What Are the Differences between SR 15-18 and SR 15-19?

What Are the Differences in Sections between SR 15-18 and SR 15-19?

In both SR 15-18 and SR 15-19, at the end of each section and appendix, the Fed attempts to summarize the differences. In some sections there is no difference or there are only minor ones.

The following sections provide a detailed comparison of each major section between SR 15-18 and SR 15-19. Use the tabs below to navigate between the different sections (A through F).

There is no difference in the expectations of boards of directors, and there are only minor differences in the expectations of senior managers. For example, senior managers at complex banks must review the capital planning process at least quarterly, whereas executives at SR 15-19 BHCs can do it semi-annually. Also, executives at the complex banks must show a higher level of engagement. We know of no objective way to measure “engagement,” and consider this to be a Category I differenceCodification or formalization of existing differences in treatment; CCAR expectations for the largest banks were already greater than for many of the regionals.–formalizing existing differences.

Risk Management has two subsections with a bit of overlap between them: (1) Risk Identification and Assessment Process and (2) Risk Measurement and Risk Materiality.

Section B.1. Risk Identification and Assessment Process

The biggest difference is that SR 15-18 BHCs must have a formal risk identification process and evaluate material risks quarterly, while that bullet/expectation is missing in SR 15-19. That seems to be an easing of expectations for the non-complex banks, which may have quarterly risk identification processes and committee meetings. That’s a Category II differenceActual relaxation or loosening of standards for the smaller CCAR banks.; however, we think the overall differences fit into Category IVGuidance (with a capital “G”) versus small “g” guidance: SR 15-19 eliminates certain explicit and specific requirements for the non-complex firms, but leaves a void. It provides no official “Guidance” of its own. For those cases, the surest and most economical way to reduce regulatory uncertainty and comply with SR 15-19 may be to follow the guidance for–the additional details provided to–large and complex banks in SR 15-18., which involves less written Guidance for the smaller banks, and, therefore, more uncertainty. Besides the missing bullet, four other sentences are missing from SR 15-19. These involve the identifying risks that are difficult to quantify, segmenting risks, seeking comments from stakeholders across the organization, and quarterly updates to risk assessments.² However, the main requirements are the same:

We’re not sure how a firm can show that material risks are represented without considering all risks, including those that are difficult to quantify, and without seeking advice from stakeholders across the organization. The problem with showing that you’ve captured your material risks is that you have to show that your other risks are immaterial–whether they are difficult to quantify, or not. So, while one could argue that SR 15-19 BHCs have fewer requirements, we would argue that the best way to satisfy those requirements is to still follow the guidance in SR 15-18. In summary, other than possibly reducing the frequency of processes and meetings, there’s no real change.

Section B.2. Risk Measurement and Risk Materiality

The main requirement–the first sentence below–is the same:

The second sentence holds for larger, complex firms, while the non-complex banks “…should employ risk measurement approaches that are appropriate for its size, complexity and risk profile.” Technically, it’s impossible to measure anything without using a quantitative approach.³ Moreover, as with the word, “engagement,” in the Governance section, the word, “appropriate,” here, is problematic. The question is: who gets to decide what’s appropriate? Hint: it’s not the firm.

The last paragraph of this subsection in 15-19 is missing in 15-18. It describes expectations for identifying and measuring risks that are inherent in the firm’s business practices as well as closely assessing assumptions about risk reduction from transfer or mitigation techniques. There’s a sentence about trading arrangements and another about asset values. (1) If your non-complex firm is reducing stress losses by applying contingent risk reduction methods, you are still going to have to defend those hypothetical actions. (2) In addition, we view this as a Category III differenceScrutiny shifted to other, non-CCAR regulatory reviews for items that are not required “in its capital planning process,” e.g., SR 12-7, which applies to every bank with more than $10B in assets.: if it hasn’t been emphasized before, anticipate defending your risk reduction techniques in other supervisory exams.

At the end of this section’s introduction, SR 15-18 includes the following statement, which is absent from SR 15-19: “…the control framework should include an evaluation of the firm’s process for integrating the separate components of the capital planning process at the enterprise-wide level.”

