AI is promoted from back-office duties to investment decisions

Asset managers are increasingly using artificial intelligence to guide investment decisions, track the habits of portfolio managers and identify moneymaking opportunities.

JPMorgan later this year plans to expand the use of a generative AI tool that flags questionable decisions by portfolio managers, such as potentially selling top-performing stocks too soon, company officials told the Financial Times.

The tool, dubbed “Moneyball”, is meant to show portfolio managers “how they and the market have behaved in similar circumstances and helps them correct for bias and improve their process”, said Kristian West, head of investment platform for JPMorgan Asset Management.

Other fund managers are using AI to complement human analysts, identify targets for litigation finance and explain allocation decisions to investors.

These disparate efforts show how the AI arms war in asset management is shifting from paperwork-intensive compliance and marketing tasks towards helping investment professionals make smarter decisions.

The JPMorgan tool — part of the $3.2tn manager’s Spectrum portfolio management platform, which is fuelled by about 40 years of data — is a pilot program that is still in development and will be made available to a broader group of portfolio managers later this year.

Voya Investment Management already has implemented a virtual analyst that can monitor stocks for potential risks, complementing the $331bn manager’s human research staff. Portfolio managers have access to a dashboard in which a human analyst’s review of securities can be viewed alongside AI feedback, such as red-flagging a stock.

So far, Voya’s AI analyst has demonstrated a good ratio of right and wrong decisions, making its alerts “a high-value signal”, said Gareth Shepherd, Voya’s co-head of machine intelligence. He likened the process to a pilot and co-pilot reading signals from an aeroplane’s flight management system.

“The flight management system augments the pilot’s decision-making, but the pilot has the final say,” Shepherd said.

Legalist, which runs a $1bn hedge fund focused on litigation finance, uses a proprietary AI search tool called “Truffle Sniffer” to find attractive investment targets among a sea of civil suits.

The “sniffer” scans court records for signs of favourable outcomes, such as friendly judges and litigation classes as well as pre-trial rulings that indicate particularly strong cases.

“We look for cases that have clear signs that they are winning but have not collected the money yet,” Legalist co-founder Eva Shang told the FT.

In some cases, AI has a good deal of say, as is the case with an AI-powered exchange traded fund from South Korean conglomerate LG and SoftBank-backed Qraft Technologies.

Their LQAI ETF, which launched in November and relies on a proprietary AI stockpicking tool, has evolved to incorporate an AI-generated monthly holdings report. A recent AI-generated report explains its decisions to favour certain companies and sectors while selling out of others.

“As LQAI’s portfolio manager, I increased investments in resilient and technologically forward companies like (Google parent Alphabet), while slightly reducing exposure to companies with traditional media struggles, reflecting a cautious yet optimistic approach to leveraging growth opportunities amid financial variances,” the AI-generated holdings report said.

Despite the developments, AI’s potential to drive long-term returns for asset managers has its sceptics.

Veteran portfolio manager David Giroux, who manages the $59bn T Rowe Price Capital Appreciation fund, argued that most of the AI-focused intellectual capital in asset management is geared towards finding a short-term edge as opposed to the trickier task of estimating earnings potential years into the future. 

“I think there is very, very little that AI is going to do to make that inefficiency go away,” Giroux said.