Concentration alert: why you should adopt better commercial real estate risk management...

Over the past decade, banks and thrifts of all sizes have significantly increased their exposure to commercial real estate (CRE) lending. The largest percentage increases have occurred at institutions with $10 billion or less in total assets. At the end of the third quarter of 2005, construction

In response to this trend, the Federal Reserve, the FDIC, the OCC, and the OTS published proposed interagency guidelines and best practices on Jan. 9, 2006. It is unclear at the time of this writing if final guidance will be issued later this year and what it will say. [The comment period was extended in March to April 13.] Regardless of the final outcome, the guidelines as initially proposed are far-reaching and make for prudent business practices. More plainly, real estate lenders that do not adhere to the principles outlined in the proposed guidelines may underestimate the risk in their portfolios and subject themselves to substandard portfolio credit performance.

The implications of enhancing risk monitoring--i.e., adopting the guidelines if and when they are finalized--are significant. First, many banks may be forced to change their current business model. Roughly one quarter to one third of all supervised institutions have portfolio CRE concentrations that exceed proposed capital thresholds. As a result, these lenders will need to meet "heightened risk management practices" and/or carry more capital to avoid enhanced scrutiny. Either way, the profitability of real estate lending may be diminished, and severely impacted banks will likely need to find alternative lending opportunities. Second, these same consequences will force many lenders to reconsider CRE pricing. Specifically, they will need to analyze and offset increased capital allocations and monitoring costs to maintain profitability.

Banks with inadequate risk monitoring must also determine how to implement an improved risk management infrastructure. While the proposed guidelines specify board and management responsibilities as well as what banks must measure, getting there--i.e., changing day-to-day practices, and, even, more so, the lending culture--can be very daunting.

The remainder of this article presents a framework for meeting two of the more esoteric yet intrinsic requirements implied in the proposed guidelines: achieving consensus on the bank's tolerance for risk and defining the model portfolio. These are also the most fundamental elements of sound risk management and critical first steps to setting loan policy and establishing effective portfolio monitoring.

Tolerance for risk

At the industry level, it is easy to distinguish the risk tolerance between sub- and superprime lenders. Ask a number of bankers at a single institution, however, and you often find inconsistency in how they articulate their own bank's tolerance for risk. This is alarming, as risk tolerance drives lending strategy and model portfolio definition, which in turn influences risk policies, procedures, systems, and controls. Lack of understanding and disagreement over risk tolerance also lead to disconnects between growth and credit quality goals; may cause frontline lenders to focus on the wrong opportunities; lead to wasted time in the credit committee; and ultimately, create unhappy management, lenders, borrowers, and shareholders.

How do you achieve consensus regarding tolerance for risk? First and foremost by involving people at all levels of the organization. Consensus is not easy. Often, the board of directors and line have polar opposite goals. It is therefore critical for centers of influence-formal and informal thought leaders across the bank--to participate in the definition process to help drive cultural acceptance and companywide adoption.

Next, while bankers may often feel more comfortable with certain types of lending, it is usually difficult to get them to verbalize clearly the institution's overall risk tolerance. One way to help cultivate more thoughtful and tangible discussion is to break the elements of credit culture into four continuums (see graphic below). It is generally easier for bankers to comment on their company's strategic priorities as well as the current and ideal risk culture, risk strategy, and level of risk control. This exercise helps to clarify how significant the thinking gap is and illuminate inconsistencies between each of the continuums.

As banks move down the left hand (priorities) continuum shown in the diagram below, implicitly, they are willing to accept greater portfolio risk and potential for increased volatility in credit quality and earnings. Consistency across the continuums is critical, and risk controls for growth-oriented companies must therefore be designed accordingly. No set of risk controls, however, can totally eliminate the risk inherent in aggressive lending businesses. Note that the regulators have identified commercial real estate as a high-risk type of lending that requires more comprehensive policies, procedures, controls, and monitoring.

