OSFI is on the Case:
Promoting Prudent Lending in Housing Finance
I am pleased to be here today to contribute to your discussions on housing markets and housing finance. The C.D. Howe Institute is to be applauded for organizing this timely event and giving you a forum to exchange ideas on these topics. They touch all Canadians, whether they own houses, aspire to own houses, or work in industries that profit from strong housing markets. Even those who do not have any direct interest in these issues understand that housing markets and housing finance matter a lot for our standard of living, which affects everyone.
Canada has not been immune from significant real estate corrections in the past and the damage they can inflict, as those of you who worked in the early 1980s and 1990s would know. We all have an interest in ensuring housing markets and the financial intermediation supporting them function smoothly.
Let me step back from your discussions to shed some light on the actions OSFI has taken over the past couple of years to promote a prudent housing finance system in Canada.
The lending environment is riskier now than it was in the past
Let’s begin with the environment in which mortgage lenders and insurers are operating.
Household indebtedness relative to income is likely to remain near record levels. While the rate of growth in household credit has slowed to about four per cent (4%) per year, growth in household income is likely to remain moderate in coming years.
The current high level of household indebtedness reflects demand for credit from three broad groups of society. The most important contributors to the growth in indebtedness have been: (i) people taking advantage of exceptionally low interest rates to leverage up and buy homes, investment properties for rental income, autos and other goods and services; and (ii) folks who are effectively punished by low interest rates, such as seniors who are borrowing against their home equity to support retirement incomes. But there is also a smaller third group whose share of household debt has been fairly stable in recent years – people whose situations are perhaps camouflaged to some extent by low interest rates. They are taking on debt to make ends meet in the wake of unfortunate life events, such as job losses or marriage breakdowns.
Now I would not presume to claim that borrowers are acting irrationally or do not know what they are doing. But, by same token, it is clear that the ability of the household sector as a whole to absorb major shocks is less now than it was a decade ago. Moreover, with interest rates near record low levels, there is not much scope for interest rates in Canada or the United States to fall further – something that helped people weather storms in the past. Governor Poloz recently noted in his testimony before the Senate that the Bank of Canada continues to expect a soft landing for the housing market and Canada’s household debt-to-income ratio to stabilize. But he also acknowledged that imbalances in the housing sector remain elevated and could pose a significant risk should economic conditions deteriorate.
One could argue that this is especially true given housing and related sectors (construction, durables, and related industries) constitute a significant share of Canada’s economy. Furthermore, recent research on the U.S. housing market collapse suggests that when the tide turns, a run-up in household debt can serve as a major drag on consumer spending if housing prices drop and trigger a decline in household net worth.
Next, whether you believe or not that housing prices are too high, you do not hear many observers arguing they are low. This matters if you are a lender or a mortgage insurer because there is a risk that the value of homes pledged as collateral may not hold up if economic conditions take a serious turn for the worse.
So from a prudential perspective, the environmental risks associated with lending to households are higher now than in the past. With interest rates expected to remain exceptionally low and household indebtedness high, these risks are likely to remain elevated for the foreseeable future.
Lenders have strong capital positions
Canadian banks have strong capital and liquidity positions. Their capital ratios are already well above Basel III requirements on an all-in basis. Moreover, those are very pure Basel III capital ratios as confirmed by the Basel Committee in its peer review of Canada that has just been published. In addition, the banking system already meets the new liquidity requirements that were published by OSFI in late May and take effect in 2015.
While there are limits on how much comfort one should take from stress test exercises given their arbitrary assumptions (they are just a model after all), the International Monetary Fund Financial Sector Assessment Program (FSAP) stress test results for Canada published earlier this year suggest that major banks carry enough capital to absorb a significant downturn in the housing sector and the economy more generally. That is to say, OSFI capital requirements and banks’ internal capital targets are such that it would take several years of continued severe losses and no mitigating action by bank management before any of the large banks would burn through the cushions they are carrying above international minimums. The results are thus comforting. But given the considerable uncertainty associated with stress test results, they are but one input into our decision-making. We cannot be complacent, encouraging outcomes notwithstanding.
So from a prudential perspective, lenders are well positioned to cope with fairly stringent stress scenarios like the one run for the FSAP. But one should not forget that in practice there is considerable uncertainty in stress test exercises due to modeling limitations and adverse behaviours in response to severe stress, which are particularly difficult to anticipate. Thus, one should not view stress test exercises as safe harbours. Judgment is required. Boards and senior management of financial institutions need to apply judgment in a forward-looking manner and not become too complacent in their capital planning exercises.
Vigilance required for mortgage lending and insurance governance, and risk management
You may wonder what more a prudential supervisor really needs to do if lenders and private mortgage insurers are well capitalized. But in stress situations, creditors and investors often lose confidence in these institutions before they run out of capital. Recall that some financial institutions lost access to funding markets in the midst of the global financial crisis even though they were reporting healthy regulatory capital ratios at the time. Sitting back and relying on capital is not enough for either financial institutions or prudential supervisors.
