Developers, banks pushing hard to put AI to work for advisors
SAN FRANCISCO — Banks and fintech startups are pushing the boundaries of artificial intelligence to try to make advisors’ working lives easier — even those advisors still unclear on the concept of machine learning.
Wells Fargo, for example, is trying to figure out how to use AI to pore over data for 1,500 mutual funds rapidly enough that advisors can have performance analyses in real time, in response to market-moving events, according to Sam Kirkland, senior vice president of global manager research for the firm.
"A lot of these advisors, they want the talking point," Kirkland told attendees at SourceMedia's In|Vest West conference. "They want to have the confidence to get these speaker points when clients call."
Right now, however, many other low-tech considerations — aside from the limits of machine learning – are preventing this from becoming a reality. Financial regulations, for one, restrict mutual fund managers from sharing data needed to complete such an analysis that fast, he said.
But with AI's help, Kirkland hopes the bank might be able to, at a minimum, get assessments of funds' quarterly results out to advisors more quickly.
A cultural swap between Wall Street and the Valley may be changing how both sides of the financial aisle operate.December 6
New platform upgrades seek to aggregate data and address client concerns.November 28
The challenges of wrangling complex data to complete such tasks were canvassed during a panel at the conference on the practical implications of AI. Kirkland and other experts discussed the difficulty of harnessing big data to save advisors time.
Their conversations unfolded both on the stage and afterwards in discussions with conference attendees.
The only way AI can help offer fund managers or advisors predictive guidance on investments right now is when a question can be answered solely with reference to a clear set of rules, Kirkland told fintech entrepreneur Robert Owen, director of business strategy at AI-based portfolio management firm Vise AI, after the session.
"If it's rules-based, then [any recommendation] is just a combination of those rules," Kirkland said. However, the fact that rules alone seldom form the sole basis for any decision demonstrates "how much intuition is actually there" in any analysis, he added.
Maybe, Kirkland mused to Owen, "at some point we would turn over the intuition to a machine."
"Who knows if it would be possible?" Owen responded. "You can't teach the machine when to break the rules."
Earlier onstage, Cory McCruden, managing director of strategy and innovation at RBC, agreed that banks are trying to figure out the best way to make use of advances in AI.
One problem banks face is fintech innovators who pitch them on solutions "in search of problems," she said. But if these innovators can discover how to solve banks’ immediate needs — such as making their advisors more profitable – then AI will start fulfilling its promise, she added.
To help with that process, banks may have to swim far "upstream," according to McCruden, to determine how to best make use of it.
That's because, she said, "the industry is also inchoate."
As a result, "we are making investments in academia" at RBC, she said. "We started a research institute. We said, 'Forget about banking. Go do really cool research.' "
Ultimately it will take "an ecosystem for us to fully realize the potential of this technology," she said. "It is going to have to be public and private, banks and academics coming together. You are just going to have to believe that this is the right thing to do for the long term."