To define a more quantitative benchmark for risk tolerance, participants must decide what loss amount constitutes a material transaction. Also known as the "threshold for pain," this is the largest loss the institution can sustain before it becomes a credit committee agenda item or draws two or more hours of board discussion. For most banks with less than $10 billion in assets, this limit is usually between $150,000 and $250,000 and often becomes the point at which credits are manually underwritten or require a more extensive loan package. Normally, deals below this material threshold make up 60-70% of transactions but only 10-15% of total balances outstanding.

The model portfolio

The next critical building block is the model portfolio. This core component of the loan policy specifies the target portfolio mix for the bank and is the benchmark against which to measure and manage portfolio performance. A bank's model portfolio must reflect the consensus of the board and executive management regarding the institution's tolerance for risk.

Specifically, the model portfolio should define portfolio limits for:

Mix of lines of business (i.e., percentage of portfolio in CRE, acquisition and development, construction, 1-4 family mortgages, etc.);

Industry (by NAICS), property type, and agriculture type concentrations;

Individual borrowers as well as the top 10 to 50 borrowers, in aggregate (the appropriate number depends upon the individual limits and the target size of the entire portfolio);

Distributions of asset quality ratings (AQRs, also known as risk grades); and

Tiered maximum individual borrower exposures by AQR.

To be effective, the asset quality rating framework must have sufficient detail to differentiate among performing borrowers and allow migration analysis among the pass (approval) categories Specifically, banks with portfolios of $100 million or more should define at least five pass categories plus the required regulatory classifications.

Banks should assign risk ratings to individual borrowers and borrower transactions to reflect potential for default and loss given default. The proposed guidelines suggest real estate lenders should also assess and assign risk grades to property classifications and establish guidelines to prevent acute covariance; that is, excessive loans to high-risk borrowers for high-risk property types. (Note the proposed guidelines pertain specifically to income-producing properties. We agree with the regulators that the risk associated with owner-occupied real estate mortgages tends to be driven much more by the industry of the property owner than the property type itself. The implication for risk managers is that they must be able to segregate their portfolio accordingly.)

Furthermore, real estate lenders should also develop perspectives and guidelines regarding the specialized criteria they need to monitor to track portfolio risk adequately. These factors are summarized in the following table above.

Model portfolio benchmarks enable risk managers to maintain a forward-looking perspective. By establishing concentration limits and later monitoring portfolio performance against those limits, banks can recognize deviations and anticipate potential credit quality deterioration. Of course, the model portfolio also should define desired ranges over the business cycle for key measures of portfolio credit quality, including delinquencies, criticized/classified assets, non-performing and non-accrual loans, and losses.

Banks should reevaluate their model portfolios--and loan policies, in general--periodically to ensure alignment with the bank's overall strategy and tolerance for risk. This practice is especially important as institutions grow and evolve to where they begin to consider more specialized types of lending. On the other hand, it is crucial to resist temptation and changing model portfolio targets in boom periods to enable higher concentrations and looser decision criteria. The model portfolio should be designed to withstand short-term trends and market cycles.

Upon defining the model portfolio, banks need to adjust loan policy, approval processes, and monitoring capabilities accordingly. These tasks are much more tactical and should follow from the activities described earlier in this article.

It remains to be seen what will come of the proposed guidelines. However, given the recent explosion in property values and the number of institutions with inadequate risk management, we hope that banks will not wait to begin adopting what we believe are critical lending practices.

Construction and Land Development

* Physical and economic occupancy by price point and market (commercial)

* Sales by price point and market (residential)

* Aging by price point and market

* Permitted construction

* Under construction (speculation vs. contract)

* Absorption

* Asking vs. actual prices/rents

* Average sales time

* Market data by market and price point

* Exposure with all borrowers in a single subdivision or development

* Builder/borrower exposure to other lending institutions.

Income Property Management

* Physical and economic occupancy

* Market data by sub-market, property type and grade:

* Physical and economic occupancy

* Permitted construction

* Under construction

* Absorption

* Asking vs. actual prices/rents.

John Barrickman, president, New Horizons Financial Group, Roswell, Ga., and Gary D. Stein, partner, Capital Performance Group, Washington, D.C.