Stressing the Stress Test: The Importance of Strong Mortgage Underwriting, Remarks by Assistant Superintendent Carolyn Rogers to the Economic Club of Canada, Toronto, Ontario, February 5, 2019


Good afternoon. Let me begin by thanking the Economic Club of Canada for the invitation to speak, and a thank you to all of you who have opted to spend your lunch hour here today.

You came here to hear from OSFI on mortgage rules. I promise I will deliver on that. But I am going to get there in a bit of a roundabout way, so I’ll start by asking for your patience while I offer a bit of context and background that we think is important.

I want to talk a bit about the job of being a bank regulator, and how we approach that job at OSFI.

I also want to talk about the economic environment, through the eyes of a bank regulator.

And then I will talk about mortgage rules.

Guideline B-20

As I am sure everyone here knows, just over a year ago OSFI updated its guideline for sound mortgage underwriting – or B-20, for short. This guideline applies to all federally regulated mortgage lenders.

Since then there have been hundreds of news articles written about it, countless commentaries, speeches, interviews, and thousands of tweets, retweets, and other social media mentions.

The pace has slowed a bit recently, but mortgage rules remain a hot topic.

What you may not know, except for those of you that work at banks, is that if you visited OSFI’s website, you will find some 50 guidelines and advisories that apply to federally regulated mortgage lenders. You would also find guidelines applying to federally regulated insurance companies and to pension plans.

Yet the combined media coverage of all of those guidelines would not come even close to what B-20 has generated. And B-20 remains the only OSFI guideline with its own hashtag, as far as we know.

We welcome this attention.

All OSFI guidelines go through a public consultation. Every single one. This is an obligation we take seriously as a regulator whose role is to serve the public interest. When we develop a new rule or adjust existing rules, we always publish a draft and invite public feedback.

And we do receive feedback.

But that feedback is primarily from businesses or trade associations, or from academics, analysts or economists in the financial sector. And of course, from banks themselves. It is rare that we get broad-based public engagement and feedback on what we do at OSFI. The average Canadian is just not that interested in how banks are regulated.

But that is exactly who bank regulation is designed to serve – Canadians.

The Role of Regulation and OSFI’s Mandate

If you have a bank account – a chequing account where your pay is deposited, a savings account for your holiday, an RSP or GIC, bank regulation is there to protect it. If you rely on a bank to lend you money to buy a car, invest in your business…and yes, to purchase a home, bank regulation also supports you.

Banks are businesses and loans and mortgages are investments that banks make. Like other businesses, banks invest when they are healthy.

However, unlike other businesses, the risks that banks take are borne, at least in part, by the public. A portion of the money they use to fund mortgages and other investments is your deposit money and is guaranteed by the government, and in turn, the taxpayer, through the Canada Deposit Insurance Corporation.

And a significant portion of banks’ investments in mortgages is also guaranteed by government-backed mortgage insurance.

This brings us to why banks are regulated.

Banks are regulated because their safety and stability is fundamental to our economy. Public confidence - your confidence - in the fact that the money you have deposited in a bank will be there when you want or need it is critical to the functioning of banks, but also to the functioning of the economy. And the ability and willingness of Canadian banks to lend through all economic cycles, not just in boom times, contributes to a strong and stable economy.

In 2018, we marked the ten-year anniversary of the global financial crisis, which was a stark reminder of just how much damage excessive risk-taking by banks and a loss of public confidence in the banking system can do to the real economy.

The job of a bank regulator is to put in place a set of rules and a system of supervision that incents or requires a level of prudence by banks through all economic cycles and therefore contributes to public confidence, and to financial stability.

In Canada, for federally regulated financial institutions, that job belongs to OSFI. Our mandate is set out in legislation. It requires us to protect the interests of depositors and other creditors while allowing financial institutions to compete and take reasonable risks.

The job necessarily requires judgement and that judgement requires us to be attuned to the environment that financial institutions are operating in.

Banks are in the business of taking risks. And Canadian banks have a good track record of managing those risks.

It’s also true that Canada and its banking system have benefitted from a long period of benign economic conditions. We did not suffer nearly the impacts of the global financial crisis that other countries did.

