how advancing artificial intelligence (AI) might challenge investment managers

"How will this grieve you,
When you shall come to clearer knowledge!"

The Winter's Tale

“Well, what are you going to do about it then – when within a few years computers will be making all the investment decisions?” So the diligent enquirer might ask. 

Here we enter the sci-fi world of artificial intelligence (AI) – except that it might no longer be -fi! Ours is already a world where computers can script speeches, direct a surgical pathway for the cardiologist’s instruments or drive a car without a driver’s intervention.  

Of course, computers have for decades been work-horses in the financial sector performing simple arithmetic calculations.  But digital technologies have moved on.  Some scientists now expect, within the visible time horizon, to develop a computer that out-thinks human intelligence.  The character and potential implications of generative AI computers (GAIs) are so weighty as to have drawn Roula Khalaf, the Editor of the FT to make a rare personal statement (27th May) setting out the paper’s response to the AI revolution.  Comfortingly she confirms that the FT’s content will continue to rely on human oversight and appraisal. Phew! 

We remind ourselves however, that almost 40% of U.S. stock market activity is already reliant on arithmetic functions performed by computers.  The functions involve machines which can tot up more or less instantly the composite value of a given set of market index constituents.  But digital capacity has already grown far beyond the rudimentary calculations required by ‘index tracking’.  

AI confronts us with the prospect of a hugely widened digital capability.  It involves nothing less than an all-encompassing computer that can assess and respond intelligently to any given question put to it – a robot that can think for itself.  Americans have already invested billions of dollars in companies delving ever deeper into the digital potential. 

Even so, a fundamental divide separates the workings of an omniscient computer from those of an investment manager.  Scientists are defining an ever widening scope for GAIs.  Their efforts are likely to lead many others on the journey towards a growing reliance on GAI outputs. 

By contrast the investment manager searches for intelligence that identifies an unfrequented pathway through the investment thicket.  It leads to an unfashionable destination.  The manager hopes to reach it alone.  

Every investment manager strives to develop a personal take on the investment prospects.  In so doing they hope to outwit other participants in the investment search.  All of them need to be able to point to a goal achieved.  But what if investment managers can’t beat the chatbot - will they lose their jobs?  

These fanciful projections refer to a computer that can answer any enquiry put to it.  Such an invention raises all sorts of complex social, governmental, and philosophical questions.  Leaving these aside, our basic question remains unanswered.  In a world where all knowledge can be accessed and shared, what value remains within an investment manager’s expertise?  

Given any set of circumstances, data supporting the optimal investment conclusion will presumably be stored in the GAI’s electronic libraries.  If that is so, how can the investment managers’ route to achievement remain distinct and valuable when nobody can generate any distinguishing characteristic?  Science is developing the means to render ‘performance’ obsolete as most currently know it.  

Certainly some investment managers would have to redefine what they are offering in response to the arrival of digitally generated omniscience.  

Nevertheless people may harbour reservations about what may prove a mixed blessing.  The development of nuclear energy in the 1940s brought consumers both a secure source of power and, arguably, a threat to their very existence.  But such threats were countered by the reality.  In the event nuclear energy plants proved to be adequately safeguarded.  Deep memory computers have not wiped out competitive chess.  The use of AI at tournaments has simply been proscribed.  Players are happy about this (apart from a few miscreants.  During an interval one recent competitor sneaked a computer into the gents).  Amateurs still enjoy competitive chess in its natural form.  But they too will probably observe that the impact of technical advance can be as destructive as it can be life enhancing.  

In no sense does AI mean that we managers have no future.  But we need to reaffirm a longstanding and different measure of our worth.  We have always sought to offer clients the value of simple satisfaction.  It affords them the prospect of a more or less carefree financial existence, an outcome built on trust and presumed competence.  

Rather than dwell on apprehensions about the advance of GAI computers we do better to consider the likely character of whatever may arrive.  The clearest mark of distinction is the soulless nature of a computer’s calculations.  It has no emotional capacity with which to weigh the cost of tears or the joy of music.  Mozart’s work enchants the listener because of its unexpected cadences, not because of any conformity with predictions suggested by an algorithm.  Yet Mozart’s music evokes a sense of indefinable beauty.  That of a tunesmith like Salieri does not.    

As to the nature of AI, the world of chess again sheds some light.  It was only possible to design “Deep Blue”, the winning computer programme, because the vast number of possible chess positions can be counted.  It is limited by the rules of the game.  Human behaviour however is not so limited.  Its outcomes are numberless.  AI projections must quickly dissolve in the mists of infinity. 

Perceptive readers may notice that we have not fully answered the question posed in the first paragraph.  How can we distinguish ourselves when computers provide all the answers?  The answer is simple.  

As human beings we possess emotions, sensitive to the moods of others.  Computers have no emotions.  Relationships of trust and loyalty, our primary goal, are shaped by emotion, a driving force unknown to any computer, even a GAI one.  We human workers alone possess the power to receive and react to all the questions and goals that make our lives what they are.  There is little reason to fear the advance of technology.  Indeed, we expect to take advantage of the new opportunities it will bring with it.  

The job of an investment manager is to investigate the life choices made by others, particularly company leaders and all who profess to have created a recognisable and distinctive company culture.  All very well you might say, but every corporate CEO claims to promote a valuable and distinctive culture of their own – how do you know which are for real and productive?  Agreed, we cannot tell for sure any more than we can predetermine the behaviour of individuals.  As for this firm we can claim that in creating a culture of our own we have achieved something visible and valuable for our clients and staff alike.  It does not rely on a steely assessment of an infinity of possibilities.  It depends on a search for the human qualities that make for satisfaction and loyalty.  This is a culture that works – it has done so here.  

1st July 2023

Back to Articles
Download PDF