GPIF to explore using AI for long-term asset management

The world’s largest pension fund has hired Sony Computer Science Laboratories to look into incorporating artificial intelligence in its investment operations.

Japan’s $1.3 trillion Government Pension Investment Fund has teamed up with Sony Computer Science Laboratories to look into using artificial intelligence in its investment decisions and manager selection.

GPIF has engaged Sony CSL following a request for proposal issued in July to study the impact of AI on asset management, with a view to use the technology for the long-term asset management of the pension fund, it said in a statement.

“Given recent developments in AI technologies, GPIF is increasingly convinced that AI would be able to assist our fund investment operation, from asset allocation to asset manager selection and evaluation processes,” Hiro Mizuno, executive managing director & chief investment officer of GPIF, said in the statement.

He added that GPIF is committed to keeping up with the rapid development of new business models in the asset management industry caused by the adaptation of new technologies such as AI.

Sony CSL’s research includes an analysis on the impact of AI on asset managers’ business models as well as manager evaluation methodologies, the use of AI to reinforce long-term investments as well as to reconcile asset manager’s investment decisions and tracking trading data.

A final report is expected to be released next spring, a spokesperson for GPIF told Private Equity International. She added, however, that the pension has no detailed timeline on its adoption of AI for its investment operations.

Richard Folsom, founder of Japan-focused private equity firm Advantage Partners, said: “Japan certainly has global leadership in AI and robotics, and GPIF saying it is going to be a model investor in using that technology is an interesting development in the asset management industry.”

Across the financial industry, machine learning is transforming asset managers investment strategies. For example, insurance firms and investment banks have employed machine learning-based systems to automate areas such as claims processing and contract validation. BlackRock is one fund manager that uses technology embedded with AI in its investment processes and in providing investors access to analytics previously only available to large institutional clients. Meanwhile, investment firms Cerebellum Capial and Sentient also use machine learning to evolve and optimise its trading algorithms.

In April this year, GPIF started its recruitment drive for private equity, real estate and infrastructure managers and received proposals from 23 private equity firms in August.

GPIF has a five percent target allocation to alternatives, of which just 0.10 percent has been allocated to alternatives as of end-September this year.