Capital Planning and Supervisory Stress Testing (Dec., 2017)

December 29, 2017

Mr. Gerard Poliquin
Secretary to the Board
National Credit Union Administration
1775 Duke Street
Alexandria, VA 22314

Re: NASCUS Comments on Proposed Rule – Capital Planning and Supervisory Stress Testing

Dear Secretary Poliquin:

The National Association of State Credit Union Supervisors (“NASCUS”), the professional association of the state credit union regulatory agencies and the nation’s state credit union system, submits the following comments in response to the National Credit Union Administration’s (“NCUA”) proposed rule on Capital Planning and Supervisory Stress Testing.

NCUA’s proposed changes are intended to “reduce regulatory burdens by removing some of the more onerous capital planning and stress testing requirements” currently applicable to covered credit unions.Furthermore, NCUA intended to promulgate a rule that more closely aligns regulatory requirements with regulatory expectations.2 To achieve these goals, NCUA proposes establishing three tiers for covered credit unions based on asset size and tenure under the supervision of the Office of National Examinations and Supervision (“ONES”). Tier I will include covered credit unions during the first three years under the supervision of ONES and the first three capital planning cycles. Tier II credit unions will be those covered credit unions in their fourth years of ONES supervision, entering their fourth capital planning cycle, with assets between $10 billion and $20 billion.  Tier III will be covered credit unions that have assets greater than $20 billion.3

With respect to capital planning, Tier I and II covered credit unions would still be required to submit capital plans annually by May 31, but would no longer have their capital plans formally approved by NCUA. Tier III credit unions would also submit their capital plan on May 31 and need to obtain formal NCUA approval. With respect to supervisory stress testing, Tier I covered credit unions would be exempted. Tier II and Tier III covered credit unions would conduct their own supervisory stress testing but only Tier III would be subject to the 5% stress test net worth requirement.

NASCUS supports NCUA’s goals in this rulemaking and agrees that many of the proposed changes represent an improvement over the existing supervisory stress testing rules. However, as detailed below, NCUA should do more to harmonize capital planning and stress testing rules with NCUA’s current regulatory reform and relief agenda and better tailor stress testing rules to actual characteristics of the covered credit unions.

When considering NCUA’s capital planning and supervisory stress testing rule and the proposed changes, we note that the Dodd-Frank Wall Street Reform and Consumer Protection Act (“the Dodd-Frank Act”) excluded credit unions from the mandate to the Federal Reserve, the Federal Deposit Insurance Corporation (“FDIC”), and the Office of the Comptroller of the Currency (“OCC”) to issue stress testing rules (“Dodd-Frank Stress Tests” or “DFASTs”).4 Because NCUA’s rule is fully within the agency’s discretion, we urge NCUA to act boldly to right-size the rule and correct the deficiencies that have become apparent in the intervening years since initial DFAST rule promulgations.

NCUA Should Increase the Thresholds for the Proposed Tiers for Covered Credit Unions to an Asset Threshold Greater than $20 Billion

NCUA’s proposal to raise the asset threshold for the most comprehensive capital planning and stress testing requirements from $10 billion to $20 billion is too conservative. Currently, Congress is considering increasing the bank stress testing thresholds to $250 billion for the most ardent requirements.5 NCUA could certainly considering emulating the banking standards and raising the credit union stress testing threshold to $250 billion. However, should NCUA seek to scale credit union prudential standards to the size of the National Credit Union Share Insurance Fund (“NCUSIF”) and the credit union system, the threshold for enhanced prudential standards should still be greater than the proposed $20 billion threshold.

The FDIC deposit insurance fund has $ 90.5 billion in assets compared to $13.2 billion in assets in the NCUSIF.6 Were NCUA to scale the Tier II/Tier III threshold to comparable sizes of the deposit insurance funds, then the credit union threshold should be 1/6.85th the bank threshold or $36.5 billion for the higher prudential stress testing standards (at a minimum). It should also be noted that in terms of exposure of the respective deposit insurance funds, the credit union system has less insured shares and overall assets concentrated in its very largest institutions than the banking system.7NASCUS supports establishing $36.5 billion (or rounded up to $40 billion) as an initial asset threshold for proposed Tiers II and III.

