David Harding dismisses investment jargon like ‘market inefficiencies’. The models developed by a 100-person strong research team, which includes 80 PhDs, aim to simply “make inferences about the future”. Winton is a London-based commodity trading advisor (CTA) hedge fund with $20 billion in funds under management. But Harding described it as a scientific research firm. Its algorithms are developed through statistical analysis that peers far back into the history of financial markets, because this provides some idea, within a range, of what will happen next. “We look for data going back to the 1950s. Most stock-pickers look back a month or two.” The manager will not attempt to predict the exact value of, say, the Australian dollar in six months’ time, but forecast the range in which it will be and place these bets through futures contracts, Harding said. This process is repeated systematically for stocks worldwide and, more often than not, the manager is right.

“When you hedge like that, year-in, year-out, you end up making money.” It’s an intensive and long-term process: it involves gathering data, making inferences from the data, and writing software to put the data to use. “You’re doing research so these inferences are based on substantial analysis rather than human feelings, which is a major influence in investment management.” This sounds like a frenetic, taxing process, but Harding said 90 per cent of Winton’s trades were executed in a room not dissimilar to any corporate office, where “people sit in front of computer screens” as models automatically interact with markets. He said it is difficult to know how many algorithms are at work at any one time, or how many algorithms Winton had deployed into markets. “An algorithm is like a house with many rooms. There is effectively one algorithm.” These methods were not breakthrough ideas, but designed to capitalise on market trends and biases.

They work because the majority of investors bestow much legitimacy in the status quo. People trained in orthodoxy find it difficult to perceive new and better ways of doing things, Harding said, but the nature of scientific discovery is to usurp previously believed truths. Harding’s passion for science and philanthropy manifested in his gift of £20 million to the Cavendish Laboratory in the physics department of Cambridge University, a donation large enough to fund 10 start-up research programmes. Harding first undertook algorithmic trading in 1987, when London emerged as a hive of CTA activity. After stints at stockbrokers and Sabre Fund Management, he co-founded CTA shop Adam, Harding and Lueck, which was later acquired by Man Group and today distributed as AHL.

Join the discussion