EUROWEEK: It is generally recognised that Canadian banks had a good — or even a very good — crisis relative to those in the US and Europe.
Dickson, OSFI: The solidity of Canada’s macroeconomic fundamentals was one contributing factor. So, too, were capital ratios in the banking sector. We had moved to a 7% tier one environment and a total requirement of 10% when many other countries had kept Basel minimums of 4% and 8% in place. We had also been pretty strict in terms of requiring that tier one capital be comprised largely of common equity, which proved to be very important at the start of the crisis. We also had a leverage ratio in place, which we think played a role.
I also like to emphasise that we were early in implementing Basel II, which forced banks to develop their risk management and data systems and to increase their risk management budgets well before the crisis. That also played a role.
We also observed that the quality of supervision is important. You can have as many capital and leverage rules as you like, but it is also essential to focus on what supervisors are doing on a day-to-day basis. Supervisors typically comprise the bulk of human resources of any regulatory agency, and little attention is paid to what they’re doing. People tend to focus far more on what the rules are.
OSFI has had a mandate since 1995 that zeroes in on our solvency role and prudential roles. There is no conflict of objectives. We were able to hire the resources we needed, and brought in a lot of people from outside — not the usual practice in some countries. And we had the independence to make decisions that were sometimes unpopular.
As a regulator, opening yourself to international assessments such as Financial Sector Assessment Programmes (FSAP) is also important.
The Canadian banks also deserve credit for how they fared in the crisis. There are far fewer people employed at OSFI than there are in the banks’ risk management and internal audit groups. Strong control functions enhanced banks’ resiliency and preparedness for the crisis.
EUROWEEK: Does OSFI share the concerns of Fitch that “the main domestic threat to the stability of the Canadian banks is the record level of consumer indebtedness and the risk of overvaluation in the housing market”? Is there anything a regulator can (or should) do to help address these threats?
Dickson, OSFI: Fitch’s comment is consistent with the views of the Bank of Canada, which have been expressed in a number of its Financial Stability reports.
As a regulator, you definitely need to heed any concerns of this nature. At OSFI, this has led to a fair amount of action. We began looking closely at the Canadian banks’ real estate loan portfolios three or 3-1/2 years ago. We were looking for clear policies on who they would lend to, and on how any exceptions would be dealt with.
We followed up by publishing guidelines, now in effect for over a year, covering a gamut of areas banks should be adhering to. The guidelines really emphasised the message that we wanted extremely close adherence to policies and wanted boards of directors involved in discussions about those policies. We also included some bright lines on areas like home equity lines of credit, where we wanted to see loan to value ratios of 65% or less. We wanted to see a lower LTV than this on any kind of credit not classified as being of prime quality.
We’ve also been spending a lot of time looking at the models banks are using to determine risk weights, especially for their uninsured books. Mortgage loans above 80% LTV must be insured, so we focus most of our efforts on the uninsured book, which in terms of LTVs are fairly conservative. We’ve taken quite a lot of comfort from the risk weights coming out of those models.
EUROWEEK: Is OSFI concerned about heightened competition in the Canadian banking industry that might encourage the banks to take more risk overseas?
Dickson, OSFI: OSFI does not express views about banks’ strategies. We do express views on the risks that accompany their strategies. If a bank is looking at expanding its global footprint, or getting into a new business — even at home — our concern would be that the bank is proactively identifying and dealing with the risks that are created as a result.
EUROWEEK: You said in a recent speech that regulation can expand rather than limit opportunities. Can you expand on this?
Dickson, OSFI: Banks have better ratings and can enjoy cheaper funding when they are based in a jurisdiction where the effectiveness and robustness of the oversight system is widely recognised. Canadian banks report that they enjoy access to consistent and cheaper funding and access to new markets as a result.
EUROWEEK: Why is a 1% surcharge on D-SIBs considered necessary?
