How to measure feed delay between two terminals
Posted: Sat Oct 31, 2026 8:11 am
Eyeballing two charts side by side and saying one is "faster" is useless. A proper comparison takes about an hour to set up and gives real numbers.
Install both terminals on the same PC, so they share a single clock. That removes the problem of syncing time between machines. Attach a small tick-logging EA to the same symbol in each, writing the local time in milliseconds and the bid and ask to a CSV on every tick. Run it for a session that includes a busy period, then match the two logs: find a price change in one file and look for the same price in the other, and take the difference in local timestamps.
Expect some noise. Feeds do not carry identical ticks, since different liquidity providers and filtering rules produce different sequences, so matching only works on distinct price moves, and a consistent offset of 100 to 300 ms may just be how the two sources work. What matters is the pattern: a stable offset is one thing, an offset that grows during news is another.
This is mostly relevant when signals come from one platform or data source and orders go out through another. Then the delay between them is a hidden cost in every trade.
Has anyone run this comparison, and how big was the gap between a demo feed and a live feed?
Install both terminals on the same PC, so they share a single clock. That removes the problem of syncing time between machines. Attach a small tick-logging EA to the same symbol in each, writing the local time in milliseconds and the bid and ask to a CSV on every tick. Run it for a session that includes a busy period, then match the two logs: find a price change in one file and look for the same price in the other, and take the difference in local timestamps.
Expect some noise. Feeds do not carry identical ticks, since different liquidity providers and filtering rules produce different sequences, so matching only works on distinct price moves, and a consistent offset of 100 to 300 ms may just be how the two sources work. What matters is the pattern: a stable offset is one thing, an offset that grows during news is another.
This is mostly relevant when signals come from one platform or data source and orders go out through another. Then the delay between them is a hidden cost in every trade.
Has anyone run this comparison, and how big was the gap between a demo feed and a live feed?