Set slippage from the swap legs’ execution risk, then check that the quoted minimum output still meets your needs; the right tolerance depends on liquidity and how long each leg waits to execute. A cross-chain quote can include a swap on the source chain, a transfer or settlement step, and a swap on the destination chain. Treating that whole route as one swap hides where the price can move.
1. Which parts of the route can reprice?
Map each exchange separately: a source-side swap can change the amount sent, and a destination-side swap can change the amount received. Some routes also involve a liquidity provider advancing destination assets before settlement; in that case, the provider’s pricing and settlement terms affect the quote, but they are not the same thing as AMM slippage.
For a like-for-like comparison, use the destination amount the quote promises after its stated fees, and identify whether that amount is guaranteed, conditional, or indicative. A crypto relay bridge can route a transfer and swap in one flow, which is useful when comparing the asset you start with against the asset you need to receive. The convenience of a single route does not mean every leg shares one slippage limit.
The usual mistake is to compare the source-chain amount or exchange rate while ignoring the destination minimum. Fix it by recording the input, expected output, minimum output, and any quoted fees for the complete route. A relay bridge can bundle these operations, but you still need to know what happens if a later swap cannot meet its output condition.
2. What does the slippage percentage actually bound?
For an exact-input swap, a common minimum-output rule is minOut = quotedOut × (1 − tolerance). If a quote is 1,000 units and the tolerance is 0.5%, the swap must return at least 995 units, before considering how the route displays or deducts fees. The tolerance is a bound on an execution result; it is not a prediction that the market will move by that amount.
AMM price impact is different: it is the effect your trade has on the pool’s price as it consumes available liquidity. Slippage is the difference between the expected and actual execution, which can arise because the pool changes between quote and execution. Uniswap’s protocol documentation describes the relationship between trade size, reserves, and price impact in constant-product pools; the same distinction helps when evaluating a routed swap.
Where both source and destination legs have independent output bounds, the total loss relative to quoted outputs can compound. For example, two legs each allowing 0.5% below their own quote can produce roughly 1% less than the combined quoted result, before fees and any other route adjustments. That calculation is illustrative: some routes expose one aggregate bound, and others execute only one swap.
3. How should you choose a tolerance?
Use the tightest limit that gives the route a realistic chance to execute under the conditions it faces. For a deep, stable pool with a prompt execution, a tolerance around 0.1%–0.5% may be workable; thinner pools, volatile assets, or delayed destination execution may need more. These are starting points, not safe defaults: pool depth, trade size, volatility, and route design decide the actual risk.
Check the implied minimum in units of the destination asset, not only as a percentage. If the quote promises 2.00 ETH-equivalent units at a 1% tolerance, the minimum is 1.98; ask whether receiving less than that still makes the transaction worthwhile. If the displayed minimum is already unacceptable, raising tolerance only permits a worse fill.
Do not widen the bound just to get past a failing quote without finding the cause. A low tolerance can cause a swap to revert when the price moves beyond its limit; a high tolerance can permit execution at a materially worse rate. The useful adjustment is the smallest increase that reflects actual execution uncertainty, followed by a fresh check of the minimum output.
4. What changes when execution spans chains?
Time and failure handling matter because the destination leg may run after the source transaction is included and the transfer or settlement step completes. The destination pool can change during that interval, so a quote that was executable when requested may no longer satisfy its minimum. Whether the user receives a refund, can retry, or must take another recovery path depends on the route’s design; check that outcome before committing value.
For example, suppose the source swap executes at its quoted minimum, but destination liquidity worsens before the second swap. If the destination leg enforces its own minimum, it may fail even though the source transaction succeeded. A route that offers a fallback asset or a recoverable destination balance handles this edge case differently from one that requires a manual recovery. Compare those outcomes alongside the quoted rate.
Before acting, ask yourself: if the route completes at its minimum output—or the destination leg fails after the source transaction—will I still receive an acceptable result, and do I understand how the route handles that failure?
Top comments (0)