Realized Spread vs Effective Spread: Two Similar Metrics, Different Use
Why these two metrics exist
Effective spread is a trade time cost metric. It compares your execution to the midquote at the moment you trade.
Realized spread answers a different question: after the market has a chance to move, how expensive was the trade in hindsight.
Effective spread recap
Effective spread measures how far your execution was from the trade time midpoint:
ES = 2 × |Pexec - Mt|
• Use VWAP for Pexec when there are multiple fills.
• Mt is the midquote at trade time.
What realized spread measures
Realized spread compares your execution to a later midpoint, after a fixed time horizon:
• Pick a horizon like 5 seconds, 30 seconds, or 1 minute
• Compute a new midpoint Mt+delta
• Compare your trade to that later midpoint
Conceptually, realized spread helps separate two effects:
• pure spread cost
• adverse selection and short term price movement
Why the two spreads differ
If prices move after your trade, realized spread changes while effective spread stays the same.
Two typical cases:
Case A: the market moves in your favor
You buy and the midpoint rises shortly after.
• Effective spread still shows the cost you paid at entry.
• Realized spread becomes smaller because the market moved toward your side.
Case B: the market moves against you
You buy and the midpoint falls shortly after.
• Effective spread still shows entry cost.
• Realized spread becomes larger because the market moved away from your side.
How to use each metric
Use effective spread for execution quality
Effective spread is the cleanest metric for how expensive your fills were relative to the trade time market. Use it to compare:
• market vs limit execution
• different sizes
• different markets and times of day
Use realized spread for market microstructure insight
Realized spread helps you understand whether trades tend to be followed by price moves. It is commonly used to study:
• adverse selection
• toxic flow vs stable flow
• how quickly liquidity providers recover after fills
Choosing the horizon
The horizon is not universal. In fast markets, 5 to 30 seconds can be meaningful. In slower prediction markets, you may need longer windows.
The key is consistency. Pick a horizon and use it across your dataset so the metric is comparable.
Data requirements
To compute realized spread you must have:
• clean trade time midpoint Mt
• clean later midpoint Mt+delta
• accurate timestamps and snapshot alignment
Takeaway
Effective spread tells you what you paid at execution. Realized spread tells you what the trade looks like after the market moves. Use effective spread for cost measurement and realized spread for understanding how trades interact with subsequent price movement.
Related
• Midquote Done Right: Snapshot Timing and Data Quality