24 May 2018
Iosif Itkin


Iosif Itkin - Exactpro

13Posts 64,867Views 6Comments
Finextra community

Banking Architecture

A community for discussing the latest happenings in banking IT. Credit Crunch impacting Risk Systems overall, revamp of mortgage backed securities, payment transformations, include business, technology, data and systems architecture capturing IT trends, 'what to dos?' concerning design of systems.

Cinnober on latency

22 October 2009  |  4776 views  |  1

Cinnober just announced the publication of an in-depth whitepaper on latency on their web site: http://www.cinnober.com/files/A_Cinnober_whitepaper_on_latency_1.pdf
The paper contains a much more detailed explanation for the items discussed in my previous post.
Taking into account that the figure (286 mks for door-to-door and 138 mks for business logic) is supported by the description of what was measured and how, it looks quite credible.
What is important is that even at this level of detail there is some room for interpretation of the door-to-door latency. Quote (page 6): “means measuring the time which elapses between a request being received at the AP and a corresponding response arriving back at the same AP after being processed by the matching engine.”
What should one consider to be the time when the AP received a request? Is it the time when the data appeared in the socket or the time when the AP actually started processing the request? Were these orders and quotes submitted against an empty book? Were any of these orders and quotes filled, acknowledged or rejected? At the microsecond scale, it makes quite a difference.
The paper contains a very good explanation of the queue theory and the service time calculation. One caveat, however, is: the obtained service time (98 mks – see page 13) is actually 4 times smaller than the real processing time, as there are 4 trading gateways that simultaneously process incoming messages. The actual processing time was 392 mks.
Overall, I would highly recommend this excellent paper to anyone interested in execution venue latencies!

TagsTrade execution

Comments: (1)

A Finextra member
A Finextra member | 25 November, 2009, 14:15

We measure by instrumenting the AP. The start time is when the AP has received the data, and the stop time is when the response is ready to be sent. The instrumentation is part of the production code, so we do always have real statistics available.

The order books were pre-filled before the tests, and we used an order mix with 2% execution rate.

Be the first to give this comment the thumbs up 0 thumb ups! (Log in to thumb up)
Comment on this story (membership required)

Latest posts from Iosif

Fearless Testing and Post-Trade Systems Disruption

10 May 2018  |  2744 views  |  0 comments | recomends Recommends 1 TagsPost-trade & opsInnovationGroupOperational Risk Management

Large Scale Infrastructure Intervention and Fake Testing

05 May 2018  |  4452 views  |  0 comments | recomends Recommends 0 TagsRisk & regulationOpen APIsGroupOperational Risk Management

The Next Generation of Testing Tools for Exchanges

17 October 2015  |  2944 views  |  0 comments | recomends Recommends 0 TagsRisk & regulationInnovationGroupMiFID

Innovation in Quality Assurance - What is the Impact on Trading Technology?

10 October 2015  |  4527 views  |  0 comments | recomends Recommends 0 TagsTrade executionRisk & regulationGroupMiFID

TTLiquidity: what could be a single big improvement in SOR?

25 November 2009  |  5707 views  |  0 comments | recomends Recommends 2 TagsTrade executionGroupMiFID

Iosif's profile

job title CEO
location London
member since 2008
Summary profile See full profile »
Iosif is a co-founder and CEO of Exactpro

Iosif's expertise

Member since 2008
11 posts6 comments

Who's commenting on Iosif's posts