When summarizing the “differences” at the end of the section, the Fed emphasizes the integration aspect. However, when read in context of the preceding two identical paragraphs and three identical bullets points, we see no difference in requirements or expectations. The first sentence of the section and the main requirement for all CCAR banks is: “A firm should have a sound internal control framework that helps ensure that all aspects of the capital planning process are functioning as designed and result in sound assessments of the firm’s capital needs.” We’re not sure how to ensure that all aspects of the process are functioning correctly without considering their integration.

There are four subsections. We’ll review the first two, where differences exist:

1. Section C.1. Comprehensive Policies, Procedures and Documentation for Capital Planning
2. Section C.2. Model Validation and Independent Review of Estimation Approaches
3. Section C.3. Management Information Systems and Change Control Processes
4. Section C.4. Internal Audit Function

Section C.1. Comprehensive Policies, Procedures and Documentation for Capital Planning

The differences in C.1. aren’t really differences. SR 15-19 is missing a sentence and paragraph related to (i) evidence of adherence and (ii) documentation, respectively. However, in our and many others’ experiences, examiners seem to have the view that if it’s not documented, it wasn’t done. So, any executive at a large, non-complex bank who views the absence of those expectations as a sign that evidentiary and documentation requirements are light-to-nonexistent and acts accordingly, is taking a large institutional risk and, we would imagine, a large personal risk.

Section C.2. Model Validation and Independent Review of Estimation Approaches

In the first two paragraphs of C.2., SR 15-19 emphasizes the application to material models, whereas SR 15-18 is comprehensive and covers all CCAR models. That is a difference. However, remember that for banking supervisors, any categorization of immaterial requires evidence. You can’t just say something is immaterial. You have to prove it. In addition, it’s crucial that the same criteria (to determine materiality) are applied consistently across the objects of interest.

Executives at SR 15-19 banks should know the examiners will be watching for firms to redefine portfolios to fall beneath the firm’s materiality threshold(s). In that regard, consider the following well-used metaphor and proverb, respectively: (i) the straw that broke the camel’s back, and (ii) the Duck Test: if it looks like a duck, swims like a duck, and quacks like a duck, then it probably is a duck. In those regards, we expect that the regulators’ will determine materiality based on the closeness of both loss behavior and the macroeconomic correlations across portfolios. A “small” portfolio that behaves like other portfolios could be that metaphorical straw that breaks the bank, and, therefore, we would expect higher scrutiny of it. In addition, if the same macro variables are used to forecast losses in a small portfolio as in other portfolios, then that makes it harder to pass the Duck Test, so-to-speak. We view materiality as relative to the firm’s systemic risk, not in isolation or with respect to absolute size.

Both attachments include similar phrases: “… a firm should conduct a conceptual soundness review of all models prior to their use in capital planning,” with “material” replacing “all” in SR 15-19. SR 11-7 (OCC 2012-11) lists the core elements of an effective model validation framework as:

• Evaluation of conceptual soundness, including developmental evidence
• Ongoing monitoring, including process verification and benchmarking
• Outcomes analysis, including back-testing

Note that these three elements are neither mutually exclusive nor equal in the effort required to complete. In our experience, evaluating conceptual soundness involves at least 75% of the time and effort of a complete validation. In fact, much of the developmental evidence reviewed during as assessment of conceptual soundness is (a) outcomes analysis, including back-testing performance across different design choices, and (b) process verification and benchmarking of alternative specifications. Note that (a) provides much of the answer to “is it the right math?” and (b) helps answer, “is the math right?” Answering those two questions provides much of the answer of whether a model in conceptually sound or not.

Please note that we are not criticizing the Fed’s validation framework. Instead, we’re saying that verifying conceptual soundness, only, will not give time and cost savings equaling two-thirds of a full validation. Also, verifying conceptual soundness does not mean that full validation need never be performed, nor does it mean that any exceptions process–for using an unvalidated model–need not be followed. So, overall, while the Fed may be more patient for all CCAR banks, there is no indication that it will be more lenient when reviewing model risk management. For that reason, this is a Category III differenceScrutiny shifted to other, non-CCAR regulatory reviews for items that are not required “in its capital planning process,” e.g., SR 12-7, which applies to every bank with more than $10B in assets. for all firms. SR 15-19 references SR 11-7 several time and each time the standards are either reiterated or simply referenced. We see no easing of requirements for model risk management.

