Explore This Section

SRC Insights: Second Quarter 2008

Stress Testing: A Risk Management Tool for Commercial Real Estate Loan Concentrations

In December 2006, the Federal Reserve issued SR Letter 07-1: Interagency Guidance on Concentrations in Commercial Real Estate (guidance) to "remind institutions that strong risk management practices and appropriate levels of capital are important elements of a sound commercial real estate (CRE) lending program."1 The guidance was issued in response to concerns surrounding the changing real estate environment and the increase in CRE lending activities over the last five years. Institutions that are actively involved in CRE lending should regularly assess the CRE portfolio to identify potential concentrations and ensure that risk management practices are in line with the size and complexity of the CRE portfolio.

This is the first part of a three-part series on stress testing and sensitivity analysis, which will both be referred to as stress testing for the purpose of this article. There are several elements in a risk management framework that identify, monitor, and control CRE concentration risk, and this part outlines the basics of stress testing and its benefits as a risk management tool. The second part in the series, to be published in the third quarter issue of SRC Insights, will address specific stress testing programs for unique portions of the CRE portfolio. The third and final part in the series will provide tips for developing a strong management oversight and contingency planning program and will be available in the fourth quarter issue of SRC Insights.

Current Stress Testing Practices
Adoption of portfolio-wide stress testing for CRE portfolios has been slow, and a few factors may be contributing to the issue. The first is that many community banks are still trying to reach or have just reached the point where they are only now able to develop consistent, meaningful, and relevant loan concentration reporting, an important first step in establishing a foundation for stress testing. Many institutions are showing meaningful progress in this area. This may be due to several factors: declining market conditions make stress testing a more relevant management tool, regulatory expectations are becoming more defined and better understood, and information system enhancements are evolving to accommodate data collection requirements.

The second reason stress testing has been slow to be implemented is that guidance specific to an individual institution's needs may be limited. The general "one size fits all" approach is not likely to produce meaningful results and may result in burdensome methodologies that are difficult to implement or are irrelevant. So why stress test? The value of stress testing, when meaningfully applied, is that it:

  • Provides a useful tool in diagnosing areas where potential risks may affect the portfolio
  • Provides meaningful insight into the durability of a loan portfolio to withstand changes in the internal and external environment
  • Develops proactive risk mitigation strategies for the future to protect financial performance and capital adequacy (a bank that finds vulnerabilities to particular risks may use the information to change its policies and strategies)

Where to Begin?
Depending on the risk characteristics of the CRE portfolio, stress testing may be as simple as analyzing the potential effect of stressed loss rates on the CRE portfolio, capital, and earnings. In its most simple form, a "break-even" scenario would identify the maximum loss rates that a bank could sustain while maintaining its capital levels in accordance with internal policy and regulatory requirements.

On the surface, a low tolerance level may cause management to forgo more detailed stress testing given the severity of this cursory analysis; however, more detailed stress testing may prove to be a very useful diagnostic tool for developing remedial strategies. Conversely, a high tolerance level, at least on the surface, may cause complacency. This may leave management with the impression that there is no need to stress test the portfolio components. Stress testing in this case, however, could provide significant insight into specific vulnerabilities that are diluted during a "broad brush" approach.

Institutions should remember that stress testing does not need to utilize sophisticated and expensive models. What is important is whether or not the stress testing program is appropriate for the size, nature, and complexity of the bank's CRE lending activities. A meaningful stress testing program, at its most basic level, is one that has:

  • A foundation (i.e., CRE loan portfolio data) for modeling that is detailed, accurate, relevant, and able to be updated easily and regularly
  • Simulations (or "what ifs?") that are appropriate, meaningful, and relevant to the size and nature of the institution and the markets that it serves
  • A process for getting to the "so what?"
  • Management oversight and review

Portfolio Concentration Reporting-The Foundation
The same data utilized in developing portfolio concentration reports can be utilized as a foundation for stress testing. These portfolio concentration data can be utilized to:

  • Determine which area of the portfolio needs to be tested first
  • Determine what data within that sector need to be tested
  • Help define appropriate testing scenarios

