Home/Metrics/Latency

Execution Latency

Round-trip time from order submission to broker server acknowledgment. Measures the speed and efficiency of order processing infrastructure.

Definition

Execution latency measures the total round-trip time (RTT) from the moment a trader submits an order to the moment the broker's server sends back a fill confirmation or rejection. This includes network transit time, server processing time, liquidity provider routing, and the LP's own response time.

In MT5, execution latency is logged in the journal and can be extracted from trading statement exports. It is typically measured in milliseconds (ms) and reported as both median and percentile distributions.

Measurement

Round-Trip Latency
RTT = T(fill_confirmed) - T(order_submitted)

Measured in milliseconds. T(order_submitted) is the client-side timestamp when the order request is sent. T(fill_confirmed) is the timestamp when the fill acknowledgment is received.

Adjusted Latency (Network-Normalized)
Adjusted Latency = RTT - (2 x Network Ping)

Removes network transit to isolate server-side processing time. Useful for comparing brokers with servers in different geographic locations.

Why It Matters

Execution latency directly determines slippage risk. The longer it takes for an order to be processed, the more the market price can move between submission and fill. For every 10ms of additional latency, the probability of adverse slippage increases measurably during volatile conditions.

  • Directly correlates with slippage probability
  • Critical for scalping and high-frequency strategies
  • Indicates infrastructure quality and LP connectivity
  • Should be analyzed at P50, P90, and P99 percentiles
  • Spikes during news events reveal true infrastructure capacity

Benchmark Ranges

Benchmark Ranges
RatingRangeAssessment
< 30msUnder 30msexcellent
30-80ms30ms -- 80msgood
80-200ms80ms -- 200msfair
> 200msOver 200mspoor

These benchmarks assume the trader's VPS or trading terminal is located within reasonable proximity (same continent) to the broker's MT5 server. Cross-continental connections will naturally add 80-150ms of network latency.

Visualization

Latency distribution across our tested broker sample. The chart shows median (P50) latency values grouped by execution model:

Execution Latency by Model (ms)

Percentile Analysis

Median latency alone can be misleading. A broker with 40ms median but 500ms P99 has serious tail-risk issues. We recommend evaluating all three percentiles:

P50 (Median)

Typical execution speed under normal conditions. Most relevant for steady-state trading.

P90

Speed during moderate stress. Reveals how the system handles normal load spikes.

P99

Worst-case scenario (excluding outliers). Critical for risk assessment during news events.