Slippage, Spread and Price Impact Are Three Different Costs

You execute a trade and receive less than the screen suggested. The gap has at least three components, and calling all of them slippage guarantees you cannot tell which venue is overcharging you, which order was placed badly, and which part of the cost was never avoidable.

Spread: the cost of immediacy

The spread is the gap between the best bid and the best offer. Buying at the offer and selling at the bid means paying it, and it is the price of demanding immediate execution rather than waiting.

It is a property of the venue and the asset rather than of your order. It compensates market makers for standing ready to trade with anyone, including people who know something they do not, and it widens with volatility and with how hard the asset is to hedge.

You can avoid it by posting rather than taking — placing a limit order and waiting for someone to come to you. That converts a certain cost into an uncertain wait, which is the trade the spread exists to price.

Price impact: the cost of size

Price impact is what your own order does to the price. Buying more than is available at the best offer means consuming that level and moving to the next, and the average price you achieve is worse than the price you saw.

This is deterministic and computable in advance from the order book or the pool formula. It is not a surprise unless you did not look, and it is a function of your size relative to available depth rather than of anyone's behaviour.

The remedy is splitting the order across time or venues, which reduces impact and introduces exposure to the price moving while you work. Every execution algorithm ever written is a way of trading one against the other.

Slippage: the cost of the delay

Slippage proper is the difference between the price when you decided and the price when you executed. It comes from time passing: the market moved, or someone else traded ahead of you.

Unlike the other two, it is genuinely uncertain and can go either way. It grows with how long the gap is between decision and execution — which on-chain means block times and congestion, and which on a centralised venue means latency.

The slippage tolerance setting in a wallet is not a cost setting. It is a limit on how much adverse movement you will accept before the transaction reverts, and setting it high does not make execution better; it makes you willing to accept worse.

The fourth cost, on-chain

There is a fourth component specific to public transaction ordering, and it hides inside what people call slippage.

A pending trade is visible before it executes. Someone can trade ahead of it, let your trade move the price, and trade back — profiting from the difference. Your fill is worse, and the loss looks exactly like ordinary slippage.

The distinguishing feature is that it scales with your declared tolerance. A wide tolerance tells everyone how much room they have to work with; a tight one limits it and risks the transaction failing. That trade-off is a real cost of transacting in public, and it is why private transaction routing exists.

Measuring the whole thing

The number that matters is implementation shortfall: the difference between the price at the moment you decided and the average price you actually achieved, including fees. It captures all four components and is the only figure comparable across venues.

Measuring it requires recording the decision price, which almost nobody does, which is why most traders have no idea what their execution costs. Recording it for a month is the cheapest available diagnostic.

• Record the price when you decide, not when you submit. Everything else is measured from there.
• Compute price impact before trading — it is deterministic on both venue types and requires no guesswork.
• Treat slippage tolerance as a limit on acceptable loss, not as a setting that improves execution.
• Compare venues on total shortfall including fees, since a venue with tight spreads and high fees is not cheap.
• If losses scale with your tolerance setting, the cost is ordering-related rather than market movement.