Northern Trust To Expand Use of Machine Learning
Dane Fannin, head of global securities lending at Northern Trust, said the firm is discussing further applications of machine learning across the trading function in the business as it launched a pricing engine using the technology.
Northern Trust announced this month that it has developed a pricing engine that uses machine learning and advanced statistical techniques in the securities lending market to forecast the rate to loan securities.
Fannin told Markets Media: “This is the first iteration of machine learning within securities lending as we saw an opportunity to improve the efficiency of pricing. We look at technology as a tool to accelerate our objective in helping clients optimise their portfolios in pursuit of increased alpha or driving out costs.“
An algorithm estimates the demand for equities in the securities lending market. Northern Trust global securities lending traders can use use these projections, with their own market intelligence, to automatically broadcast lending rates for 34 global markets.
Northern Trust held approximately $1.2 (€1.1) trillion in lendable assets for more than 450 clients worldwide at the end of June this year. Fannin said that maintaining optimal pricing on all of these at point of trade and continuously through the life of the loan is intensive.
“Deploying machine learning helps to replicate and enhance the manual decision making process traders use to project the demand for a subset of securities, which can then be leveraged to optimize and sustain efficient pricing to drive increased revenues,” he added.
Chris Price, data scientist at Northern Trust, told Markets Media that the firm gathers internal data and mixes it with external market indicators across asset classes for use in the machine learning model.
“That is our secret sauce,” said Price. “We track results against internal benchmarks and make necessary adjustments.”
Fannin continued that Northern Trust has created an infrastructure and analytical framework that can intelligently adapt to changing market conditions.
“Having already invested time and effort to automate our trading process, we are in a strong position to benefit from increasingly automated borrower demand,” he added. “We continue to discuss a roadmap for further applications of machine learning across the trading function, where it makes sense, to allow traders to focus on high touch, value-added transactions.”
Importance of data
Matt Wolfe, vice president of business development at the US OCC, said on the options clearing corporation’s blog that there is a growing emphasis on data collection in securities lending.
“Decisions are being influenced by data analytics,” added Wolfe. “In my view, the future of securities lending will be data-driven and the leaders will be those that make the most effective use of data.”
He continued that one of the catalysts for the increased importance of data is the European Union’s Securities Financing Transaction Regulation that will mandate regulatory reporting within specific timeframes.
“An increased quality and breadth of data enables advanced data analytics, such as machine learning,” wrote Wolfe. “There’s little doubt that such a lucrative business (driving nearly $10bn in revenues according to a recent DataLend announcement) is going to attract advanced data science.”
Wolfe predicted that the use of machine learning and distributed ledger technology will benefit beneficial owners over the coming years.
“For example, programs that can anticipate changes in the demand for securities enable firms to make more informed decisions about when to lend and rerate securities,” he added. “Similarly, distributed ledger technology has the potential to not only improve the transparency for beneficial owners, but also to potentially enable them to take a more active role in their lending programs.”
Fannin agreed that there are a lot of exciting ideas about using new technology in securities lending to drive greater efficiencies. He said: “We are being disrupted for the right reasons.”
RBC GAM finds anti-corruption, cybersecurity and climate change are top ESG concerns.
The firm's total assets under management reached $9.5 trillion.
Institutional investors nearly doubled their expected required rate of return in 2020.
A consolidated tape will significantly improve transparency and create a level playing field.
In the past 12 months ETFs globally had $1.2 trillion in net inflows.