However, we also agree with NCUA that capital planning and stress testing thresholds should consider a covered credit union’s specific characteristics beyond mere size, notably complexity and financial condition.8 Therefore, in addition to an initial $36.5 billion to $40 billion asset threshold, NCUA should provide for a waiver delaying stress testing until a covered credit union’s assets reach $50 billion for credit unions with non-complex balance sheets or net worth ratio in excess of 10%. Other characteristics that may recommend delaying stress testing requirements could include lack of concentration risk, or the existing ratios in NCUA Part 702.106.

Providing higher asset thresholds for the application of supervisory stress testing and a waiver process for substantially capitalized credit unions as well as those with non-complex balance sheets allows for a supervisory standard more closely tailored to characteristics that might pose a material risk to the NCUSIF. Furthermore, the higher thresholds provide a longer runway for Tier I credit unions to develop the necessary infrastructure and governance to develop meaningful and effective stress testing models and analysis and incorporate those results into their enterprise wide risk mitigation programs.

We also note the discrepancy between NCUA’s definition of a covered credit union and the definition of a covered credit union pursuant to FDIC’s rules. NCUA uses a single threshold of assets as of March 31 each calendar year.9 FDIC however uses a four quarter average of assets before a bank becomes a covered institution.10 NCUA should consider whether the four quarter method for determining a covered credit union might better accommodate any seasonal fluctuations in a credit union’s assets.

NCUA Should Consider Whether Supervisory Stress Tests Could be Conducted Periodically Rather than Annually

Earlier this year, legislation was introduced in the Senate to provide DFAST regulators the discretion to require stress tests less frequently than annually.11 We encourage NCUA to establish criteria that would, within the agency’s discretion, allow for Tier II credit unions to conduct stress tests every other, or every third year. Such a discretionary provision could provide significant regulatory relief in a manner tailored to a reduced risk profile of a specific qualifying Tier II covered credit union.

Frequency of Data Submissions Associated with Supervisory Stress Testing Should be Reduced

It is our understanding that currently, covered credit union are subject to significant quarterly data submission requirements to support NCUA supervisory stress testing. Not only are these data submission requests burdensome in and of themselves, it is our further understanding that they are subject to change from quarter to quarter, complicating the ability of covered credit unions to standardize their process for responding to the requests.12 NCUA should reconsider the breadth, and frequency, of the data submission requirements that accompany the supervisory stress testing process. We recommend that the data submission fields be tailored to the proposed Tiers, with more data required of the higher Tier covered credit unions. Furthermore, NASCUS recommends that NCUA match the frequency of the data submissions to the frequency of the stress tests.

Redundant Requirements Should Be Eliminated for Covered Credit Unions

The DFAST model upon which NCUA’s capital planning and stress testing rules are based are designed to examine how a covered credit union’s balance sheet would react to adverse economic conditions such as negative interest rates, steep market declines, or major losses in lending portfolios. The ensuing qualitative and quantitative analysis evaluates not just the resulting effect on the balance sheet (capital position) but the credit union’s understanding of its unique risk profile and its policies, procedures, and protocols for managing through the crisis scenario.

Given the enterprise-wide nature of the capital planning and supervisory stress testing regime, NCUA should consider whether certain generally applicable share insurance rules are unnecessarily redundant when applied to covered credit unions.

For example, NCUA states that the “overarching intent [of the risk-based capital rule] is to reduce the likelihood of a relatively small number of high-risk outliers exhausting their capital and causing systemic losses.”13 The rule attempts to achieve that supervisory goal by risk weighting a complex credit union’s balance sheet to match capital levels with risk exposure.14 However, where a risk-based capital rule attempts to match a balance sheet’s inherent risk with capital levels through arbitrary risk weightings, the capital planning and stress testing rule provides performance based metrics on the covered credit union’s actual balance sheet. In light of the data based metrics on capital planning and stress testing, NCUA should consider whether application of risk-based capital to covered credit unions provides any materially beneficial safeguard.

