Resources being ploughed into artificial intelligence (AI) funds that utilise big data are set to result in an array of new investment products, which could put a swathe of the active asset management industry out of business, say experts.*
Today, robo advisers are pretty basic: in short, investors input their risk parameters and the algorithms behind the platforms assign the money into applicable exchange-traded funds (ETFs).
“So far artificial intelligence has been utilised in very basic decision-making, such as asset allocation,” said Katsuhiko Okada, chief executive and chief investment officer of Japan's Magne-Max Capital Management. “What [US financial services firm] Wells Fargo and others are doing is [replicating] the financial adviser’s role [in] machines, but it’s very simple stuff; not even machine-learning algorithms."
Robo ripe for development
Today’s leading market players are quite small. Some $4 trillion globally is invested in ETFs, according to research firm ETFGI; but the two largest US robo advisers — Betterment and Wealthfront — have just $12.4 billion in assets under management between them. And while around 200 robo advisers operate in the US, the Asia-Pacific region has only about 50, of which 14 are in Japan and eight in Australia.
But robo advisers are likely to become more popular and more sophisticated, using AI algorithms to analyse masses of data on market movements, economic conditions and asset valuations, before making appropriate investment decisions. In essence, they will act more like active fund managers.
“Over the next 10 years we may see financial technology move from robotics to machine learning to AI, and from there to the holy grail: cognitive science,” said Justin Ong, asset and wealth management leader for Asia Pacific at consultancy PwC.
Cognitive science employs machines that can crunch enormous quantities of data to analyse trends and make predictions. It already exists in some industries, such as manufacturing and healthcare, but has yet to gain traction in investing because of regulatory barriers.
That is likely to change, with potentially huge ramifications for the fund management industry. “In 10 years, I think the competition in the fund management industry will be in designing the best AIs, not among human fund managers…and trading will be filled with lots of engineers instead of fund managers,” predicted Okada.
Yet there will always be a place for humans, he argued. “AIs are extraordinarily good at predicting patterns humans cannot detect. But [US president] Donald Trump’s election could not be predicted [by AIs]. The noise of trading behaviour is likely to remain.”
AI pros and cons
The prospect of big data and better AI analysis tools also presents possibilities for active fund houses. They could amass far more data on the financial health of companies and how well they are run. That would help embed environmental, social and governance (ESG) practices into portfolios — and help astute investors make money.
Meta-data analysis — statistical analysis combining findings from multiple scientific studies – would also support more in-depth factor investing strategies and could flag when it was time to shift from one to another.
A downside for asset managers is that AI and automated funds will further erode fees. Okada believes this is inevitable: “AIs will make the cost structure of funds more transparent.”
PwC's Ong agreed, but noted that the operating profit margin of fund houses was often above 30%, so they could cope even if this dropped by half. “It’s not capital-intensive, so you can run a good business whether you are managing $10 billion or $100 billion.”
* This is the second article in a series on how financial technology will reshape investment management, of which the first piece appeared on June 22. Look out for further articles on other topics in the coming days. The full feature appears in the latest (June/July) issue of AsianInvestor.