Measuring Institutional Trading Latency, Throughput, and Market Data
Summary
The article explains execution latency as a chain spanning market data delivery, strategy processing, network routing, exchange gateway checks, matching, and confirmation. It distinguishes latency from throughput and recommends measuring more than an average, including median and tail latency, jitter, packet loss, round-trip time, and behavior during volatile periods. It also describes connectivity choices, higher-frequency market data, binary feeds, timestamps, and request capacity as components institutions may evaluate when designing execution systems.
The text reports venue-specific infrastructure figures and examples, including a claimed order-processing improvement and request limits, but those claims are not independently verified in the document. It notes that observed performance varies with account tier, server location, route, system configuration, and market conditions. Its most transferable guidance is to benchmark the complete path from the same infrastructure intended for live use and to distinguish exchange-side processing from client network and strategy delays.
Key ideas
- Observed execution latency combines network, client processing, gateway, matching, and return-path delays.
- Throughput describes request capacity and should not be confused with the time required to execute one order.
- Latency testing should track tail percentiles, jitter, packet loss, and performance under changing market conditions.
- Market data speed and depth affect whether a strategy acts on current information.
- Benchmarks are route-specific and should use the infrastructure and configuration planned for production.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.