Latency in trading
Family V · Platforms
Not to be confused with latency in trading.
Latency in trading is the delay between an order being generated and its execution or acknowledgement by a trading venue. It is measured in milliseconds or microseconds and is the sum of network transmission, internal processing and exchange matching times. Latency is a property of the entire path an order takes, not of any single component.
Where latency arises
Total latency is cumulative across several stages:
- Market data latency — time for a price update to travel from the venue to the trader.
- Decision latency — time taken by software or a human to generate an order.
- Network latency — transmission time between the trader's system and the venue.
- Gateway and risk-check latency — time spent in pre-trade validation and order routing.
- Matching latency — time the venue's engine takes to process and confirm the order.
Each stage adds to the round-trip figure that determines how quickly a strategy can react and how much slippage it may incur.
Worked example
A trader sends a market order and measures the time from order generation to exchange acknowledgement.
The round-trip figure, which also includes the return path, would be higher. Figures vary by venue, infrastructure and location.
Why it matters
Latency affects the price at which an order is filled and the likelihood that a quote is still available. Strategies that depend on speed, such as market making or arbitrage, are sensitive to small differences. Latency is not a single universal number: it depends on the venue, the participant's technology, physical distance and prevailing market conditions.
Often confused with
- latency in trading
- Latency in trading is the elapsed time between an order being generated and its arrival at the execution venue, encompassing network, processing and matching delays.