Jacqui Henderson | 17 Aug, 2016

Why robo advice is a furphy (and Pokémon Go is not an Olympic sport)

Linking the terms ‘robo’ and ‘advice’ together is to me the same as saying Pokémon Go qualifies as an Olympic sport.

Some might consider the popular smartphone app a worthy athletic pursuit…but it all depends on your definition of sport.

The elasticity of definitions is the central point of this blog, and the reason why I believe the hype around so-called ‘robo advice’ has gone unchecked and needs a referral to the fact department.

For the record, then, let’s understand the fundamental elements of ‘robo - advice’.

Robo is using artificial intelligence to mimic and predict human behaviour. Advice, in it’s true nature, is goals oriented and holistic with interrelating elements of a consumer’s cashflow, investment, super, risk and estate planning etc.

No so-called ‘robo advice’ offer I have seen, and I’ve seen a few, including the Betterments of this world, have solved the challenges of a) behavioural science and b) advice that is goals based and holistic.

Human beings differ in their financial decision-making abilities in distinguishing the effects of risk (known probabilities) and ambiguity (unknown probabilities). Human behaviour has ambiguity, aversion, time, knowledge, beliefs and influence from others and there are many theories like Elsberg, Savage and even quantum cognition being used to develop models to predict human behaviour, but the closest we have seen in the finance world is using the ‘trade off’ theory of risk vs rewards, money today vs money tomorrow and the impact of self vs other, by Professor Shachar Kariv. This is a fundamental element for any ’robo’.

I am not the first to say it, but surely from the point of view of supporting ‘high quality financial advice’ – shouldn’t we first agree to a functional definition of robo advice, if not ditch the term altogether?

 

Digital advice automation

So what is the alternative language for so-called ‘robo advice’? I submit my vote for digital advice automation.

This is because not one clever FinTech ’robo’, IT lab or multinational software corporation like Microsoft has yet developed a deterministic algorithmic application that predicts "uncertainty" - the irrational part of the human brain - alongside "certainty". 

It’s not to say that we aren’t headed in this direction, because when scientists, technologists and professors of our world solve how to mimic and predict human behaviour, that is when ‘robo’ will become true robo.

I prefer the expression digital advice automation because it more accurately reflects our current position.  We human beings will always need human connection, augmented by smart, semi and fully-automated digital tools.

Clients still need to "validate" their thinking, and will always value the human element and relationship of a trusted adviser. The fact is that to a consumer, financial advice is holistic, it all interrelates, it needs to take into consideration proper ‘needs analysis’, goals, client’s actual position in regards to cashflow, (income & debt), risk protection, tax, structures, estate planning and so on.

There is no ‘robo’ that does this.

Some so-called ‘robo’ offers deploy human advisers who provide validation of their advice. It’s just delivered via virtual means like phone and online chat. Personally, I'd like to see some of the robo models under the event of a GFC and see how many "5 question" risk profiles (which is supposed to make up the “needs analysis with best interest duty” component of the advice process), how adequately they predicted their client’s "preference" towards risk and asset allocation, the human behavioural component. 

Maybe then we could also answer the question of whether or not it’s a good idea to allow Pokémon Go players into the Olympics.

Footnote:

I had some interesting reaction to an earlier blog on this subject earlier this month. In that post I argued that the current software development race to ‘robo’ the advice process in financial planning falls short as the essential stakeholder – the client – has largely been ignored. The technology medium, not the end user, has somehow become the central point of focus in public discourse.

This observation received a warm and supportive response.

And, bless his contrariness, one detractor. Mr Kidd from Omniwealth cited Betterment, in his words, one of the ‘top robo providers’ in the US as an example of ‘real substance’.

Again, the hype belies the reality. Yes, Betterment is a pioneer in this space, now managing over $4.8bn AUM. The company recently raised $100m in capital funding, valuing it at $700m, or 14.6% of the value of its AUM. 

Let's put this into context: this is 12x the valuation of Blackrock, that is valued at 1.2% of its AUM. Analysts forecast Blackrock to make profits of $4.6bn this year. Betterment does not appear to be making a profit.

Betterment’s average account size is $28,571 and 0.25% pa fees implies $12m pa of revenues. However, it's reported their total costs are estimated at $50m a year, consisting of high cost of acquisition, administration and development costs. They’ve hit hard competition against Schwab’s ‘free offer’ and Vanguard’s 0.30% robo offerings, reducing their rate of growth in this competitive landscape.

Morningstar estimated last year that robo-advisors need at least $16B and as much as $40B of AUM just to cover core operating costs and recoup advertising expenses, and that robo-advisers may need $50B – $80B of AUM or more to justify their $500M – $700M company valuations, the current linear growth pace of even $150M per month implies that Betterment and similar business Wealthfront may not even reach $10B of AUM by 2020, a mere 1/200th the size of what a sensationalist A.T. Kearney study projected for robo-advisers just a year ago!

Read more about financial advice automation.