How Will Machine Learning Affect the Financial Services Industry?

Michael Brown, CSI

Since the dawn of the Industrial Revolution, society has grappled with what man could produce versus a machine, and with each mechanical invention, man ultimately benefited. This century will be no different, and one of the key business disrupters will be machine learning. Today’s machines not only produce faster than humans, but also learn faster. The financial services industry has the potential to significantly benefit from this breakthrough, but first, bankers must do a bit of learning themselves about this emerging phenomenon.

A Data Analysis Breakthrough
Machines streamlined factories in the 19th century, and computers expanded that reach to the workplace in the 20th century. And now, computer systems with the ability to learn and adjust as they go are poised to take businesses into previously uncharted territory.  In “An Executive’s Guide to Machine Learning,” the consultants at McKinsey & Company explain that “machine learning is based on algorithms that can learn from data without relying on rules-based programming.” It is software that not only gathers and analyzes data, but then learns from that data analysis to adjust itself for better performance results or to predict future behaviors. McKinsey notes that what machine learning “already does extraordinarily well—and will get better at—is relentlessly chewing through any amount of data and every combination of variables.” So a task that could take an employee weeks or months can be completed in seconds or minutes by software with machine learning.  

Boundless Potential for Endless Data 
Since their introduction, computers have enabled businesses to effortlessly capture and store large amounts of data, at first benefiting the largest organizations. Cloud computing exponentially expanded that storage capacity and made it available to even the smallest organizations. However, the deluge of data is almost impossible for employees alone to fully wade through and harness. While employees can analyze aspects of that data, significant amounts of useful information have remained untapped inside the computer. 

That changes with machine learning. In “Machine Learning Is Redefining the Enterprise in 2016,” Forbes explained the benefit of this application: “The good news for businesses is that all the data they have been saving for years can now be turned into a competitive advantage and lead to strategic goals being accomplished.” In the not-too-distant future (now in some industries), experts believe machine learning will be a key differentiator between profitable and thriving organizations and their unsuccessful and stagnant counterparts.


How can financial institutions use machine learning? 
The financial services industry has quickly taken to this idea. From large financial institutions to smaller banks, and across the spectrum of businesses that serve them all, machine learning is being applied to various tasks with promising results. Global Banking and Finance Review attributes “this appetite to embrace machine learning” as being “driven by a strong belief that the future of the industry will be decided by financial computing engineers, and their algorithms.”  

Here are five areas in which our industry is exploring the potential of machine learning to harness data for a significant competitive advantage:


Risk Management
For decades, financial institutions used a silo-based risk management approach. That began changing in the ‘90s as larger institutions converted to an enterprise risk management approach, which collectively views and scores risk data from across the organization. After the 2008 financial crisis, smaller institutions followed suit. Now, imagine adding the ability to quickly and seamlessly analyze all your risk data from various systems through machine learning software, and that holistic risk view is magnified tenfold for better decision making and risk mitigation.  McKinsey & Company, in predicting “The Future of Bank Risk Management,” notes the absolute importance of the risk function to the overall health and success of institutions. It identifies machine learning as a critical element of that function going forward because it “improves the accuracy of risk models by identifying complex, nonlinear patterns in large data sets.” And so, “Every bit of new information is used to increase the predictive power of the model.”


Compliance
An institution’s compliance office could also benefit from machine learning moving into the mainstream. For example, USA PATRIOT Act compliance and such related regulations as OFAC screening require significant data analysis. This can be especially challenging for larger institutions with multiple systems and even for smaller institutions with disparate systems as a result of mergers.  With machine learning software, the compliance office would be able to feed data from all its various current systems (deposits, cards, mortgages, other loans, etc.) to generate a real-time and complete representation of its customers based on their latest transactions and interactions with the institution. This takes “know your customer” to a whole new, unprecedented level. 


Financial Crime, Fraud Detection and Cybersecurity
In “Clever Banking with Artificial Intelligence,” Banking Technology points out that, “Banks and fintech companies already use machine learning to detect fraud by flagging unusual transactions.” Such anomalies are investigated, with the result being fed back into the system so it can “learn” and thus further build the customer profile. This, the industry news source says, is “far more efficient than human manual monitoring and is expected to become the norm in banking and finance.” Even smaller institutions are experimenting with this concept. Network World tells the story of a community bank that “wanted a way of tackling fraud in an ongoing way, but within the context of their budget and technology constraints.” The initial upshot of implementing a machine learning-based solution to identify fraud related to card use stopped 250 fraudulent transactions, leading to the bank’s decision to expand the technology into other areas. Tech Emergence, an AI market research firm, explains that, “While previous financial fraud detection systems depended heavily on complex and robust sets of rules, modern fraud detection goes beyond following a checklist of risk factors—it actively learns and calibrates to new potential (or real) security threats.” In addition to fraud detection, the firm predicts that machine learning also will prove essential in cybersecurity, noting that “given the incalculably high number of ways that security can be breached, genuinely ‘learning’ systems will be a necessity in the five to ten years ahead.”

Credit Underwriting and Portfolio Monitoring
Consulting firm Barclay Simpson anticipates another use for machine learning. “Credit applications and underwriting are the key areas where machine learning, and data analytics in general, will have an initial impact.” It predicts that “the outcomes will include cost reductions, increased efficiency and less onerous customer experiences.” To that point, the McKinsey research included an example of European banks that “replaced older statistical-modeling approaches with machine-learning techniques” and reported significant results. These banks were able to build “micro-targeted models that more accurately forecast who will cancel service or default on their loans, and how best to intervene.” The success of these early adopters suggests that the predictive power of machine learning is going to significantly enhance future credit underwriting, portfolio monitoring and collection processes throughout the industry.

Customer Sales and Service
Machine learning also is likely the only way for organizations to provide the instant, personalized offers and assistance that consumers are fast coming to expect, according to TechEmergence.“In the future, increasingly personalized and calibrated apps and personal assistants may be perceived (not just by millennials) as more trustworthy, objective and reliable than in-person advisors. Just as Amazon and Netflix can recommend books and movies better than any living human ‘expert,’ ongoing conversations with financial personal assistants might do the same for financial products.” Forbes agrees, noting that “machine learning is proving to be efficient at handling predictive tasks including defining which behaviors have the highest propensity to drive desired sales and marketing outcomes.”

Man and Machine Take the Leap Together
Some fear technology’s potential impact on man, but resisting machine learning and other AI applications won’t stop their proliferation. The potential benefits described above, as well as those related to energy efficiency, disease diagnosis and treatment, and other industries, are just too great.  Rather than digging in their heels, McKinsey advises executives that “now is the time to grapple with these issues, because the competitive significance of business models turbocharged by machine learning is poised to surge.” As for fear of how machine learning will impact human jobs, the consulting firm quickly points out that, “The role of humans will be to direct and guide the algorithms as they attempt to achieve the objectives they are given.” Don’t think of this phenomenon as “man versus machine.” This is man and machine leaping into a future, with untold possibilities for benefiting man through the power of machine. 

Michael Brown is vice president of product strategy for CSI Regulatory Compliance. Michael holds nearly two decades of IT and project management experience, and currently leads strategic product development for a wide variety of risk management solutions. His expertise includes governance, risk and compliance; enterprise risk management; GLBA and AML compliance; ERP and CRM systems; and master data management.

 



 

 Back to March 2017 Technology & Security Digest


 

   
Issue #32 - March 2017