Traders can secure better Solana (SOL) swap prices while passive pool depositors remain vulnerable to market participants exploiting outdated quotes, according to a preprint released on Sept. 29.
When analyzing quiet-market SOL/USDC fills, the study found that propAMMs—pools managed by professional operators—had a reference-relative execution cost proxy of 0.26 basis points. In contrast, public automated market makers (AMMs) registered 2.59 basis points.
The research period spans from Sept. 1, 2025, to Aug. 31, 2026, accompanied by shorter sample windows for Base and Monad. Fills are weighted by notional against Bybit’s size-weighted top-of-book USDT microprice, which is converted using its USDC/USDT midpoint. The paper’s authors are affiliated with ETH Zurich and Category Labs.
While swappers look to maximize token output for a given input, depositors provide the inventory that facilitates these trades and require proper compensation for the associated risks. Consequently, low execution costs may attract traders without necessarily establishing a viable investment case for liquidity providers.
Swap prices and depositor returns on Solana
The paper highlights that across the Solana dataset, two-second gross maker markouts averaged +0.37 basis points for propAMMs and −0.22 basis points for public AMMs. A markout evaluates a fill against a subsequent reference price, where a positive value benefits the market maker.
Quiet-flow execution assesses the concessions a trader makes relative to a relatively stable reference price. This proxy requires the reference price to move less than 1 basis point during the window from five seconds before to one second after a fill.
Conversely, maker markouts track the trajectory of a trade’s value after a pool accepts it. Combining these two metrics would mistakenly transform data regarding pricing dynamics and adverse selection into definitive claims about profitability that the figures do not substantiate.
If an external market shifts first, a pool continuing to quote an outdated price risks selling too cheaply or purchasing too dearly. Although an arbitrageur eventually realigns the prices, this correction occurs via a trade executed against the liquidity currently residing in the pool.
Research into loss-versus-rebalancing views such arbitrage expenses as a specific element of liquidity provider (LP) economics. Because total returns also depend on asset exposure and accumulated fees, evaluating an investment requires analyzing a specific position, holding period, and its associated income and costs.
Proper accounting requires appropriate allocation of trading fees, alongside considerations for inventory adjustments, hedging, operating expenses, and transaction costs. The brief evaluation timeframe leaves this accounting incomplete, and venue-wide averages cannot prove that professional pools directly caused aggregate losses for passive LPs.
Depositors require a comprehensive balance sheet evaluation to gauge returns, whereas swappers benefit from liquidity managed by operators actively mitigating pricing risk.
An overview published by Jump Crypto in April details how propAMMs—including its own BisonFi—adjust pricing and available liquidity based on inventory levels, quote freshness, and incoming flow quality. Jump is an interested market participant, and implementation methods vary across operators.
A maker holding an excess inventory of a particular asset can discourage trades that add more of it, whereas a stale price may prompt the maker to withdraw liquidity depth or increase fees. Furthermore, routing paths associated with higher adverse selection may encounter different terms than order flow deemed less risky by the maker.
From an economic standpoint, these mechanisms allow makers to offer tighter quotes when anticipating lower risk. Mandating identical terms for every counterparty would eliminate a vital tool for managing that risk, leaving the final price received by an ordinary swapper still subject to measurement.
Documentation from Jupiter’s AMM integration outlines a dedicated signer that identifies trades originating from its frontend, classifying that flow as retail and non-toxic. Nevertheless, identifying the origin of a trade is distinct from independently verifying that every transaction is completely harmless to the market maker.
Private market-making logic frequently operates behind public settlement layers. While the capacity to defend a specific price can enable a firm to provide cheaper liquidity, access to that price relies heavily on the specific execution route and counterparty.
Quote reliability is a separate test
Regarding Tessera on Base, execution outcomes averaged 1.08 basis points worse per trade and 0.56 basis points worse by volume compared to reconstructed previous-block-end quotes—a block-timed fee pattern researchers refer to as “spoofing.”
Because the researchers compare reconstructed pool outputs against actual executions rather than evaluating individual screen quotes directly, the observed patterns do not offer conclusive evidence of operator intent. Consequently, achieving better execution relative to an external market reference can coexist with worse execution relative to an earlier quote.
A March 20 report by routing provider 0x highlighted that prices on Base degraded between the moment of quote selection and final settlement due to block timing and spread shifts. Because the operators were unnamed in that report, Tessera cannot be definitively identified as the subject. The report also noted 0x’s policy of cutting off data sources until execution deficiencies are resolved.
When an advertised price attracts an order but a different execution output is delivered, competition based purely on the advertised figure can reward the wrong venue. The key issue is whether routers accurately compare what a trader actually receives under the specific conditions of that transaction.
Jump argued that routing engines capable of selecting executable prices dynamically during transaction execution can substantially bridge the display-to-fill gap. From a design perspective, this suggests a maker could preserve inventory, freshness, and counterparty protections, provided routers evaluate outputs that already incorporate those factors.
Overview documentation for Jupiter’s Swap API details competition among routing engines alongside a mechanism that sidelines underperforming sources. Its integration guide also mandates quote-to-execution parity checks measured against identical pool snapshots.
While a parity check verifies agreement on a single snapshot, maintaining consistency through subsequent updates remains a separate challenge. Engine competition also leaves open the question of whether each individual venue is reassessed during an executing transaction.
Although comparing executable outputs provides a clear conceptual direction, its real-world effectiveness requires empirical measurement.
To achieve a meaningful comparison, the evaluated executable output must reflect identical trade sizes, callers, current pool states, and applicable fees. Failing this, a price accessible via one routing path might be mistakenly assumed available to another.
Jupiter’s documentation outlines a platform swap fee on its Meta-Aggregator path alongside zero fees on its Router path, noting that integrator fees and landing arrangements can also vary. A protocol-level spread cannot substitute for the final net amount received after factoring in all applicable charges.
Future evidence should compare quoted versus delivered outputs across matched transactions, clearly detail which costs are included, and demonstrate how routing systems handle persistently underperforming liquidity sources.
For passive liquidity providers, assessing returns requires a separate position-level evaluation. Ultimately, improved routing can optimize the swapper’s decisions while leaving the depositor’s investment analysis unaddressed.