AMMs are execution primitives, not endpoints. Their core innovation—permissionless liquidity pools—solved bootstrapping but created a fragmented, inefficient market. Users now interact with intent-based solvers like UniswapX and CowSwap, which treat AMMs as just another liquidity source.
Why AMMs Must Evolve or Be Abstracted Into Irrelevance
The static constant product AMM is being reduced to a commodity liquidity source. This analysis explores the rise of intent-based solvers, the need for dynamic curves, and the existential pressure on legacy DEX designs.
Introduction
The current AMM model is a liquidity bottleneck that will be abstracted away by superior execution layers.
The value is shifting from liquidity provision to execution intelligence. Protocols like Across and 1inch Fusion demonstrate that finding the best price across venues is more valuable than owning a single pool. This abstracts the AMM into a commodity.
Evidence: Over 60% of DEX volume on Ethereum now routes through aggregators. The AMM's role is being reduced to a settlement layer for smarter systems.
The Core Argument
The current AMM model is a bottleneck for capital efficiency and user experience, creating an architectural opening for intent-based systems.
AMMs are inefficient price finders. They require constant liquidity to facilitate trades, locking capital in static pools while intent-based solvers like those in UniswapX and CowSwap source liquidity dynamically across venues.
The user experience is adversarial. Users battle MEV and slippage with each swap. Intent abstraction inverts this model: users declare a desired outcome, and a competitive network of solvers fulfills it optimally.
Evidence: Over 70% of swap volume on CowSwap is settled via off-chain solvers, demonstrating demand for this model. Protocols like Across use intents to unify bridging and swapping into a single declarative transaction.
The Three Forces Commoditizing AMMs
AMMs are being unbundled into specialized, commoditized layers. The generic v2/v3 pool is now a low-margin utility.
The Problem: Intent-Based Abstraction
Users no longer care about the pool. Protocols like UniswapX, CowSwap, and Across treat AMMs as just another liquidity source in a solver network. The winning DEX is the one with the best routing, not the deepest pool.
- Key Benefit: Users get better prices via MEV capture and gasless orders.
- Key Benefit: Liquidity becomes a commodity, with solvers competing to fill intents from the cheapest source.
The Problem: Modular Liquidity Layers
Liquidity is being separated from execution. Morpho Blue and Aerodrome separate risk parameters and incentives from core swap logic. This creates a market for capital efficiency where AMMs are just one permissionless plugin.
- Key Benefit: Capital efficiency can reach 95%+ LTV for lending, far beyond AMM LP.
- Key Benefit: Innovation shifts to risk modules and incentive flywheels, not the AMM curve.
The Problem: Cross-Chain Native Assets
Native yield-bearing assets (e.g., stETH, weETH) and omnichain tokens (via LayerZero, Wormhole) fragment liquidity. A single-chain pool is obsolete. The new battleground is canonical bridging and yield aggregation across chains.
- Key Benefit: Users demand native yield across chains, not wrapped derivatives.
- Key Benefit: Liquidity follows the canonical asset, forcing AMMs to become cross-chain routers or die.
AMM vs. Intent-Based Volume Share (Hypothetical Snapshot)
A first-principles comparison of incumbent Automated Market Maker (AMM) architecture versus emerging intent-based systems, quantifying the trade-offs in capital, user experience, and composability.
| Core Architectural Metric | Classic AMM (Uniswap V3) | Hybrid Solver (CowSwap) | Fully Expressive Intent (UniswapX, Across) |
|---|---|---|---|
Capital Efficiency (Utilization) | ~20-40% (idle LP capital) | ~60-80% (batch auctions) | ~95%+ (solver competition) |
User Execution Cost (vs. Quote) | +10-50 bps (slippage + fee) | -5 to +5 bps (surplus) | -10 to +30 bps (MEV capture) |
Settlement Finality Time | 1 block (12 sec on Ethereum) | ~5-30 min (batch window) | 1 block to 30 min (flexible) |
Cross-Chain Native | |||
Composability (Multi-Step Logic) | |||
Required User Knowledge | Pool selection, slippage, gas | None (quote & sign) | None (declare outcome) |
Liquidity Source Fragmentation | |||
Protocol Fee Take Rate | 0.05% - 1.0% | ~0.0% (surplus-based) | 0.0% - 0.1% (success fee) |
The Path to Relevance: Dynamic Curves & LP Protection
Static AMMs are being commoditized by intent-based architectures, forcing a shift to dynamic liquidity management or obsolescence.
Static AMMs are commodities. The core constant product formula is a solved problem, offering zero defensibility. Protocols like Uniswap V3 and Curve have already optimized for capital efficiency, leaving only marginal gains on the table. The next frontier is not the curve itself, but the system managing it.
