Trade when the book is deep
The same order costs less in the middle of the session than overnight. If your strategy is indifferent to the hour, it should not be.
Liquidity is a scheduleExecution
Slippage is not bad luck.
It gets treated as noise — a tick here or there that averages out. It does not average out, it is not symmetric, and on a strategy averaging thirty dollars a trade it is frequently the whole difference between profitable and not. It is also largely predictable, which means it is partly controllable.
What it actually is
Slippage is the gap between the price you expected and the price you got. You saw 5812.25, your order filled at 5812.50, and the quarter point is slippage.
It happens because a price on your screen is a snapshot of a queue, and by the time your order arrives the queue has changed. Someone else took the contracts that were resting there, or the market moved in the interval, or there were never as many contracts at that price as the display implied.
The important property: it is not symmetric. A market order is filled by whatever is available, which means you get the good fills when nobody wanted the other side and the bad fills when everyone did. The bad fills cluster in exactly the moments you most wanted to trade.
Where it comes from
Four causes, and knowing which one you are experiencing tells you whether it is fixable.
- Thin liquidity
- Fewer contracts resting at each price, so a normal order eats through several levels. Worst in overnight hours, in the minutes around a session boundary, and always worse on micros than on their full-size equivalent relative to size.
- Speed of the move
- The market is moving faster than the round trip to the exchange. Around a scheduled release the price can be several ticks away by the time your order is acknowledged — and that is not a broker problem, it is distance and time.
- Order type
- A market order accepts whatever it finds. A stop becomes a market order when it triggers, which is why stops slip most in the conditions you set them for. Limits do not slip at all — they just do not fill, which is a different cost.
- Your own size
- If you are large relative to what is resting, you move through the book yourself. Ten micros filled one at a time is a very different experience from ten E-minis at once.
The stop-loss paradox
The single most expensive interaction, and the one people build strategies around without noticing.
A stop order rests inert until price reaches it, then becomes a market order. So it converts to a market order at precisely the moment the market is moving quickly through that price — which is the moment liquidity is thinnest and slippage is worst.
This is why measured slippage on stops is routinely several times slippage on entries, and why a backtest that assumes stops fill at the stop price is not describing anything real. It is also why a strategy with a very tight stop can test beautifully and lose money live: the tighter the stop, the more often it is hit, and every hit pays the worst slippage in the strategy.
None of that means do not use stops. It means the cost of the stop belongs in the arithmetic before you decide the strategy works, not after.
Stop ordersConvert to market exactly when the market is fastest and thinnest.Worst
Scheduled releasesPrice can be several ticks away by the time an order is acknowledged.Worst
Session openWide spreads and a shallow book while the day finds a price.Bad
Overnight hoursThin book. A size that is nothing at midday walks the levels at 2am.Bad
Resting limitsDo not slip. They fail to fill instead, which is a different cost.None
What it costs, concretely
Take a strategy averaging $30 a trade on MES, trading ten contracts, with a stop that gets hit 40% of the time.
One tick of entry slippage is $1.25 per contract — $12.50 on ten. One tick on the exit is another $12.50. That is $25 a round trip before commission, on an edge of $30 a trade, and this is the optimistic version because it assumes one tick on the stop.
Add commission at, say, $0.50 per contract per side and you have another $10. The $30 edge is now $-5.
This arithmetic is why so many strategies are profitable in testing and not in practice, and it is entirely mechanical. Nothing about it requires the market to behave unexpectedly.
Run your own numbers through the expectancy calculator with and without costs. If the strategy only survives at zero cost, you have found that out for free.
What you can actually do about it
Some of it is controllable and some of it is the price of participating.
Size to the book, not the account
Ten micros often slip less than one E-mini, because they can be filled across levels rather than needing depth at one.
Micros are not just smallerModel it before you trust the test
Half a tick each way as a floor, more if you use stops or trade the open. A test without costs is a description of a market that does not exist.
Per side, not per tradeWhere copying does and does not add slippage
A fair question if you mirror one account onto others: does the follower get a worse fill than the leader?
Slightly, and for a reason that has nothing to do with the copier. The follower's order cannot be placed until the leader's fill has been reported, so it arrives some tens of milliseconds later. In a fast market that is occasionally a tick. In an ordinary one it is nothing.
What matters is that the delay is your broker's round trip in each direction, not the platform's arithmetic. Quanify's own contribution to that gap is microseconds — under a tenth of a percent of the round trip — and no copier can do better, because none can place an order before the broker has confirmed the fill it is copying.
The larger effect is the one people miss: costs multiply across accounts. Four followers pay four sets of commission and four sets of slippage on a strategy that was marginal on one. That is worth putting through the arithmetic before adding the fourth account, not after.
Common questions
What is slippage in trading?
The difference between the price you expected and the price you actually got. It happens because the price on screen is a snapshot of a queue that has changed by the time your order arrives.
Why do stop losses slip so much?
Because a stop becomes a market order at the moment it triggers — which is the moment price is moving quickly through that level and liquidity is thinnest. Stops routinely slip several times more than entries, which is why a backtest assuming they fill at the stop price is not describing anything real.
How much slippage should I expect in futures?
It depends on the product, the hour and your size relative to what is resting. Half a tick each way is a reasonable floor for a liquid contract in session hours; stops, the open and scheduled releases are all worse. Measure your own fills rather than trusting a rule of thumb.
Does slippage average out over time?
No. Market orders are filled by whatever is available, so favourable fills happen when nobody wanted the other side and unfavourable ones when everyone did. It is a persistent cost, not noise around zero.
How do I reduce slippage?
Trade when the book is deep rather than overnight or at the open, use limits where the strategy tolerates a missed fill, and size relative to what is actually resting — ten micros frequently slip less than one E-mini. Some of it is simply the cost of participating.
Does copy trading add slippage?
A little. A follower's order cannot be placed until the leader's fill is reported, so it arrives tens of milliseconds later — occasionally a tick in a fast market, usually nothing. The larger effect is that costs multiply across accounts: four followers pay four sets of commission and slippage.
Start on a simulated account.
Connect a sim or evaluation account and run the whole product against it before you point anything at live money. Nothing about the setup changes when you do.