What is slippage in trading?
Slippage is the difference between the price a strategy expected to trade at and the price its order actually filled at. It is a real, recurring cost of execution, and because it applies to every trade, ignoring it flatters every backtest and quietly breaks many live strategies.
Slippage is not a malfunction. Markets move between decision and fill, and orders consume liquidity as they execute. The question is never whether a strategy pays slippage, only how much and whether the edge survives it.
Where does slippage come from?
Three main sources. Time: prices move in the interval between the signal and the order landing at the exchange. Liquidity: a market order eats through the order book, and if the best price has less size than the order, the remainder fills at worse levels. Volatility: in fast markets, quotes are stale by the time they are acted on, and everyone's fills degrade at once.
Order size scales all three. A small order in a liquid market slips by fractions; a large order in a thin one can move the price it is trying to trade at.
How should backtests model slippage?
The simple, honest approach is a fixed penalty per fill, calibrated to the market's spread and depth, applied to every simulated trade alongside fees. More refined models scale the penalty with volatility or order size. What matters most is that the number is nonzero and defensible.
A useful stress test: rerun the backtest with slippage doubled. A robust strategy gets modestly worse; a fragile one flips to losing, revealing that its edge was smaller than its costs all along. High-frequency strategies fail this test most often, because more trades mean more times the toll is paid.
How can slippage be reduced?
Trade liquid markets, where spreads are tight and books are deep. Prefer limit orders where the strategy allows, since they cap the fill price at the cost of sometimes not filling. Avoid executing into news spikes and thin hours, and keep order sizes small relative to the market's typical volume.
There is a real trade-off inside that list: limit orders eliminate negative slippage but introduce missed trades, and which cost is worse depends on the strategy. Testing both execution styles is the only way to know.
Frequently asked questions
- Is slippage always bad?
- No. Positive slippage, filling at a better price than expected, does happen, particularly with limit orders in fluctuating markets. On average, though, aggressive orders pay rather than collect, which is why backtests should assume slippage costs money.
- Do limit orders eliminate slippage?
- They cap it: a limit order cannot fill worse than its stated price. The cost moves elsewhere, into missed fills when the market never reaches the limit. For strategies whose profits depend on catching every signal, missed trades can cost more than slippage did.
- How much slippage should I assume in a backtest?
- At minimum the typical spread of the market being traded, applied to every fill, plus fees. For volatile or thinner markets, meaningfully more. If results only look good with slippage set near zero, the strategy does not have an edge at realistic costs.
- Why were my live fills worse than my backtest fills?
- Backtests fill at recorded historical prices, while live orders pay the spread, wait out latency, and consume real liquidity. Some gap is structural and expected. Persistent large gaps usually mean the backtest's slippage assumption was too optimistic for the market and order sizes involved.