Intent-based architectures abstract liquidity. Solvers for UniswapX and CowSwap treat on-chain pools as just another liquidity source, not a destination. This commoditizes the AMM, reducing LPs to passive yield farmers while solvers capture the value of order flow and cross-domain execution.
Survival requires dynamic parameterization. AMMs must evolve into reactive systems. This means curves that auto-adjust fees and curvature based on volatility, akin to dynamic automated market makers (dAMMs), or integrate proactive LP protection like MEV-aware vaults. The goal is to make the pool the optimal venue, not just a fallback.
Evidence: The rise of solver dominance. Over 70% of CowSwap's volume is settled via on-chain AMMs, but the economic surplus accrues to solvers, not LPs. This demonstrates the abstraction layer's power and the urgent need for AMMs to capture more value from their own liquidity.
Steelman: The Bull Case for the Dumb Pipe
AMMs are becoming commoditized liquidity infrastructure, and their value will be captured by superior intent-based execution layers.
AMMs are commoditized infrastructure. The core innovation—constant product formulas, permissionless pools—is now a solved problem. The differentiated value has shifted from the liquidity source to the execution layer that can best route and settle user intents.
Intent-based architectures abstract AMMs. Protocols like UniswapX and CowSwap treat the AMM as a dumb liquidity pipe. Their solvers compete to find the best execution path across AMMs, RFQ systems, and private liquidity, rendering the specific AMM irrelevant to the user.
Liquidity follows volume, not protocol. The flywheel of intent attracts users with better prices and UX, which attracts more solvers and private liquidity, which further improves execution. This creates a winner-take-most market for the aggregation layer, not the underlying AMM.
Evidence: UniswapX now routes over 30% of Uniswap's volume, often settling trades off-chain or on alternative AMMs. This proves the value accrual is shifting from the AMM's LP fees to the solver network and the protocol capturing the user's intent.
TL;DR for Protocol Architects
The constant-function AMM is a foundational primitive, but its rigid, on-chain execution model is being outflanked by more efficient and user-centric architectures.
The Problem: Lazy, Expensive Liquidity
Passive, generalized LPs in pools like Uniswap V3 suffer from massive adverse selection and impermanent loss, requiring unsustainable >100% APY to compete with just holding. This creates a fragile, mercenary capital base.
- Inefficient Capital: >90% of a Uniswap V3 LP's capital sits idle, unused for trading.
- Negative Sum Game: MEV bots and arbitrageurs extract value, making retail LPs net losers.
The Solution: Intents & Solver Networks
Separate expression of user preference (the intent) from execution. Let off-chain solvers (CowSwap, UniswapX) compete to find optimal routing across all liquidity venues, including private order flow and AMM pools.
- Better Prices: Solvers internalize MEV for user benefit, finding paths AMMs can't see.
- Gasless UX: Users sign a message, solvers pay gas and handle complexity.
The Problem: Fragmented, Inefficient Execution
AMMs force execution onto a single chain/layer, ignoring better prices or liquidity elsewhere. Cross-chain swaps via canonical bridges and AMMs are slow (10+ mins) and capital inefficient, locking $10B+ in bridge liquidity.
- Siloed Liquidity: LPs must manually deploy and manage capital across dozens of chains.
- Slow Settlement: Users wait for block confirmations and bridge finality.
The Solution: Universal Liquidity Layers
Abstract liquidity into a verifiable intent layer. Protocols like Across and Circle's CCTP use fast, optimistic messaging with on-chain attestation. LayerZero and CCIP enable generalized cross-chain state.
- Unified Liquidity: A single liquidity position can service trades across any connected chain.
- Near-Instant Finality: Users get assets on destination chain in ~1-2 mins.
The Problem: Inflexible, One-Size-Fits-All Logic
AMM logic is hardcoded (x*y=k). It cannot adapt to market regimes, offer advanced order types (TWAP, limit), or incorporate real-world data without complex, expensive peripheral contracts.
- Static Pricing: Vulnerable to flash crashes and manipulation in low-liquidity pools.
- No Composability: Difficult to integrate with lending or options protocols without trust assumptions.
The Solution: Modular AMM Cores & App-Chains
Decompose the AMM into specialized, upgradeable components: a settlement layer, a risk engine, and a liquidity manager. App-chains like dYdX v4 and Injective demonstrate the performance gains.
- Tailored Logic: Optimize fee models, oracle integration, and MEV resistance for specific assets.
- Sovereign Execution: Achieve ~10k TPS and sub-second block times for derivatives or spot trading.
Get In Touch
today.
Our experts will offer a free quote and a 30min call to discuss your project.