UNconventional Wisdom is a periodic guest blog, where authors hold up the so-called conventional wisdom to a fresh perspective, or apply common principles in new ways. To propose a guest blog, email Steve Cocheo, executive editor & digital content manager.

By Michelle Lucci, risk management consultant, Bankers Toolbox. Bankers Toolbox produces Commercial Real Estate Loan Portfolio Stress Testing (CREST) software, which is endorsed by ABA’s Corporation for American Banking. Lucci is a former federal examiner.

 

 

In recent conference appearances I’ve been presenting the elements of commercial real estate concentration analysis and stress testing. However, it turns out some bankers have questions and concerns about the treatment of concentrations of residential real estate loans during their regulatory examinations.

Most such questions that came up in recent sessions came from bankers formerly regulated by the Office of Thrift Supervision, now regulated by the Comptroller’s Office. These institutions have historically originated the majority of their loans in the residential category in part because they were subject to the thrift lender test on the Thrift Financial Report.

However, this portfolio composition is now being viewed as a concentration of credit, which even at 200% of capital placed one institution in the group of at best a moderate risk by OCC.

This banker was understandably upset at this news but it really should not be much of a surprise. The OCC’s Handbook on Concentration of Credit, issued in December 2011, states: 

“Other credit concentrations, such as loans secured by first liens on residential real estate, have historically posed fewer problems. However, during the recession of 2007–2009, the banking industry experienced significant losses in these exposures when the national housing market suffered broad declines in home values. This experience indicates that although a concentration has not proven problematic in the past does not mean that it is precluded from becoming a problem in the future.”

 

 

Action in the face of differing policy
So what can you do if your institution now falls into this category? These loans are typically long-term in nature, so short of selling them the best you can do is actively manage them.

I see this as both an opportunity and a challenge to demonstrate your knowledge of the local real estate market and your portfolio. What I think regulators are looking for is the bank to acknowledge the risks and demonstrate that every effort is being made to mitigate the risks, while protecting the bank’s earnings stream and capital base.

The steps you take along the way will also demonstrate that the bank is in compliance with several pieces of key regulatory guidance. Fortunately with a concentration in residential real estate I think this can be accomplished fairly easily.

 

 

Establishing active monitoring of the portfolio
A key step is to actively monitor the collateral values securing these loans based on their locations on both an individual loan basis as well as on the portfolio level. This does not mean that the bank has to reappraise or even reevaluate each loan when there is no circumstance triggering the regulatory requirement.

The 2010 Interagency Guidance on Appraisal and Evaluation Guidelines states that:

“Consistent with sound collateral valuation monitoring practices, an institution can use a variety of techniques for monitoring the effect of collateral valuation trends on portfolio risk. Sources of relevant information may include external market data, internal data, or reviews of recently obtained appraisals and evaluations.” 

To accomplish this task for residential real estate there are a couple of options that I know, of which are free:

• The Housing Price Index (HPI) is a broad measure of the movement of single-family house prices measuring average price changes in repeat sales and refinancings on the same property. The HPI weights price trends equally for all properties and includes house price figures for the nine Census Bureau divisions, for the 50 states and the District of Columbia, and for Metropolitan Statistical Areas (MSA) and Divisions. The Downloadable Data portion of the HPI page on the Federal Housing Finance Agency’s website will allow you to access a txt or csv file with many geographic areas from which to choose.

 

 

• The S&P/Case-Shiller Home Price Indices are also leading measures for the U.S. residential housing market. The U.S. National Home Price Index is a quarterly index of single-family home prices for the nine U.S. Census divisions. The 10- and 20-city composite indices are calculated monthly. The condominium indices track condominium prices in Boston, Chicago, Los Angeles, New York, and San Francisco. These indices measure repeat-sales on the same property on a value-weighted basis meaning that price trends for more expensive home have greater influence on estimated price changes than other homes. Like the index above, the Download Data area of the website will allow you to access the files and these are in an excel format.

Which of these indices are right for your portfolio? That depends on a several factors.

In my opinion both have significant value and the decision is yours. But I will tell you some of my experiences. I work with bankers from different parts of the country and after conferring we chose the appropriate index for each institution based on their individual needs.

For example, one of our customers is based in Idaho. The Case-Shiller Indices don’t offer a trend specifically for this area. So we chose the HPI which has several trends based on MSAs (Boise City, Coeur d’Alene, Idaho Falls, and Pocatello) which encompass this institution’s lending area. In another case our customer operates within metropolitan areas of Florida for which there are two trends (Miami & Tampa) from which to choose from in the Case-Shiller sets.

