Discoveries in retail credit risk management

RMA Journal, The, April, 2003 by Mikkalya Murray

I have sometimes felt like Christopher Columbus must have felt exploring the New World. When I began my quest to discover the New World of risk management, I didn't know where I was going or how I would get there. I only knew there was something out there worth the effort. During the journey, I never knew exactly where I was or what I had. I only knew that the results were worth far more than the price paid. When I returned to my day-to-day world, I knew only that the discovery was beyond my expectations and the possibilities were endless.

The New World of risk management has much in common with the Age of Discovery. As credit risk managers, we face new challenges and work diligently to reveal the unseen and unknown. We enjoy our new accomplishments, and we get much satisfaction from providing a practical framework to better manage the day-to-day risks faced in all financial services venues. My quest is to explore, develop, and expand retail credit risk management (RCRM). As community bankers, we can build a better model by consistently employing the basic tools and then exploring more advanced tools.

Welcome to the New World

In the past decade, risk managers have become the champions of their credit products. No longer are we simply the guardians of asset quality. Nonetheless, managing asset quality is the imperative. Failure to execute a sound RCRM process is, indeed, a fatal error.

In the ever-changing sea of retail risk management, we are relying more frequently on analytical tools to adjust the composition and the acquisition of new credits (velocity of the portfolio). As risk managers we are less product focused and more customer focused as we search for efficiencies and improvement in the performance of the portfolio segments. We use all the basic tools we have always used, but have added forecasting to predict performance results. Given the fragile condition of the consumer balance sheet, managing embedded, emerging, and acquired risk requires real agility. But a smooth sea never made a skilled mariner.

Blending the Old with the New

A decade ago, we relied heavily on professional experience and intuition. Today, our goal is to create a profitable, well-managed retail portfolio, one where we maximize earnings, identify growth opportunities, and promptly mitigate emerging risks.

New techniques help us say yes while retaining asset quality. New tools and processes have replaced or enhanced the traditional process. Instead of managing approvals, credit by credit, we now accept an identified range of origination risk. Charting the benchmark exception levels and carefully identifying risk attributes in different terms were two of the critical first steps. In this New World, we create:

* Exception-to-policy benchmarks measured in number, dollar, and percentage of capital, by product line.

* FICO ranges/arrays to visually display the consumer repayment risk attributes from a historical perspective.

* LTV ranges, captured on a regular basis, which set a baseline and act as an early-warning signal should a sea change occur.

* D/I arrays, which are good indicators of adherence to existing underwriting criteria and repayment capacity.

* Risk-based pricing, which becomes much easier to employ on a systematic basis when the risk components are reviewed consistently with an eye toward true earnings potential and sound asset quality.

Charting with MIS

Loan accounting systems and ancillary databases provided much of the data we needed to manage traditional risk components such as delinquency, nonaccruals, charge-off activity, and risk rating. The newer techniques place a much heavier reliance on credit-related MIS. To skillfully employ management-by-exception, we needed to automatically track a wide array of exceptions and risk attributes using a core processing system and/or ancillary databases. This data must be input in many cases, and be scrubbed, to provide a clear view of present and emerging risk within the portfolio segments.

A strong retail risk management process starts with a comprehensive retail credit policy. Reflecting the institution's appetite for risk, the credit policy identifies the product lines offered, their basic terms, and acceptable attributes for retail products. Policy is then applied tactically via procedures used consistently by both the retail delivery staff and support staff. Many have found it beneficial to train both line staff and support staff on policy and procedures to reinforce the credit culture and focus on corporate goals.

Exploring Portfolio Segmentation

Risk managers should carefully review each product line to determine if it accomplishes the profit plan and achieves volume and the desired product mix while maintaining asset quality. Segments in the portfolio might include:

* Residential real estate.

* Auto.

* Unsecured.

* Revolving vs. closed-end loan.

* Populations of exceptions-to-policy (ETPs).

* High LTV, high D/I low FICO (high risk to default, etc.).

After appropriate and measurable levels of risk are expressed in terms of percentage volume, dollars, and/or number of notes, by segmented product line, it is easier to track portfolio performance relative to both the profit plan and asset quality.