Dickson, OSFI: The idea of having a surcharge is definitely consistent with the Basel Committee’s position, which is that each jurisdiction should assess which banks within their sectors are systemically important and determine what the consequences of that should be.
When we analysed the Canadian system, we reached the view that we have six systemically important banks, and that the failure of one of those banks could have serious implications for the entire economy. That suggests that action needs to be taken to prevent failure, and in the Canadian context the 1% surcharge increases the probability that banks will remain solvent when the unexpected happens.
We’ve announced a number of other measures combined with the surcharge. We’re following the guidelines of the Enhanced Disclosure Taskforce set up by the Financial Stability Board, which is recommending that banks release a lot more information. We’re also continuing to implement more robust supervision of systemically important banks.
EUROWEEK: What are the prospects for the issuance of non-viability capital contingent capital (NVCC) in Canada, and how does this differ from European Cocos?
Dickson, OSFI: Everyone has a different definition of what a Coco is. The Basel Committee’s final agreement on tier one capital allows banks to include instruments that qualify as non-viability contingent capital (NVCC) as tier one. Our view — and the Basel Committee accepts this — is that if you’re going to qualify an instrument as capital, the trigger has to come very late in the process.
But institutions are free, as are supervisors around the world, to think of other instruments. We have chosen to adopt the NVCC route.
Our stance on non-viability triggers has been very clear for a long time, which is that if a bank reaches the point at which it is non-viable, it ought to be closed. This is all part of the aim of the statute that the supervisor be able to take early action to resolve problems, and that a bank failure does not mean the supervisor has failed.
Given that investors have been buying shares and subordinated debt of Canadian banks for a long time, and that legislation has always defined to some degree what non-viability is, we decided that it would make a lot of sense to use it as a trigger for this type of contingent capital instrument. People are familiar with it, and it is important that the market has confidence that the supervisor is not going to trigger it on a whim.
EUROWEEK: Where do you stand on the debate over the computation of RWAs?
Dickson, OSFI: We believe the computation of RWAs is extremely important. It’s an outcome of people not emphasising the importance of supervision enough.
As I said earlier, there has not been enough of a focus on what supervisors are doing. When banks use their own models for risk-weighted assets, supervisors have to be all over them. When people suggest we go to a simpler system and not use models at all, supervisors would need to be even more intrusive.
EUROWEEK: What are your views on stress testing?
Dickson, OSFI: OSFI takes stress testing very seriously and agrees it is extremely important. We’ve been very focused on it over the last few years, as has everyone else.
Where there is some difference of view is on whether it is appropriate to conduct stress tests every year with individual banks named and their results disclosed publicly. Our view is this is not appropriate, because it can potentially increase risk if people use the numbers inappropriately.
We have to remember that stress testing is a giant model and that a lot of assumptions have to be made. You’re trying to estimate the impact of what would happen in a tail event, and because tail events are few and far between, history may not be a reliable guide.
It is the second round impacts of stress testing that are really difficult to sort out. If you extend stress tests from two or three years to five, the results can look very different, because they will depend on the rates of change of key variables such as interest rates and inflation.
We’ve been emphasising that stress tests are a very good platform for supervisory discussions with a bank about the reasons for its results, especially if it is an outlier in terms of its performance. We think you can get a lot more out of those discussions than if you use them as a tool to base public dividend decisions on.
EUROWEEK: Was this influenced by the market’s response to stress testing in Europe?
Dickson, OSFI: Our views have certainly been informed by what we’ve seen around the world. We’re also asking if the time it takes to publish these results wouldn’t be better spent zeroing in on detailed discussions with banks. If the results aren’t published, you can have much more open discussions with banks.
Other supervisors in other jurisdictions have chosen different paths. The path each supervisor chooses has to be based on its own experience. We would prefer that banks issue information. They’re the ones that should be informing the market about how they would fare under various stress scenarios. That’s why we’re such big fans of more disclosure.