Why I Support the Final Stress Test Rule
This is the most difficult and complicated final rule since I began my service on the NCUA Board. All of these covered credit unions should be proud of the fact that they made it through our most recent financial crisis. History suggests, however, that the next crisis is likely different from the last crisis, and the one before that.
Our job is to set these institutions up to succeed in the next crisis because the consequences, if they do not succeed, are systemic risk for the Share Insurance Fund and for the entire credit union movement.
No matter how remote the likelihood of failure us we have to remember that the cost of failure is borne not just by these institutions, but is borne by all insured credit unions through their Share Insurance Fund assessments. A failure by any one of them would have a material impact on the SIF, which is why we all have a stake in the outcome and need to work with them to make sure this never happens.
As someone who hails from the Pacific Northwest, which is subject to the risk of earthquakes, the analogy I make is that the credit union community has to be ready for, “the big one.” Just Thursday morning there was an earthquake that measured 5.7 on the Richter Scale off of Vancouver Island.
This doesn’t mean that “the big one” will happen, or is even necessarily likely, but rather that we have to be prepared for it.
We prepare for earthquakes by having building codes.
We prepare for financial disasters by having safety and soundness regulations.
Building codes differ based on the size and nature of a building, its life expectancy, and where it is located and the risk it faces.
The requirements for multi-family housing are typically more stringent than for single-family housing, the requirements for commercial and very large buildings are more stringent, and the requirements for buildings in precarious locations are even more stringent.
Similarly, when this board issued its liquidity regulation, it differentiated between credit unions of different sizes.
The rule we issue today is a tool designed to help ensure the survival of the largest credit unions during even the most stressful financial circumstances, even if the chances of such events occurring is remote.
Because it is a tool to ensure future success, and not a measure of current success or health, and because this process is new both for the NCUA and for covered credit unions, I have been persuaded by the arguments that during an initial period, stress test results should remain private.
My personal preference is for transparency and openness, but during this shake-down period disclosure might generate more heat than light.
Like the other financial regulators, once this system is up and running smoothly, the agency may want to release stress test results, as the Fed and the FDIC are now doing. This rule permits such flexibility.
But that is a decision which this rule leaves to a future board when the rule is reviewed every third year.
It has been suggested that this rule should be delayed, and in particular that it should wait until after a final risk-based capital rule is issued.
I would note that some of the people asking us to delay this rule until after a risk-based capital rule is issued also don’t want a RBC rule to be issued. That would mean that we would never issue a stress testing rule.
It has also been suggested that we issue guidance instead of a rule. While guidance will have to be issued after a rule is adopted, it is not a substitute for a rule.
To return to the building code analogy, building codes are usually mandatory, not simply a set of “best practices.” For these reasons, I support enactment of a final rule, and do not believe it makes sense to delay its issuance.
I want to emphasize, however, that we have listened to the comments and have made significant changes to the final rule to address many of the concerns that have been raised.
First, while we are proceeding with adoption of a rule today, we are providing more time, as requested, for covered credit unions to modify their capital plans if modifications are necessary. The final rule triples from 30 to 90 days the length of time a covered credit union has to modify its capital plan if modifications are necessary.
In addition, we have recognized that implementation the first year may take longer than in subsequent years, and we have built flexibility into the final rule to provide additional time if needed to get systems up and running.
Second, rather than just issuing supervisory guidance to implement this proposed rule, Director (Scott) Hunt and the Office of National Examinations and Supervision team will consult with covered credit unions before final supervisory guidance is issued under this rule.
Since this regulation and the guidance we issue will impact state chartered, federally insured credit unions, I am also recommending that the agency consult with and work cooperatively with state supervisors before issuing final guidance.
I also want to note that when the Fed, the FDIC, and the OCC issued their proposed stress test supervisory guidance last August they put that guidance out for formal comment for about 60 days. While this rule does not require a formal comment period, my hope and expectation is that the draft guidance will be circulated widely for comment and that the comments will be taken into consideration before final supervisory guidance is issued.
Both the flexibility in the first year timeline, and the consultation that will take place with covered entities on the supervisory guidance are designed to address concerns such as whether the NCUA and the third-party it retains to conduct tests will be ready, and whether the covered credit unions will understand their responsibilities and the relationship, between the stress test and a credit union’s exam, its CAMEL rating, and
its RBC requirement.
Stress testing, the CAMEL rating and supervisory examinations are done for different purposes and it is important to remember that a change in one does not necessarily require a change in the others. While a change in an institution’s risk profile, or its capital, could impact one or more of these measures, the change is due to the underlying change in risk or capital, not because an institution’s stress test results, its CAMEL rating, or its exam is directly tied to one of the other measures.
