In a constant-product pool, adding 5% of the input-side reserve moves the marginal price by about 10.25% before fees. For a team making regular payouts, that curve means a token’s quoted spot price is not the price you get for the whole transfer: each unit of output depletes the pool and makes the next unit more expensive.
Pool imbalance raises the marginal price
In a classic constant-product automated market maker (AMM), the two reserves are x and y, and the pool maintains x × y = k. If traders add token y to buy token x, the pool’s x reserve falls while its y reserve rises; the marginal price of x in y is the reserve ratio y/x, so it increases as the trade proceeds.
Picture a treasury buying foreign currency from a small cash desk: the first notes come at the posted rate, but each note removed leaves the desk with less to sell. An AMM encodes that worsening rate in its reserve curve rather than waiting for a dealer to update a quote. Each trade also updates the state that the next trader faces.
For operational due diligence, see four checks before a base swap; this article focuses on estimating execution cost as pool depth changes. BaseSwap is an AMM on Base, Coinbase’s Ethereum layer 2, and its official app is one venue for token swaps and liquidity pools. The exact pool design and fee are pool-specific, so check the relevant market rather than assuming every pool uses the same curve or rate.
Reserve share predicts the curve cost
For a constant-product pool, an exact-input swap of Δy into reserves x and y returns Δx = x × Δy / (y + Δy), before accounting for the swap fee. The spot quote starts at x/y, but the average execution price is worse because the output amount is calculated against the changing reserve ratio across the full trade.
For example, suppose a pool holds 1,000 WETH and 2,000,000 USDC, a spot ratio of 2,000 USDC per WETH. A 100,000 USDC trade is 5% of the USDC reserve. Ignoring fees, it returns about 47.62 WETH, or an average price of about 2,100 USDC per WETH; the post-trade marginal price is 2,205. That roughly 5% average-price impact is the curve cost, separate from fees and transaction costs.
To estimate a treasury flow, compare the input amount with the reserve on that side of the pair, then calculate output and average execution price. A displayed total-value figure can hide the relevant constraint: the depth in the direction of your trade. Repeated payouts in one direction keep taking from the same output reserve until new deposits or opposing trades rebalance it.
Fees, slippage and pool design add distinct costs
The swap fee is charged according to the pool’s rules and is usually a percentage of trade value; AMM fee tiers commonly range from 0.05% to 1%, with the actual tier set by the pool design. In a familiar v2-style example, a 0.30% fee is applied to the input before enforcing the curve, but that figure should not be assumed for BaseSwap or any particular pool.
Price impact is the cost caused by your trade moving the pool along its curve. Slippage is the additional change possible between quoting and execution because other transactions may land first; a minimum-output bound on an exact-input trade limits the worst accepted result, but too tight a bound can cause a revert. Add network execution cost separately, especially when comparing small frequent transactions with fewer larger ones.
Pool design changes how depletion appears. A concentrated-liquidity pool may offer deeper liquidity near the current price, then encounter a sharp loss of active liquidity at a tick boundary; trades crossing initialized ticks can also require more execution work. A quote that looks inexpensive at the starting price can therefore deteriorate quickly if the trade consumes liquidity across several ranges.
Choose trade size by total execution cost
Splitting a trade into several swaps against the same unchanged pool does not remove the constant-product curve cost: in a frictionless pool, the combined reserve movement is effectively the same. Splitting also adds transaction costs and exposure to price changes between executions. It helps when the pool replenishes, opposing flow arrives, or your schedule can avoid thin periods—not simply because each individual swap is smaller.
For a regular transfer program, compare the expected total cost at the required cadence with the cost of batching. Use the reserve balances and pool fee for each route, include network costs, and estimate the average execution price against an external reference price. The deciding measure is the total cost for the treasury’s required amount and timing, not the quoted spot rate or pool TVL alone.
Before committing a recurring flow, check:
- Output-side reserves and your trade as a share of input-side reserves.
- Pool fee tier and whether the route crosses concentrated-liquidity ticks.
- Average execution price against a trusted reference, including fees.
- Minimum output, network cost and whether batching can wait for replenishment.