Two Clocks: Why Exchange Time and Receive Time Diverge
Every market event carries two timestamps: when the venue says it happened and when you learned about it. Treating them as one silently corrupts research.
When a venue publishes a trade, it attaches a timestamp: the moment its matching engine recorded the event. When that message arrives at a data consumer, a second moment exists — the time it was received. In an idealized world the two would be nearly identical. In practice the gap between them varies with distance, network conditions, venue load, and the accuracy of the clocks on both ends.
Figure 1Your arrival delay and clock offset, measured live
Live dataWhat the gap contains
- Physical and network transit time between the venue and the receiver.
- Venue-side processing and publication delay, which can grow during bursts of activity.
- Clock offset: neither the venue’s clock nor the receiver’s is perfectly synchronized.
- Batching, where several events are published together with a single delivery time.
Why it matters for research
Suppose two venues are compared to see which leads price discovery. If one venue’s data is captured close to its servers and the other far away, receive-time ordering will favor the nearby venue even if both moved simultaneously. The effect is not subtle: at high frequency, delivery delay can be larger than the true lead-lag being measured.
Keeping both clocks also makes data quality measurable. Sudden growth in the gap flags capture problems or venue stress; negative gaps reveal clock offset that must be corrected before any sequencing analysis. What looks like a housekeeping detail turns out to be a precondition for defensible conclusions.
This publication is provided for informational purposes only and does not constitute investment, legal, or tax advice, or an offer or solicitation to buy or sell any asset. Live figures are computed from third-party public market data and may be delayed, incomplete, or inaccurate.