A third major change in the rule is that while the original proposed rule would have had the NCUA conduct stress tests indefinitely, the final rule allows credit unions to apply to conduct the tests themselves, under NCUA supervision, after three years.
This is a compromise between having the NCUA run the tests forever, which is the process the Fed follows for institutions over $50 billion that present the possibility of systemic risk to the economy, and the system of company-run tests the Fed and the FDIC use for institutions between $10 billion and $50 billion.
I originally thought it would be sufficient for the NCUA to follow the FDIC model of company-run tests. Given the history of what has occurred during the corporate crisis, and the first year of the FDIC tests, I am convinced it is prudent to start with NCUA-run tests so we establish public confidence in the process, and commonality in how the scenarios are applied to the covered credit unions.
Once both the NCUA and covered credit unions are familiar with the process and on the same page, and there is confidence in the validity of the tests, I believe we can, and should, transition to credit union-run stress tests, which this rule allows under NCUA supervision.
Another difference between this rule and the Fed’s rule for large banks is that the Fed requires two stress tests a year, and our rule, like the FDIC’s rule, only requires one stress test a year. That is another effort in this rule to reduce the burden on covered credit unions.
A major reason why I support an eventual transition to credit-union run stress tests is that while the agency may have to rely on a third-party to conduct the tests in the early years, I believe it is imperative that we develop the talent we need in-house to run and supervise stress tests. ONES is a new office, so it is appropriate to bring in additional expertise in the early years, but in the long run we need to develop expertise within our own staff to run and/or supervise stress tests.
As an agency, we need to start preparing today for the risks we will be facing tomorrow.
This is particularly necessary because while there are currently only four, soon to be five, credit unions over the $10 billion threshold, there are another 15 or so “waiting in the wings” who are over $5 billion and who at their current rates of growth may exceed $10 billion during the next decade.
In the long run, as the number of covered credit unions grows, it will be more cost-effective to invest in our internal capability to run and/or supervise stress tests, than to continue contracting with a third party.
A fourth major change in the rule is that we responded to criticisms that we hard-wired into the rule certain economic assumptions. The final rule removes those hard-wired assumptions and requires instead that reverse stress-test analysis be done. This change makes the rule more flexible, and some may say that it makes the rule too vague. We have to choose, however, between specifying assumptions in the rule, and providing greater flexibility so that the tests can appropriately recognize changing economic circumstances, differences among covered credit unions, and alternative models and assumptions.
A fifth major change is that we have added language to ensure that we will work cooperatively with state credit union supervisors on this issue. If, as I have suggested, we also work cooperatively with them on the proposed supervisory guidance, we can ensure that the dual chartering system is protected.
In summary, we have:
- Accommodated commenters requests that stress test results remain confidential in the early years with board flexibility to disclose results in future years;
- Provided a path that allows covered credit unions to run their own tests after three years;
- Provided additional flexibility on the timeline for implementation, particularly in the first year;
- Removed hard-wired assumptions which some commenters felt were too inflexible; and
- Strengthened language protecting the dual chartering system.
I also want to address the cost of the proposed rule.
I, too, am concerned about cost, but the cost of action must be balanced against the cost of inaction. The corporate crisis is evidence that the cost of inaction may be orders of magnitude greater than the cost of action. We need to view this challenge through the prism of an insurance provider, which is what the NCUSIF is. While the cost of the proposed rule is not insignificant, I would note that the covered credit unions have approximately 7.5 million members. At $250,000 per insured member, the estimated cost of the proposed rule is equal to the exposure the Share Insurance Fund has on the accounts of just 20 of those 7.5 million members. As an insurance premium, that is a very low cost compared to the covered risk. We must, however, work to keep that cost as low as possible, which is why I believe it is preferable in the long run for us to develop the expertise we need in-house, rather than continuing to contract this work out to a third party.
I would be remiss if I did not acknowledge the contributions of my two fellow board members, former Chairman Fryzel for leading the Agency during a very difficult period, and our current Chairman, Ms. Matz, for building a stronger regulatory framework that is capable of addressing the challenges of the future as well as the challenges of the past.
I also want to thank Chairman Matz for working with me and other stakeholders to address many of the concerns that were raised during the comment period.
Finally, I want to express my willingness to consider additional changes as we proceed with implementation if further changes are warranted. It is important to the covered credit unions, to the agency, and to the Share Insurance Fund that we all succeed in this endeavor.