How do asset managers see AI?

Once the playground of quant-based asset managers, artificial intelligence is seeing much greater adoption across the industry, according to new research from Mercer Investments.

According to the report, which draws from a global survey of 150 asset managers, nine out of 10 managers are either currently using AI (54 percent) or have plans to use it (37 percent) to support their investment strategies or their research into specific asset classes.

Despite some confusion around the exact definition of AI, the survey revealed clear consensus on the “core capabilities” of the technology, including generative AI (93 percent), large language models and natural language processing (93 percent), and machine learning (85 percent).

Among the investment teams that have already integrated AI into their processes, 43 percent say they’re using the technology to incorporate alternative data into their decisions, while over a third (37 percent) are using AI to develop forward-looking signals. Another third (33 percent) are using AI as a tool to analyze market indicators.

More than half of that cohort (56 percent) say AI informs their decisions – as opposed making the final call – while 20 percent say they let AI propose decisions, which they can then overturn.

“We are mainly seeing investment managers applying AI to augment existing capabilities, but there is opportunity to deploy AI in more complex aspects of portfolio management, such as portfolio construction and rebalancing,” said Joanne Holden, global head of investment research and consulting at Mercer Investments.

When asked about the potential for AI to drive value across asset classes, the respondents were most likely to see a “significant” or “very significant” opportunity in equities (61 percent), followed by hedge funds (53 percent), fixed income (47 percent), and digital assets (45 percent).

Through a sector lens, managers already using AI were most likely to say it can drive value for tech investments (83 percent), followed by health care (72 percent) and financial services (70 percent).

The survey also looked at the risks and challenges around the use of AI. Among managers already utilizing AI, more than two-thirds (68 percent) cited data quality and availability, followed by integration and compatibility (54 percent) and ethical and legal considerations (51 percent).

The report also revealed concerns around divergent AI regulation, with 48 percent of AI-integrated managers saying it’s a “very significant” or “significant” risk.