Sageworks Advisory Services PRACTICAL CECL TRANSITION QUALITATIVE POLICY

Transcription

1 Sageworks Advisory Services PRACTICAL CECL TRANSITION QUALITATIVE POLICY

2 Use of this content constitutes acceptance of the license terms incorporated at This content license confers certain rights to you, the user, of this material, and reserves others to Sageworks, Inc. Further, this content license disclaims liabilities for use of the content. It should be carefully reviewed before transmitting or using this content. The following summary is not a substitute for, nor supercedes, the content license agreement referenced above: You may share this content in the exact form it was provided to you, so long as it is shared unaltered and in its entirety, including these terms, and our brands and marks -- any user of the content will be bound by the same Agreement by using the content. While we encourage sharing of this content, you are prohibited from profiting from the use of the content or any derivative works. We make no warranty of any kind for this content -- it is provided in good faith as a courtesy to the industry we are privileged to serve. As this content is provided to an industry, it may have absolutely no relevance or applicability to your particular institution or use case, and your critical thinking is required to successfully use (or not use) the content. We undertake no obligation to keep the content up to date, answer inquiries regarding the content, or support it in any other way. 2

3 EXECUTIVE SUMMARY Current Allowance practices, especially for regulated banks, lean heavily on the use of qualitative factors to achieve reserve levels deemed appropriate by management and acceptable by bank examiners. While Practical CECL is designed to create quantitative justification for the majority of an institution s allocation, this does not mean that qualitative adjustment is no longer appropriate. Rather, institutions should use qualitative adjustment for its original intent a less formal framework to adjust, in a consistent and directionally sensible manner, allocation based on the defensible application of management s judgment. We recommend a qualitative policy result in adjustments that are Minor, Independent, Disclosable, Auditable, and Specific MIDAS. The MIDAS framework for qualitative allocation, combined with the more rigorous consideration of available internal and external information in the Loss Driver Analysis exercise, should result in a justified allowance reserve in a manner optimized for the broadest interpretation of users of financial statements. APPLICABLE GUIDANCE 1 FASB Accounting Standards Update Topic 326 (CECL): When developing an estimate of expected credit losses on financial asset(s), an entity shall consider available information relevant to assessing the collectibility of cash flows. This information may include internal information, external information, or a combination of both relating to past events, current conditions, and reasonable and supportable forecasts. An entity shall consider relevant qualitative and quantitative factors that relate to the environment in which the entity operates and are specific to the borrower(s). p110, , emphasis ours To adjust historical credit loss information for current conditions and reasonable and supportable forecasts, an entity should consider significant factors that are relevant to determining the expected collectibility. Examples of factors an entity may consider include any of the following, depending on the nature of the asset (not all of these may be relevant to every situation, and other factors not on the list may be relevant): a. The borrower s financial condition, credit rating, credit score, asset quality, or business prospects b. The borrower s ability to make scheduled interest or principal payments c. The remaining payment terms of the financial asset(s) d. The remaining time to maturity and the timing and extent of prepayments on the financial asset(s) e. The nature and volume of the entity s financial asset(s) f. The volume and severity of past due financial asset(s) and the volume and severity of adversely classified or rated financial asset(s) 1 3

