The financial crisis triggered a paradigm shift in financial institutions’ assumptions about credit risk, centering around the manner in—and frequency at—which they monitor it, and the value they place on proactively influencing it.
The shift rests on this fundamental truth, which was summed up quite eloquently in a 2014 article by credit union training consultants Dennis Child and Randy Thompson: “We in the lending business witnessed how dramatically loan portfolios can change over relatively short time frames, and how detrimental those changes can be.”
Credit risk modeling has advanced considerably in the last decade, driven by the need for credit unions and banks to address that reality — and the development of tools to do the job more thoroughly than ever before, thanks to the remarkable rise of analytics.
At its heart, credit risk modeling represents the due diligence a financial institution should perform to assess borrowers’ risk, individually and collectively, from origination until repayment. It’s not a crystal ball, but it is a forward-looking exercise, forecasting a loan portfolio’s degree of value fluctuation based on changes in borrowers’ underlying credit.
The growing pool of consumer data — combined with the benefit of hindsight in the form of measurable outcomes — has allowed program designers and lending executives to refine the process, elevating or decreasing the importance of indicators that determine credit risk.
The goal is to arm portfolio managers and executives with a clear picture of the market-based risks and regulatory risks their financial institutions face. This knowledge allows them to optimally allocate capital to seize on opportunities and brace for potential losses, and make operational adjustments such as adjusting underwriting, loan pricing, and servicing standards, and tweaking concentration limits.
Credit risk modeling can address any number of concerns, including:
Internal capital adequacy assessment. Many credit union boards have pushed their institutions to remain or become well-capitalized, for actual and perceived assurance of their overall strength. The National Credit Union Administration’s revised risk-based capital rule would mandate credit unions to increase their capitalization buffer to remain even adequately capitalized.
External reporting. Federal and state regulators continually emphasize they’ll apply more of their resources toward ensuring financial institutions better manage risk in their loan portfolios and investments than in the past.
Risk-based pricing. Tiered loan pricing programs reward members who have established strong credit, but they also allow a credit union to offer loans to people with substandard credit who wouldn’t otherwise qualify.
Performance management. Monitoring the progress of outstanding loans allows credit unions to take proactive measures that mitigate losses. Industry experts estimate that more than 80% of delinquencies and charge-offs can be traced to loans that dropped two or more credit grades from the original score, and that loans with unchanged credit scores contribute to less than 10% of delinquencies and charge-offs.
Acquisition/divestiture analysis. Concentration risk is a prime concern at many credit unions, especially those with homogeneous fields of membership, as through single select employee group (SEG) relationships or a handful of such arrangements with individual companies.
Stress-testing and what-if analysis. Credit risk modeling provides an opportunity for creative and intuitive loan portfolio managers to project the impact of a full range of hypothetical scenarios — including those driven by external forces — and design contingency plans accordingly.
Credit risk modeling is not a static exercise. As the Federal Reserve Bank of San Francisco noted, “An important question for both banks and their regulators is evaluating the accuracy of a model’s forecasts of credit losses … specifically, models (must be) evaluated not only on their forecasts over time, but also on their forecasts at a given point in time for simulated credit portfolios.”
Also, credit risk modeling is not an exact science, notes quantitative risk analyst Mikhail Voropeav: “Risk-based, real-time pricing remains the ultimate challenge,” he writes, because of the challenge of providing accurate, stable, and time-efficient input. Voropeav highlights another, often-overlooked issue, “the sometimes overly complex structure of the models as perceived by end users. Very often the complexity of the models makes them hard to be understood and, hence, affects their acceptance within an organization.”
As a result, credit unions should realize the importance of acquiring user-friendly credit risk modeling software, establishing internal reporting protocols and standards, and allowing substantial time for skilled employees to regularly review and adjust modeling parameters and triggers to the benefit of the credit union’s loan portfolio.