Robo-advisors: Disciplined but Not Intelligent (Just Yet)

September 22, 2016
Robo-advisors: Disciplined but Not Intelligent (Just Yet)

Einstein dies and goes to heaven only to be informed that his room is not yet ready. “I hope you will not mind waiting in a dormitory and you will have to share the room with others” he is told by the doorman. “See, here is your first roommate. He has an IQ of 180!” – “Why that’s wonderful!”, says Albert. “We can discuss relativity theory!” “And here is your second roommate. His IQ is 120!” – “Why that’s wonderful!”, says Albert. “We can play Go!” “And here is your third roommate. His IQ is 80!” – “Why that’s wonderful!”, says Albert. “Say, where do you think the market is headed?”

Excelling at investment strategy would seem to be a low bar for an artificially intelligent machine: Earlier this year, a machine trained by deep learning beat Lee Sedol, one of the world’s best and most experienced players, at the ancient board game Go. Yet, digital wealth managers, or robo-advisors, are not beating humans at the investment strategy game; they are not even playing. And here is why:

Today’s digital wealth management solutions are based on the same investment principles and models used for the past three decades. 21th century technology is using 20th century concepts to decide on asset allocation and portfolio diversification. And there is nothing wrong with that – those models are based on timeless insights. Technological advance can refine them and bring more computing power to the table but the core concepts remain the same. It is telling that Wealthfront, a leading digital wealth manager, has Burton G. Malkiel serve as its chief investment officer. Dr. Malkiel is an 84 year old finance professor, best known for his book “A random walk down wall street”, first published in 1973.

So, if not in investment strategy, where did the digital revolution take place in wealth management?

Technology has made it possible to efficiently roll out portfolio management solutions to a large number of accounts. This brought down cost, resulting in low fees and a reduced minimum deposit requirement. Thus wealth management can be delivered to a client segment that did not have access to such services under traditional wealth management offerings.

In addition, cleverly designed investment tools and easy-to-use interfaces create a user experience rivalling and even surpassing that of a human advisor. What the ATM has done for bank branches and online banking to off-line money transfer is now being done for wealth management: a service that’s available anytime, anyplace and on any device. And while risk tolerance questionnaires, low cost diversified investments, such as ETFs, and immediate market access through online brokers have existed for at least a decade – it is only with digital wealth managers that all these elements are integrated on a single platform.

There is another area in which robo-advisors can outperform a human investment manager: digital wealth managers rely on rules-driven investment decisions, governed by mathematical formulae and quantitative criteria. While a human advisor consults similar models, automation brings discipline to the table. Human bias is taken out of the equation and decisions are consistent and unemotional. Nobel-prize winning psychologist Daniel Kahneman notes in his book ‘Thinking, fast and slow’ that, in general, formulae trump human judgement. And digital wealth managers rely on formulae.

Low cost, an integrated user experience and discipline are forecasted to attract up to US$ 300Bn* in assets by the end of the year to firms such as Betterment, Wealthfront, Motif and Personal Capital. This success lays the ground for further sustained technological advance in the sector. Digital wealth managers are already beginning to experiment with the application of artificial intelligence (AI) to their marketing (if not their investment) strategy. Salesforce already applies AI to customer relationship management. Their product is called Einstein.

*Source: AT Kearney