A major considerations to be aware of is that the HPI All Transactions set also includes refinance appraisals. If you are of the opinion, as I am, that these transactions are a reliable source of home value, then that would be the choice for you.

 

 

Using the data, group by group
Now that you have chosen which index to use, hopefully within your institution you have a piece of software that you can enter these values into (or be really good with Excel!). The collateral value of the property will be adjusted according to the trend of the index.

Doing so will refresh collateral values on every loan in your portfolio, which is a very valuable piece of information from both an individual loan and portfolio level perspective.

From this point I recommend isolating and grouping the loans in several ways:

• Loans that are “underwater” from a collateral perspective: While many homeowners find themselves in this situation and continue to pay their monthly mortgage payments, we know that “walkaway behavior” increases dramatically under these circumstances. If the loans are still current from a payment perspective, a qualitative adjustment to the ALLL would be appropriate to compensate for the heightened risk.

 

 

• Loans by property type: Property type for residential loans has a different meaning, i.e. investor owned or owner-occupied, than commercial real estate loans, i.e. retail, office, warehouse, etc. Residential investor loans experience a much higher default rate and applying a qualitative adjustment to the ALLL would most likely be appropriate in this case as well. Your board may also want to revisit and possibly revise the bank’s concentrations policy, focusing on allowable concentration ranges for this type of asset, as this group will be heavily scrutinized during your regulatory examination.

 

 

• Loans that have a junior lien: Depending on the size of your institution’s junior lien portfolio, this group can and should be segmented in several different ways. You already have a refreshed combined loan to value ratio and now you can take the opportunity to obtain current credit scores. The software should be able to group loans by credit score ranges, and this will give you the opportunity to adjust the ALLL for  declining credit quality. Also consider loans with a junior lien that also have a first lien that has been modified and/or is delinquent. Other segments include junior lien loans where the borrower is making only the minimum payment due and those that will be subject to interest rate shock once rates start to rise. Now is the opportunity to perform a search to ensure that all real estate taxes and homeowner association dues have been paid to date.

The newly released “Interagency Guidance on Allowance Estimation Practices for Junior Lien Loans and Lines of Credit” reminds bankers to monitor all credit quality indicators relevant to credit portfolios.

Adequate segmentation of the residential portfolio will facilitate the ability to track default rates and severity, and can ensure the ALLL is not being maintained lower than it should be if the higher risk segments were not isolated and analyzed separately. The segmentation should include both loans with a first lien only and loans containing a junior lien. The December 2006 Interagency Policy Statement on the Allowance for Loan and Lease Losses states that institutions should adjust a loan group’s historical loss rate for the effect of qualitative or environmental factors that are likely to cause estimated credit losses as of the evaluation date to differ from the group’s historical loss experience.

 

 

• Stress test the entire group of residential loans to analyze what impact a further decline in collateral values would have. The Case-Shiller indices reported that all three composites ended 2011 at new lows. At the ABA’s recent Real Estate Lending Conference, Dr. Celia Chen, senior director at Moody’s Analytics, stated that housing prices won’t stabilize until the end of 2012. She predicts further declines of 3%-5%.

The stress test factors should be decided upon based on the knowledge you have in your lending area. For example, an appropriate decline in Florida, where there remains an oversupply of units and a backlog in foreclosures, will not require the same factors as other locations.

The 2006 Interagency Commercial Real Estate Guidance (officially, the “Guidance on Concentrations in Commercial Real Estate Lending, Sound Risk Management Practices”) requires a “multi-variable” stress test which is slightly easier to perform on CRE loans because bankers make every effort to obtain current financials on income-producing real estate.

However, with residential real estate loans it is not common practice to obtain updated salary information from individuals, so a payment stress test is difficult to accomplish. But you can incorporate another level of stress test when rolling up the results of the portfolio level test. Normally the analyst applies the institution’s average historical loan loss rate to the loan balances that have failed the stress test and recalculates the capital and earnings ratios. To compensate for a portfolio that is displaying heightened risk, the historical loss rate factor applied can be increased, which is in and of itself another form of stress testing.

 

 

• Finally, are the stressed financials within your institution’s risk appetite? There are industry standards available for ratio ranges, however, this determination can only be made on a case-by-case basis. Good luck!