AI could help Asian fund firms win global mandates, says fund veteran Gerard Lee

VETERAN fund manager Gerard Lee worked for less than two years marketing real estate after university in the 1980s – then he landed a spot in GIC’s traineeship programme.

That coveted stint, which he found “more cerebral than expected’’, launched him into a four-decade career in asset management, of which nearly two decades were in the C-suite.

Just last month, Lee, 64, stepped down from Lion Global Investors (LGI), where he was chief executive until last November. “I’m not retiring from life. I’m retiring from a full-time job. I think I have a runway of around 20 years before I kick the bucket. I’m still very useful, still full of ideas,’’ he says.

“My wish list is to be able to get into one of the fund management entities, maybe from Japan or China, which aim to build something global. I can help them think strategically to find a path of least resistance.’’

In his reckoning, fintech – specifically artificial intelligence (AI) and machine learning – could help Asian fund managers leapfrog growth challenges, which have become even more daunting since the emergence of large-bracket managers with trillions in assets under management.

Lee’s track record includes 10 years at GIC where he honed his skills as a fixed income manager, then seen as a poor cousin to equities. He spent another decade at Temasek, and eventually Fullerton Fund Management, where he was CEO. He joined LGI in 2010.

His 12 years at LGI’s helm, in addition to six years at Fullerton Fund Management, have also deepened his convictions on what ails the retail investment landscape today, particularly when it comes to the CPF Investment Scheme (CPFIS).

On revamping the CPFIS

He believes the investing public has evolved far ahead of the current decades-old structure of the CPFIS, where funds have to satisfy certain criteria to be included. CPFIS, he says, needs a “big revamp’’.

The current stable of CPFIS-included funds is overpopulated by active funds – and equity funds in particular. There are six exchange-traded funds (ETFs), all of which are classified as “narrowly focused’’. None of them give globally diversified exposure to either equities or bonds.

Among unit trusts, there are four index-tracking funds. Two by Amundi were recently included: one tracks the MSCI World index (total expense ratio 0.1 per cent) and the other tracks a US large and mid-cap index (total expense ratio 0.05 per cent). The funds can be accessed via Endowus’ platform.

Today, thanks to the ubiquity of the Internet and emergence of robo-advisors , Lee believes investor education has become less of an issue. Even retail fund fees have declined – if funds are purchased through robos. Among digital advisors, Endowus and MoneyOwl currently offer investment advisory for CPF savings.

On the CPFIS, Lee says: “The need for funds to be evaluated by Morningstar is costly. Many firms have decided they don’t want to be part of this, that it doesn’t make sense. CPFIS needs a refresh because technology has made it possible for investors to be more self-directed.’’

Robos such as Endowus and Stashaway have done a “fabulous’’ job, he adds. Endowus now boasts assets under advisory of about US$4 billion, of which Singapore has the lion’s share.

Lee believes “active’’ ETFs are a way to overcome retail distribution fees. Retail fund fees include an upfront sales charge and a trail fee, which is the share of funds’ annual management fees paid to distributors. Endowus’ model is to onboard institutional share classes of funds which do not carry trail fees, or rebate trail fees, back to investors.

Lee says: “I think we should migrate from unit trusts to active ETFs, and that would solve a lot of problems that the average retail customers face.’’

Active ETFs are actively managed funds listed on an exchange. Cerulli Associates says they are a growing segment, accounting for about a third of the total ETF market of US$7 trillion. ETFs listed on SGX are passive index trackers.

For now, a proposal by a CPF advisory panel in 2015 for the creation of an optional Lifetime Retirement Investment Scheme (LRIS) for the CPF appears to be at a standstill.

The LRIS was conceived as a way to simplify choices for members, and help them to achieve higher returns by providing access to low-cost and passively managed funds.

Lee believes one possible avenue for higher returns above the default CPF interest rates is to enable members to participate in the long-term returns generated by GIC, possibly via the creation of an index of GIC’s moving-average returns, for instance.

Last year, GIC’s portfolio generated an annualised nominal rate of return of 7 per cent in US dollar terms, and an annualised real return (above inflation) of 4.2 per cent over a 20-year period to end-March 2022.

Harnessing AI for fund management

Meanwhile, he believes that AI and machine learning could enable fund management firms in Asia to manage global portfolios efficiently, giving them a leg up as they compete for institutional mandates.

Under his watch, LGI last year rolled out a revamp for its Disruptive Innovation Fund, using AI and machine learning to create a “dynamic’’ smart-beta stock-screening model.

Smart beta overlays factors such as company size and quality onto a market index as an efficient means to reap index-plus returns.

Previously, the Disruptive Innovation Fund used a static five-factor model. The revamp uses a dynamic model to consider over 30 factors, including metrics such as earnings yield and cash flow, to systematically select stocks. Year to mid-April, the fund has underperformed its peer group based on Morningstar data.

LGI chief investment strategist Lim Yuin says the fund’s core AI engine uses a “lookback’’ window to evaluate recent market data. “As such, the AI engine would take time to learn new data presented by volatile market conditions with a regime change, like what we are experiencing in the recent months. We expect the performance to pick up as it incorporates the latest data.’’

The firm also has a segregated mandate investing in Asia ex-Japan equities, using an AI engine since January 2022.

This strategy has outperformed the MSCI Asia ex-Japan index by 4.3 per cent between January 2022 and end-March 2023.

Lee says: “We discovered we could use machine learning to improve on the factors. That gives fund management companies in Asia the opportunity to manage global portfolios in one location, or just a few locations.”

He believes the potential is large for firms in China and Japan, which have large populations and a critical mass of long-term savings.