Strategic Economics

Market Impact and the Cost of Trading Size

The cost of a large order is not the spread. It is the price move your own trading causes, it grows with the square root of your participation, and it is the reason a good idea has a maximum size. Compute it once and you can compute the capacity of any strategy, the real cost of any buyback, and the true price of any exit.

  • Expert
  • 14 min total
  • 14 chapters

What decision this helps you make: How fast to execute a large order, and how large an order to contemplate at all — trading off the cost you cause by going quickly against the risk you accept by going slowly, with both sides denominated in money.

What this topic is

Market impact is the amount by which your own trading moves the price against you. Buy a lot of something and the price rises while you are buying, so the average you pay is worse than the price you saw when you decided. That gap is not a fee anybody charges and not a spread anybody quotes — it is the market reacting to the fact that you are there. Across essentially every market that has been measured, the size of the effect follows a strikingly simple rule: it scales with the asset's volatility and with the square root of the fraction of daily volume you represent.

Why it matters

Impact is usually the largest cost of trading and almost always the least measured. In the example worked below, an order for a tenth of a day's volume costs sixteen times more in impact than it does in spread. It is also the binding constraint on scale: because total cost grows faster than order size, every idea has a size beyond which it stops making money, and that size can be computed from three numbers you already have. Anyone running a buyback, exiting a position, rebalancing a fund, or wondering why a strategy that worked at small size stopped working at large size is looking at this arithmetic.

Who should learn it

Portfolio managers and traders sizing positions, CFOs and boards running repurchase programmes, founders and funds planning an exit from a concentrated holding, and anyone who has ever been told that a strategy "does not scale" and wanted the number rather than the assertion.

What you will understand

  • The square-root law: why impact scales with volatility times the square root of participation, and what that implies about doubling an order
  • How to compute the capacity of a strategy — the size at which your own impact consumes the edge entirely
  • The Almgren-Chriss trade-off between impact cost and timing risk, and how to choose a horizon deliberately
  • Why measuring your own impact is genuinely hard, and how it gets confounded with the signal that made you trade

Prerequisites

Common misconception

"Trading costs are the spread plus commission." For a small order, roughly. For an institutional order, the spread is a rounding error. In the worked example, a 200,000-share order in a stock trading two million shares a day pays about $2,000 in half-spread and about $31,600 in impact. Worse, the relationship is not proportional: doubling the order size does not double the cost, it multiplies it by about 2.8, because cost per share also rises. Any budget, any expected return, and any strategy capacity built on a linear cost assumption is wrong in the direction that hurts most as you get bigger.