Wealth Management: Strategy blog – Putting the AI in finance

This is the alt text

Strategy team - Kevin Gardiner and Victor Balfour (Wealth Management)

We're told Artificial Intelligence (AI) will change finance beyond recognition. Jobs are at risk:  elsewhere, in US manufacturing, one estimate suggested that new technology cut headcount by a quarter between 2000 and 2010. Is finance equally vulnerable?   

It will depend partly on how big the 'processing' element in finance is. A recent AI seminar at Centre for the Study of Financial Innovation (CSFI) reminded me that some non-practitioners can have an unusual view.     

As a City-based economist, I was first quizzed about markets by AI programmers in the late 1980s. The focus then, as now, was on trading and information. Could trading be automated? Could software spot trends, and second-guess market responses to data releases? 

Progress since then has been dramatic in some ways - and underwhelming in others. 

Better communications, and lots more data, have slashed informational inefficiencies. Algorithmic (computer rule-driven) and high frequency trading have taken off, thanks to computing power and off-exchange dealing pools. 

Mountains have been tunneled to shave nanoseconds (a billionth of a second) from stock trading times. Costs and spreads have collapsed.   

The main winners have been specialised traders and funds, and established institutional investors. There has been talk of (allegedly) sharp practices in order execution, but trade times and costs have fallen for retail investors too. 

As with bitcoin 'mining', we wonder whether society really needs those tunnels, those faster trades. But we feel the same about lots of things - and in other contexts the narrow-mindedness of free enterprise is a strength. 

The point though is that the programmers' view of finance - like so many others' - is a caricature. Automatable tasks may not be as big a proportion of  financial services' cost base as in manufacturing. 

Trading and data-mining is not where most people in financial services work. Few of us think anyone - person or machine - can mechanically beat the market, or predict the future. City economists do not spend all their time forecasting. 

For sure, AI will cut other processing costs, such as administration. Robots are reading forms - and not just in finance: legal services are also benefiting. Improved encryption may yet transform custody (in a way that blockchain hasn't and, we can now see, probably won't). 

Nobody knows how big the net job losses will be, but the room for labour saving may be overstated.  And some cost savings may be reinvested to grow businesses - even, perhaps, to provide better financial 'service'. History suggests new industries will come along to help absorb those displaced, even if we can't see them yet.  

But again, the point we wanted to make is not so much about the scale of the pending changes in transactions and processes. Rather, it's whether there is more to finance than those activities - and whether the 'I' in 'AI' is quite what it's cracked up to be.   

The discussion of AI in finance is focused on processes, from security research and trading through to settlement, account opening and stewardship. Its wider relevance is less clear. 

Pundits who see financial services as one big foreign exchange dealing desk - not as uncommon a view as you might think - might be forgiven for thinking wheeling and dealing around the short-term direction of the dollar is central to finance. But it isn't (luckily: it is unpredictable anyway). 

Much financial activity is focused on providing advice and connections, and the big questions we face tend to be conceptual, about causality rather than correlation. Such questions are not easily answered - by us, or machines.     

For example, what really drives interest and exchange rates? Is there too much debt? Is a PE of 16 too high? Will investors fund a new issue? Does a merger make commercial sense? Can we insure against catastrophe? What is a 'true and fair' picture of a company's health? How might we best safeguard the real value of our savings?  

Like all big questions, these can't be solved by throwing more computing power at them. Deep Thought said that 'the answer to life, the universe and everything' was 42, remember. The A in AI can variously stand for Artificial, Augmented or Autonomous - but it's the Intelligence that will create lasting value. Machines may not have it yet. 



Past performance is not a guide to future performance and nothing in this blog constitutes advice. Although the information and data herein are obtained from sources believed to be reliable, no representation or warranty, expressed or implied, is or will be made and, save in the case of fraud, no responsibility or liability is or will be accepted by Rothschild & Co Wealth Management UK Limited as to or in relation to the fairness, accuracy or completeness of this document or the information forming the basis of this document or for any reliance placed on this document by any person whatsoever. In particular, no representation or warranty is given as to the achievement or reasonableness of any future projections, targets, estimates or forecasts contained in this document. Furthermore, all opinions and data used in this document are subject to change without prior notice.

More Information

View more Wealth Insights

Back to top