Instead, we believe it makes sense to work with mortgage lenders and insurers to reduce the likelihood of serious problems in the first place by promoting strong governance and risk management controls around mortgage lending and insurance underwriting activities. This is especially true given residential real estate lending represents more than 60 per cent of bank lending in Canada. External auditors also have a role to play by working with lenders and mortgage insurers to ensure that loan loss provisioning levels remain adequate for any emerging credit quality deterioration.
In an environment where indebtedness is high, the underlying real estate collateral is not cheap, and demand for new credit is tapering off, the value of strong governance, risk management controls and provisioning practices becomes even more apparent.
This is especially true now, given the relaxation of some lending parameters in the Canadian mortgage market over the past 20 years. For example: mortgages with amortization terms of 30 years became more readily available than they were in the past; and maximum debt service scores for GDS and TDS ratios are higher now than they were back then when 32/40 was more common. I think you would agree the mortgage market of today is rather different from the one in which our parents borrowed money. So while underwriting practices may be good today, past experience suggests that it could become very tempting in the current environment for mortgage lenders and insurers to ease up under the enchanting lull of the siren song of market share.
Siren songs can be bewitching. So OSFI is on the case. As Julie Dickson recently noted, we have:
- combed through mortgage portfolios to test the quality of the underwriting;
- issued guidance on mortgage underwriting (Guideline B-20) as well as draft guidance applicable to mortgage insurers (Guideline B-21);
- collected more loan data in the course of our supervisory work;
- required more disclosure of key mortgage lending data by institutions;
- assessed retail risk governance and risk appetite frameworks in retail banking; and
- reviewed the capital models backing real estate portfolios.
In the wake of the global financial crisis, many observers are suggesting that bank regulators need to think about their tool kit and employ macro‑prudential tools like changes in loan‑to‑value limits to lean against rising environmental risks. But at OSFI we believe it makes more sense to promote prudent lending all of the time. Hence, the 80 per cent loan‑to‑value limit on conventional mortgages enshrined in the federal legislation; and, where necessary, deep dives like the ones I just described in the current environment.
By the same token, let me note the focus in the B-20 and B-21 guidelines on governance and risk management principles. Such principles are meant to stand the test of time. They do not lend themselves to hard limits that one can vary in response to changing economic and financial conditions.
Frankly, OSFI generally prefers to take a principles-based approach in setting our regulatory and supervisory expectations. Hard limits like the 65 per cent LTV limit on Home Equity Lines of Credit (HELOCs) are more the exception than the rule. The key advantage of a principles-based approach is that it provides us the flexibility we need to tailor supervisory expectations to the situation at hand. This avoids safe harbours and compliance mentalities that breed complacency on the part of regulated entities, not to mention supervisors. Instead, principles help to underscore the point that regulated institutions are expected to use judgment and apply the guidelines to the situations they face on the ground within their own organizations.
Let me summarize a few key points before offering some final thoughts.
- The high level of household indebtedness is potentially a real problem. Banks and other deposit-taking institutions that are granting the loans, as well as the mortgage insurers who are insuring some of them, have introduced strong internal risk management practices and internal controls to ensure they have prudent underwriting procedures as well as controls to make sure staff are following those procedures.
- OSFI has been promoting the adoption of strong, prudent practices directly through our regulatory guidance, as well as by engaging with bank management and boards of directors. Similarly, we also have encouraged prudent risk management practices on the part of mortgage insurers.
- As a supplementary measure, OSFI has also encouraged banks and mortgage insurers to carry capital and liquidity commensurate with the risks that currently exist in housing finance. Exercises like the recent FSAP stress test give us comfort that they are well positioned to cope with fairly stringent stress scenarios.
At the end of the day, mortgage lenders and insurers must accept that they are responsible for the loans they are granting and insuring, and thus the risks they are running. After all, they are the ones that willingly granted the loans and insurance in the first place. Indeed, they are best placed to do so given their employees are on the front lines dealing with the customers and thus know them best. Collectively, those organizations have tens of thousands of staff involved in this process and in operating the multiple lines of defence within their organizations in the form of controls surrounding these activities.
No public sector authority, including a prudential supervisory agency like OSFI, can or should aspire to substitute for that expertise. For one thing, we do not employ thousands of people, nor should we. Instead, what we, at OSFI, can best do is work with boards and senior management to promote good governance and management of the risks by their institutions, conduct deep dives to test the quality of underwriting and closely monitor the evolution of loan portfolios, promote good disclosure of lending activities, and ensure the institutions carry capital commensurate with the risks they are running. This is very much in keeping with the statutory mandate OSFI has been granted by Parliament. A mandate that emphasizes that our activities must be carried out with regard to the fact that boards of directors (not OSFI) are ultimately responsible for the management of financial institutions.
So, while OSFI is on the case, lenders and mortgage insurers are ultimately responsible for their actions.
Thank you for having invited me to speak with you today.