SR 15-19 excludes any mention of benchmark models, which seems to be an easing of expectations–a Category II differenceActual relaxation or loosening of standards for the smaller CCAR banks.. However, we consider this to be a Category IV differenceGuidance (with a capital “G”) versus small “g” guidance: SR 15-19 eliminates certain explicit and specific requirements for the non-complex firms, but leaves a void. It provides no official “Guidance” of its own. For those cases, the surest and most economical way to reduce regulatory uncertainty and comply with SR 15-19 may be to follow the guidance for–the additional details provided to–large and complex banks in SR 15-18., especially for non-complex firms that already have multiple models. Benchmark models help show conceptual soundness of the model of record and provide very useful justifications for overlays and other adjustments. Depending upon one’s interpretation, they are still expected for important models in SR 11-7. For 15-19 firms that have a benchmark, unless it is no longer sound, we would maintaining it as it be silly to throw one away.

SR 15-18 includes three bullet points that describe the elements to include in the capital policy. That is a difference, but it is a Category IV differenceGuidance (with a capital “G”) versus small “g” guidance: SR 15-19 eliminates certain explicit and specific requirements for the non-complex firms, but leaves a void. It provides no official “Guidance” of its own. For those cases, the surest and most economical way to reduce regulatory uncertainty and comply with SR 15-19 may be to follow the guidance for–the additional details provided to–large and complex banks in SR 15-18.. When a 15-19 firm writes or rewrites its capital policy, we suggest it uses those three bullet points as additional guidance about what to include.

There are several differences in this section, but we classify them as either Category IIIScrutiny shifted to other, non-CCAR regulatory reviews for items that are not required “in its capital planning process,” e.g., SR 12-7, which applies to every bank with more than $10B in assets. or Category IVGuidance (with a capital “G”) versus small “g” guidance: SR 15-19 eliminates certain explicit and specific requirements for the non-complex firms, but leaves a void. It provides no official “Guidance” of its own. For those cases, the surest and most economical way to reduce regulatory uncertainty and comply with SR 15-19 may be to follow the guidance for–the additional details provided to–large and complex banks in SR 15-18.. For example, SR 15-18 includes a sentence at the end of first paragraph that does not appear in SR 15-19:

In the box at the end of the section, that difference is stated:

We interpret the italicized phrases to be qualifiers: SR 15-19 banks do not need to assess multiple scenarios for capital adequacy, but that requirement is lifted only for capital planning. The sentence does not mean that 15-19 banks do not need to assess multiple scenarios for risk management to comply with the separate stress testing requirements in SR 12-7.⁴

The subsections on Scenario Design (E.1) and Scenario Narrative are both much shorter in SR 15-19 than in SR 15-18. However, these are Category IV differencesGuidance (with a capital “G”) versus small “g” guidance: SR 15-19 eliminates certain explicit and specific requirements for the non-complex firms, but leaves a void. It provides no official “Guidance” of its own. For those cases, the surest and most economical way to reduce regulatory uncertainty and comply with SR 15-19 may be to follow the guidance for–the additional details provided to–large and complex banks in SR 15-18.: there is less formal guidance for 15-19 firms, so following the additional SR 15-18 guidance is probably the easiest way to avoid unwanted findings. For example, in the Differences box, the Fed writes that it “expects these (Large and Complex) firms to articulate how risks not captured by scenario analysis are otherwise addressed in the capital planning process.” That expectation is not in SR 15-19. However, given the purpose and nature of the CCAR exercise, would any responsible executive at a 15-19 firm exclude such elaboration and analysis? We think not. Moreover, this relaxation seems to be contradicted by the entire first paragraph in Section F.

In SR 15-18, this section is about one page longer than in SR 15-19; however, except for some of the expectations related to operational risk, we don’t see real differences or changes. They are Category ICodification or formalization of existing differences in treatment; CCAR expectations for the largest banks were already greater than for many of the regionals. (codification of existing differences), and Category IVGuidance (with a capital “G”) versus small “g” guidance: SR 15-19 eliminates certain explicit and specific requirements for the non-complex firms, but leaves a void. It provides no official “Guidance” of its own. For those cases, the surest and most economical way to reduce regulatory uncertainty and comply with SR 15-19 may be to follow the guidance for–the additional details provided to–large and complex banks in SR 15-18. as 15-19 firms will likely follow 15-18 guidance to reduce regulatory risk.

