Can You Pick Managers Better Than an Algorithm?

Prove it, in the name of science.

Photos by Bigstock

Photos by Bigstock

Stanford University is challenging asset allocators to demonstrate their ability to pick superior fund managers for a new study, which pits human investors against an algorithm.

The school’s Global Projects Center, led by executive director Ashby Monk, developed the algorithm which it says can “select managers with very little information in very little time.” The study, which launched Tuesday, aims to assess whether this algorithm beats, matches, or underperforms real-world professionals.

“We are extremely interested in how technology is going to change the way people allocate capital,” said Monk, who is also an Institutional Investor columnist. “We’ve come to realize that technology is one of those great opportunities to improve investment decision-making and performance. And one of the components that you can’t ignore is artificial intelligence.”

[II Deep Dive: And the Artificial Intelligence Winner Is…]

The survey asks respondents to compare and rate two external managers based on one-page summaries of performance history, team experience, or other details. The managers are real. Names of firms and individuals have been anonymized, but the biographical and historical data reflect actual asset management teams and their track records.


Researchers encouraged anyone who influences capital allocation decisions to take the survey, including consultants, service providers, and institutional allocators like pension funds, endowments, foundations, and sovereign wealth funds.

“If you participate in the decision of an LP choosing a GP, we’re interested in hearing from you,” Monk said.

The survey will remain open for two weeks. The Global Projects Center expects to publish its findings later this year.