Market orders vs limit orders: spread, slippage and the real cost…

A market order buys or sells immediately at the best prices available, so it pays the bid–ask spread and, if it is large, walks through several price levels. A limit order sets the worst price you will accept, so it can save the spread but may never fill. The choice is a trade-off between a cost you can see now and a risk you can't: the market moving away while you wait.

The cost of a trade is easy to underestimate. The quoted spread is only the cost of the first share. This guide measures execution cost against a single benchmark, walks a real-looking order book, and puts a number on when waiting with a limit order beats crossing the spread. Every example can be reproduced in the free limit order book simulator.

What does a market order actually pay?

Start with an order book whose best bid is

99.95 and whose best ask is

100.05. The bid–ask spread is 10 cents and the mid-price is $100.00. On the ask side, sellers are waiting at four levels:

Ask price	Shares offered
$100.05	300
$100.10	500
$100.15	800
$100.25	1,500

A market order to buy consumes these levels from the top. Measured against the $100.00 mid, its cost is:

200 shares: all fill at $100.05. Average price$ 100.05, a cost of 5 basis points: half the spread.
1,000 shares: 300 at $100.05, 500 at$ 100.10 and 200 at $100.15. Average$ 100.095, a cost of 9.5 basis points.
2,500 shares: through the fourth level. Average $100.164, a cost of 16.4 basis points.

The ask ladder and the cost of a market buy against the mid as order size grows

The extra cost beyond half the spread is slippage: the difference between the price you hoped for and the average you got. It grows with size because each extra share comes from a deeper, worse level. A tight spread can sit on top of a thin book, so a small spread does not guarantee a cheap large trade.

Why measure cost against the mid?

You need one fixed reference price chosen before you trade, or you cannot compare decisions. Three references are common:

The mid at decision time, as above. It charges you for half the spread and for any impact.
The best quote, here $100.05. It shows only the impact beyond the touch and makes small orders look free.
The arrival price used in implementation-shortfall accounting: the mid when the decision was made. It also charges for the price drifting while you work the order.

Say which reference you use. A cost of 5 basis points against the mid and zero against the best ask describe the same fill.

When is a limit order better?

A limit order to buy at $100.00, inside the spread, would save 5 cents a share compared with buying at the ask. The catch is that it might not fill. If the market rises, you end up buying later at a worse price.

Put that into a simple model. Rest a buy at $100.00. With probability$ p $it fills, at zero cost against the mid. With probability$ 1 - p $it doesn't, and you buy later at$ 100.12, a cost of 12 cents. Buying now at the ask costs 5 cents. The expected cost of waiting is

(1 - p) \times 12 \text{ cents},

which beats 5 cents only when $p > 1 - 5/12 \approx 58\%$ .

Expected cost of resting a limit order against the fill probability, compared with buying at the ask

The useful lesson is the shape, not the specific numbers. Waiting is worth it when fills are likely and the cost of missing is small. It is a bad trade when the price tends to run away from unfilled orders.

Why do limit orders get filled at the worst times?

The model above treats the fill as a coin flip, independent of where the price goes next. In real markets it is not.

A resting buy order is most likely to fill when sellers are pushing the price down. Some of those sellers may know something you don't. So, conditional on being filled, the price tends to keep falling. When the price rises instead, your order sits there unfilled and you chase. This is adverse selection, the same force a market maker faces: the trades you get are biased towards the ones you would rather not have.

Measured properly, a limit order's saving is the spread captured minus the adverse move after fills, minus the cost of chasing misses. Research on execution consistently finds that the headline saving shrinks once those effects are counted.

How does order size change the decision?

For a small order the choice is mostly about the spread. For a large one, impact dominates. Empirical studies of large institutional trades find that impact grows roughly with the square root of order size relative to daily volume (Tóth et al., 2011). Doubling the size raises the cost per share by about 40%, not 100%.

That non-linear impact is why large orders are split over time. Execution algorithms balance two costs:

Trading fast pays more impact.
Trading slowly leaves the order exposed to the price drifting in the meantime.

The execution exercise in the order book simulator gives you a target of up to 200 shares and a deadline, so you can feel that trade-off directly. Rest limit orders, replace them or cross the spread. At the deadline, remaining shares trade at whatever prices are available, and any still unfilled carry a $1-per-share penalty.

Common mistakes

Using the best quote as the cost of a large order. Walk the book: the average fill, not the top price, is what you pay.
Ignoring missed fills in a backtest. Assuming every limit order fills at its price makes passive strategies look far better than they are. Model fills explicitly, with queue position and adverse selection.
Comparing costs with different references. Basis points against the mid and against the arrival price are different numbers.
Forgetting fees and rebates. Many venues charge takers and pay makers. That can tilt the decision but rarely reverses a large adverse move.

Where to go next

Trading costs and slippage in backtesting
Market making with Python and public order book data
What is market making? A quoting game in five stages. Try the other side of the trade in the market making game.
The transaction cost models lesson.

References

Tóth, B., Lempérière, Y., Deremble, C., de Lataillade, J., Kockelkoren, J. and Bouchaud, J.-P. (2011). Anomalous price impact and the critical nature of liquidity in financial markets. Physical Review X, 1, 021006.

Perold, A. F. (1988). The implementation shortfall: paper versus reality. Journal of Portfolio Management, 14(3), 4–9.

The order book and the limit-order model in this article are illustrative. Real fill probabilities depend on queue position, order flow and volatility.