K & K GLOBAL CONSULTING
RESEARCH
Statistical evidence based trading tools and stochastic broker selection
One of the positive outcomes of the new mandatory reporting regimes will hopefully be a significant improvement on the quality of big data , better visualisation tools and more informed decision making . We already see automation tools based on varying levels of third party statistical data , some with stochastic broker selection . Most automation tools would still be “ exception based ” so if the order doesn ’ t meet the “ normal ” criteria , the system will forward the order for human intervention . We are excited to hear about the progress of “ exceptions based ” technology as it frees up time for the buyside trader . To what extent this evolution will lead us toward the purest definition of Artificial Intelligence ( AI ) is still to be proven . Our definition of AI suggests that the machine will question and propose to override its policies programmed by a human when it finds a better way of doing things . We understand that activities that overrides policies are still better suited for the human trader .
We expect the interest in the Algo Wheel type of technology to increase not only within equities trading but we also see the interest within foreign exchange trading . The interconnectivity of smarter TCA technology and independent thirdparty data will support the automation and randomisation of broker selection to remove human bias . Ultimately it will require a smarter broker review and classification process . Within equities , the buy side will need to understand who are their range of preferred brokers within each strategy . Within foreign exchange , the buy side are discussing the future of classifying counterparties based on their performance within currency pairs . More buy side are now evaluating if they should start using foreign exchange trading algorithms for Spot and derivatives .
The fixed income trader will be more interested in statistical evidence based tools that visualise historical counterparties of ISINs to minimise the footprint and market impact . We have already heard one buy-side success story with full automation of bond trading of highly liquid government bonds . We are eager to hear more stories like this in 2018 .
The discussions about pre-trade pricing tools within the equities ATF highlighted the consensus that technology can be useful up to 10 % or a maximum of 20 % Average Daily Trading Volume ( ADV ). The machines are still not smart enough to replace the human intelligence of the sell-side trader .
16 www . buysideintel . com Winter 2017