For example, consider the first paragraph in F.1., Loss Estimation, in SR 15-19:

Now, how could or should a firm provide support for the assumed relationships between risk drivers and losses? Look at the first paragraph in SR 15-18:

Consider the first sentence in 15-19.

• How can a 15-19 bank provide support without using a sound method? Would sufficient support consist of using an unsound method?
• Would the sound method empirically demonstrate that a relationship exists economic variables and losses, or is actual evidence no longer needed?
• Many of the SR 15-19 banks are regional. Would it make sense to for those firms to use only national variables that the Fed provides for their own bank’s regional portfolios or exposures, especially if they view CCAR as more than a regulatory exercise? We think not.

Consider the second sentence in 15-19, is there a difference between estimating losses by type of business activity and losses based on the firm’s exposures and activities, including retail and wholesale credit, securities, market risk, operational risk losses, etc?

Reading the paragraphs immediately after the italicized ones in Sections F.1.a. shows only one minor difference related to credit losses on loans and securities; 15-18 firms should account for loss timing each quarter. Otherwise, the requirements are the same:

The first sentence says it all, which is why we interpret the italicized paragraphs of F.1., above, to have the same meaning.

SR 15-18 does include a paragraph on fair-value losses on loans and securities as well as a paragraph on market risk and trading, including counterparty exposures. If these are material for 15-19 firms, we would recommend they follow the 15-18 guidance.

Regarding operational-risk losses, both constituencies are to have sound processes, but only 15-18 firms need “a structured, transparent and repeatable framework to develop credible loss projections…” Really? And only 15-18 firms need approaches that are “well supported.” It would be great to hear an explanation of how does one know if losses are sensitive to different drivers unless one does the analysis? First, and most importantly, how does one know if losses are sensitive to different drivers unless one does the analysis? Second, how should the losses be estimated? Anyone?

As in the main sections, the word “material” appears in the SR 15-19 appendices, but tends to be absent in SR 15-18. Our caution remains the same: if a 15-19 bank wants to treat an item as immaterial, be prepared to document it, provide evidence of its immateriality, and ensure consistency in the threshold across the objects of interest, e.g., loan portfolios. Sometimes that’s harder to do than treating the item as material.

Continue to Appendices Differences

What Are the Differences in Appendices between SR 15-18 and SR 15-19?

As in the main sections, the word “material” appears in the SR 15-19 appendices, but tends to be absent in SR 15-18. Our caution remains the same: if a 15-19 bank wants to treat an item as immaterial, be prepared to document it, provide evidence of its immateriality, and ensure consistency in the threshold across the objects of interest, e.g., loan portfolios. Sometimes that’s harder to do than treating the item as material.

Use the tabs below to navigate between the different appendices (A through I) and their specific analysis.

Appendix A.1. Quantitative Approaches

In SR 15-19, this section is a single, medium-sized paragraph. In SR 15-18, it’s about five times as long. What’s the difference? The single paragraph in 15-19 seems to be a summary of the four paragraphs in 15-18. In other words, how would we comprehend the meaning of the 15-19 paragraph? We would read the details in 15-18. For example, 15-19 states:

First, and most importantly, how does one know if losses are sensitive to different drivers unless one does the analysis? Second, how should the losses be estimated? SR 15-18 states:

The same holds true for subsections 1.a., on the Use of Data, and 1.b., on the Use of Vendor Models. However, there is a Category III differenceScrutiny shifted to other, non-CCAR regulatory reviews for items that are not required “in its capital planning process,” e.g., SR 12-7, which applies to every bank with more than $10B in assets. with vendor models. The missing paragraph on vendor management in 15-19 will be covered during vendor management compliance exams, rather than as part of the CCAR review.

