Since the release of ChatGPT and a subsequent rally in AI-related stocks, fund managers have been vocal about the disruptive potential of artificial intelligence (AI) on their investments.
However, the investment industry itself is also being transformed by artificial intelligence tools, according to BlackRock, the world’s largest asset manager with $9.4trn of assets under management.
“I think we are in the middle of a bit of a revolution,” said Jeff Shen, co-CIO and co-head of systematic active equity at BlackRock. “Generative AI and big data is transforming pretty much every industry out there. I don’t think investment will be immune to it.”
He warned that investment firms may be facing a more abrupt change going forward as AI tools become more and more implemented into the investment process by market participants.
“I think that there will be a bit of a step function,” he said. “We will have more data available for us to measure a lot of different things and there will be better algorithms for us to capture the data.”
Although Shen (pictured) has been working with big data and AI tools for almost 20 years at BlackRock, he said: “It’s actually at this particular moment that I’m most excited about what’s going to be coming up on the investment horizon.”
Describing his experience starting out in the investment industry more than three decades ago, he observed how senior investment staff were largely doing the same tasks: listening to conference calls, looking at news, reading broker reports, doing field visits, talking to experts, looking at government statistics and occasionally discussing the animal spirits of the market.
“All of these things today, I think can be measured using data,” he added. “We are in this world where big data is upon us; a lot of things can be measured that could be relevant from an investment perspective.”
One example Shen pointed to was in the use of geospatial information. When investors would previously have to physically visit a location or warehouses to gather information, now they can use data from satellite images, or Wi-Fi or GPS data instead.
He highlighted how today investors now can track anonymised truck movement data in the United States to get a real time report on how trucks are moving and in what volume.
“To understand where the economy is, sometimes it helps to know where the trucks are going,” he said. “Now we can almost go down to the extraordinary micro level to measure where the economy is.”
Another example Shen pointed to was in the use of BlackRock’s finance-specific AI tool based on its own large language model (LLM) trained on texts specific to investing.
Shen’s BlackRock’s Systematic Active Equity team – comprised of computer scientists, engineers, physicists and mathematicians – have developed their own LLM using financial and investment related text to better summarise and extract key information from earnings calls.
Shen emphasized that while their LLM is “simpler” than the latest ChatGPT iteration, it has better predictive power for investment analysis because it is trained on specific financial information only.
He said: “The large language model is very important, but at the same time, what kind of input you put into the machine-learning algorithm to train it is equally important.”
