What it means
A trade can occur at a price different from the midpoint between the best bid and ask. Realized spread compares that execution price with the midpoint at a specified time after the trade. The interval matters: a five-minute realized spread and a 30-minute realized spread need not be the same.
Market-microstructure analysis commonly uses this measure to examine the portion of a trade’s effective spread not offset by a subsequent midpoint move. It is usually calculated from the perspective of a liquidity provider: a positive value indicates that the later midpoint remains on the favorable side of the execution price. It is not the provider’s net profit, because it excludes other costs and inventory risk.
How it is calculated
Realized spread per share = 2 × side × (execution price − midpoint after interval)
- Execution price: the price paid or received for the trade, in US$ per share.
- Midpoint after interval:
(best bid + best ask) ÷ 2at the specified later time, in US$ per share. - Side:
+1for a buyer-initiated trade and−1for a seller-initiated trade. This classifies which side demanded liquidity, not whether the investor ultimately gained or lost. - 2: doubles the signed price difference to express a full-spread measure. Omitting it produces the realized half-spread.
The result is in US$ per share. To express it in basis points, use realized spread (bps) = 10,000 × realized spread per share ÷ trade-time midpoint, with both prices in the same currency. The trade-time midpoint is the midpoint when the trade occurred; one basis point is 0.01%.
Worked example
Suppose, hypothetically, a buyer-initiated trade executes for 100 shares at US$100.05. At the trade time, the best bid is US$99.95 and the best ask is US$100.05, so the midpoint is US$100.00. Five minutes later, the midpoint is US$100.02.
- Set
side = +1because the buyer initiated the trade. - Subtract the later midpoint from the execution price:
US$100.05 − US$100.02 = US$0.03per share. - Double the difference:
2 × (+1) × US$0.03 = US$0.06per share of five-minute realized spread. - Convert to basis points using the trade-time midpoint:
10,000 × US$0.06 ÷ US$100.00 = 6 bps.
For comparison, the effective spread at execution was 2 × (US$100.05 − US$100.00) = US$0.10 per share, or 10 bps. The midpoint rose by US$0.02 after the trade, accounting for the US$0.04-per-share difference between those two full-spread measures. Multiplying US$0.06 by 100 gives US$6, but that is not a complete trading-cost or profit figure.
Why it matters before you trade
Realized spread requires a future midpoint, so the realized value for a proposed trade cannot be known before execution. Historical measurements can provide context about how much of an initial spread has tended to persist over a chosen interval, but they cannot determine what will happen to a particular order.
A pre-trade analysis instead works with information available before an order: the estimated bid-ask spread, order size relative to daily volume, volatility and estimated market impact. Those estimates concern possible execution conditions; they are not realized-spread observations. After a fill, a realized-spread calculation can add context to a post-trade review, provided the trade direction, quotes and time interval are recorded consistently.
How PreTrAIde uses it
PreTrAIde does not list realized spread as one of its calculated measures. Its pre-trade analysis estimates the bid-ask spread using a liquidity-tier model with a volatility uplift and a one-tick floor; it also estimates participation in daily volume and market impact. These estimates help inform its basket and per-stock decision-support verdicts, not a promise about subsequent midpoint movements.
PreTrAIde does not receive or execute orders. A realized-spread calculation would require an actual execution price and a later market midpoint, so it should not be confused with PreTrAIde’s pre-trade spread estimate.
Common mistakes
- Calling the quoted bid-ask spread the realized spread. Quoted spread uses bid and ask at one point in time; realized spread uses an execution price and a later midpoint.
- Ignoring trade direction. The sign changes for buyer- versus seller-initiated trades.
- Leaving out the measurement interval. Different later midpoints produce different results.
- Treating a negative value as an error. A sufficiently large midpoint move can make realized spread negative.
- Equating it with investor returns or dealer profit. It is a price-based execution measure, not a full accounting of costs, risks or returns.