4 g. The value of underlying collateral on financial assets in which the collateral-dependent practical expedient has not been utilized h. The entity s lending policies and procedures, including changes in lending strategies, underwriting standards, collection, writeoff, and recovery practices, as well as knowledge of the borrower s operations or the borrower s standing in the community i. The quality of the entity s credit review system j. The experience, ability, and depth of the entity s management, lending staff, and other relevant staff k. The environmental factors of a borrower and the areas in which the entity s credit is concentrated, such as: 1. Regulatory, legal, or technological environment to which the entity has exposure 2. Changes and expected changes in the general market condition of either the geographical area or the industry to which the entity has exposure 3. Changes and expected changes in international, national, regional, and local economic and business conditions and developments in which the entity operates, including the condition and expected condition of various market segments. p124, , emphasis ours INTRODUCTION An informal Sageworks study conducted in late 2016 on our proprietary database of Allowance implementations sampled a few hundred calculations. We found that for every dollar of calculated reserves for banks in our sample set, approximately ten cents were specific reserves, twenty cents were reserves based on some historical rate calculation, and seventy cents were allocated through qualitative policy. For credit unions, the qualitative and quantitative allocations were inverse. In designing the Practical CECL transition methodology, one of our primary goals was to create an approach to measurement under the CECL standard that would address this problem for all manner of financial institutions. Under current practices, credit unions have an easier lift not only do they typically have greater statistical power in their portfolios than many banking institutions, the increased risk of retail lending makes risk detection and measurement more straightforward. As a consequence, credit unions tend to use qualitative allocation for its intended purpose small adjustments to the base level of measured risk, whereas banking institutions must resort more heavily to their use to achieve a level of allocation that is satisfactory to management and external audiences. We find fundamentally unsatisfying the idea that an institution would invest the time and resources to transition to a forward-looking, lifetime-loss standard, only to face the same qualitative justification problems; this scenario is what we seek to avoid. Principle: Qualitative Adjustments should not comprise a majority of an institution s allocation in the ordinary course. 4

5 Readers of this chapter should: Believe that qualitative policy under the future standard can justifiably be different than qualitative policy under the current standard Be comfortable using the MIDAS framework to determine and defend appropriate policy Understand that future policy can and should inform current policy. QUALITATIVE VERSUS QUANTITATIVE We have described in the Loss Driver Analysis chapter an approach to implement management s judgment and tendency to reserve pessimistically or conservatively in a defensible, auditable manner by including economic indicators in the quantitative modeling effort. The factors used in that effort, and the leverage of available external information where appropriate, may make many aspects of current qualitative policy redundant. We advise our clients to then consider this exercise of qualitative policy tuning the final step in the policy-formation exercise; it should be performed only with a good understanding of how the quantitative model is already likely to behave, and should be informed by some of the regression work performed in the loss driver analysis activity. Principle: Address qualitative policy as the final step in a transition effort. Recall our base and peak load operating model from the Loss Driver Analysis discussion: 5

6 Our qualitative policy implementation should be proportionate to our level of rigor in the adjustments: If a factor in our qualitative policy has a significant and measurable impact on default or loss behavior, then it rightfully belongs in the quantitative component. As discussed in the forecasting chapter, however, it may not be possible to make a meaningful forecast for a factor, or its impact might be intuitive but not detectable numerically. Regardless, it stands to reason that for these remaining factors, their impact should be small relative to the rigorous, quantitative measurement. Principle: Qualitative adjustments should be Minor when our quantitative effort is rigorous. It is our belief that keeping qualitative adjustments limited in the context of the overall allocated reserve (and therefore less material to net income) will serve to reduce the level of scrutiny and justification required. AVOID DOUBLE-COUNTING Depending on the results of a loss driver analysis, many economic or environmental factors may be reflected in the quantitative allocation. It is conceptually important to ensure that factors included in qualitative policy not be redundant to already modeled macroeconomic indicators. Consider Rocket Bank, a rural institution that has aligned on a macroeconomic loss driver model driven primarily by regional unemployment. This institution has a significant exposure to certain outdoor and leisure activities; they are the lender to a notable regional ski park and its attendant hospitality providers, as well as a lender to many expedition outfitters. This institution included tourism and weather statistics in the loss driver analysis but could not find good forecast providers for these figures. 6

