人工智能不仅仅适用于宇宙了

Artificial intelligence is no longer simply an experiment for traditional asset managers.

公司实现了一个full range of artificial intelligence techniques — including data mining, signal generation, and optimization — are generating 3 percent in alpha, according to a new Accenture report.

“Mature firms that are true leaders in AI, meaning they have industrialized and centralized these techniques across the investment process, see 300 basis points in alpha coming from that scaled activity,” said Michael Spellacy, who leads Accenture’s capital markets group, in an interview.

“The leaders have a clear culture of AI. They centralize the capabilities and then let portfolio managers figure out where and how they want to use it,” he added.

Accenture interviewed CEOs, chief information officers, portfolio managers, and data scientists at more than 40 asset managers, alternative investment firms, asset owners, and others to determine what results the industry is getting with artificial intelligence. The consultant found that 16 percent of those firms were “leaders” who had implemented advanced applications of AI with results that could be directly tied back to the technologies. Twenty-five percent of firms are in the early stages of data analytics and AI; 31 percent are progressing with applications; and 28 percent have developed sophisticated applications of AI, but are not yet using them at scale.

“But there are leaders and then there are the rest. There’s a huge gap in capabilities,” Spellacy added. “Many firms are still running science experiments.”

Still, four in 10 of the firms Accenture interviewed have successfully used AI to make predictions that ultimately affected their investment decisions.

Forty-five percent said it was possible to capture and attribute that performance to AI.

“It wasn’t a one shot deal. With AI, they told us they have better research, better portfolio allocation, better portfolio construction, and better trading. All that adds up to better alpha generation possibilities systematically,” said Spellacy.

[IIDeep Dive:AI-Powered Hedge Funds Vastly Outperformed, Research Shows]

Firms are facing a number of challenges to effectively implementing AI in their organizations, including separate investment teams developing their own apps with little leadership from the top.

Almost all of the executives interviewed (90 percent) by Accenture are actively applying artificial intelligence. But less than half are confident enough to attribute the alpha to AI.

“Given the discussion frequency and today’s data availability, we are not surprised to report that nearly nine out of every 10 investment managers are actively applying AI in their front office investment processes,” according to the Accenture report. “These activities include investment opportunity identification, research and due diligence, portfolio construction and management, trading, and performance and attribution reporting.”

但顾问发现了AI的许多障碍。

“Several firms are also grappling with challenges related to culture, their operating model, talent and skill building, and the inability to put AI into production and scale it. This explains the wide range of AI capability across the industry and the various stages of maturity of projects,” said Spellacy.