Appendix A.2 Assessing Model Performance & Appendix A.3. Qualitative Approaches

15-19 ignores any mention of benchmark models, and it also excludes the following paragraph from 15-18:

As we stated above, if an executive at a 15-19 bank interprets the absence of this paragraph as a sign that CCAR model validations need not be thorough, we strongly recommend that he or she reread SR 11-07 or OCC 2011-12 as many times as necessary. Likewise, if one believes that qualitative approaches are easier to defend than models, please read the first paragraph on page 12 of SR 11-07. Moreover, if the qualitative approach generates a numerical output, than we would anticipate that it would need to be on the firm’s CCAR model inventory and, therefore, would need to be validated. Qualitative approaches may initially seem cheap, but they can get very expensive, especially relative to available and defendable quantitative approaches.

SR 15-19 excludes one and two-thirds paragraphs that appear in SR 15-18 related to (1) avoiding the extensive reliance on overlays and (2) identifying factors necessitating the use of an overlay (and how the overlay compensates for those factors or weaknesses), respectively. First, extensive reliance on overlays can indicate that a model is not conceptually sound. So, while that may be permissible for 15-19 firms, that doesn’t mean it should be done. In addition, 15-19 does require support; it uses words like, “well-supported” and “appropriate.” We’re not sure how to show the appropriateness of an overlay without identifying the factors necessitating it or how the overlay compensates for those shortcomings. Therefore, we view these as Category IV differencesGuidance (with a capital “G”) versus small “g” guidance: SR 15-19 eliminates certain explicit and specific requirements for the non-complex firms, but leaves a void. It provides no official “Guidance” of its own. For those cases, the surest and most economical way to reduce regulatory uncertainty and comply with SR 15-19 may be to follow the guidance for–the additional details provided to–large and complex banks in SR 15-18.: how should a 15-19 show support and appropriateness? Understand and follow the more detailed 15-18 guidance.

The sections in Appendix B, (1) Process of Applying Overlays and (2) Governance of Overlays, follow the same pattern. SR 15-19 is shorter and less prescriptive, which makes it harder to interpret and increases uncertainty, if not anxiety. So, our recommendation remains the same: to understand what the terser 15-19 is requiring, read 15-18.

SR 15-18’s appendix is one page long. SR 15-19’s contains two sentences:

Given that benchmarking is mentioned 11 times in SR 11-7, the need for it as part of model validation has been interpreted somewhat differently than the need for them in SR 15-19. For example, we’ve heard regulators frequently mention that methods to assess performance would include building or reviewing benchmark models. Therefore, we, again, caution 15-19 executives to ignore SR 11-7 at their own peril.

Appendix D.1. Sensitivity Analysis

There is no real difference. SR 15-19 includes “material” three times, and SR 15-18 includes a few examples. This sentence says it all for both types of CCAR firms:

Appendix D.2. Assumptions Management

The following warning is not in SR 15-19. This is a Category IV differenceGuidance (with a capital “G”) versus small “g” guidance: SR 15-19 eliminates certain explicit and specific requirements for the non-complex firms, but leaves a void. It provides no official “Guidance” of its own. For those cases, the surest and most economical way to reduce regulatory uncertainty and comply with SR 15-19 may be to follow the guidance for–the additional details provided to–large and complex banks in SR 15-18. as we can’t imagine that examiners would allow 15-19 banks to behave otherwise.

It might seem like a large, Category II differenceActual relaxation or loosening of standards for the smaller CCAR banks., but we would be extremely wary of that interpretation.

Appendix E.1. Responsibilities of Audit Function

There is one relatively large difference: audit departments of 15-18 firms are expected to have internal, quantitative expertise, which we consider a new requirement for complex firms and not a relaxation of expectations for non-complex firms.

Six bullets of supervisory expectations that are in 15-18 are missing from 15-19. Again, we would consider these Category IV differencesGuidance (with a capital “G”) versus small “g” guidance: SR 15-19 eliminates certain explicit and specific requirements for the non-complex firms, but leaves a void. It provides no official “Guidance” of its own. For those cases, the surest and most economical way to reduce regulatory uncertainty and comply with SR 15-19 may be to follow the guidance for–the additional details provided to–large and complex banks in SR 15-18. and would urge 15-19 banks to consider the bullets as small “g” guidance to follow.

