Technology will not replace humans in the financial markets - Financial Times

Recently I watched a computer playing a game that I’d enjoyed as a boy — Pong. It is one of the oldest computer games, in which a bat bounces a ball against blocks, attempting to destroy them all. The computer was terrible at first, but by the end of 600 iterations, it was unbeatable, learning from its mistakes not only how to win, but how to win in the swiftest possible time.

It’s an innocuous example, but one which illustrates the significant advantages that computers have over humans in the performance of routine tasks.

Goldman Sachs recently announced that two-thirds of all trades on major stock exchanges are now carried out by computers. Despite this, we strongly believe that humans still have a central role to play in the financial markets, and will do for the foreseeable future.

Let’s consider one element of our fundamental research process — the visit to a company’s management. Human beings are deeply social creatures, with the ability to use instinct, to read body language and to judge atmosphere.

It’s hard to imagine a computer knowing precisely when to ask the killer question of a chief executive, and how to judge the response. The portfolio manager will have seen numerous similar management teams and will be able to measure the chief executive’s answers versus his or her experience of similar situations.

Will computers be able to get a feel for the culture of a company, for how the employees are being motivated, for the ambience on a factory floor?

The answer is maybe — but not yet. Deep learning is moving quickly and is making advances which, until only recently, seemed unimaginable. There is technology in development that will analyse company conference calls, letting portfolio managers know not only whether the management is telling the truth, but also how confident they are in what they’re saying.

In our view, however, such analysis will for some time yet, still need to go through a human portfolio manager who can synthesise it into an effective investment decision.

We were recently told by the head of research at a leading investment bank that 2016 would be the first year that he had not hired a single analyst. All of his recruitment budget went on technology and people to operate that technology.

Humans will have an integral part to play but it is not unfair to say that the investment firm of the future is likely to look very different to its current incarnation.

We think that there will be fewer people, and many of those who are there will be tasked not with executing trades and analysing results, but with designing and validating algorithms, with maintaining and optimising the vast amount of technology that will be needed. At the same time, we believe that it will still be humans, rather than machines, making the most important decisions of all.

All this technology will merely support human functions in the efficient execution of routine tasks, rather than replacing them wholesale.

For the time being, humans do fundamental analysis better than machines and I believe this will remain the case for some years, perhaps decades, to come. People judge the calibre of soft data about companies better than computers.

Chief executives are human — illogical, impulsive, unpredictable — and it takes a human analyst to predict the vagaries and eccentricities of their behaviour. As such, we believe that humans will continue to be the ultimate decision makers in the investment process, utilising the advantages that technology confers.

Pierre-Henri Flamand is the chief investment officer of Man GLG

Sample the FT’s top stories for a week

You select the topic, we deliver the news.

http://ift.tt/2hc3cJf