So if you’re a bank, or a bank regulator, it’s hard not to be optimistic…perhaps even complacent.

We have not chosen that path at OSFI. Instead we try to learn from others and learn from the past, while at the same time remaining vigilant and proactive.

That brings me to B-20.

Strong Mortgage Underwriting

The first version of B-20 was issued in 2012. At that time, the role that weak mortgage underwriting had played in the global financial crisis was clear. While we did not see evidence of the extreme risk-taking that occurred elsewhere, OSFI chose to be proactive and to set out clear expectations for strong mortgage underwriting in Canadian banks.

Some years later, in 2016, we reminded banks of our expectations in the form of an public letter, in the face of what we observed as escalating competition, historically low interest rates and rapidly increasing property values leading to some deterioration in underwriting practices. 

That reminder had some impact, but, in our judgement, not enough. The prevailing sentiment that house prices would only go up and interest rates would only go down was, in our view, contributing to an over reliance on collateral value and not enough scrutiny of a borrower’s ability to repay a loan, particularly if conditions were to change.

Against a backdrop of record levels of consumer debt, this was a level of risk-taking that OSFI decided needed to be reigned in.

That brings us to the changes to B-20 that came into effect in January of last year; the most significant of those being the stress test.

The stress test requires a borrower be qualified for their mortgage with a buffer of affordability built in to ensure they can continue to pay their mortgage if conditions change. Those conditions could be a rise in interest rates that increase their payment obligation, or they could be a loss or reduction of income or an increase in other, non-mortgage expenses.

The stress test is, quite simply, a safety buffer that ensures a borrower doesn’t stretch their borrowing capacity to its maximum, leaving no room to absorb unforeseen events. This is simply prudent. It’s prudent for the bank and it’s prudent for the borrower too.

Criticisms of B-20

Let me return to the attention that B-20 received.

Some of the attention has been positive and we welcome that. It should come as no surprise that regulators don’t get a lot of fan mail, but we have received some positive feedback on B-20.

There has also been criticism. And we welcome that too.

House prices

The most common criticism is that we implemented a national policy to deal with a localized problem – that of extreme price escalation in the residential property market that only exists in Toronto and Vancouver.

This criticism is often accompanied by the suggestion that borrowers in other cities should not be subject to the stress test. Borrowers in Calgary for example, who may be facing tough economic conditions but not the same level of house price escalation, or borrowers in Winnipeg who are facing neither extreme price escalation nor difficult economic conditions.

This criticism assumes that B-20 was designed to target escalating home prices. Which it was not.

B-20 was designed to target mortgage underwriting standards. And sound underwritings look the same no matter what city or province you live in. Ensuring a borrower is not over-leveraged and can withstand a change in circumstances, including a change in interest rates, is sensible regardless of what city you’re in.

And when interest rates rise, they will go up in Calgary, and Winnipeg, at the same time and by the same amount that they will go up in Vancouver and Toronto.

Interest rates

Which brings me to another common criticism: the stress test was meant to offset low interest rates and now that interest rates have gone up, the stress test should be reduced or eliminated.

This is understandable. The simple design of the stress test - adjusting the interest rate upward for the purposes of qualifying a borrower’s capacity - might make it look like its simply there to front run a potential interest rate increase. But borrowers face other risks that can impact their ability to pay their mortgage that I mentioned earlier: changes to income or changes to expenses other than their mortgage. It’s prudent to have a buffer for these changes as well.

And interest rates, although they have adjusted upward, remain at historically low levels.  And personal debt levels remain historically high. A margin of safety in these conditions is prudent.

Should that margin of safety be monitored, and should changes be considered if conditions in the environment change? Of course they should.

The introduction of the stress test was, itself, an adjustment to B-20 in response to a shift in the environment – a shift in risks facing the financial sector. OSFI monitors the environment on a continual basis and when we determine that adjustments to our standards and guidelines are warranted, we make them.

We also work closely with our colleagues at the Bank of Canada and benefit from their research on the impact and interaction between our prudential standards and interest rates. This analysis has, and will continue to inform our guideline development process. 

