Algo Testing

MiFID II enacts a range of organisational requirements on trading venues that will also impact Participants.

Cboe must require Participants to test algorithms and must provide some facility to support that testing. Cboe intends to implement the requirement on Participants through its rules. Cboe will seek to facilitate Participant testing by simulating a number of conditions as accurately as possible in its UAT environment, including disorderly market conditions.

Specifically, the following conditions will be configured on a range of symbols:

• Amended Executions (Trade Reports): Confirmed trade reports will have their price and/or size amended up to a maximum of a half hour after confirmation.
• Cancelled Executions (Orders): Order book executions will be cancelled up to a maximum of a half hour after execution.

• Cancelled Executions (Trade Reports): Confirmed trade reports will be cancelled up to a maximum of 30 minutes after confirmation.
• Halted All Day: These symbols will be halted simulating a regulatory halt for the entire day.
• Halt state toggling: These symbols will toggle between halted and unhalted states on a periodic basis.
• Low Load: These symbols will be configured with a regular load throughout the entire day of one order per second.
• Medium Load: These symbols will be configured with a load alternating between one order per second and twenty orders per second, with the higher load having a one-minute interval and five-second duration.
• High Load: These symbols will be configured with a load varying between one order per second and one hundred orders per second, with the higher load having a three-minute interval and a three-second duration.
• Burst Load: These symbols will have a base load of one order per second and periodically spike to one thousand orders per second, with the higher load having a ten-minute interval and a one-second duration.
• Dark and Lit Order Book Loaded: These symbols will have liquidity available in both lit and dark order books, which will be periodically refreshed on execution.

Cboe will continue to enhance the capabilities of its UAT system over time, making it easier for Participants to test a range of algorithms. For example, Cboe plans to introduce a simulated behaviour with large price movements, where execution prices vary +/- 10% from the reference price published in our reference data file.

Cboe recognises that not all Participants operate on a pan-European basis and may specialise within certain geographies. Accordingly, Cboe intends to configure one symbol per listing market for such behaviour. Refer to the following table: