Thanks to everyone who attend our inaugural QB | The Best Execution Conference 2017. We appreciated your interest and active participation. Thanks also to our guest speakers, our co-founder Robert Almgren, and our Research team. Here is a recap of some of the topics:
The great level of interest in all things AI / Machine Learning is certainly a takeaway from this event. There is no doubt that it’s going to affect everything we do. However, the intelligent symbiosis of man and machine is just as relevant in our industry as it is for others – we mustn’t get too caught up in the hype that machines and models will solve all our problems without the oversight and context that humans can provide. What we do at QB is augment a trader’s abilities, and our adoption of these new technologies will further that purpose. For us, machine learning is for making better order placement decisions and analyzing troves of data for new insights in TCA. As stated in our July blog, allowing an algorithm to effectively change its own code is not something we envisage anytime soon.
The Evolution of Best Execution
One of Robert’s presentations focused on the history of QB and the evolution of our own approach to best execution. We have over time evolved to be much more opportunistic. We believe that optimal execution is opportunistic, but opportunities can only be assessed with intelligence, and so we have gone to great efforts to build such intelligence into our algorithms. Mathematical models give insight and are important, but these alone do not respond to real-time opportunities in the market.
Robert described best execution as the implementation of trade decisions and the analysis around trading. A cycle of pre-trade analytics, trade execution and post-trade TCA. It is a virtuous circle: each part drives improvements in others. We are excited to announce that in 2018, pre-trade analytics will formally become a part of the QB offering.
November 30, 2017