Originally published by Emma Wallis on Trustnet.
A growing number of asset managers are experimenting with artificial intelligence (AI) to see whether a combination of human ingenuity and machine learning can deliver enhanced results. One such company is Momentum Global Investment Management. For the past 18 months, the £6.8bn firm has been working with a fintech called MDOTM Ltd and its AI asset allocation tool, Sphere. During this trial period, using Sphere resulted in an information ratio advantage – generating a similar level of returns but for less risk.
The beating heart of Sphere – its engine – is called ALICE, which stands for adaptive learning in complex environments. This assimilates economic, fundamental and valuation data for a vast range of asset classes and securities. It looks at correlations and whether they are changing, as well as return expectations. ALICE uses all of this data to determine what regime the market is currently in – growth, stressed or lateral – and the probability of each of these states occurring in 30 days’ time. At the moment, stressed market conditions are most likely to occur. It then recommends an optimal asset allocation for different risk parameters and gives views on more than 100 asset classes.
Andrew Hardy, Momentum’s director of investment management, said Sphere can detect far more changes in the relationships between asset classes than Momentum’s investment team would naturally observe and it looks at a different time horizon to the firm’s investment professionals.
“We like to think of it as a different analyst sat over there introducing a different view. They think differently and they’re wired differently but we shouldn’t be ignoring them. ALICE is less emotional and its arguably more transparent than people’s minds.”
Tommaso Migliore, Chief Executive Officer and Co-Founder of MDOTM, agreed.
“Humans and technology are working together to make better informed investment decisions and I do think this is the future. Investment professionals and AI both add value and “capture reality” in different ways. The machine doesn’t get tired or bored, it can run very complicated calculations multiple times. Humans meanwhile notice things that are not in the data, can respond emotionally and can capture surprises."
Momentum has spent such a long time researching Sphere because Hardy and his colleagues wanted to understand how and why ALICE makes decisions and build trust. Recognising that transparency is critical, MDOTM built a Generative AI layer into Sphere that explains ALICE’s recommendations.
Sphere has a dashboard with a page for each asset class. It gives a recommendation (e.g. a strong overweight to large-cap European equity), then summarizes in a paragraph the reasons for its decision. In this case, European equities have performed well in similar macro environments, volatility is low, European GDP is holding up well and the economy is resilient. Sphere also has a Q&A section enabling users to ask questions and interact with it. This facilitates a debate, not dissimilar to the dialogue between colleagues, Migliore said.
Momentum has found that ALICE concurs with some of the team’s asset allocation decisions. At present, Momentum is overweight government bonds and underweight credit, which is consistent with ALICE’s forecast for stressed market conditions. Hardy said the market got ahead of itself in November and December and became excessively optimistic.
“History tells us when inflation has been this high and persistent it can take a while to come down."
He observed.
Momentum bought duration in October, which was well timed, and took duration back off the table in December. The firm has regional tilts towards UK and Japanese equities, which are more attractively valued than other markets, and is commensurately underweight the expensive US market. ALICE also recommends going overweight Japan. Momentum has been overweight Japan for about three years as corporate governance reform has gained traction, with companies paying dividends, buying back shares and reducing cross shareholdings. After last year’s rally, Momentum moderated its exposure but the firm is still overweight.
Japan is hard to access from a geography, culture and language standpoint so had become easy for global equity managers to ignore and thus became very cheap, Hardy added. This is a “great environment” for active managers to add value but they need to be on the ground with knowledge of the language and culture. Momentum invests with Morant Wright, a Japan specialist that has outperformed its benchmark “quite handsomely”.
ALICE favours Japan for a different reason – namely, how efficient Japanese equity exposure is from a risk diversification angle because Japan trades on a different cycle to other developed markets.