources.
Ultimately, passive liquidity demands an independent, position-level return review. Enhanced routing may improve the swapper’s outcome while leaving the depositor’s core investment inquiry unresolved.
Frequently Asked Questions
What is the main difference between propAMMs and public AMMs?
PropAMMs are pools managed by professional operators that achieved lower execution costs for quiet-market trades, while public AMMs leave passive pool depositors more exposed to traders picking off stale quotes.
What study period and affiliations are associated with the preprint?
The study covers the period from Sept. 1, 2025, through Aug. 31, 2026, along with shorter samples for Base and Monad, and its authors are affiliated with ETH Zurich and Category Labs.
Why can’t maker markouts and execution costs be used to claim passive LP losses?
Returns also depend on asset exposure, collected fees, inventory management, hedging, and operating costs. Venue averages and short horizons cannot establish that professional pools caused aggregate passive-LP losses without a complete position-level balance sheet review.
What is “spoofing” in the context of the study?
Researchers use the term to describe a block-timed fee pattern observed on Base (specifically Tessera), where execution averaged worse than reconstructed previous-block-end quotes.
