Artificial intelligence is shaking up how the finance industry operates, not only by changing how institutions invest but possibly where financial firms are located, according to new research from Stanford University.
In a paper entitled “Deep Geography: Implications of the Socio-Spatial Structure in Artificial-Intelligence Research for Financial Institutions,” authors Ashby Monk and Dane Rook of Stanford's Global Projects Center found that the “concentration of cutting-edge research in Canada and California could – to some degree – erode the hegemony of New York and London as the joint epicenter of global financial innovation.”
Much of the cutting-edge deep learning technology is being created by either smaller academic organizations or technology companies, rather than financial technology companies or financial institutions, Rook said in a phone interview Tuesday. As institutions begin to incorporate more AI programs into their investment process, he said it is important for them to understand where the industry is going.
“The fact is that there is a massive shakeup going on in the global financial and goods system,” said Rook, who works as a post-doctoral researcher at the Global Projects Center. “Institutional investors need to be aware of how the industrial shakeup will affect their partners and their tech providers.”
Deep learning, which is “a broad class of self-adapting algorithms whose purpose is to characterize functional relationships between sets of variables,” allows researchers to test many millions of hypotheses efficiently, according to the paper.
The technology can be used for fraud detection, efficiencies in market making, and liquidity management, said Rook, adding that deep learning is a major area of growth in AI.
The need for more professionals experienced in deep learning and artificial intelligence is driving financial firms to consider opening offices in California's Silicon Valley and Canada, where they can be closer to tech companies and academic organizations working in AI, according to the paper.
“The sort of supposition that has existed so far is that these new market participants are nimbler,” Rook said. “They can more easily embrace change, particularly with new findings. They are better at responding to new developments and then deploying them to the public.”
Given that the number of professionals well-versed in deep learning and artificial intelligence is small, financial institutions are having to compete with academic organizations and tech companies for talent, Rook said.
According to Rook, there’s a “scarcity” of “good quality folks” in major financial centers like London and New York.
“That is not where the deep learning research is occurring,” he said. Instead, those researchers are in Silicon Valley, Montreal, and Toronto, and to a lesser extent Edmonton and Pittsburgh, he added.
According to the paper, the top-cited and -downloaded research is originating from companies located in the San Francisco Bay Area, Toronto, and Montreal, rather than in New York and London. And the research that is published in major financial centers often involves “substantial input” from professionals in the Bay Area, Toronto, and Montreal.
“These are issues that institutional investors should be aware of when they’re putting out an RFP,” Rook said. He added that institutions should consider whether a younger, nimbler partner or an older institution would be a better fit for them.