The third time’s the charm for Bloomberg Tradebook, which has been named best agency broker in the Waters Rankings for three years in a row with its win in the 2017 awards. This year has seen big developments for the broker, but looking forward, it is aiming for innovative heights in machine learning and natural-language processing to continue its success.
As with most other global firms, Tradebook is heavily focused on January 3, 2018—the date when the revised Markets in Financial Instruments Directive (Mifid II) is scheduled to go live. The launch of its exchange-traded fund request-for-quote (ETF RFQ) service and its Optimal Execution (OPTX) offering have proven to be key in this regard.
“OPTX will combine a real-time quantitative approach to the broker wheel with Bloomberg’s desktop distribution for analytics,” says Kapil Phadnis, global head of quantitative research at Bloomberg Tradebook. “We also released our ETF RFQ for Asia-listed ETFs, providing regional customization of trading tools for Asia, all while continuing to keep our business aligned for Mifid II compliance.”
The OPTX service has been running for around nine months, Phadnis says, during which time Tradebook has “significantly” improved its latency to exchanges, along with developing its broker-partnership program for the provision of trading algorithms, infrastructure and quantitative performance metrics.
Tradebook’s service has also seen some realignment this year, with the broker closing its foreign-exchange (FX) business in February 2017. Phadnis says that the decision was taken to provide focus to its core agency business, but reiterates that Bloomberg as a whole is still committed to FX.
Looking ahead, the broker is exploring emerging technologies for ways to keep its edges honed. These include, among other areas, broadening its approach to machine learning, along with solving hurdles related to natural-language processing.
“Machine learning is a key driver of OPTX and we are leveraging powerful analytics to provide us with the best real-time optimization engine,” Phadnis says. “Bloomberg Tradebook’s quant team is actively using deep learning and other machine-learning methods in the financial-trading domain to solve problems that would otherwise have a low success rate with traditional methods. These include natural-language processing problems to simplify workflows and identifying differences between trading algorithms using sophisticated pattern recognition.”
Jesse Lund talks about real uses for DLT in the capital markets, lessons learned while rolling out IBM's blockchain platform, and what’s ahead for 2018, and into 2019.Subscribe to Weekly Wrap emails