Last 5 posts


23 March 2016

March of the Robots

The automation of the financial markets continues apace. Robo-advice is the coming thing in money management, apparently. I may have come a bit late to this, although the statistics suggest a gathering momentum: in December 2014, around $19 billion was managed by robo-advisors, and by 2020 that number is predicted to reach around $250 billion. Now, obviously in global terms those numbers are peanuts, but the percentage increase suggests a phenomenon which is not going to go away.

Run a Mile...

My first instinct would be to run a mile from this stuff; if my money is being managed, I like to have access to the human who is responsible for it. Naively, perhaps, I tend to think that gives a greater control. However, if we look a little more closely, things are perhaps a bit more nuanced than that. 

First, it’s important to understand that - at the moment - robo-advice is focussed on relatively small portfolios. The traditional advisor or manager will not look at that level of investment, because the cost would make it prohibitive, so the automated systems are providing a service which would otherwise not be there. 

Secondly, one has to take the name with a pinch of salt, at least for now. The word ‘robot’ almost features, but in truth it’s a bit of a misnomer. The way these things work is to use a series of algorithms to extract the basic information about the putative investment from the client.Then, using that information, the algorithms will lead the investor to one (or a choice of several) investment portfolio(s). At that point, when the investor has made his choice, either the funds will be automatically invested or there will be a human intervention to make the investment. From there, on an ongoing basis, the algorithm will continually rebalance the portfolio to ensure asset allocation remains as desired. So as I say, there isn’t really too much of the ‘robot’ involved, strictly speaking. What there is is an on-line equivalent of a meeting with a mid-level financial advisor, and then a continuing automated balancing of the resultant investment. 

On-Line Activity 

We buy all sorts of things on-line now, so it’s difficult to see why basic financial management should not be one of them. If the cost then makes it a service more widely available across the spectrum of wealth, then that’s probably a good thing. (Incidentally, when I started looking at this, I thought it was really interesting that one fact that came out of this development is that people are more likely - in the circumstances of revealing their wealth - to be honest with the computer that they are to be honest with a human advisor; apparently, we’re all less tempted to exaggerate to impress the computer, while we really like the human to be impressed………..)

So that’s where we are with robo-advice right now. Not really terribly revolutionary, and given that the changing nature of financial provision (pensions and so on) is pushing individuals more and more to take personal responsibility, it’s not difficult to see why rapid growth is anticipated.

Artificial Intelligence

What’s much more interesting to consider, though, is whether or not the use of the word ‘robot’ will come to be accurate. In other words, right now it’s just a sequence of fixed algorithms; but are we approaching a time when the use of artificial intelligence will enable those algorithms to be modified by the machine itself to take account of changing circumstances? Look at driverless cars, for example; they  are a reality. They may not be available for us punters to use on the public roads yet, but one would be naive not to accept that they will be within the foreseeable future. 

I am not - sadly - a denizen of the digital world, so I can only look at this at the level of principle, not execution. But that’s fine; it’s just like I haven’t a clue how you actually stop the driverless cars from colliding, while fully understanding that it can be done. So, in investment decisions, information is ever the key factor. Information used to be the prerogative of the elite in any sector - that’s where specialisms come from. But I would argue now that access to information is far broader than it has ever been, and it’s available far more quickly than before. As an example, Twitter is generally a faster source of news than TV, radio or the printed newspapers; the corollary of that, of course, is that the information isn’t filtered or checked by any outside agency. 

Information and Knowledge

So what if the artificial intelligence is able to gather news and information from the internet and then learn how that information affects markets? At that point, it’s probably not so much robo-advice as actual robo-investment. I realise that this is the world represented in a book I have recommended here a few times before - Robert Harris’ ‘The Fear Index’ - but in the story, it all goes horribly wrong. But assume it doesn’t all go wrong; surely it’s reasonable to propose that, just as AI is being developed in many other fields, it will become accepted in financial markets? 

And then, of course, we can forget all the debate about whether London, New York, Shanghai or Frankfurt is the world’s financial centre: it will actually be in those black boxes sitting in air-conditioned warehouses far away from capital cities, their lights blinking as they scan the airwaves to gather information, the most valuable commodity of all.

Enter text here]

comments powered by Disqus