Published on the 31/01/2018 | Written by Jonathan Cotton

The rise of the roboadvisor is upon us and they’re coming for your assets (or the management thereof at least)…

Experts agree: We’re entering the age of the roboadvisor. According to new research from Juniper, assets under management by fully AI-controlled roboadvisors will reach a staggering US$987 billion per annum by 2022.

That figure is a significant chunk of the global asset pie, representing 25 percent of the total roboadvisor-managed pot, and means that fully-automated roboadvisors will outpace semi-automated (i.e. slightly supervised) deployment types significantly over the next few years.

To question is, are we ready for such a state of affairs? Though intelligent systems for managing assets have been around since the 80s (generally to detect fraudulent behaviour) many of us still feel a tinge of fear at the thought of handing over control of our finances to the algorithms entirely – clever though they may be.

“The emotional nature of personal finance means that the presence of a qualified contactable human behind the service is a powerful ‘peace of mind’ weapon,” says Juniper, “leading many in the industry to claim that hybrid roboadvisors will prevail over fully automated services”.

With rapid gains being made in the tech and new platforms popping up daily, their researchers believe however that full automation will ultimately become the new norm.

“Market consolidation is likely to occur,” continues the report, “which will lead to larger players offering various levels of services, with fully automated services being targeted at the mass market. Increased regulatory scrutiny meanwhile is likely to drive service providers to a better understanding of an individual’s circumstances, which should lead to improved performance.”

And with financial advisors (the human kind) looking increasingly favourably on the business opportunities offered by AI-enabled wealth management platforms – expect to see more and more asset managers and independent advisors partnering with tech firms – there’s little standing in the way of something of a robo-revolution in the industry.

Overall, Juniper forecasts that ‘hybrid’ roboadvisors will dominate the market in the short term, managing 66 percent of global roboadvisory by 2022. Human advisor input is also likely to play a key role here, allaying consumer fears of handing management of their cash over to an algorithm.

But that too is likely to change as younger (and less affluent) investors enter the market, more willing to hand over control of their assets to the machines.

“Digital-savvy millennials are rapidly reaching the age where the idea of financial planning is an important consideration,” notes JUniper researcher and author Steffen Sorrell.

“This demographics’ greater inherent trust in algorithms, alongside demand for ‘fire-and-forget’ convenience will drive take-up for AI fully-managed services.”

And the rise of roboadvisory is likely drive down service fees, opening the market to just such investors.

“The traditional financial advisory market has, in essence, been limited by the quantity of human resources that can be provided to offer services,” says Sorrell.

“For this reason, services have normally been restricted to customers with large quantities of capital to invest, which in turn has left out the vast proportion of middle income individuals who, previously, had insufficient capital to meet the minimum service provider requirements.”

Which all sounds great on paper, but there’s a fundamental challenge created by using complex algorithms to make decisions on humans’ behalf, especially those with such serious consequences.

When it all goes wrong, who’s responsible?

This unanswered question is up for debate now, with the European Union’s GDPR (General Data Protection Regulation) – which includes a right to obtain an explanation of decisions made by algorithms – due to come into force from May 2018.

“The complex nature of artificial neural networks, particularly when they are applied as multi-layered ‘deep learning’ algorithms, means that meeting this regulatory challenge will be difficult without significant investment,” says Juniper.

“Neural and deep learning networks’ primary strength lies in their ability to predict outcomes from a range of input variables.”

“The Regulation requires ‘meaningful information about the logic involved, as well as the significance and the envisaged consequences of such processing for the data subject.’ Thus the real test for service providers will be in what ‘meaningful information’ actually constitutes.”

Download Juniper research, Should You Trust an AI with Your Cash.