7 Regional economic outlooks begin to turn dire, and so Rocket Bank s indicated allocation rises. It would be conceptually difficult to defend additional allocation based on national economic outlooks, which are less specific and cover the same underlying theoretical concerns as the regional outlook. Making an adjustment for the national outlook would not be independent. If, instead, regional outlooks stayed generally rosy, but the national picture or the outlook in surrounding metro areas began to turn, Rocket Bank might justifiably conclude that their borrowers will suffer from an impact on tourism, and make an independent forward-looking adjustment. Similarly, if the snowfall is observed or projected to be wildly below seasonal expectations, Rocket Bank might come to a similar conclusion about its tourism and hospitality borrowers and make an additional, independent allocation. Conversely, if snowfall in a competing region is problematic, but excellent for Rocket Bank s borrowers, Rocket Bank might be justified in making a downward qualitative adjustment. Principle: Qualitative adjustments should be independent of factors that are already included in the modeling efforts. An example of double-counting we often see, and seek to curtail by the MIDAS framework, is reliance on the factor Volume and severity of past due and adversely classified or rated financial assets. A number of quantitative modeling approaches under the current standard, and certainly a risk-rated DCF model under the future standard, automatically create additional allocation for such credit deterioration. Further qualitative adjustment may not be appropriate in such cases. PRESENTATION MATTERS Conceptually, qualitative adjustment could be layered into the quantitative modeling; an institution could use a default rate from their measured conditions and a loss driver analysis, and by applying further management judgment determine that due to e.g. changes in management, additional (or lower) default rates should be expected, and thereby adjust their default curve. Legacy expectations from common ASC practices make this approach unwieldy. Users of financial statements will want these qualitative impacts disclosed and reportable separately, both for sensible assurance reasons and out of tradition. We therefore recommend that these minor, independent adjustments be recorded on top of the quantitative work; this less rigorous application of management s judgment should be separately disclosable. Note that if performing a DCF projection for internal consumption (e.g. to estimate fair value of a portfolio) we would recommend adjustments to the modeled rates be performed at the input level, and not the output level as described for the purposes of CECL measurement. AUDITABILITY AND SPECIFICITY Regulatory audiences will generally exert an upward pressure on management s allocation; we have discussed in our Loss Driver Analysis concept how management can create a base layer of allocation to satisfy this pressure. Our goal in qualitative policy should be conceptual soundness when subjected to independent audit scrutiny. We further this goal by electing policy adjustments that are minor and independent. 7

8 We can advance the Auditability of our framework by performing some checks on our adjustments: Directional Consistency Simply put, the sign of our adjustment should be intuitively consistent with our justification for that adjustment. In the Rocket Bank example, it would not stand to scrutiny to reserve less because of a soft outlook for tourism. Intuitive Leaps Would a reasonable person conclude the qualitative factor deserves allocation? Or must we construct a complex conceptual chain to arrive at the intended adjustment? Recall the core purpose of the qualitative adjustment is to account for conditions that did not obtain during the period considered for historical or current conditions framework. Numerical Consistency Adjustments should be consistently applied. If, in management s judgment, a certain market is softening (but not in a manner detectable through the quantitative analysis), and a small adjustment of 5 bp is applied, we would expect that other similar small adjustments would be of 5 bp. The range of adjustments should be specified in policy. Specificity By far the most important contributor to the justification of a qualitative adjustment is specificity. Qualitative adjustments should not be of indefinite life. For example, we commonly see large adjustments for the factor changes in management that have not been changed in several years. Principle: When providing narrative support for a qualitative adjustment, be specific about the conditions under which the adjustment would be changed or removed. This criterion of specificity will make the use of minor, independent adjustments much more palatable to an assurance audience. We observe increased market volatility and allocate an additional 10 bp is more concerning than We observe increased market volatility and allocate an additional 10 bp. We plan to remove this allocation when the 30-day moving average of VIX falls below is a far superior narrative formulation. There is an excellent name for cookie-jar accounting where management only has limited discretion to open and close the cookie jar: accounting. 8

9 USE OF FACTORS The guidance is clear that use of all of the enumerated factors is not expected and, in practice, we rarely find institutions with material differing conditions in more than a few of them. By setting qualitative allocation in a MIDAS framework, institutions will likely find themselves using fewer of these factors, or even factors not on the enumerated list. We would find it remarkable that an institution were experiencing new management, material policy changes, changing economic conditions, changing competitive pressures, changing staff experience, portfolio degradation, changing loan terms, etc. all at the same time in the ordinary course. By privileging our quantitative work and making minor, independent, disclosable, auditable, and specific adjustments, our financial statements can reflect that. 9