Artificial intelligence coming to life in financial advice
Imaginearriving to work in the morning — the lights already powered on, the temperature set the way you like it, the coffee brewed — and being greeted by an assistant that spent all night preparing a prioritized list of opportunities for you and your clients.
Source: Ryan W. Neal
"Your client, Jessica, had questions last night about recent volatility affecting her portfolio," Alexa tells you. "I provided an update on her balance and shared your analysis of the stock market, but she has additional questions for you. You have a call scheduled for 10:30.
"I also detected an opportunity for Richard to save $200 a month by refinancing his mortgage. You haven't spoken with him since your quarterly review. Would you like me to generate an email about this opportunity, or would you like to call him personally?"
This is what financial advice could look like in the age of artificial intelligence. With a computer handling everything from administrative back-office tasks to compliance, marketing and investment management, advisers would be free to spend their time expanding their businesses and serving their clients even better.
What was once the stuff of science fiction is now being brought to life and heavily invested in by the nation's banks, brokerages and custodians.
Technology leaders at the biggest firms view AI as the key to helping advisers get more done with less.
"We see artificial intelligence as a quickly emerging, and very strong, opportunity type of capability. It's something that we have a heavy focus on," said Andrew Brzezinski, vice president of Fidelity Institutional. "We could fundamentally improve our interactions with our advisory business clients and in turn help them be better advisers to their investor clients."
Optimism abounds. However, we are still very much in the early stages of deploying the technology to advisers.
The technology exists. California now allows driverless cars, and media outlets such as the Associated Press automatically generate articles about fantasy football and earnings reports. But there have been few practical applications in wealth management to show advisers how AI can work in their firms.
"The systems are not yet capable of providing the depth of analysis and evaluation that clients depend on from their financial planner," said Ric Edelman, founder and executive chairman of Edelman Financial Services. "When it comes down to the question of how do I counsel my client who is saving for college for his child, when in fact that client needs to first save for his retirement, the AI programs don't exist yet to replace the adviser's intuition, personality, deep understanding of the client's family dynamics and the motivation to recommend assistance."
Complex industry regulations, misinformation and a simple fear of change all play a role in preventing widespread adoption of the technology. But the biggest obstacle is the sordid state of data at many firms.
"Most wealth management firms still have a constellation of different tech tools," said Doug Fritz, CEO and founder of F2 Strategy, a technology and consulting firm, in explaining why most firms can't deploy this type of technology. "There's not one place you can point the AI tool to."
Some firms have made data a top priority since the startup robo-advisers sounded a wake-up call for the wealth management industry to modernize technology. Those that heeded the alarm are best-positioned to adapt to the coming AI era.
Those that didn't may be left behind.
Advisers who keep hitting the snooze button may remain oblivious to artificial intelligence, but the technology is already affecting the industry far more than they realize.
BlackRock Inc., for instance, has a new suite of exchange-traded funds managed entirely by AI. And most large financial firms through data in search of anomalies that could indicate fraud, cybersecurity lapses or a compliance violation.
Jeff McMillan, Morgan Stanley Wealth Management's chief analytics officer, said his firm is employing AI to automate many tasks that used to require manual labor, such as calculating tax withholdings for international clients.
"Artificial intelligence automates the entire 12-step process, taking something that took 15 to 20 minutes in the past down to the click of a button," Mr. McMillan said. "There are massive opportunities to improve the velocity of transaction processing."
But he's most interested in how AI can augment and improve the way advisers work with clients using a new product called Next Best Action. That platform, being rolled out to Morgan Stanley's advisers, uses machine learning — a more advanced version of AI that can teach itself to perform tasks instead of simply executing a human's programming — to upload thousands of possible adviser-client interactions each night to determine which will be most beneficial.
Next Best Action factors in the adviser's individual investment strategy and the firm's internal research to suggest a prioritized list of activities tailored to each client's unique needs and preferences.
So if an adviser wants a client to have a 6% exposure to Japanese equities, Next Best Action can detect when a portfolio drifts away from that target and analyze historical data to determine which investment products the client is likely to prefer. It can then present the options to the adviser as a high-priority task. If the adviser is unable to make the phone call personally, the program can populate a digital message with the information and send it to the client on the adviser's behalf.
