Skip to main content
Learn — order book latency

Order book latency — what the queue looks like before your fill arrives.

Every order you submit joins a queue at the broker's matching engine. Your fill depends on where in that queue your order arrives. Order book latency is the delay between when you click and when your order joins the queue — and during fast markets, milliseconds of queue position matter.

Network latency determines when your order arrives at the matching engine. A cleaner route means an earlier arrival, a better queue position, and a fill closer to the price you intended.

See the plansTick data latency
Queue
Orders are filled in the order they arrive at the matching engine.
Arrival
Network latency determines when your order joins the queue.
Position
Earlier arrival means better fill quality during busy market events.

How the matching engine queue works

A broker's matching engine processes orders in the order they arrive — specifically, the order they arrive with the correct timestamp. When you submit a buy order, your platform creates a packet and sends it to the broker's server. At the server, the engine assigns a timestamp to your order based on when it arrived, then processes the queue in arrival order.

During quiet periods, you are likely the only order at that price level. Queue position makes no practical difference. During a news event or a market open, hundreds or thousands of orders arrive within the same second. The engine processes them in arrival order. An order that arrived 50ms after yours gets a worse queue position. If the available liquidity at your price level runs out before the engine reaches a later-arriving order, that order gets requoted or filled at a different level.

How network latency affects queue position

The gap between your click and your order's arrival at the matching engine is your network latency. A 40ms round-trip means your order takes 20ms to reach the engine one-way. A 200ms round-trip means 100ms. During a fast market where prices reprice every 100-200ms, that 80ms difference in delivery time means your order arrives in a fundamentally different market state — and at a different position in the queue.

Jitter makes this worse. If your delivery time varies between 20ms and 100ms per order, your queue positions vary unpredictably. Orders that arrive at the 20ms end are filled at fair levels. Orders that arrive at the 100ms end get the fills that were left over after faster-arriving orders were processed. That unpredictability is why consistent latency matters more than fast latency for most retail traders.

What a proxy changes about queue position

A proxy node at or near the broker's data centre reduces the delivery time from the node to the matching engine to just a few milliseconds — essentially the same as a co-located VPS, for the segment of the path from the node to the engine. The larger residential transit path still exists, but the variable, high-latency portion of the journey is replaced with a short, direct segment. Your order's queue position consistently reflects the network latency from your region plus a fixed data centre segment, rather than the sum of all variable residential transit hops.

Read it like a trader

If your fills are consistently worse at news events than they are during quiet sessions — same strategy, same broker, same instrument — order book queue position is a candidate explanation. The question is not just "how long does my order take?" but "where does it arrive in the queue relative to everyone else at the same moment?"

Queue position vs broker execution model

Queue position matters most with true ECN/STP execution. Market makers handle fills differently.

With a market maker broker, the broker itself takes the other side of your trade. Queue position in a traditional order book sense may matter less — the broker has more discretion over fill timing and requoting. With ECN or STP brokers, your order enters a real order book where time priority matters directly. A proxy has the most visible fill impact on ECN/STP accounts at active brokers.

  • Order book latency = network delivery time + engine processing time.
  • Proxy reduces network delivery time — engine processing time is fixed.
  • Impact is most visible on ECN/STP accounts during high-volume events.
FAQ

Order book latency is the delay between when an order enters the broker's matching engine queue and when it is processed and filled. It has two components: the network latency of getting your order to the matching engine, and the processing time within the engine itself once it arrives.

A proxy improves the network latency component — it reduces the time your order takes to travel from your machine to the matching engine, which means your order joins the queue earlier. Queue position is time-stamped by arrival at the matching engine. Arriving earlier means a better position in the queue during busy periods.

Tick data latency is the delay in receiving price updates from the broker to your screen — it affects what you see. Order book latency is the delay in your orders arriving at the matching engine — it affects what you get filled at. Both improve with a cleaner, lower-latency route to the broker's data centre.