Timed Batch Matching as an Alternative to Continuous Order Books

Summary

The document raises a market design question: could an exchange collect orders during fixed intervals and match them together at each interval’s end, instead of processing each incoming order immediately? It contrasts this proposed batch process with continuous electronic order books, where the matching engine typically acts on new orders in near real time.

The motivation is to reduce the value of minimizing network latency and locating trading systems close to an exchange, so participants could compete more on decisions based on public information. The document presents this as a question and rationale, not as an account of particular exchanges or a tested market design. It gives no details about interval length, price formation, order priority, cancellations, liquidity, or execution outcomes. Consequently, it introduces the latency and matching-timing tradeoff but does not show whether batch matching achieves the stated goal or how it affects other aspects of trading.

Key ideas

• Continuous order books attempt to match incoming orders as they arrive.
• Timed batch matching would collect orders and match them at scheduled intervals.
• Batch timing is proposed as a way to reduce the advantage of lower communication latency.
• The document does not assess the effects of batch matching on pricing, liquidity, or execution.

Tags

# Are there any exchanges which match orders in timed intervals?


# Are there any exchanges which match orders in timed intervals?












When an exchange works with an electronic order book and a matching engine, it usually tries to match orders in realtime during trading hours. That means when a new order comes in, it tries to match this order to existing ones as soon as possible, usually within microseconds.

Are there any exchanges which don't do this in realtime, but instead employ a timed interval, which gives the participants a certain time (e.g. some seconds) to decide about their orders and place them, and then matching all known orders after the interval has ended?

This would eliminate the advantage of hosting trading algorithms nearby to the exchange, thus eliminate the need for co-hosting and the rally to host closest, reducing the trading to what it should actually be, a decision based on publicly available information.

Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.