The program will also monitor how the client interacts with the message and use that information to improve itself in the future.
"These tools make the adviser smarter, identify when you haven't touched clients enough and … scale that engagement," Mr. McMillan said. "The quality of wealth management is going to get better for every single client, both big and small. Advisers are going to be able to provide a higher quality of advice and service for clients."
Eventually, Mr. McMillan expects Next Best Action to help advisers connect with clients beyond investment products by recommending information relevant to a client's hobbies, health care or major life events, he said. Such interactions with a financial adviser are the most powerful driver of client satisfaction, according to Morgan Stanley research, even beating out investment performance.
"It's no longer a science experiment," Mr. McMillan said. "It's about businesses engaging with these tools."
And these tools only scratch the surface of what machine-learning technology is capable of doing. In May 2017, researchers at Google Brain announced AutoML, a program designed to create its own AI system. By December, AutoML produced a "child" AI tasked with object recognition that outperformed all similar systems designed by humans.
Such a shift from conventional computing to machines that can learn, decide and create is as dramatic as the move from gas lighting, when Edison and Westinghouse brought electricity into our homes, offices and factories, according to the Future Today Institute, a research group that publishes an annual report on emerging technology trends.
"AI is the new electricity, and our personal data is what's generating the current," the FTI's 2018 report said.
STORING THE DATA
If that's the case, many firms are still relying on steam power.
Brokerages, custodians and individual advisers have spent the past decade bolting on new digital tools in a desperate attempt to modernize legacy systems without having any plan for what to do with all the digital information. The problem isn't collecting the data, it's storing it in a central location in a manner that will be useful to an AI system.
"One of the challenges is, we are an industry that has existed for some time," Fidelity's Mr. Brzezinski said. "Assembling the large, centralized, linked-up data sets that are needed to accomplish artificial intelligence is an incremental process."
Financial services firms face an uphill battle in recruiting talent to help them modernize their data infrastructures. As capital pours into AI technology, firms find themselves competing not just with Silicon Valley for the top information technology talent, but also with nearly every other industry.
There's also the ever-present challenge that new regulations could slow the adoption of the technology, as well as the added cybersecurity risks that all digital technology solutions raise.
Some firms are comfortable with sacrificing time to market in the name of security.
"It doesn't take long to go down a list and find recent headlines of companies that have had [client] data exposed," said Todd Slawson, director of financial adviser platform development at Wells Fargo Advisors. "Our industry gets saddled with having to make sure we meet all of the rigorous standards that we have set forth."
'LIFEBLOOD' OF THE BIZ
Morgan Stanley Wealth Management's database accounts for every trade, call, email and client login on its platform and is one of the most important things Mr. McMillan has worked on over the past few years.
The firm also has a team dedicated to analyzing the information and ensuring data quality. Mr. McMillan called data "the lifeblood of this business" and believes it will be critical for firms in the AI era as technology becomes increasingly commoditized.
"The winners in this space will not be because they have the best AI solution," he said. "They'll have the most historical data together with the richest intellectual capital, and they will use that to drive the best outcomes."
With brokerages and custodians racing to get their data houses in order, it won't be long before AI is widely adopted across the wealth management landscape.
The final obstacle may simply be convincing advisers of the technology's benefits, which is no small task. Even with the amount of capital investment flooding the AI ecosystem, there is also misinformation polluting perceptions of what the technology is and will become.
Some advisers will inevitably see AI as a threat to their business and fight it. Others may simply find themselves too outdated to adapt.
"Advisers who are only providing investment management services are under threat, and rapidly so," Mr. Edelman said. He estimates that AI could fully replace these advisers in as little as two years.
Those who embrace the technology and use it to improve the quality and profitability of comprehensive, relationship-based advice will be better able to adapt to the AI era. Machines may evolve to the point of being able to totally replicate the experience of a human adviser, but that doesn't mean people won't still seek out that human connection.
"Women have the ability to color their hair at home, and yet the hair salon remains very popular," Mr. Edelman said.