Christiane Perry-Roy L'actualité internationale

Man Group Switches Focus of Oxford University Department

The Wall Street Journal The Wall Street Journal

The hedge fund has pumped millions of dollars into the Oxford Man Institute


By Laurence Fletcher May 11, 2016LONDON—
Man Group, one of the world’s biggest hedge fund firms, is changing the focus of an Oxford University research department it funds to “machine learning” as firms increasingly embrace technology to gain a competitive edge.

Man, which runs $78.6 billion in assets, has put millions of dollars into the Oxford Man Institute, or OMI, since setting it up in 2007, with the aim of tapping some of its research ideas.

The unit, which has been researching a range of subjects such as statistics, economics and financial asset prices, will now focus solely on machine learning, in which computers learn from data rather than simply acting on it, and data analytics.

Computer-driven hedge funds, which employ physicists and mathematicians to build complex trading algorithms, are increasingly focusing on areas such as artificial intelligence and big data as they attempt to gain any trading edge they can.

The idea behind using machine learning—a branch of artificial intelligence in which computer algorithms learn from data rather just following instructions—is that a computer will spot subtle possible money-making patterns among the data that would be undetectable to humans.

Sandy Rattray, chief executive of Man’s computer-driven AHL unit, said the company’s allocation to machine learning has been growing.

The firm made around half of last year’s profit in its $4.1 billion AHL Dimension fund from machine learning, said a person familiar with the fund’s performance, higher than the previous year. Machine learning algorithms made money for the fund late last summer when they started taking positions, shortly before markets rebounded.

A spokeswoman for Man Group declined to comment on performance.

Man, which was founded in 1783, initially gave £13.75 million ($19.88 million) to fund the OMI and said in 2013 it was extending the funding by five years. It declined to reveal further financial details.

Mr. Rattray said that while human traders may say that “markets feel a bit like” a certain period in the past, machine learning algorithms could look at different characteristics of markets from different periods of history at the same time.

“In five years’ time will we have machine learning funds?” he said. “I don’t know but I certainly expect so if things continue at the current rate.”