BlackRock’s AI is watching – and helping to make stock picks based on what it sees

Name: SCMP
Telephone: +852-2680-8888

How asset managers are exploring the use of artificial intelligence, while regulators including Hong Kong’s watchdog try to set guard rails

Reading Time:3 minutes

Why you can trust SCMP

K-pop group Enhypen performs in Macau on January 26, 2024. Photo: Belift Lab

Aileen Chuang

Published: 7:30am, 18 Nov 2024

When K-pop group Enhypen enthusiastically Instagrammed a Prada fashion show in Milan in September, the posts might have led to a spike in website activity for the Hong Kong-listed Italian luxury brand. If so, an artificial intelligence (AI) tool built by a fund house might have noticed that “signal” and logged it to inform the firm’s view of the fashion label’s stock.

This hypothetical example illustrates just one of many ways asset managers – including the world’s largest, BlackRock, JPMorgan and ChinaAMC – are exploring the use of generative AI (GenAI) for alpha generation and operational efficiency.

Meanwhile, Hong Kong’s securities regulator has established guidelines to ensure responsible use of the technology.

BlackRock’s systematic investing platform – what it calls a “human-machine team” – has used such technology to come up with more than 1,000 signals based on over 300 alternative data sets, including social media content, blogs and internet searches.

It looks out for spikes in social-media chatter about a company followed by increased activity on the company’s website, which can correlate with a climbing stock price. “For one given stock, we have many signals,” Ahmed Talhaoui, head of BlackRock’s Systematic Group for the Asia-Pacific and Europe, Middle East and Africa regions, said in a media briefing in Hong Kong last month. “Some could contradict each other. So we have to be really good at how we combine these signals.”

Machine learning and other techniques help in that process, but humans still have a critical role to play.