There’s been a great deal of discussion recently over the mortgage stress tests and whether it makes sense to maintain the current policy parameters in this period of elevated interest rates.
In this two-part series, I’ll explore the premise of the mortgage stress tests and their current relevance in the context of today’s sharply higher interest rates and worsening economic conditions.
A recap: What is the mortgage stress test?
Most new residential mortgages in Canada are subject to “stress testing.” There are two stress tests at present, firstly, for insured mortgages (starting in late 2016) and secondly, for mortgages that are uninsured but are issued by federally-regulated financial institutions (since the start of 2018). Regulations are broadly similar for the two sets of tests.
The objective of the tests is to reduce risks within the financial system: the tests assess whether the borrower would be able to afford the payments not just at present, but in the future. Since interest rates might be higher in the future, calculations are made at a “qualifying rate,” which is the greater of:
- An interest rate that is specified by regulations (the rate is currently 5.25%, but it will be re-assessed at year-end, and it seems quite likely that it will be raised), or
- The contracted interest rate plus a “buffer,” which is currently 2 percentage points. With actual “special offer” rates from major lenders currently in the vicinity of 5%, stress testing will now often use qualifying rates in the vicinity of 7%.
Impact of the tests on current risks
With mortgage interest rates now at the highest levels in over a decade, this is a good time to consider whether the stress tests are reducing risk related to mortgage borrowing, and also to discuss whether there should be changes to the qualifying interest rates that are used in the tests.
Since the stress test for insured mortgages started late in 2016, some of those are now being renewed, but that represents a minority of mortgages. According to Bank of Canada data, 29% of funds advanced for new mortgages were insured in 2017.
The largest share of new mortgages is non-insured, federally-regulated mortgages. Since the stress test for them started at the beginning of 2018, and since the largest group within these mortgages is fixed-rate with 5-year terms, renewals of stress-tested mortgages won’t start in earnest until 2023 and beyond.
Analyzing the first round of renewals for stress-tested mortgages
Here are some calculations that I think will be typical for mortgages renewed during 2023. The initial mortgage rates (during 2018) will have been in the area of 3.25% (or higher).
During the initial 5-year term, and assuming a 25-year amortization, 14% of the principal will have been repaid (or more, if any pre-payments were made). Assuming a 5% interest rate at renewal (and applied to the remaining principal and the remaining amortization period), the monthly payment will increase by 16%.
What has happened to incomes will be an important factor for those future renewals. In most situations for people renewing mortgages, incomes will have increased. At present, growth in average weekly wages versus five years ago is slightly over 20%, so a lot of renewing mortgage borrowers will have seen income growth larger than their payment increase.
On the other hand, it is possible that for some borrowers, incomes will have been substantially reduced as a result of job losses that are induced by high interest rates.
Recent buyers are more vulnerable
There is a very different scenario for people who have bought recently and took variable rate mortgages at the very low rates that were available from mid-2020 until early this year: they are much more vulnerable, considering typical variable rates have gone from a low of about 1.3-1.4% a year ago to the current 5%.
Since the start of 2020, there have been about 1.5 million home sales in Canada. The data on mortgage choices for those purchases is murky: the Bank of Canada has data on “funds advanced,” but this includes renewals and equity take-out via HELOCs, in addition to purchases. I have noticed in the past that people renewing sometimes make different choices than are made by purchasers.
So, we don’t know for sure what mortgage choices have been made by recent buyers. That said, according to the BoC data, since the start of 2020 about 37% of funds advanced (for purchases, HELOCS and renewals) were in variable-rate mortgages and 63% were fixed rate.
This data implies that there might have been about 500,000 to 600,000 mortgage holders (out of about 10 million total owners and 6 million total mortgage holders) who bought during the Covid period with a variable rate mortgage. Some of them may have switched to fixed rates.
But many of those borrowers may now be highly vulnerable. In addition, a smaller number of people who bought in the Covid period with short-term (1- or 2-year) fixed rates, will be renewing this year and next at sharply higher payments.
The remainder of recent buyers have fixed rates that are mostly not due for renewal for a while – therefore, they are not immediately vulnerable. And when they do renew, they will have significantly higher incomes and lower remaining principals.
In this situation, I can repeat something I’ve said many times: the greatest risk to mortgage borrowers is not an unaffordable rise in payments, it is a loss of ability to pay—and the stress tests did nothing to control for this, let alone reduce this risk.
That said (as was discussed above) there is now a minority of mortgage holders who are currently extremely vulnerable, because they have bought recently with variable rates or short-term fixed rates, and they have not had enough time for income growth or principal repayment.
Stay tuned for Part 2, where I will explore some issues that have been created by the policy designs, and implications for stress testing in the future.