The Death of the AMM: Why Intent-Based Systems Hollow Out Liquidity Pools

For years, the automated market maker was the backbone of on-chain liquidity. Today, a structural shift in how trades are routed is quietly eroding the revenue that made passive LP positions viable — and the implications for DEX revenue quality, intent-based DeFi liquidity, and protocol revenue sustainability on Solana are more significant than most participants realize.

From Price Discovery to Residual Counterparty: What Intent-Based Routing Actually Does

Traditional AMMs operate on a simple premise: liquidity providers deposit assets, traders swap against the pool, and LPs earn fees proportional to volume. The AMM is the price-discovery venue and the execution venue simultaneously.

Intent-based systems break this coupling. Under an intent-based model, a user broadcasts what they want (e.g., “sell 10,000 USDC for maximum SOL”) rather than how to execute it. A network of solvers — specialized off-chain agents — compete to fill that intent optimally. The winning solver may source liquidity from a private inventory, a request-for-quote (RFQ) system, a cross-chain bridge, or an aggregated route that touches an AMM pool only as a last resort.

The critical consequence: the most profitable, high-volume trades are captured by solvers before they ever reach an AMM pool. What remains in the pool is the flow that solvers declined — typically smaller, less profitable, or more complex trades that don’t fit clean solver inventory.

This is not a marginal effect. As intent-based infrastructure matures on Solana, passive LPs increasingly find themselves acting as residual counterparties rather than primary market makers. The fee revenue that once justified impermanent loss exposure is migrating to solver networks, not staying in the pool.

The Adverse Selection Trap: Why Remaining Flow Is Structurally Worse

The flow that solvers leave for AMM pools is not a random sample of all trading activity — it is adversely selected. Solvers rationally cherry-pick trades where they can offer price improvement and capture the spread. What they pass to AMMs tends to be:

• Informationally toxic flow: trades where the trader has a directional edge (e.g., reacting to a price feed update before the AMM oracle refreshes)
• Residual complexity: multi-hop routes that solvers cannot fill cleanly from inventory
• Low-margin volume: small trades where solver operational costs exceed potential profit

For passive LPs, this creates a structural deterioration in the quality of fee income. Volume metrics may remain stable or even grow — because aggregators still route through AMM pools when necessary — but the composition of that volume shifts toward trades that are more likely to leave LPs worse off on a net basis.

This is the core of the DEX revenue quality problem on Solana: headline volume figures can mask a fundamental degradation in LP economics. A pool processing $500M in monthly volume where the majority is solver-rejected residual flow is not equivalent to the same pool processing $500M of organic, balanced two-way flow.

Protocol Revenue Quality: The Metric DEX Dashboards Don’t Show

Protocol revenue quality — the sustainability and predictability of fee income relative to the risks borne by capital providers — is becoming the defining metric for evaluating DEX health on Solana. Yet most DEX dashboards surface only total fees collected and TVL, neither of which captures the adverse selection dynamic described above.

As covered in Protocol Revenue vs. Token Inflation: The Sustainable DeFi P&L, the distinction between genuine protocol revenue and inflation-subsidized yield is critical for long-term sustainability. The same logic applies here: fee revenue generated from adversely selected residual flow is structurally less sustainable than fee revenue from balanced, organic volume.

The practical implication for Solana DeFi participants: TVL in AMM pools is no longer a reliable proxy for LP profitability or protocol revenue quality. A pool can maintain high TVL while its LP base silently bleeds to impermanent loss and adverse selection — because LPs are slow to recognize that flow composition has changed even when volume numbers look healthy.

The Composability Pressure Point: LSTs in a Hollowed AMM Environment

Liquid staking tokens depend on AMM liquidity for one critical function: instant exit. When a JSOL holder wants to convert to SOL immediately rather than waiting for a delayed unstake settlement, the DEX swap route is the primary alternative. The depth and quality of JSOL/SOL liquidity in AMM pools directly affects the cost of that exit.

As intent-based systems hollow out AMM liquidity — concentrating the best flow in solver networks while leaving residual flow in pools — the economics for LPs providing JSOL/SOL liquidity deteriorate. Fewer high-quality trades flow through the pool, fee income per unit of TVL declines, and rational LPs reduce their positions. The result is thinner on-chain liquidity for LST pairs, wider spreads on instant exits, and higher effective costs for users who need immediate liquidity.

This is not a hypothetical: it is the structural direction that intent-based adoption creates. The question for any LST is whether its primary yield source is exposed to this dynamic — or insulated from it.

JSOL’s Structural Insulation: Why the Yield Source Matters

Here the architecture of JSOL becomes directly relevant. JSOL’s yield is generated by Solana validator staking rewards — a function of network consensus participation, not AMM fee income. SOL deposited into JPool is staked across a validator set managed by the on-chain Stake Pool Program; JPool never has access to user funds, and the yield accrual mechanism operates entirely at the protocol layer, independent of DEX activity.

This means JSOL’s core yield — the rate at which the JSOL↔SOL exchange rate appreciates — is structurally decoupled from AMM fee revenue dynamics. Whether solver networks capture a small or dominant share of Solana DEX volume, the staking reward mechanism that drives JSOL appreciation is unaffected.

The composability layer — using JSOL in AMM pools, as lending collateral, or in leveraged positions — does interact with DEX liquidity conditions. But the base yield is not contingent on LP fee income. This is a meaningful distinction when evaluating protocol revenue quality on Solana: JSOL’s yield source is natively backed by network consensus, not by the increasingly contested fee revenue of on-chain liquidity pools.

JPool’s reserve liquidity mechanism — maintaining a minimum reserve to support instant unstake — provides an additional buffer: users have a protocol-native exit path that does not require routing through an AMM pool at all, partially insulating the instant-exit experience from AMM liquidity deterioration.

What This Means for Solana DeFi Participants

The transition from AMM-centric to intent-based DeFi liquidity is not a temporary market condition — it reflects a structural preference for execution quality over passive market-making. For participants evaluating assets and protocols on Solana, three implications follow:

1. Evaluate yield source, not just yield rate. An LST or DeFi token whose yield depends on AMM fee income is exposed to the adverse selection dynamic described above. Yield sourced from network consensus (staking rewards) is structurally more predictable and insulated from solver-driven flow migration.
2. Distinguish volume from revenue quality. High DEX volume on Solana does not guarantee LP profitability. As noted in Geo-Blocking at the RPC Layer: Solana’s Hidden Infrastructure Chokepoint, infrastructure-layer dynamics can concentrate activity in ways that aggregate metrics obscure. The same principle applies to flow quality in AMM pools: the number that matters is not total volume, but the composition of that volume.
3. Composability depth requires liquidity depth. As AMM pools thin out for LST pairs, the practical utility of LST composability — particularly instant exits and LP provision — becomes more dependent on the protocol’s native liquidity mechanisms. Protocols with robust on-chain reserve systems are better positioned than those relying entirely on external AMM depth.

The death of the AMM as the primary execution venue is not the death of on-chain liquidity. But it does restructure who captures DEX revenue, who bears the residual risk, and which assets are positioned to maintain composability as the market evolves. For a deeper look at how hidden infrastructure dependencies can create analogous structural vulnerabilities, see Canonical Bridge Design: Security Models After Wormhole and Ronin.

Explore JPool’s liquid staking infrastructure and JSOL token at jpool.one.