Impact cost is the part of your trading cost that you cause yourself. Your order consumes the available depth on one side of the book, and the price moves against you while you are still filling.
NSE defines it as the percentage markup suffered while buying or selling a desired quantity of a security, measured against the ideal price:
ideal price = (best bid + best ask) / 2
impact cost = (execution price - ideal price) / ideal price
The important property is that impact cost is defined for a specific order size. A quote of “impact cost is 0.1%” is meaningless without the quantity it refers to.
How NSE uses it
Impact cost is not a theoretical construct here. NSE uses it as a hard liquidity criterion for index inclusion. For a stock to be eligible for the Nifty 50, it must have traded at an average impact cost of 0.50% or less for 90% of observations over the last six months, measured for a portfolio of ₹10 crore (NSE, Impact Cost).
That gives you a usable benchmark. A large-cap index constituent is a stock where an institutional-sized order costs under half a percent. Anything outside that set is, by NSE’s own measure, less liquid than that.
What raises it
| Factor | Effect |
|---|---|
| Order size relative to book depth | The dominant driver. Impact rises non-linearly as the order consumes deeper levels |
| Market capitalisation | Small and micro caps carry wider spreads and thinner books |
| Time of day | Depth is thinnest at the open and around the intraday square-off window |
| Surveillance measures | Stocks under ASM or GSM face restrictions and reduced participation |
| Market stress | The book empties exactly when everyone wants out |
Why it caps strategy capacity
A strategy that returns 20% on ₹5 lakh may return far less on ₹5 crore, purely because the same rules now demand orders the market cannot absorb cheaply. This is capacity, and it is the reason a smallcap strategy that looks excellent in a backtest can be untradeable at size.
The saral.money tradeability audit flags securities a backtest traded that would have been hard or impossible to trade in reality, including thin liquidity and surveillance bans, so the gap between the simulated portfolio and an executable one is visible rather than assumed.
Caveats
Published impact cost is an average over normal sessions. It tells you what a stock cost in the median case, not what it will cost on the day you need to sell it alongside everyone else.
Backtests generally assume you are a price taker with no effect on the market. That assumption is reasonable at retail size in liquid names and steadily less reasonable everywhere else.