Just how far can AI go in transforming wealth management?

The impact of Artificial Intelligence (AI) on the wealth and asset management industries has been a topic of furious discussion at a number of recent conferences, both among the speakers onstage and between delegates on the sidelines. On one the one hand, many can envisage the efficiency savings and data analytic capabilities that could be on offer, yet others fear job losses and the end of face-to-face relationship management.

Talk of the game-changing potential of AI has been going on for decades, said Alessandro Tonchia, co-founder of Finantix, the global software vendor, speaking on a panel at the PIMFA Summit in London. “I studied AI 30 years ago at college and back then, people said we were five years away from a fully intelligent, cognitive computer. Today, people say we are five years away from a fully intelligent, cognitive computer. There has been hype all along.”

Nonetheless, he claimed that by taking “concrete” problems in wealth management, and looking at them in detail, AI can achieve real results.

There are lots of aspects of wealth management that technology cannot replace, said Mr Tonchia, but a great deal of what goes on in the background is very mechanical in nature. For example technology can help build up a more complete profile of clients, along with helping to understand how markets are behaving and the likely impact it will have on portfolios. This is all useful information and can also help relationship managers be better prepared for face-to-face conversations with clients.

There has been a tendency for wealth managers to dismiss AI, said Wendy Spires, global head of research at Wealth Briefing, with many believing they can ignore it for a few decades more. “This is wrong. It is having a huge impact in financial services right now.”

But incorporating AI would not mean a wholesale replacement of the human element, she said, rather it can be used to take care of the day to day grind all firms need to undertake, leaving the value-add element to humans. Ms Spires also highlighted the fact that profitability was falling for many wealth managers, and the efficiencies technology can bring would help reverse that trend, while robo-advice platforms are also bringing advice to those clients who cannot afford more traditional wealth management services.

The best use of AI is when you do not realise you are using it, said James Howell, wealth management lead at Accenture. “The majority of the work we do is outside of financial services, and in our daily lives we all touch products which have AI behind the scenes.”

He agreed that when it comes to wealth management, human interaction is absolutely critical. “All of our research shows that. So it is what enables those humans to do their jobs well that is the key for technology providers like us. It is about augmenting the adviser not replacing them. The end client should really be unaware that these tools are being used.”

Many investment managers are already using alternative data to help them make decisions, claimed Mr Howell. He gave multiple examples of the kind of data they were using, for instance satellite imaging, the monitoring of retail car parks, crop production and yields, activity levels at industrial facilities and more, all of which can help managers make investment decisions.

The message is, if you don’t think this is real, believe it, it is. Investment firms are using it today. Some are very open about this. Others less so.”

For many people, when we hear the phrase AI, not only do we fail to really understand what it means, as no one really tells us, but we also instinctively fear it, for what it could mean for our jobs, said John Beckett, the author of New Fund Order, speaking at Inside ETFs Europe. 

“We still seem to be somewhat blindsided by the rapid advances in technology that are coming,” he said. “I think that is because most of those advances are happening outside of our industry. According to a new report, global GDP from AI will be the equivalent by 2030 of $15.7tn, more than the combined GDP of India and China today.”

Mr Beckett discussed the impact that AI could have on the ETF industry. That industry might be in rude health right now, with volumes at all times high and growth forecast to continue, but the next evolution will involve some difficult choices. He painted a picture where the ETF industry could keep going along the path it is on now, dominated by funds tracking traditional industries, or future products could adopt ever “smarter beta”.

Or, he predicts, the future could be one of increased use of smarter beta, which effectively gives way to AI beta, underpinned by Blockchain technology.

ETFs are not yet reaching their full potential, claimed Mr Beckett, because so many are tied to indices. “Yes, using indices was a simple solution to a problem, but we have become ever more obsessed by benchmarks. The concentration of assets into a handful of market cap indices is not going to end well. Indices do serve a purpose, they used to tell us how capital had been allocated in the market, but they have begun to tell us how to allocate capital and they are not meant to do that.”

So if market-led indices are not the future of ETFs, what is? We are already going beyond the index with the rise of smart beta products and factor investing, he said.

And the levels of complexity in some of these products will require AI and automation, said Mr Beckett. “In my view, the advent of Blockchain architecture combined with AI machine learning will create a form of ETF. ETF 2.0 if you will. The programming language and skills required to do this already exist. It is just the ideas and the impetus that are required,” he said, adding that the launch of the first AI ETF, using IBM Watson technology, has already occurred, and this is just the start.

AI is yet to be transformational to the wealth management industry, says Dominic Snell, head of consulting at WealthDynamix, but its potential to be a differentiator in the market is making firms seriously consider how this technology could be employed, and the main challenges in doing this.

“Firms looking to establish a capability in this area are keen to identify some suitable use cases which can be used to prove its value before investing further,” he says, though he cautions that choosing a partner is key as the skills and methodology required are quite different from more traditional software implementations.

One barrier to any firm adopting innovative technologies such as AI is the data that they have to power it, says Mr Snell. Traditional firms who have not taken steps to address lack of integration in legacy system and data issues are likely to find that time to market is longer than it will be for those companies which grew up being technology-led.

“The culture of the firm is also vital – if the view of technology is positive, and it is seen as a trusted partner in driving the business forward then the sponsorship, skills and resource needed to succeed in this area will be easier to acquire.”

Nonetheless, AI also provides technology departments in traditional companies with the opportunity to change perceptions held both internally and by their customers, “by delivering cutting edge solutions which position the firm as innovative”.