<!-- Swiftype Variables -->

Special Report: Artificial Intelligence in money management

Broad adoption will mean better beta for all investors

Eric Nierenberg questioned whether broad adoption will leave ‘any alpha left.’

The prospect of artificial intelligence and machine learning morphing from a hobby horse for a small circle of hedge funds to a standard investment tool for big asset management firms could make markets more efficient and less volatile over time, some observers speculate.

It could also make alpha harder to come by.

Hedge funds have led the charge when it comes to artificial intelligence, but in recent years "we've seen it come much more into the mainstream," said Richard Dell, a London-based partner with Mercer and head of the firm's global equity manager research team.

And that's where, potentially, it could have "bigger implications, because there you're generally looking at firms with an awful lot of capital who are driving price discovery," he said.

Some observers say the growing weight of capital might shorten the time it takes to arbitrage away the investment opportunities — predominantly short-term and capacity constrained in nature — that artificial intelligence and machine learning programs are offering.

The number of fundamental stock managers in recent years citing their use of satellite data counting cars in Walmart Inc. parking lots is more reason to invest in the firm selling the data than with the managers using it, said Eric Nierenberg, chief strategy officer with the $67 billion Massachusetts Pension Reserves Investment Management Board, Boston.

With so many managers availing themselves of that data, "I don't know that there's going to be any alpha left," Mr. Nierenberg said.

The question of scale — in terms of the investment opportunities identified by AI and machine learning programs — "will become more relevant as players such as BlackRock (BLK) and Fidelity step up, with the resources they have to put behind this (far exceeding those that) a $10 billion hedge fund can bring to bear," said Timothy Bruce, director of traditional research with Boston-based investment consultant NEPC LLC.

Mercer's Mr. Dell said it's too early to determine if "everyone is going to be using the same machine learning algorithms and coming to similar conclusions, or whether it's going to be people with baked-in advantages because they've developed their own capabilities."

But with a growing number of big asset managers likely to move in the direction BlackRock Inc. (BLK) has gone, relying more on quantitative and data-focused professionals and relatively less on fundamental analysts, the question of what that means for price discovery will arise, he said.

"Does it become commoditized as everyone comes to the same conclusions? Do prices better reflect reality, and does volatility become potentially less because there are less behavioral biases driving markets," asked Mr. Dell.

The enhanced insights money managers can glean by analyzing big data should lead to a more efficiently priced market in the short term, making it harder for stock market bubbles to develop, said Tony Ho, co-founder and managing director of Hong Kong-based Sandalwood Advisors, an alternative data platform focused on Chinese consumer spending trends.

The broad use of big data should better ensure that resources are allocated to companies that are performing well, said Mr. Ho. Managers will know almost instantly when companies are doing well or badly, and bad companies no longer will have much room to talk up their share prices, he said.

"All the information will be out there, people will know," said Mr. Ho, adding that such a transformation could be underway now.