Likewise, NCUA’s general interest rate risk and liquidity rules for federally insured credit unions may be redundant to the more expansive requirements of capital planning and supervisory stress testing.15 NASCUS has supported NCUA’s emphasis on the importance of both interest rate risk and liquidity management.16 In addition, state regulators’ emphasis on the importance of interest rate risk and liquidity management is demonstrated by the fact that nearly twenty states have adopted the market sensitivity rating, or “S” component, to the CAMEL(S) rating system for use with their credit unions. However, as discussed above, the breadth and width of the NCUA’s capital planning and supervisory stress testing rules encompass all of the elements of those rules. Consistent with NCUA’s “principle based” approach to regulation and supervision, exempting covered credit unions from compliance with those rules streamlines the regulatory process.

Conclusion

NASCUS commends NCUA for re-evaluating the capital planning and supervisory stress testing rules and proposing changes to improve the supervisory utility and reducing unproductive regulatory burden. The changes as proposed by NCUA represent a good start to reforming these rules, and we urge NCUA to take this opportunity to make further improvements. There is a general consensus that the bank DFAST rules upon which NCUA modeled its capital planning and stress testing rules are flawed in their application to regional institutions with limited footprints and assets less than $50 billion.17 However, where the bank regulators are unfortunately constrained by the Dodd-Frank Act, NCUA faces no such constraints. Furthermore, as is clear from the numerous initiatives to raise thresholds on the bank side, NCUA can pursue our recommended improvements to the rule without sacrificing prudential safety and soundness.

Thank you for the opportunity to comment on NCUA’s proposed changes to the rules regarding supervisory stress tests. We would be happy to discuss our comments in more detail at your convenience.

Sincerely,

– signature redacted for electronic publication –

Brian Knight
Executive Vice President and General Counsel

1. Capital Planning and Supervisory Stress Testing, 82 Fed. Reg. 50095 (October 30, 2017).

2. Ibid.

3. 82 Fed. Reg. 50095 (October 30, 2017).

4. 12 U.S.C. 5365(a).

5. Economic Growth, Regulatory Relief, and Consumer Protection Act, S. 2155, § 401, 115th Congress (2017).

6. See FDIC Quarterly Banking Profile, Third Quarter 2017. Available at https://www.fdic.gov/bank/analytical/qbp/2017sep/qbpdep.html. See also NCUA NCUSIF Financial Statistics, September 30, 2017. Available at https://www.ncua.gov/About/Documents/Agenda%20Items/AG20171019Item1a.pdf.

7. Based upon review of quarter-end September 2017 call report data.

8. 82 Fed. Reg. 50095 (October 30, 2017).

9. 12 CFR §702.502.

10. 12 CFR §352.202

11. Main Street Regulatory Fairness Act, S. 1139, § 2, 115th Congress (2017).

12. There may also be some utility to subjecting the data submission fields themselves to stakeholder feedback.

13. Final Rule, Risk-Based Capital, 80 Fed. Reg. 66626 (October 29, 2015).

14. Proposed Rule, Prompt Corrective Action – Risk-Based Capital, 79 Fed. Reg. 11184 (February 27, 2014).

15. See 12 CFR part 741.12 Liquidity and contingency funding plans; and 12 CFR part 741.3 Criteria and Appendix B to Part 741 Guidance for an Interest Rate Risk Policy and an Effective Program.

16. In fact, in 2011, NASCUS’ comments expressed concern that NCUA’s proposed interest rate risk rule unintentionally de-emphasized the importance of interest rate risk management for modest sized credit unions. See NASCUS Comments – Proposed Interest Rate Risk Rule (May 23, 2011). Available at https://www.nascus.org/legacy-url/Regulatory/RegCommentLetters/5-23-11-NCUA-Proposed-Rule-Interest-Rate-Risk.pdf.

17. See testimony of Federal Reserve Board Governor Jerome Powell, November 21, 2017.