Will robo-advice mean checkmate for human financial advisers?
With continuing advances in artificial intelligence and the rise in 'robo-advisers', is financial advice following in the footsteps of chess, where computers have long since outstripped humans in ability? Are human financial advisers nearing checkmate or is the endgame still unclear?
The rise in the power of artificial intelligence is well illustrated by advancements in computers' ability to play chess over the last 40 years.
The complexity of chess and sheer number of possible positions and moves presented a huge challenge to programmers. As recently as the 1970s chess computers were very weak players. In 1976 Eliot Hearst of Indiana University, a master level chess player and professor of psychology noted that "the only way a current computer program could ever win a single game against a master player would be for the master, perhaps in a drunken stupor while playing 50 games simultaneously, to commit some once-in-a-year blunder". The idea of a computer being stronger than the human world champion would have seemed absurd.
By the early 1980s, however, computers had become average players and by the late 1980s they could win against master level players. Then in 1996 world champion Gary Kasparov, considered one of the strongest human players in history, lost a game to IBM's dedicated chess-playing supercomputer Deep Blue, albeit while winning the overall match. In a rematch the following year, however, he was defeated by an upgraded version of the supercomputer that scored a narrow match victory which would have seemed impossible 20 years earlier.
Chess programs continued to become more and more sophisticated and efficient and now, another 20 years on, anyone can download a smartphone app that would comfortably be capable of beating the human world champion.
Is the field of financial advice following a similar trajectory with the rise of robo-advisers? Robo-advice, the provision of financial advice with minimal human intervention by reference to mathematical rules or algorithms, has been around as a concept for almost a decade but has become a particular buzzword in recent years. In the FCA's 2016 Financial Advice Market Review Final Report, for example, the FCA was enthusiastic about the use of "mass-market automated advice models" and it subsequently established an Automated Advice Unit to assist firms in implementing robo-advice solutions. Encouraged by the FCA's attitude, many firms have been developing and launching robo-advice offerings over the last twelve months and these are becoming more and more an accepted tool for the provision of advice.
To date, robo-advice offerings have tended to focus more on the less complex end of the financial advice spectrum. The FCA's enthusiasm for robo-advisers stemmed largely from concerns about an 'advice gap', situations where consumers (particularly those with limited assets) are unable to get advice and guidance on a need they have at a price they are willing to pay. The FCA has therefore seen robo-advisers as a way of providing quick and inexpensive advice primarily in relation to lower value or less complex scenarios. Robo-advice is still in its infancy and is generally not yet at the stage where it can provide sophisticated advice in relation to complex circumstances. Certain external factors are also slowing the spread of robo-advice. For example, recent research sponsored by ING suggests that the majority of the public remains uncomfortable with the idea of automated advice and this attitude will take time to change. In addition, PI Insurers may feel, by necessity, that they need to take a relatively cautious view in relation to new and untested technology. Firms implementing robo-advice solutions may therefore face increased insurance premiums, which is a further potential brake on innovation.
For the foreseeable future then, it appears that human advisers have no need to be concerned. Robo-advice is in the process of automating only one part of the wider industry and even then at a relatively restricted pace. But what about the longer term? If we make the reasonable assumption that advances in robo-adviser technology are inevitable does that mean that robo-advisers will one day become so sophisticated and accepted as to render human input redundant?
Returning to the analogy of chess, the strongest players today are often considered so called 'centaurs', where a sophisticated program and a strong human player work together in collaboration. Such players are considered to be stronger than either computers or humans playing alone as it appears that human input still improves overall performance, regardless of how strong computers become. Similarly, perhaps financial advice will one day see a situation in which collaborative human and AI input produces the best possible overall outcome for customers. If so, the role of the human financial adviser looks to remain central to financial advice no matter how capable robo-advisers become, while the overall standard of advice will only improve.