Trading & AMMs

Liquidity Aggregation

Combining liquidity from multiple sources for better trading execution

Definition

Liquidity aggregation means splitting one trade across several venues and pools so that the total execution cost is lower than any single route. It works because price impact is convex: three partial fills of a third each cost meaningfully less than one full fill in the deepest pool. Aggregators such as 1inch, Paraswap, Odos and LlamaSwap compute these routes across dozens of DEXes, while CoW Swap goes further by batching orders into auctions where independent solvers compete to fill them, matching users directly against each other when a coincidence of wants exists and settling the remainder on-chain. The cost is complexity: a split route touches more contracts, costs more gas, and the improvement has to survive that overhead to be worth taking.

Instead of dumping your whole trade into one pool, an aggregator chops it up across several so you move each pool's price less and get a better average fill.

Example

Swapping 500 ETH into USDC might route 60% through a Uniswap v3 0.05% pool, 25% through Curve's tricrypto pool and 15% through Balancer, producing something like 0.35% total price impact instead of the roughly 0.9% the deepest single pool would have charged. On $1.5M of notional, that is about $8,000 saved for maybe $30 of extra gas. Now do the same for a $200 swap: the split route saves 0.2%, which is $0.40, while the extra hops cost several dollars in gas. The aggregator's quote was better and the trade was worse.

How it works

A router searches paths across DEXes, splits the order across them, and settles in one transaction. Intent-based aggregators go further and let competing solvers bid to fill your order, including by matching it against other users.

Why it matters

For any sizeable swap, routing is the difference between losing 0.3% and losing 1% or more, and that gap compounds fast for anyone trading regularly.

What to check

Compare quotes net of gas, not on headline rate. Watch the number of hops and the approvals a route requires, set a sensible slippage tolerance, and consider an intent-based venue when trade size makes sandwiching likely.

Risks to Consider

Multi-hop routes touch more contracts, each of which is an approval and an exploit surface
Quotes decay: the route was optimal against the reserves at quote time, and reverts or worse fills follow if the pools move
Some aggregators route through pools or order flow arrangements they benefit from, so 'best price' deserves a sanity check against a second quote
Headline savings are often quoted before gas, which inverts the conclusion entirely for small trades

Common Questions

Is an aggregator always the better choice?

For any trade large enough that price impact dominates gas, yes. For small trades on an expensive chain, no: a direct swap in the deepest pool is often cheaper once gas is included, because the aggregator's edge is a percentage of a small number while its overhead is fixed. The crossover on Ethereum mainnet is usually somewhere in the low thousands of dollars, and much lower on L2s.

Why do two aggregators quote different prices for the same swap?

Different venue coverage, different splitting algorithms, different gas assumptions, and different treatment of their own order flow. They also quote at slightly different moments, and on a volatile pair that alone explains a few basis points. For a large trade it is worth pulling two or three quotes and comparing the net amount received after gas rather than the headline rate.

What about MEV protection?

Routing and MEV protection are separate problems that some aggregators solve together. Splitting a trade reduces price impact but does not stop a sandwich. Intent-based venues such as CoW Swap or 1inch Fusion move settlement into a batch auction where a solver bears the execution risk, which removes the ordinary sandwich vector. If you are trading size on a volatile pair, that protection is frequently worth more than the last few basis points of routing.