Slippage and market impact
The quote said 40,000. The fill came back at 40,036 average. Two entirely different things could have produced that, and they have different causes, different sizes, and different behaviour as your order grows.
The words get used interchangeably, which makes the whole subject muddier than it needs to be. They aren’t interchangeable.
The two definitions
Slippage is the difference between the price you expected and the price you got. It’s an outcome measure and it’s agnostic about cause. The market moving between your decision and your fill is slippage. Your own order consuming the book is also slippage.
Market impact is specifically the price movement caused by your own order. It’s a component of slippage, not a synonym for it, and it’s the component that scales with your size.
The distinction matters because they respond to different things. Impact is a function of your order size against available depth, and it’s the same whether markets are calm or violent. The rest of slippage is a function of how fast the market is moving and how long your order takes to arrive, and it’s roughly independent of your size.
Where impact comes from
Directly from the structure of the book. An order larger than the quantity at the touch consumes that level and continues to the next, at a worse price, and so on until filled.
Illustrative, with synthetic figures. Suppose the ask side is:
40,004 × 1.2
40,010 × 2.5
40,025 × 5.0
40,060 × 8.0A market buy for 1.0 fills entirely at 40,004. Impact: none — the touch is unchanged in the sense that the level still has quantity behind your fill.
A market buy for 5.0 takes 1.2 at 40,004, 2.5 at 40,010, and 1.3 at 40,025. Total cost is 1.2 × 40,004 + 2.5 × 40,010 + 1.3 × 40,025, giving an average of about 40,013.6. That’s roughly 9.6 above the touch, and every bit of it is impact.
A market buy for 15.0 exhausts all four visible levels and continues into whatever is above.
Notice the shape: impact per unit increases with size, because you’re consuming progressively worse prices. Doubling your order more than doubles your impact cost. This is the single most important property of the subject and it’s why execution behaves qualitatively differently at different sizes.
Temporary and permanent
A refinement worth having, because they behave differently afterwards.
Temporary impact is the price moving because you consumed depth, followed by the book refilling as other participants post new orders. The displacement fades.
Permanent impact is the price staying moved, because your order conveyed information that other participants updated on.
The practical consequence: a large order executed all at once pays the full walk up the book. The same order executed in pieces over time pays less impact per piece and lets the book refill in between, at the cost of exposure to the market moving during that time — which is the other component of slippage.
That’s a genuine trade-off with no free side, and how anyone resolves it is a strategy question this site doesn’t address.
The other components
Impact is one source. The rest of slippage comes from several places, and they add.
Latency. Between your decision and your order arriving at the matching engine there is a delay. The book can change in that window: levels consumed, orders cancelled, the touch moved. Nothing about this is avoidable; it can only be made smaller.
Stale quotes. The price you’re looking at is a snapshot, and depending on how it reached your screen it may already be old. This is more pronounced on aggregated or delayed displays than on a direct feed.
Spread. If you were quoting the mid-price to yourself and then crossed, half the spread is immediately “slippage” — but it isn’t really; it’s just the cost of crossing, and it was always going to be there. Measuring slippage against the mid rather than against the relevant touch inflates the number and obscures where the cost came from.
Fees, which aren’t slippage at all but end up in the same mental bucket when people compare expected against realised cost. They’re separate and worth separating.
The mechanism
THE MECHANISM — where the gap comes from
· Order smaller than the touch quantity
→ fills at the touch. No impact.
· Order larger than the touch quantity
→ walks outward. Average fill
degrades with size, faster than
linearly.
· Same order, thinner book
→ same size, more impact. Impact
is size against depth, not size
alone.
· Delay between decision and arrival
→ slippage independent of your
size; scales with how fast the
market is moving.
· Splitting an order over time
→ less impact per piece, more
exposure to market movement.
NO FREE SIDE.
· Displayed depth
→ a lower bound only. Hidden and
iceberg orders mean real depth
can exceed it.
· Protective caps and slippage tolerances
→ VENUE-SPECIFIC. Some cap how far
a market order may walk; some
don't.Measuring it honestly
If you want to know what execution actually cost, the comparison has to be against a defined reference, and the choice of reference changes the answer.
Against the touch on your side at decision time measures what you gave up by not being able to trade instantly at the best available price. This is usually the most informative for a taker.
Against the mid includes half the spread, which was never available to you as a taker. It makes crossing look worse than it is, though it’s the right reference if you’re comparing crossing against resting.
Against the last trade is the weakest, because the last trade may have been on either side and at any time.
The one thing worth avoiding is comparing against the best price printed anywhere during the interval. That’s a price you could only have got by knowing the future, and measuring against it produces a number that’s always bad and never actionable.
The failure mode
Slippage is not an error, a fee, or something a venue is doing to you. It’s the difference between a displayed number that describes one unit of available liquidity and an order that wanted more than one unit, plus the ordinary fact that time passes.
It cannot be eliminated. A limit order removes the price uncertainty and replaces it with execution uncertainty; splitting an order reduces impact and increases exposure to movement; trading smaller reduces both and gets you a smaller position. Every route trades one uncertainty for another, and the useful thing is knowing which one you’ve chosen rather than expecting a route with neither.