European fund performance at the height of the region's Covid-19 pandemic shows that artificial intelligence could continue to play a greater role in managing active equity funds, according to a new report by Cerulli Associates.
Although AI-led active equity funds represent a tiny slice of the fund market, an analysis of fund activity in Europe shows that active equity funds run by artificial intelligence grew at a faster rate than their peers from January to April and did a better job of limiting declines, according to the latest issue of "The Cerulli Edge―Global Edition."
The Boston-based research firm's analysis also showed that the cumulative return of AI-led hedge funds in Europe was three times higher than the overall hedge fund universe from 2016 to 2019, with AI funds returning 33.9% compared with 12.1% for all hedge funds.
For example, the data found that funds with "AI" or "artificial intelligence" in their name, comprising about 4 billion euros in AUM, showed a market appreciation of 15.7% in April, versus 10.3% for all other active equity funds.
Data such as this could be an indication that AI funds have a place in the industry despite continued resistance to using the technology for fund management, Cerulli said. In 2019, the report noted, a CFA Institute survey of about 400 analysts and portfolio managers found that 69% of the managers and 75% of the analysts did not use either AI or machine-learning techniques for trading or equity analysis.
"There has long been suspicion of the ability of AI to react to unexpected events, such as the coronavirus pandemic, but there is now a sense that the technology has advanced to the point where it is better able to adapt to unforeseen scenarios via the ever growing amount of market data available," Justina Deveikyte, associate director of European institutional research at Cerulli, said in a statement.
In issuing its findings, Cerulli cautioned that the value of AI in reacting to a crises like the pandemic can't be truly assessed until the Covid epidemic subsides. The report noted that most machine-learning algorithms used in finance are "supervised," meaning the models learn to recognize patterns by analyzing historical data.
"The Covid-19 crisis being a once-in-a-century event means it is very difficult for the models to adapt to lockdown scenarios dynamically," Emmanuel Hauptmann, partner and senior fund manager at Geneva-based RAM Active Investments, was quoted as saying in the report. "Despite tremendous advances in technology, machine learning still relies on a huge amount of data to be efficiently trained."
In addition to having to deal with unprecedented events, AI models are also hampered by the fact they're built on only a few decades of usable data, meaning the algorithms are blind to most of human history—including past pandemics, the report said.
Nonetheless, proponents of AI fund management said advancement in artificial intelligence continue to show that the technology will continue to have a place in the industry, whether it be AI-led funds or AI tools for managers.
Speaking of the latter, Rani Piputri, head of automated intelligence investing at NN Investment Partners in The Hague, said in the report, “In fact, this is where we believe outperformance will materialize over the long term—from seemingly traditional funds that successfully adopt AI insights and technologies to strengthen their investment expertise and gain competitive advantage in a cost-effective manner.”