Learn / Risk & Position Sizing
Slippage and market impact
Slippage is the difference between the price you expected and the price you received. Unlike fees, it is not quoted anywhere, does not appear on a schedule, and is worst in exactly the conditions where it costs you most.
It is the third cost between gross and net — after fees and funding — and the only one no calculator can give you in advance.
Where it comes from
Crossing the spread. A market order buys at the ask and sells at the bid. The spread is the immediate cost, and it widens when liquidity thins.
Walking the book. Your order consumes the best price, then the next, then the next. A large order fills at a volume-weighted average worse than the top of book. This is market impact, and it scales with size relative to depth rather than with size alone.
Latency. Between deciding and arriving, the book moves. In a fast market it moves a lot.
Gaps. Price jumps past your level without trading there. Stops do not execute at the stop price; they execute at the first available price after it.
Why stop-outs are the worst case
This is the part worth internalising, because it compounds with everything else.
A stop triggers when price moves against you, which is disproportionately during fast, one-directional moves. In those moves:
- Spreads widen.
- Depth on your side thins — everyone wants the same exit.
- Other stops trigger, adding to the same flow.
So the moment your stop fires is the moment the book is least able to absorb it at a decent price. Slippage is not evenly distributed across your trades. It concentrates on your losers.
The practical consequence: a −1R planned loss is routinely a −1.1R or −1.3R realised loss, while wins rarely overshoot in the same proportion. That asymmetry quietly degrades your expectancy in a way that neither your fee schedule nor your R multiples will show unless you record both planned and realised.
Size, depth, and the compounding problem
Impact is about your order relative to the book, not your order in absolute terms. The same size is negligible in a deep market and expensive in a thin one.
Which is why this interacts badly with tight stops. As covered in why tight stops can cost more, a tighter stop means a proportionally larger position for the same risk. So the tight-stop configuration produces:
- A larger order, taking more of the book.
- Triggered more often, because the stop sits inside ordinary noise.
- Exiting into the thinnest conditions, because that is when stops fire.
Three multiplicative effects, all pointing the same way, none of them visible in the sizing formula.
Reducing it
Use limit orders where time is not critical. This removes spread-crossing and moves you to maker fees — at the cost of adverse selection, which is a real trade-off rather than a free win.
Split large orders. Working a position in over time reduces instantaneous impact, at the cost of price movement during the fill.
Trade liquid instruments and liquid hours. Thin books and thin sessions are where impact lives, and the difference is often larger than any fee optimisation.
Do not use limit orders for stops. A stop that becomes a resting limit order in a gap is a stop that does not execute, and you remain in a position you decided to leave. Pay the slippage.
Size for the book, not just for the account. If your position is a meaningful fraction of visible depth, your exit is going to be expensive regardless of what your risk calculation said.
Measuring it
Slippage is the one cost you have to measure because nobody quotes it.
Record, per fill: the price you intended and the price you got. The difference, aggregated, tells you:
- Average slippage on entries versus exits — expect exits to be worse.
- Average slippage on stop-outs specifically — expect these to be worst.
- Whether it scales with your size, which tells you whether you have outgrown the instrument.
- Whether it clusters by time of day.
This is the same record that makes realised-versus-planned R computable, which is a strong argument for keeping it — see logging tool calls for a trading audit trail. Without it, slippage is invisible, and invisible costs do not get managed.
Where it fits with the other costs
| Cost | Scales with | Quoted in advance? |
|---|---|---|
| Fees | Turnover × notional | Yes — your venue’s schedule |
| Funding | Holding time × notional | Yes — the current rate |
| Slippage | Size relative to depth | No |
Fees and funding can be computed before the trade using the fee and funding calculators. Slippage cannot, which is why it is the one people leave out of the plan and discover in the results.
FAQ
What causes slippage?
Four things, usually in combination: crossing the bid-ask spread, consuming multiple levels of the order book on a large order, price moving between your decision and your order arriving, and price gapping past your level without trading there. The last is why stop orders fill below the stop price rather than at it.
Why is slippage worse on stop-losses?
Because stops trigger during fast moves against you, which is when spreads are widest and depth on your side is thinnest — and when other stops are triggering into the same flow. The result is that slippage concentrates on losing trades, making realised losses systematically larger than planned ones while wins do not overshoot to the same degree.
Can I avoid slippage with limit orders?
On entries, largely yes, and you get the lower maker fee too. The cost is adverse selection: a resting order fills when price comes to your level and continues through it, which selects for the trades moving against you. On exits and stops, limit orders are a bad idea — an unfilled stop leaves you in a position you decided to leave.
How much slippage should I expect?
There is no general number, because it depends on the instrument, the session, your size relative to depth, and the conditions at the moment you trade. That makes it the one cost you have to measure rather than look up. Record intended versus filled price on every order and the distribution will tell you more than any estimate.