Home Equity Lending: A HELOC Hangover Helper — Part 2*
by Michael Webb, Managing Examiner, Federal Reserve Bank of Richmond
The first article of this two-part series provided an overview of the risks inherent in home equity line of credit (HELOC) lending activity, especially for institutions with large HELOC portfolios. This article discusses some risk management elements — specifically, internal controls, management information systems (MIS), policies and procedures, board reporting, and loss mitigation strategies — that can help minimize these risks.
Internal Controls and Management Information Systems
For many community banks, MIS reports are often transactional rather than portfolio based. While transactional reports are effective for overseeing the activity of individual borrowers, these reports typically are not sufficient to determine whether the composition and overall risk profile of the HELOC portfolio remain within the acceptable risk levels defined by the bank's board of directors. Unfortunately, some banks are tracking little more than the delinquency status for their HELOC portfolios. This is particularly troubling given the recent shifts in consumer payment hierarchies, as discussed in the first article. Ideally, MIS should capture credit risk in the HELOC portfolio both in terms of the likelihood of customer default (often referred to as probability of default, or PD) and the potential loss in the event of a default, or loss given default (LGD).1
Supervision & Regulation (SR) Letters 05-11 and 12-3 highlight key factors to consider when performing risk analysis of HELOC lending.2 Those factors most applicable to community bank HELOC lending are discussed below.
Probability of Default Analysis
Credit Scores and Line Utilization
Monitoring changes in credit scores can be a very effective predictive tool. Although individual results vary, declining credit scores generally have a strong correlation with higher delinquency rates. However, consumer repayment assessment should begin, but not stop, with an analysis of credit scores. Ultimately, the accuracy of any predictive model is best determined by validation and backtesting; solely assessing the relationship between declining credit scores and default rates can leave lenders and risk management officers with an incomplete picture of portfolio risk.
Academic and industry studies also suggest a strong correlation between high credit line utilization rates and defaults on open-ended credit products.3 Tracking line utilization rates and segmenting HELOC portfolios accordingly is a useful risk management tool, particularly if these metrics are leveraged against other data such as credit scores and recent delinquency rates on the first-lien position. Subsegments of portfolios based on multiple high-risk variables could provide a solid base for robust analysis. For example, knowing that 21 percent of the borrowers in the HELOC portfolio have a credit score below 700 is beneficial. However, knowing that half of that segment also has a line utilization rate above 90 percent adds far more context that could trigger deeper reviews. Tracking line utilization is fairly simple to do with most loan system software.
Origination Terms, Initial Underwriting Standards,
Community bank HELOC portfolios are subject to many of the systemic risk issues that affect the HELOC portfolios of their larger counterparts, such as reset volumes and changes in consumer payment behavior. Some community banks nevertheless appear to have avoided the underwriting pitfalls associated with the precrisis credit boom by, for example, underwriting their own loans and avoiding the low-doc and no-doc products originated by brokers. Furthermore, because these community banks originated loans to hold in their portfolios rather than to sell, they ensured that appropriate appraisals were used at loan origination.
However, many community banks failed to factor in borrowers' ability to service their HELOC lines on an amortizing basis at origination. As highlighted in the first article, the difference in debt service requirements can be dramatic when a HELOC converts to amortizing terms. However, on an encouraging note, examiners in one Federal Reserve District noted that a number of community banks had already identified this risk and modified lending terms to require debt-to-income (DTI) analyses on an amortizing basis. These banks also seemed to have addressed, in part, the risk of higher interest rates by assuming a 6 or 7 percent interest rate in their DTI calculations.
Although these changes in banks' policies are expected to improve the quality of future HELOC originations, they do not address risk in existing portfolios. This issue could have a greater impact as HELOC portfolios approach peak reset years for the 2004–2008 vintage originations. Banks that originated HELOC loans with underwriting terms that accounted solely for interest-only payments should incorporate revised amortizing DTI estimates into routine risk management processes and default calculations. Borrowers with a high potential for distressed cash flows, high utilization rates, or increased loan-to-value (LTV) ratios should be considered for additional analysis. The results of these efforts could identify high-risk HELOC relationships that may be considered for nonaccrual treatment, inclusion into analyses of the adequacy of the allowance for loan and lease losses (ALLL), and other workout or remediation options.
Other Probability of Default Issues
Other practices that may improve PD analysis include segmentation of HELOC lines by the underlying property type, such as liens on rental or vacation properties; origination channel (if applicable); and combined loan-to-value (CLTV) ratios. The CLTV ratio, in particular, is critical and warrants further discussion in the following section.
