Morgan Stanley Is Pitching Big Data Against Robos
Morgan Stanley’s use of machine learning to assist human advisors may soon put an end to the term “robo-advisor,” academic Tom Davenport and big data expert Randy Bean write in the Harvard Business Review.
The brokerage’s wealth management model is based on personal relationships, so advice without human input goes against its culture and business model, according to Davenport and Bean. But Morgan Stanley’s so-called “next best action” system aims to enable its reps to deliver advice more effectively and efficiently, say Davenport and Bean. And while at first the system merely suggested investment options derived by a rules-based model, the current iteration matches investment options to investor preferences, according to Davenport and Bean. Thus, the system allows advisors to narrow down their choice of recommendations, they say.
But it also lets advisors go further than most robo-advisors: while most pure robos typically recommend only mutual funds and ETFs, the next best action system can aid advisors in recommending a specific stock using Morgan Stanley’s own research, according to the writers. It also lets advisors quickly reach out to numerous clients in the event of a major shock in the market, such as the Brexit vote in the U.K. And advisors can still personalize their message for each client, say Davenport and Bean.
Morgan Stanley’s new system, which is being tested now and will roll out to 500 advisors in September, takes personalization one step further by considering a client’s life events, claim the authors. And not just simple things like remembering a client's birthday -- for example, the system can use information such a client's child's illness to make recommendations on hospitals, schools and financial strategies to help the client handle the situation, according to the publication. That type of machine-assisted advice can help human advisors build a more trusting relationship with more clients, say Davenport and Bean. And more advice firms are likely to follow this machine-enhanced human advice rather than rely on robos alone, according to Davenport and Bean.