Solana study finds wider spreads outside aggregators at five of six sampled pools

A recent Solana study has highlighted differences in trading costs depending on how orders are routed. Researchers Ozan Solmaz, Lioba Heimbach, and Jason Milionis examined proprietary automated market makers (propAMMs) on the Solana blockchain, finding that realized round-trip spreads were generally wider for trades routed outside of aggregators at five out of six major liquidity venues. This research, which analyzed data from September 2025 through August 2026, suggests that accessing the same liquidity pool does not necessarily guarantee identical trading terms for all users.

Understanding the Solana study

An aggregator, such as Jupiter, acts as a routing provider that searches across various liquidity venues and can split orders to optimize execution. In contrast, a propAMM is a venue where a single operator provides inventory and can update prices directly. The researchers focused on SOL/USDC pairs, comparing flow from aggregators against non-aggregator traffic. By matching consecutive fills on opposite sides of a pool within specific time and size constraints, the team controlled for variables to better isolate the impact of the routing class on realized spreads.

For five of the six venues studied, non-aggregator spreads were wider in 68% to 83% of matched cells. The study notes that these findings represent a historical association between routing methods and realized spreads rather than a definitive measurement of current interface performance. The economic rationale for these differences may stem from how operators manage stale prices; by updating prices before incoming trades, operators can reduce their exposure to losses caused by market movements.

Trades routed outside of aggregators faced wider realized round-trip spreads at five of six Solana venues.

It is important to distinguish these round-trip price gaps from individual swap fees or wallet charges. Wallet-level fees, such as those disclosed by platforms like Phantom, are separate from the underlying pool execution costs. Furthermore, the study cautions that its findings do not constitute a randomized test of what occurs when a single user switches interfaces. The research also highlights that proprietary venues can potentially incorporate caller identity into execution pricing, as observed in a separate analysis of a Base implementation.

Why it matters

This research clarifies that the route taken to execute a trade can significantly impact the final price received, independent of the liquidity pool itself. For investors, it underscores that simply selecting a pool is not the only factor in achieving optimal execution. These findings demonstrate that routing providers play a critical role in price discovery and cost management, which is essential for users navigating the decentralized finance ecosystem.