Determining which area of the portfolio needs to be tested first can be as simple as reviewing the portion of the CRE portfolio with the highest dollar exposure or the segment that is most likely to be affected by current or prospective external factors. A portfolio that is diversified by sector may have geographic concentrations. Interdependencies in sectors may also call for stress testing multiple sections of the CRE portfolio. Regardless, the analysis should focus on the more vulnerable segments of a bank's CRE portfolio, taking into consideration the prevailing market environment and the bank's business strategy.2

Raw data inputs for testing will depend on the nature of the sector being tested. Common raw data utilized in stress testing include: loan outstandings, interest rates, interest rate spreads, collateral values, revenues, adjusted gross income (AGI), vacancies, expenses, net operating income (NOI), targeted sales prices, interest reserves, and other types of loan information. Raw data or loan inputs will largely depend on the unique nature of the segment to be tested and the anticipated events and external factors that are of concern. We will discuss the relationship between raw data inputs and specific stress testing scenarios in the second part of this series in more detail.

Some institutions are able to download pre-coded loan data directly from their loan accounting systems into spreadsheets that can be manipulated manually or through macros. Other institutions choose a more rigorous approach and utilize more comprehensive and integrated stress testing software or programs such as Argus Asset Management, Moody's KMV, etc., particularly when portfolios are larger, more complex, and affected by multiple variables-or where significant interdependencies apply. Other institutions choose to manually input loan data into spreadsheets and work from there. Some institutions, based on their size and systems limitations, may even aggregate individual loan information from loan files or credit write-ups in order to gather information for testing.

Regardless of the type of platform used, it is important that it be flexible and able to accommodate changes in information without considerable burden. Changes in strategic direction, loan growth, market conditions, competitive pressures, demand shifts, economic conditions, and other factors will need to be considered. The model's ability to adapt to these changes is imperative if stress testing is to remain relevant.

Development of "What If?" Scenarios
Developing "what if" scenarios is key in stress testing. Management should start by asking two basic questions-what keeps us up at night and what do we worry about? Management can begin by stressing only one or two variables and determining how significant the impact could be to individual loans or groups of loans. At the loan level, the question becomes what might cause this borrower's loan to default-tenant loss, absorption rate decline, or cost of construction increases? And while at the portfolio level, the question becomes what might cause many borrowers to default-increasing unemployment, increasing interest rates, or increasing cap rates and vacancy rates-the key is determining what is most appropriate based on the unique characteristics and resources of your institution.

We will discuss stress testing scenarios for specific loan sectors, such as income-producing commercial property, land acquisition and development (LAD), and construction loans, as well as other sectors, in the second part of this series in more detail.

Getting to the "So What?"
Once stress testing is completed and the quantitative results are produced, management needs to analyze the data and draw conclusions, answering the "so what?" question. How could the results of the various stresses employed impact the portfolio, financial performance, competitive position, market, etc.? Based on the results of the stress tests, management will be able to answer questions like: Should capital levels be adjusted? Are more provisions needed to ensure the adequacy of the ALLL? Should more credit enhancements be required going forward to strengthen the portfolio? In general, getting to the "so what?" will enable management to strengthen the strategic decision-making process.

Board and Management Oversight and Review
Regardless of the structure of the stress testing program, no program is effective when it is not fully supported by senior management. Management can emphasize the value of the program in many ways, including:

  • Developing a policy for stress testing
  • Assigning responsibility (and, ultimately, accountability) for the stress testing program
  • Establishing ongoing periodic management reporting requirements
  • Utilizing the results as part of the strategic decision-making process
  • Developing a contingency plan to mitigate areas of weakness

Whatever decisions are made during this process, in the beginning, it is important that management accept the fact that the policies, reporting frequency, data utilization, and contingency planning elements may be somewhat fluid. This will enable management to better understand the portfolio and become more proactively focused on market information retrieval and develop the means and ability to react accordingly. Management must also consider its own unique markets and individual risks. The key is to start somewhere, stick with a program, and be open to what works and what does not.

Suggestions for a strong stress testing oversight program will be included in the last of this three-part series. If you have questions related to stress testing or other CRE risk management inquiries, contact Jim Adams at (215) 574-4325 or Sharon Wells at (215) 574-2548.

The views expressed in this article are those of the author and are not necessarily those of this Reserve Bank or the Federal Reserve System.