In this appendix’s other sections, related to E.2. Development of the Audit Plan and E.3. Briefings to Senior Management and the Board, we see analogous Category IIIScrutiny shifted to other, non-CCAR regulatory reviews for items that are not required “in its capital planning process,” e.g., SR 12-7, which applies to every bank with more than $10B in assets. and Category IV differencesGuidance (with a capital “G”) versus small “g” guidance: SR 15-19 eliminates certain explicit and specific requirements for the non-complex firms, but leaves a void. It provides no official “Guidance” of its own. For those cases, the surest and most economical way to reduce regulatory uncertainty and comply with SR 15-19 may be to follow the guidance for–the additional details provided to–large and complex banks in SR 15-18.. For example, the audit departments of 15-18 banks must report any material deficiencies, limitations, or weaknesses related to the firm’s capital planning process to the board. That expectation is missing for 15-19 firms; however, such issues would be part of a general internal audit board report.

There is no difference in the introduction or in Section 1, Post-Stress Capital Goals. In Section 2, Dividends and Stock Repurchases, 15-18 has an additional paragraph that we classify at Category IVGuidance (with a capital “G”) versus small “g” guidance: SR 15-19 eliminates certain explicit and specific requirements for the non-complex firms, but leaves a void. It provides no official “Guidance” of its own. For those cases, the surest and most economical way to reduce regulatory uncertainty and comply with SR 15-19 may be to follow the guidance for–the additional details provided to–large and complex banks in SR 15-18..

In Section 3, Contingency Plans for Capital Shortfalls, 15-19 firms need not link triggers in the capital plan to triggers in the firm’s recovery plan. That is a difference. Also, 15-18 has an additional paragraph related to detailed explanations of circumstances in which the firm would implement contingency plans as well as the ranking of possible actions across several dimensions. That is a real difference; however, if such a situation were to arise, we would imagine that regulators would require the same type of analysis before approving any capital action for any bank. Therefore, while less administratively less burdensome for 15-19 firms, planning in accordance to the paragraph makes sense to us, i.e., a Category IV differenceGuidance (with a capital “G”) versus small “g” guidance: SR 15-19 eliminates certain explicit and specific requirements for the non-complex firms, but leaves a void. It provides no official “Guidance” of its own. For those cases, the surest and most economical way to reduce regulatory uncertainty and comply with SR 15-19 may be to follow the guidance for–the additional details provided to–large and complex banks in SR 15-18..

SR 15-18’s appendix is a little over three times longer than is 15-19’s. However, and again, much of the extra guidance in 15-18 instructs how to perform the work required in both documents. In addition, what may not be required for capital planning purposes for 15-19 banks, e.g., multiple scenarios, is still required to satisfy SR 12-7, which we anticipate will be enforced more than it has been in the past. Again, in the Differences box, the Fed states, “A Large and Noncomplex Firm is not expected to use multiple scenarios in its capital planning process.” That qualifier, “in its capital planning process,” is very specific. SR 15-19 does not absolve firms from complying with SR 12-7’s principles.

SR 15-18’s appendix is almost five times longer than is 15-19’s. We would classify much of the difference as Category IVGuidance (with a capital “G”) versus small “g” guidance: SR 15-19 eliminates certain explicit and specific requirements for the non-complex firms, but leaves a void. It provides no official “Guidance” of its own. For those cases, the surest and most economical way to reduce regulatory uncertainty and comply with SR 15-19 may be to follow the guidance for–the additional details provided to–large and complex banks in SR 15-18.: 15-18 provides guidance or direction that 15-19 banks are likely to follow, too. However, 15-18 firms do have additional burdens related to providing information to assess the effect of potential changes to eight items; see page 39 of 42 for the details.

Again, SR 15-18 provides more details than SR 15-19, but for the most part, they are Category IV differencesGuidance (with a capital “G”) versus small “g” guidance: SR 15-19 eliminates certain explicit and specific requirements for the non-complex firms, but leaves a void. It provides no official “Guidance” of its own. For those cases, the surest and most economical way to reduce regulatory uncertainty and comply with SR 15-19 may be to follow the guidance for–the additional details provided to–large and complex banks in SR 15-18.. 15-19 banks are well-advised to read 15-18 when wondering how to interpret the shorter (and seemingly easier) requirements in SR 15-19.

See the next section that lists the 15-18 and 15-19 banks; return to the differences in the bodies of the guidance statements; or return to the background tab.

Continue to Bank Lists

SR 15-18 & SR 15-19 Bank Lists