Artificial intelligence will continue to be buzzing in wealth management in 2018. But there’s a short list of professionals who actually understand AI and can clearly explain how advisors and wealth management firms will benefit from it now and in the future.
To help break it down, WealthMangement.com interviewed Doug Fritz, the CEO and founder of F2 Strategy, a technology and marketing consulting firm to the wealth management industry. We asked Fritz to unpack AI in a way anyone in the industry can understand and even act on it.
Prior to founding F2 Strategy, Fritz was the CTO for First Republic Private Wealth Management. He also was a senior vice president at Wells Fargo Wealth Management, where he ran strategy, implementation and support for the company’s Portfolio Management Desktop, and he spent years as a management consultant for KPMG Consulting.
Responses have been edited for length and clarity.
WealthManagement.com: Can you briefly unpack AI? What are the subsets of it?
Doug Fritz: There are as many different definitions and subsets as there are consultants pretending to know them all. I like the following descriptions, but even these overlap and confound.
Deep Learning: Computer finds correlations across multiple data sets and tasks and builds its own output from its own logic.
Recommendation Engines: Computer picks the right next step/product/task/etc. for a human based on learned behavior and outcome.
Predictive Analytics: Analyzing past and current trends to spot and highlight potential high-conviction outcomes.
Natural Language Generation: Computer analyzes data and writes natural, readable context/sentences about the data (early benefit for advisors).
Natural Language Processing: Humans use context/sentences to request data from the computer (early benefit for advisors).
Evidence Based: Computer scans through zillions of records and outcomes to help prove/disprove causality or correlation (mostly for medicine).
Machine Learning Systems: Computer creates native structures from seemingly uncorrelated and unstructured data to find trends or predict outcomes (early benefit for advisors).
Prescriptive Analytics: Computer analyzes current situations and provides the human with the most optimal next step.
WM: Which subsets are advisors using already and how?
DF: Advisors are probably already using natural language generation when they read fund and securities research reports. Those are increasingly written by a computer. Thompson Icon has a version of natural language processing which lets users create a data query by typing in a search box (for example, S&P 500 vs. MSCI last five years) and the graph and data pop-up. More comprehensive use cases are still out of reach. Why? Mostly because of data disorganization and a slow trend for an industry to see the benefits of some of this technology.
WM: Data disorganization?
DF: Data disorganization is basically the technical version of a messy filing cabinet. Most of the newer wealth-tech tools, including AI tools, require a minimum amount of organization to the data. If your CRM data is stuck in Salesforce and your custodial data is stuck at Schwab and your performance data is stuck at BlackDiamond … it’s going to be very hard to pull all this together. Firms that have invested, or are investing, in their own data will find it much easier to adopt and use innovation like AI.
WM: Which subsets should they or will they be using more in the future?
DF: Compliance teams are going to love some of the machine learning and deep learning tools because they can spot risks and fraud which have historically been impossible to track. I think NLG will continue to gain traction as advisors and clients start feeling better about consuming content which is focused, relevant and real-time, but not necessarily written by a human.
WM: Which subset does IBM’s Watson fall into? And besides Watson, are there any other for-hire AI computers out there right now?
DF: Watson is a Machine Learning tool. Its outputs and answers are primarily based on the data that it gathers or is fed and can train itself to change its algorithms as the process repeats. In effect, it can learn. Other AI firms’ outputs are based on rules that are provided to it by a human (or a combination of both).
The downside to tools like Watson for wealth management is that they may require significant investment in talent to operate or use, which few wealth firms possess and the talent is predominantly (and not surprisingly) sourced from IBM.
There are other tools on the market that, like Watson, possess machine learning ability but they all require a human to configure and train them to produce relevant outputs. Specifically, Microsoft Oxford and Google Deepmind.
Unlike Watson, tools that are focused on a specific industry or function are far easier to set up and maintain. They also deliver ROI faster and are, frankly, easier to understand for most wealth firms. Natural Language Search firms like Insight Engines and Natural Language Generation firms like Narrative Science are great tools that quickly prove value and can save advisors time.
WM: Why else aren’t advisors seeing the benefits and, perhaps, becoming skeptical about AI helping or impacting them?
DF: Over the last five years or so, some firms have chosen to ignore the signs that client expectations, regulation and need to increase scalability have demanded smarter use of technology. Unfortunately, many of these firms have dug themselves sizable holes and will struggle to catch up and compete. However, it’s great to see there are a growing set of advisory firms that are leaning forward to the future and have used technology to propel their growth.
WM: So some advisors or firms are stuck in their ways and ignoring the possible benefits of AI?
DF: Yes. At the risk of sounding ageist, the generational demographics of our average advisor is probably not helping with adoption of innovative technology.
WM: Is it too late for them to change their minds?
DF: Of course not. But it gets harder each year. Advisors get older. Clients are expecting more. The gaps widen ... possibly the most critical issue with waiting to begin is that by the time they decide to make changes, the internal culture and experience with technology change isn’t there and it becomes more disruptive than it should be.
WM: Occasionally you see people in the media raise concern about AI “self-awareness” or a Skynet scenario. How probable is that? Or have I seen “Terminator” too many times?
DF: Ha! I’m sure the worst thing our advisor AI will do when it takes over our office is transfer all our clients’ assets over to Wealthfront!
Seriously though, these tools are just that: tools. If we use them to be better and more compassionate humans, philanthropists, advisors, planners and life coaches, we raise the bar for our industry and our society. Tools that give us better answers, and enable us to know and service our clients more deeply, do just that.