Loss Given Default Analysis
Real estate values in most markets have at least stabilized, and, in many cases, they have experienced appreciation in recent months. Whether this trend is the beginning of a meaningful recovery in real estate values or a temporary anomaly based on a surge in cash-based investor purchases is the subject of some debate. Regardless, community banks need to be aware of the LGD profile of their HELOC portfolios, particularly if the business line represents a substantial concentration of assets. As a general principle, lenders should not wait until a HELOC loan is seriously delinquent to update collateral values. Indeed, updating CLTVs is a critical principle in understanding HELOC credit risk, ALLL analyses, and the need for active loss mitigation strategies. Revaluing real estate collateral on performing assets can be a hard sell for banks with large HELOC portfolios. For community bankers already pressed by tight net interest margins, it is not difficult to understand why senior management might be reluctant to do so. However, updating collateral values does not need to be an expensive exercise.
A "Prox" on Thy House…
Automated valuation models (AVMs) can provide a much more cost-effective alternative to full appraisals. Some valuation model applications can be fairly robust in terms of analysis, and, if well documented and supported, a population of updated AVMs can form the basis for updating internal evaluations based on a general proxy adjustment to market values. For example, a population of AVM updates on 50 homes in a particular area in a bank's lending footprint can support the conclusion that market values in that area have declined an average of 11 percent since 2006. Vintage 2006 origination values can then be adjusted downward by 11 percent, or some factor thereof, to support ALLL analyses.4 Moreover, banks have the discretion to order updated appraisals or AVMs based on an assessment of the other PD variables previously discussed, such as all HELOC loans with originating CLTV ratios greater than 85 percent and average line utilization of 90 percent or more. With some well-detailed AVM models costing as little as $20 per property, such measures can keep the costs of updating collateral values well contained.
For community banks in rural areas, sales volume can be insufficient to form a basis for an AVM approach. The best alternative for such organizations may be to perform similar proxy analyses using tax-assessed values for each county in the lending footprint. The sales in a given rural county over the course of a year will typically be far fewer than those in an urban area, but documenting and updating the relationship between appraised values and tax-assessed values can still form a solid, cost-effective basis for internal analysis.
Bank management should also be mindful of the minimum requirements set forth in SR Letter 10-16, "Interagency Appraisal and Evaluation Guidelines," regarding real estate valuation methods that support credit decisions.5 Many AVMs contain sufficient detail to substantiate updates to existing lines of credit for internal analyses, but they may not meet all of the acceptable evaluation criteria to support loan originations or renewals. In addition, LGD analysis should also factor in the cost of selling the property, whether the bank uses an AVM or a full appraisal to value collateral.
With appropriate collateral valuations, lien position status is an easy, effective metric for capturing LGD exposure for management reporting and ALLL analysis, and it relates directly to CLTV ratios. Many banks capture this exposure by segmenting the lien position into three categories: first-lien HELOCs, second-lien HELOCs behind their own first positions, and second liens behind third parties. Second-lien positions behind third parties are obviously the most susceptible to loss given declines in property values, and ALLL factors can be targeted to capture this exposure. Similarly, this segment should be subject to far more ongoing monitoring should management need to pursue higher allowances and active loss mitigation strategies.
Other Odds and Ends — Back-End Control Functions
When considering the scope of the internal audit and loan review functions for most community banks, retail lending, in general, is rarely the first business line that comes to mind. The inherent credit risk in loans collateralized by residential real estate has traditionally been far lower than commercial credits, and the relatively low dollar exposure on a per-event basis has usually been sufficient justification to limit coverage or scopes for both disciplines. From an audit standpoint, scope is often limited to an assessment of loan administration and accounting processes, postings, and similar back-office functions often embodied in an overall loan administration audit. Loan review personnel typically conduct a large portion of their sampling based on the dollar amount of the subject loans, and retail loans are usually too small to warrant assignment of internal loan grades. Community banks with significant HELOC exposure should assess the roles of internal audit and loan review to determine whether those control functions could enhance their analysis with modified coverage that focuses on risk management processes rather than standard "sample-and-grade" techniques. Aside from compliance-related reviews, traditional transaction testing may not provide as much value as conducting reviews of risk management practices, policies and procedures, and controls.
