Robots vs. Humans: How Machine Learning is Impacting the Financial Services Industry
Disruptive Technologies: How will machine learning alter society?
Machine learning is critical to the financial services industry and provides a unique opportunity to develop products in a more deliberate and sophisticated manner. Machine learning also allows for financial services firms to augment current products or create a differentiated customer experience. One firm that is leveraging machine learning is Morgan Stanley. Machine learning is particularly important for Morgan Stanley as it can provide a competitive advantage in a highly competitive financial services marketplace. The ability to customize the customer experience and deliver exceptional customer service is paramount and with machine learning higher levels of product differentiation and customer service can be achieved. If Morgan Stanley were to take a passive approach with machine learning and technology, they could find themselves quickly obsolete in the financial services industry. Machine learning, coupled with other disruptive technologies, are revolutionizing the process improvement cycle and product development cycle to ultimately create value for shareholders. It would be a disservice to all Morgan Stanley stakeholders to not adopt such technologies as the potential to drive revenue (Columbus) and reshape pricing strategies (Rizzi) is a compelling value creation. Morgan Stanley can emerge as the thought leader for machine learning in the financial services industry if they continue to invest in disruptive technologies and take the necessary actions to capitalize on this opportunity.
Morgan Stanley has taken a lot of steps to ensure that they are well equipped to address the importance of machine learning. In the short term, Morgan Stanley has launched a few key projects that leverage machine learning. Using machine learning, Morgan Stanley launched a robust program that arms their financial advisors with a unique ‘personal assistant’ (Clary), one that provides insightful customer information to better serve Morgan Stanley clients by providing them with products that customers need. This program allows for a “personalized and engaging” (DiCamillo) relationship, a unique value proposition for financial advisers. In addition, Morgan Stanley has acted from an organizational perspective. Morgan Stanley created an entire department, the Analytics and Data Organization (Davenport), to ensure that the company can effectively implement and leverage machine learning. They are also investing in top industry talent (Bielski), which will create value for the organization over the long-term horizon. Another differentiated long-term approach Morgan Stanley has taken is to not simply employ machine learning but become a hub for machine learning research. Instead of taking a complacent approach and outsourcing machine learning techniques, Morgan Stanley employs a team focused on cutting edge machine learning research and technology upgrades. Morgan Stanley creates content i.e. article publications (Helfstein) and is becoming a thought leader in the industry.
Despite a promising technology with immense benefits, there are some challenges with machine learning adoption in the financial services sector. A heavily regulated financial services industry could potentially have stringent regulations imposed (Machine Learning). I would recommend Morgan Stanley work closely with the regulatory bodies to ensure they are ahead of the curve and proactive instead of reactive. Machine learning will be best leveraged with other technological advancements such as digital, blockchain, and data analytics. From an organizational perspective, to best optimize disruptive technologies, it would be prudent for Morgan Stanley to create a cross-functional team that oversees all technology and disruptive efforts. Education can be a critical element to the success of machine learning customer adoption. It would be beneficial for Morgan Stanley to have internal ‘translators’ i.e. someone extensively trained in the machine learning field that can teach internal and external stakeholders. Machine learning is an entirely new language, one that can be confusing and misunderstood. I would recommend Morgan Stanley educate their customers and consumers on the powerful benefits of machine learning, but not entirely replace the “human feel”. Eventually the technology will evolve such that humans can be entirely replaced. I would recommend being cautious with this and allow for some human interaction to be embedded in the customer journey. Another pitfall Morgan Stanley should be cautious with is creating a machine learning environment that is ‘bias’ in its coding (Baer). Proper machine teaching techniques and clean data integration is a critical component of a successful utilization of machine learning technology.
Two questions I have are as follows:
• As machine learning technology progresses and humans become obsolete in a customer service capacity, do you believe firm’s will and should continue to employ some humans?
• Do you believe organizations should adopt machine learning in-house or outsource? Should all organizations strive to become technology/software companies that just happen to sell financial products?
Total Words: 746
Works Cited
Baer, Tobias, and Vishnu Kamalnath. “Controlling Machine-Learning Algorithms and Their Biases.” McKinsey & Company, Nov. 2017, www.mckinsey.com/business-functions/risk/our-insights/controlling-machine-learning-algorithms-and-their-biases.
Bielski, Vincent. “Morgan Stanley Hires Ex-SAC Capital Artificial Intelligence Expert.” Bloomberg.com, Bloomberg, 26 June 2018, www.bloomberg.com/news/articles/2018-06-26/morgan-stanley-hires-michael-kearns-to-lead-ai-research-effort.
Clary, Timothy A. “Morgan Stanley Arms Advisors With Machine Learning.” Barron’s, Barrons, 31 May 2017, www.barrons.com/articles/morgan-stanley-arms-advisors-with-machine-learning-1496248706.
Columbus, Louis. “Machine Learning’s Greatest Potential Is Driving Revenue In The Enterprise.” Forbes, Forbes Magazine, 20 Nov. 2017, www.forbes.com/sites/louiscolumbus/2017/10/23/machine-learnings-greatest-potential-is-driving-revenue-in-the-enterprise/#6dbaa4ae41db.
Davenport, Thomas H, and Randy Bean. “How Machine Learning Is Helping Morgan Stanley Better Understand Client Needs.” Harvard Business Review, 3 Aug. 2017, hbr.org/2017/08/how-machine-learning-is-helping-morgan-stanley-better-understand-client-needs.
DiCamillo, Nathan. “Morgan Stanley Draws from ‘Hundreds of Conversations’ with Experts to Build Its AI.” American Banker, 12 July 2018, www.americanbanker.com/news/morgan-stanley-draws-from-millions-of-conversations-to-build-its-ai.
Helfstein, Scott. “Investing in Artificial Intelligence and Automation.” Morgan Stanley, Morgan Stanley Wealth Management, 21 May 2018, www.morganstanley.com/ideas/artificial-intelligence-and-automation.
“Machine Learning.” The Edge, Morgan Stanley Investment Management, 2017, www.morganstanley.com/im/publication/insights/investment-insights/ii_theedgemachinelearning_us.pdf.
Rizzi, Walter, et al. “How Machine Learning Can Improve Pricing Performance.” McKinsey & Company, Sept. 2018, www.mckinsey.com/industries/financial-services/our-insights/how-machine-learning-can-improve-pricing-performance.
I don’t believe all organizations should building machine learning capabilities in-house; rather, I think deploying machine learning should be achieved via acquisition, outsourcing, or partnership. Morgan Stanley has a clear competitive advantage and core capability as a retail banking institution – i.e., in servicing loans, pricing risk, etc. They do not have existing talent to build out their own algorithms to handle customer service functions, for example; they would be better suited partnering with an expert in cloud-based AI like AWS/Amazon’s Lex product suite.
Yes, I believe that firms will continue to employ humans in customer service. However, I think it is likely that the number of humans employed will be reduced, as some tasks can be automated. On the other hand, certain tasks that require human-to-human interaction and/or on the spot judgement (without historical precedent) will never be automated among institutions that offer excellent customer service.
As firms begin to use more machine learning, it will be critical that they maintain control over the development of machine learning algorithms. Therefore, machine learning should be developed in house to ensure that firms – particularly financial institutions – can exercise sufficient influence and regulation over their business.