Unregulated lenders

Next I want to respond to the concern that B-20 has contributed to increased competition from unregulated lenders. More specifically, that borrowers who may no longer qualify at federally regulated lenders will turn to unregulated lenders to avoid the stress test and may pay much higher interest rates or, worse, risk being taken advantage of or otherwise treated poorly. 

This is a legitimate concern. The risk that increasing standards for regulated institutions may shift more business to unregulated channels is always present and it’s not unique to mortgage standards or even to financial regulation. This is a balance that all regulators must grapple with. But it cannot be a reason not to act, or not to do our job.

There are sound options available for some borrowers in the private or unregulated lending space. But we understand the concerns raised by real estate agents, mortgage brokers and other financial advisors that some more vulnerable borrowers such as first time home buyers, risk making bad decisions or being taken advantage of by less scrupulous lenders.

The mortgage broker and the real estate industry are well placed to help manage this risk. If you see risks, if you think these options put your borrower in a vulnerable position, you can steer them away. That would be the right thing to do.

Reduced competition

Another concern we hear is that B-20 has contributed to reduced competition. This is usually attributed to the fact that borrowers who are renewing a mortgage and cannot pass the stress test can be exempted when they remain with their current lender.

The proposed solution is that borrowers in this position should be allowed to shop their mortgage to other lenders to secure a more competitive rate, and be exempted by those lenders as well.

We see two problems with this suggestion. First, we do not want a change in regulatory expectations on underwriting standards to displace borrowers who are meeting their obligations to their lender. Allowing an exemption to the stress test for mortgage renewals mitigates this risk.

On the other hand, we also do not want borrowers who do not meet the increased underwriting standard to become the focus of price competition among lenders. Limiting the exemption to renewals with existing lenders also mitigates this risk. 

We acknowledge that this exemption can be seen to reduce the incentive for a bank to offer its most competitive rate to a borrower renewing their mortgage who does not pass the stress test. To guard against this, OSFI put in place a tracking system to monitor this risk and we shared our early observations in our newsletter, The Pillar, in October.

To date we haven’t seen any evidence that banks are taking advantage of this situation to the detriment of borrowers.

We will continue to monitor this and will continue to report our observations publicly. If we see the need to take action, we will.  

Unintended consequences

Finally, there is a broad group of challenges that some commentators have attributed to B-20, often characterized as “unintended consequences”. These include: first time home buyers being locked out, having to defer their purchase or purchase less house than they planned. Rents increasing because there are fewer buyers and more renters. Borrowers relying on the bank of Mom and Dad who are, in turn, depleting their savings. Some have even gone as far to say that mortgage rules are exacerbating economic inequality.

The escalating cost of homeownership in Canada, and its knock-on effects to the economy and to society, is a problem. And it’s a problem that is proving very challenging to address.

But the answer to this important problem, cannot be more debt. Particularly, it cannot be more consumer debt, fuelled by lower underwriting standards.

Recent history has shown that relaxing bank underwriting standards can lead to extreme and persistent levels of financial instability that more than undo any economic gains they were intended to support.

What to expect from OSFI

So what changes, if any, can you expect from OSFI?

Let me start with what you should expect will not change.

We will remain focused on our mandate – protecting the interests of depositors and other creditors of financial institutions, while allowing them to compete and take reasonable risks.

We will continue to guard against the complacency that can creep in with long periods of benign economic conditions and financial stability.

We will continue to remain vigilant and proactive – we will regulate by looking through the windshield, not the rear view mirror.

And we will continue to welcome the public engagement on mortgage underwriting standards. We appreciate that mortgage rules have a broad and direct impact on many people’s lives and we understand that they have become part of a broader, important public dialogue on how to address the rising cost of owning a home in Canada.  

This brings me to what will change.

Increasing transparency is a goal we have set for ourselves at OSFI and we see an opportunity to support that goal when it comes to setting and maintaining sound mortgage underwriting standards.

By providing OSFI’s perspective today and by addressing some of the criticisms head-on, I hope we have taken an important step in that direction.

I appreciate your time and attention and I look forward to your questions.