Policies and Procedures
Based upon a recent assessment by one of the Federal Reserve Districts, even those banks with the strongest HELOC risk management practices needed more robust policies governing the business line. Policies focused almost exclusively on initial underwriting terms often collapsed into little more than underwriting crib sheets containing DTI, LTV, and credit score targets. Little was noted in regard to ongoing risk management practices. HELOC policies should describe procedures needed to assess overall portfolio risk after origination, such as updating credit scores or LTV ratios and assessing DTI ratios, line utilization, origination vintage, property type, and lien position. Banks should address the suggested frequency of these recurring practices and clearly detail possible remediation plans, such as freezing or closing the line in a manner compliant with Regulation Z, segmenting the portfolio by risk, and provisioning for riskier segments.6 They should also address a protocol for board reporting of key risk metrics.
Charge-Off and Uniform Retail Credit Classification Policies
Although many banks have focused on the Financial Accounting Standards Board's Accounting Standards Codification (ASC) sections 310 and 450, they are still expected to comply with the standards in the Uniform Retail Credit Classification (URC) and Account Management Policy.7 Banks may perceive the ASC requirements as a more conservative approach; however, in some cases this can be a misperception based on an incomplete interpretation of the regulatory guidance.
In response to the financial crisis, many banks revised their standards for internal loan risk grading, including lowering the dollar threshold for relationships that would be individually analyzed, graded, and measured for potential impairment. However, what was generally viewed as more conservative treatment of these individual credits often resulted in untimely charge-off practices for some retail loans, including HELOCs, as banks in many cases did not distinguish between retail and commercial credits for charge-off purposes. Commercial grading and impairment processes were misapplied to certain retail loans; this involved taking specific allowances against impaired, but not collateral-dependent, loans regardless of delinquency status. If those loans were in a state of severe delinquency, this treatment would be inconsistent with URC guidance, which contains specific, delinquency-based thresholds that trigger a charge-off of a retail loan.
For example, a bank could have a $300,000 HELOC that was individually analyzed for impairment, and the bank could have assigned an internal rating of "doubtful," with specific impairment taken in full in the ALLL. Since the URC methodology does not employ a "doubtful" rating, the bank's treatment could result in higher credit risk classification numbers if the loan was within the 120- or 180-day delinquency period (as it would be "substandard" under URC with certain LTV ratios). Such a loan may not have been charged off even after breaching the URC delinquency thresholds. This can result in an overstatement of credit risk classifications, the ALLL, interest income, and possibly nonaccrual numbers, as well as an understatement of period loan losses. A bank can adopt a more conservative credit risk classification approach on a retail loan, but regardless of the internal classification, it must adhere to URC charge-off triggers. Consider that SR Letter 00-8 clearly states:
Actual credit losses on individual retail credits should be recorded when the institution becomes aware of the loss, but in no case should the charge-off exceed the time frames stated in the policy. This policy does not preclude an institution from adopting a more conservative internal policy.
And in regard to home equity lending, the letter states:
Home equity loans to the same borrower at the same institution as the senior mortgage loan with a combined loan-to-value ratio equal to or less than 60 percent need not be classified. However, home equity loans where the institution does not hold the senior mortgages that are past due 90 days or more should be classified Substandard, even if the loan-to-value ratio is equal to, or less than, 60 percent.
The Federal Reserve does not discourage banks from adopting more conservative treatment on credit risk classifications, provided that the subject loans are always charged off by delinquency in accordance with guidance. Banks need to charge off retail loans in compliance with SR Letter 00-8, regardless of internal loan grades and ALLL analyses.
A sound HELOC program does not simply end with the implementation of robust internal controls, MIS, and policies. An article in a previous issue of Community Banking Connections underscored the responsibility of the bank's board of directors for establishing the risk philosophy of the organization and for holding management accountable for implementing sound policies and procedures.8 The quality of information presented to the board of directors is paramount to its ability to fulfill that responsibility. The information presented on HELOC exposure should be similar in nature to other risks; reports should contain consistent, salient, timely information in a format that allows for period-to-period comparisons. The metrics should be compared against policy limitations and benchmarks. Exceptions to policies and procedures should be detailed and aggregated based upon common exceptions, such as breaches of internal LTV ratio limitations. The volume and the formatting of information are critical considerations; presenting a stack of data-intense reports to each director is unlikely to impart a sense of overall risk, particularly if no synopsis of the data has been prepared. Beyond these considerations, management has substantial discretion over the nature of the metrics to be presented, as long as those metrics capture the risk in an adequate manner.
Loss Mitigation Strategies
In banking, the term loss mitigation is frequently used interchangeably with collection, workout, or special assets. However, in the context of this article, loss mitigation refers to an active process in which risks in HELOC structures, cash flows, or collateral are identified and addressed well before those risks manifest themselves in delinquencies. Freezing further advances on credit lines seems to be a prominent, well-understood strategy for most retail credit managers, particularly in regard to Regulation Z requirements. However, this practice has legal and reputational risks, and management may wish to act on a problem before a significant decline in collateral value occurs.
