FinServ Ramps Up Machine Learning
Two thirds of financial services firms currently deploy machine learning and expect to increase their use of the technology within the next three years.
The Bank of England and the UK Financial Conduct Authority conducted a joint survey this year on the current use of machine learning, a methodology where computer programmes fit a model or recognise patterns from data, without being explicitly programmed and with limited or no human intervention. This contrasts with ‘rules-based algorithms’ where the human programmer explicitly decides what decisions are being taken under which states of the world.
The study said the median firm uses live machine learning applications in two business areas and this is expected to more than double within the next three years.
— Virginie O'Shea (@virginieoshea) October 17, 2019
The survey was sent to almost 300 firms, including banks, credit brokers, e-money institutions, financial market infrastructure firms, investment managers, insurers, non-bank lenders and principal trading firms, and received 106 responses.
Mark Carney, governor of the Bank of England, said in his annual speech at the Lord Mayor’s Banquet for Bankers and Merchants of the City of London in June this year that a new economy is emerging which is driven by changes in technology, demographics and the environment and which requires a new finance.
Carney said: “With its leadership in fintech and green finance, the UK private sector is creating the new finance, but your efforts will be more effective with the right conditions in which to innovate and the level playing fields on which to compete.”
The study said firms were using machine learning in cases ranging from equity trading to optimize order-routing and deal execution to anti-money laundering where the technology is used to analyse millions of documents for ‘know-your-customer’ checks. Insurance and banking had the most live cases in the sample with the median bank having 5.5 machine learning applications.
“Larger firms may possibly be more advanced in their ML deployment due to benefits of scale, access to data, ability to attract ML talent, or greater resources,” said the study. “However, more research would be needed to shed light on the specific reasons for sectoral differences.”
The median respondent expects their number of machine learning applications to more than double over the next three years although banking expects growth to almost triple to 15.5 applications.
“This underlines growing interest in ML and the prospect of increasing use across the financial sector in coming years,” said the survey.
The survey found that machine learning is used in sales and trading to increase speed and accuracy of processing orders; in pricing through combining a large number of market time-series to arrive at an estimate of a short-term fair value; and in execution to evaluate venue, timing and order size choices.
Did you know that two thirds of financial firms use artificial intelligence? Discover how the financial sector uses machine learning in our new report, published with the @TheFCA. https://t.co/nHieRpJbjb#fintech #AI #machinelearning pic.twitter.com/2BN8bOgsDd
— Bank of England (@bankofengland) October 16, 2019
“Within this, ML may also be used for intermediate steps of the process; for instance, for calculating the probability of an order being filled given the available characteristics of the order,” added the study. “Firms use ML techniques to determine order routing logic, this is often contained within systems called smart order routers or broker/algo wheels.”
Data used is still largely of a traditional, structured type but some firms also use unstructured data, such as text data, to estimate prices in illiquid markets.
In asset management machine learning often plays a supporting role according to the survey. Applications are used to analyse large amounts of data from diverse sources and in different formats; to assist in establishing a fair market price for a security; support decision-making processes by linking data points and finding relationships across a large number of sources; and sift through vast amounts of news feeds to extract useful insights.
The CFA Institute said in a new report that AI and big data allow analysts to perform more thorough analysis and for portfolio managers to make better informed decisions.
— EI (@Enterprising) October 22, 2019
Larry Cao said in a blog that, for example, analysts can estimate staffing levels at Tesla using publicly available cell-phone data.
“In fact, that’s precisely what Thasos Group did,” he wrote. “By gauging the number of cell phones present near Tesla’s plant, they independently verified that Tesla was running around the clock with three full shifts.”
He also gave the example of analysts at Goldman Sachs overlaying publicly available labor information on top of the geometric data of production sites to estimate the market power of manufacturers in aggregate.
Cao concluded that AI will transform investment management, but it will not lead to the mass extinction of human investment managers.
“Rather those investment teams that successfully adapt to the evolving landscape will persevere,” said Cao. “Those that don’t, will render themselves obsolete.”
Consultancy Greenwich Associates said in a survey this month that 44% of capital markets firms are already using artificial intelligence in their trading processes.
— Greenwich Associates (@GreenwichAssoc) October 1, 2019
Almost one-fifth, 17%, of firms reported plans to implement AI in trading in the next 12 to 24 months with four out of five expect AI and/or machine learning to be fully integrated into the trading process in three to five years.
Kevin McPartland, head of research in Greenwich Associates market structure and technology group, said in a statement: “Many buy-side institutions feel strongly that the best way to advance an emerging technology is to leave long-term development to IT providers, and then implement resulting innovations in AI and other areas to improve their businesses.”
The Bank of England and the FCA said their survey is the first step towards better understanding the impact of machine learning. The Bank of England and the FCA are also establishing a public-private working group to explore some of the questions raised by the study and will consider repeating the survey next year.
Data extraction and integration is the second stage of a digitization process.
Increased electronification has created useable and accessible real-time and historic trade data.
The fintech uses data so institutions can assess the environment impact of their portfolios.
Working with Riskfuel will reduce reliance on slow and expensive financial models.
The platform uses AI to match experts and users and digitize the research process.