The importance of sound residential mortgage underwriting

Remarks by OSFI Assistant Superintendent Tolga Yalkin, C.D. Howe Institute, Toronto, Ontario, April 13, 2023

Please check against delivery

Before I begin, I would like to acknowledge that we find ourselves today on the traditional and unceded territory of the Mississaugas of the Credit, the Anishinabeg, the Chippewa, the Wendat and the Haudenosaunee.

For me, recognizing the unique link of indigenous peoples to the land that we occupy shows awareness of the benefits we’ve received from being where we are but also our commitment to reconciliation.

Now, to the theme of my speech today, the measures OSFI has taken—and is taking—to ensure sound residential mortgage underwriting and credit quality in times of:

• high inflation;
• high interest rates; and,
• high household indebtedness.

I’m going to talk a today a bit about:

• the risks we’ve been seeing as an organization;
• what we’ve been doing about them over the years; and,
• why we’re continuing to explore strengthening our approaches.

In our Annual Risk Outlook, otherwise known as the ARO, we flagged risks associated with unprecedented house price increases, which have outstripped household savings and income growth, and ultimately resulted in more highly indebted borrowers—a concerning trend that is even more so given the rapid shift we have seen from historically low interest rates to the highest levels since late 2008.

Over the past year, we, at OSFI, have continued to track various indicators of mortgage risk for signs of borrower vulnerability. We’ve been seeing that the sharp change in the cost of borrowing is posing short and long-term risk to mortgage holders and lenders.

• The short-term risk is related primarily to two types of existing mortgage holders: those with variable rate mortgages—both adjustable and fixed—and those with fixed rates who are close to renewal.
• The longer-term risk is related to the build-up of household indebtedness.

We’ve done several things to address these risks over the last few years:

• In 2018—just over 5 years ago, we implemented the Minimum Qualifying Rate (MQR). Commonly called the “stress test,” it requires new uninsured mortgages to be qualified at a higher rate than that of the contract. In 2021, we adjusted that qualifying level to be the contract rate plus 2 percent or 5.25 percent—whichever is higher.
• As many of you know, the MQR tests whether a borrower could manage higher mortgage payments in the event of adverse financial shocks. This ensures that borrowers have a buffer that will add to their financial resilience in the face of unanticipated events, such as:
  • a rise in interest rates—which we have seen as of late;
  • a loss of income—potentially precipitated by a slowdown in the economy; or,
  • other unexpected expenses.
• In addition to adjusting the qualifying level in 2021, we also worked with lenders to reinforce our mortgage underwriting guidance. Among other expectations, we underscored the need for a more prudent approach to collateral valuations for loans in markets that have experienced rapid property price increases.
• And, in early 2022, we updated our capital adequacy guidelines to fine-tune capital requirements for household credit and align risk weights for exposures to private mortgage insurers.
• Also in 2022, we clarified our expectations around Combined Loan Plans (CLPs) to address concerns about the re-advanceability of credit above the 65 percent loan-to-value (LTV) limit given impacts on prolonged debt and associated risks to lenders and households.
• Finally, over the past year, we’ve been assessing the impact of rising interest rates on borrowers. We’ve been working with lenders who offer different types of variable rate mortgages to understand their plans for identifying and managing potentially vulnerable borrowers who might be stressed by rising mortgage payments.

One strategy—especially attractive to those with variable-rates—is to temporarily increase the mortgage amortization period. While this is one way to cope with higher interest rates in the short term, it’s not without risk, given extended amortization periods keep borrowers in debt longer and lead to higher interest payments.

Our conversations with financial institutions have emphasized proactively managing accounts, acting before borrower stress become unmanageable. We have been clear that we appreciate lenders working with clients to help them stay in their homes while ensuring actions taken remain within the institution’s risk appetite, while recognizing this can sometimes be a challenging path for lenders to navigate.

All this said, at the same time, it won’t surprise you to hear that we are not wearing rose-coloured glasses: The growth in highly-leveraged borrowers increases the risk of weaker credit performance. While lending institutions are well-capitalized and financially resilient, the higher cost of borrowing—and any potential economic downturn—could lead to more borrower defaults, potentially a disorderly market reaction, and even broader economic uncertainty and volatility.

Guideline B-20 on residential mortgage underwriting is part of the way in which we help manage these risks. Sound mortgage underwriting is an important contributor to the stability of Canada’s financial system. And, it’s through this guideline that we ensure federally regulated lenders are managing risks associated with mortgage lending in a manner that encourages sound credit quality.

First introduced in 2012, this piece of guidance is now more than ten years old. And, much has changed in our risk environment since then, including the heightened risks arising from high household indebtedness to which I referred. That’s why we committed in our ARO update in the fall to launching a holistic review.

To do our job right, we must be proactive not reactive and balanced in our approach, ensuring the safety and soundness of federally regulated entities while allowing institutions to compete effectively and take reasonable risks.

And, as many of you will know, in helping inform how we go about doing this, we launched an initial consultation on B-20 in January of this year, focusing on debt serviceability measures, which, incidentally, closes tomorrow.

Now, some have questioned the need for this consultation, arguing that our underwriting standards are even too restrictive given low delinquency rates.

I’d note, however, that arrears are a lagging indicator of risk. We take the view that we wouldn’t be doing our job as the prudential regulator if we assume past credit performance will be future. And, low delinquency rates can quickly turn, as we saw in the US through the 2008 global financial crisis.

We, accordingly, asked, as part of this consultation, for feedback on several debt serviceability proposals, including:

• Loan-to-income (LTI) and debt-to-income (DTI) restrictions, which would limit mortgage debt or total indebtedness as a multiple, or percentage, of borrower income;
• Debt-service coverage restrictions—such as gross debt service ratio (GDS) and total debt service ratio (TDS)—that limit ongoing debt-service obligations as a percentage of borrower income; and,
• Interest rate affordability stress tests—like the MQR—that assess the minimum interest rate applied in debt-service coverage calculations.

These proposals could be introduced at the loan level and applied to each borrower’s mortgage application, like the existing MQR. Alternatively, they could be implemented at the lender level, applied to the lender’s aggregate volume or portfolio of mortgage business.

In our paper, we indicate an inclination towards lender level approaches, which could allow lenders the discretion to underwrite a certain number of loans that exceed any new thresholds.

We welcome stakeholder views on these proposals and how they might be implemented, if adopted. And we look forward to reviewing the submissions we have received.

In addition to consulting with stakeholders, we’re considering the mortgage underwriting rules of our international peers. Several other jurisdictions have moved forward with measures like those we’re looking at, so we have much to learn from their experiences.

We plan to issue a summary report on what we heard in the first phase of the consultation as quickly as possible. Feedback from this initial phase will inform proposed next steps, including possible revisions to B-20, which would be issued—in draft form—for an additional consultation. It’s also possible that we may publish interim guidance in the meantime based, of course, on this initial consultation.

This leads me to an important point: We, at OSFI, understand that the decisions we make around B-20 affect a large number of Canadians. It’s why we have committed to transparency in this work, and a key reason that before coming out with a draft guideline with a decided way forward, we initiated a consultation to solicit views on the range of options that may be considered. We believe this form of early consultation will help us better land in an optimal place where Canadians have access to mortgages and we have a resilient lending industry.

Time to wrap up.

Sound mortgage underwriting and credit quality supports a well-functioning mortgage market and is critical to the financial wellbeing of Canadians.

We know many of you along with our other stakeholders have important input and ideas that can help us navigate these issues, and we look forward to carefully studying the submissions we receive to help us chart a way through the current economic environment to support, overall, the ongoing stability of our financial system.

Thank you.