Renegotiating problematic HELOC loans may provide a path of lesser resistance with fringe benefits. HELOCs in lien positions junior to third parties are prime candidates for such treatment, particularly if they are approaching conversion to amortizing status. Renegotiated structures tend to vary based on individual circumstances, but a common objective is to capture the first-lien position, along with the existing second, into a new amortizing structure. Some community banks view this effort as a risk management initiative, rather than an income generator, and waive some fees and points to provide borrowers with additional incentive to refinance. Such a program can be a "win-win" for all parties.
With appropriate underwriting and risk management practices, second-lien residential lending can be a highly profitable business line and serve as a ready investment alternative to banks with heavy concentrations in other asset classes. HELOCs can also serve the personal financial needs of the bank's customers. Most banks with substantial second-lien exposure appear to be gradually, but steadily, acclimating to regulatory guidance, and risk management practices in general are in varying stages of development. Much remains to be done, however, and community bank managers should feel free to contact their supervisor for additional insight and perspective on HELOC risk management strategies.
Back to top
- * This article is the second of a two-part series that explores home equity lending. The first article, titled "Home Equity Lending: A HELOC Hangover Helper," appeared in the Second Quarter 2013 issue of Community Banking Connections and is available at www.cbcfrs.org/articles/2013/Q2/Home-Equity-Lending-A-HELOC-Hangover-Helper.
- 1 When this article discusses PD and LGD — which are parameters that are widely used by many large banking organizations in credit risk modeling — the intent is to refer to their conceptual meaning rather than to their use as credit metrics. Community banks are not expected to calculate quantitative PD or LGD estimates for extensions of credit.
- 2 SR Letter 05-11, "Interagency Credit Risk Management Guidance for Home Equity Lending," is available at www.federalreserve.gov/boarddocs/srletters/2005/sr0511.htm. SR Letter 12-3, "Interagency Guidance on Allowance Estimation Practices for Junior Lien Loans and Lines of Credit," is available at www.federalreserve.gov/bankinforeg/srletters/sr1203.htm.
- 3 For example, see Ronel Elul, Nicholas S. Souleles, Souphala Chomsisengphet, et al., "What 'Triggers' Mortgage Default?" Federal Reserve Bank of Philadelphia Working Paper 10-13 (2010); Julapa Jagtiani and William W. Lang, "Strategic Default on First and Second Lien Mortgages During the Financial Crisis," Federal Reserve Bank of Philadelphia Working Paper 11-3 (2010); and Anne-Sophie Bergerès, Philippe D'Astous, and Georges Dionne, "Is There Any Dependence Between Consumer Credit Line Utilization and Default Probability on a Term Loan? Evidence from Bank-Level Data," Canada: Interuniversity Research Centre on Enterprise Networks, Logistics and Transportation (CIRRELT) Research Paper CIRRELT-2011-45 (2011).
- 4 The Truth in Lending Act (TILA) and Regulation Z, TILA's implementing regulation, prohibit a creditor from freezing or reducing a HELOC unless an exception applies. See 12 CFR section 1026.40(f). One exception under section 1026.40(f)(3)(vi)(A) is when "the value of the dwelling that secures the plan declines significantly below the dwelling's appraised value for purposes of the plan." The Federal Reserve System's compliance publication, Consumer Compliance Outlook, included an article that discussed these requirements. See Jason Lew, "HELOCs: Consumer Compliance Implications," Consumer Compliance Outlook (Third Quarter 2008), available at www.philadelphiafed.org/bank-resources/publications/consumer-compliance-outlook/2008/third-quarter/q3_02.cfm.
- 5 SR Letter 10-16 is available at www.federalreserve.gov/boarddocs/srletters/2010/sr1016.htm.
- 6 See Regulation Z, Truth in Lending Act (12 CFR section 1026) at www.ecfr.gov/cgi-bin/text-idx?SID=76578ad310d99586ead5c0ae4e43df94&tpl=/ecfrbrowse/Title12/12cfr1026_main_02.tpl.
- 7 SR Letter 00-8, "Revised Uniform Retail Credit Classification and Account Management Policy," is available at www.federalreserve.gov/boarddocs/srletters/2000/SR0008.htm.
- 8 See Kevin Moore, "View from the District: The Importance of Effective Corporate Governance," Community Banking Connections, Fourth Quarter 2012, at www.cbcfrs.org/articles/2012/Q4/Importance-of-Effective-Corporate-Governance.