“Credit is both the oldest financial industry in the world and the least mature,” according to Christopher Mann, a part-time lecturer in the Enterprise Risk Management (ERM) program at Columbia University School of Professional Studies (SPS). “Most of the industry is still done using rules of thumb and expert knowledge. Only the largest institutions actively use data analytics.”
Mann currently teaches Credit Risk Management, Credit Risk Analytics, Quantitative Risk Management, and Machine Learning for Risk Management at the University, and was previously the head of wholesale credit risk modeling at MUFG Bank and U.S. Bank.
On June 3 and 4, 2024, Mann chaired the Marcus Evans GFMI Credit Modeling conference, where industry leaders discussed strategies in credit risk management and modeling, attempts to gain data-driven insights into which borrowers might default on payments and how to better predict potential losses. This notable event saw participation of key representatives from the Federal Reserve, Federal Deposit Insurance Corporation (FDIC), and a variety of major banks, credit unions, and insurance companies.
Mann's presentation focused on the integration of advanced analytics and machine learning to enhance credit risk modeling. His insights centered around:
- Leverage Advanced Analytics and Machine Learning: Mann emphasized how advanced analytics could drive more timely and effective insights for credit risk modeling. He explored the integration of these techniques to profile credit risk more accurately.
- Dynamic Modeling and Real-Time Forecasting: Discussing the benefits of machine learning, Mann highlighted how techniques like neural networks, random forests, and gradient boosting can adapt to changing market conditions, enabling dynamic modeling and real-time forecasting.
- Trends and Future Applications: Mann delved into current trends in machine learning within the credit industry, offering a forward-looking perspective on the advancements expected over the next five years. He also examined the potential of large language models in credit modeling and the increasing use of alternative data types.
“My presentation focused on advanced analytics and tools that might help practitioners achieve more accurate, insightful, and timely assessments,” Mann said. “Under the current regulatory regime, models are considered to be static and rarely updated, since updating a model usually requires significant resources and between six and 12 months of work, mainly to support the regulatory requirements. This contrasts significantly with modern software companies where daily software commitments are normal.”
Before joining academia, Mann led wholesale credit risk modeling at MUFG Bank and U.S. Bank. His role encompassed overseeing research and development for credit underwriting for MUFG Bank’s global business and stress testing, and reserve estimation for Latin America, the United States, and Canada.
"Since coming to Columbia, I’ve done my best to bring analytics, data science, and technology to the students,” says Mann. “Our students are the best placed to take advantage of technology changes, with significant capabilities to work on modeling projects and opportunities at the intersection between analytics and leadership."
The Master of Science in Enterprise Risk Management (ERM) program at Columbia University prepares graduates to inform better risk-reward decisions by providing a complete, robust, and integrated picture of both upside and downside volatility across an entire enterprise. The program focuses on all aspects of ERM, including frameworks, risk governance, risk identification, risk quantification, risk-reward decision-making and risk messaging.
Beyond his professional achievements, Mann is a veteran of the U.S. Navy and a prolific researcher with several refereed articles published in top journals. His recent projects include automating credit underwriting for small businesses and leveraging machine learning to improve credit forecasts. He also dedicates much of his free time to exploring banking data using resources like the Federal Reserve’s Federal Financial Institutions Examination Council public data and the FDIC’s failed bank dataset.
Mann’s role as chair of the Marcus Evans GFMI Credit Modeling conference underscores his commitment to advancing the field of credit risk modeling through the application of cutting-edge technologies and his dedication to sharing knowledge with industry professionals and students alike.
“The key takeaway from the conference is that everyone is worried about technological change, but few have the resources, time, or support from their management to make changes,” explained Mann. “Despite this, technological change is inevitable, even if it takes another 10 to 15 years.”
About the Program
The Master of Science in Enterprise Risk Management (ERM) program at Columbia University prepares graduates to inform better risk-reward decisions by providing a complete, robust, and integrated picture of both upside and downside volatility across an entire enterprise.
