Skip to main content

The Mortgage Stress Test Changes

breaking news mortgage stress test benchmark rateThe Department of Finance today announced changes to Canada’s benchmark qualifying rate, a key component used in stress-testing insured mortgages.

The new benchmark rate will come into effect on April 6, 2020. It means insured borrowers (including those buying with less than a 20% down payment) will have to prove they can afford a monthly payment based on a rate that equals the weekly median 5-year fixed insured mortgage rate, plus 2%.

“As of February 18, 2020, based on the weekly median 5-year fixed insured mortgage rate from insured mortgage applications received by the Canada Mortgage and Housing Corporation, the new Benchmark Rate would be roughly 4.89%,” a Department of Finance official told us.

That’s 30 bps less than today’s minimum stress test rate of 5.19%.

The Bank of Canada will calculate this new benchmark weekly, based on actual rates from mortgage insurance applications, as underwritten by Canada’s three default insurers.

Why is the Stress Test Changing?

“This adjustment to the stress test will allow it to be more representative of the mortgage rates offered by lenders and more responsive to market conditions,” the DoF said in its announcement.

Critics say the old stress test was being artificially propped up by the Big 6 banks, which refused to lower their posted 5-year fixed rates in line with falling market rates. That meant the old stress test was not able to adapt to falling rates and act as an economic shock absorber (significantly falling rates generally imply declining economic prospects).

The news comes just weeks after Prime Minister Justin Trudeau formally asked Finance Minister Bill Morneau in December to review recommendations from financial agencies related to “making the borrower stress test more dynamic.”

Changes to the Uninsured Stress Test Coming Too

The Office of the Superintendent of Financial Institutions (OSFI) confirmed today that it too is considering the new benchmark rate for its minimum stress test rate on uninsured mortgages (mortgages with at least 20% equity).

“The proposed new benchmark for uninsured mortgages is based on rates from mortgage applications submitted by a wide variety of lenders, which makes it more representative of both the broader market and fluctuations in actual contract rates,” OSFI said in its release.

“In addition to introducing a more accurate floor, OSFI’s proposal maintains cohesion between the benchmarks used to qualify both uninsured and insured mortgages.” (Thank goodness, as the last thing the mortgage market needs is more complexity.)

OSFI has been dropping some not-so-subtle hints that changes to the benchmark rate would be forthcoming. In a recent speech, Assistant OSFI Superintendent Ben Gully said, “the posted rate is not playing the role that we intended.”

The problem (which has been growing since the first half of 2019) is that today’s benchmark rate (currently 5.19%) is too high relative to actual mortgage rates, making it unnecessarily difficult for homebuyers to qualify for a mortgage. Meanwhile, the actual rates available to borrowers have been diving.

OSFI is accepting input from stakeholders by email until March 17, 2020. But we’re guessing its mind is already mostly made up.

Outcomes to Expect

Will a looser stress test stoke the market? Absolutely.

While it might boost buying power by just 3% or less (depending on what the new benchmark turns out to be, come April 6), the psychological boost will be material. Homebuyersparticularly younger buyersare already worried about prices running away from them, given the double-digit gains of the last 12 months. News of an easier mortgage stress test won’t help.

As a result, the government might have been wiser to defer the new test a few months, so as not to fuel spring market activity. But time will tell.

Today’s news will be met with criticism by those who think easing the stress test will lead to overvaluation. But the government can’t effectively time home prices. Anyone who looks at government budget and GDP forecasts knows that. So they might as well correct the poorly designed stress test now, the industry would counter.

compare button


  • Ray McDonald says:

    The stress test was comically out of touch with reality under the thumb of the big-6 banks.

    Regulators used the wrong tool to begin with and have now corrected their mistake, albeit two years late.

    Arguably the 2% threshold is still very high, perhaps unrealistically so?

  • Jim says:

    I question why our rule makers couldn’t foresee the problem of tying the stress to posted rates in the first place.

  • Erratic1 says:

    The more lenders and brokers buy down insured rates, the more the stress test rate drops. Go buydowns!

  • The Spy says:

    Hi Ray,

    With the estimated long-term “neutral” rate just a point higher than today, and rising incomes, and the fact mortgages are amortized (balances fall) over the term, an argument could be made that 200 bps is an excessive increment above actual rates.

    On the other hand, rates have and can increase that much from their trough to peak during the business cycle. Moreover, there’s always the chance that inflation expectations become unanchored in the future.

    Either way, barring a change in government and considerable political pressure, regulators won’t budge on this point. They’re too worried about rising debt levels to reduce this buffer, and with good reason.

  • david says:

    Erratic1: I noticed that the new formula uses the median rate, not the mean. Not sure how easy it will be to affect the median rate since it will presumably be driven largely by the banks. the lowest rates going even lower doesn’t affect the median, especially if there are thousands?? of loan applications every week.

  • Ray McDonald says:

    Thanks Spy,

    I understand the buffer and that it’s for a good reason. Unfortunately, the side affects appear to be the creation of a two-tiered level of qualified buyers and very high rents.

    It seems there are now qualified buyers and the super-qualified, those that can out-income and out-bid other would be buyers.

    Some unsuccessful would-be buyers then choose to rent and are able to out-income and out-bid other would be renters.

    Perhaps the medicine is killing the patient ever so slowly, especially in the GTA and GVA.

    Is there a middle ground?

  • Jason says:

    It’s about time. Now OSFI just needs to make the same change for the uninsured test and we’ll have some semblance of common sense back in the process.

  • Ray McDonald says:

    P.S. Forgot to mention what appears to be another nasty side effect of the 2% stress test.

    Which is well-qualified buyers purchasing downstream and sending condo prices to the moon, currently running at +20% y/y.

    As evidenced by TRREB mid-month numbers as per John Pasalis

  • Vincent says:

    Hello there I’m looking to get a mortgage in the next two months but I’m unsure if a 2 year or 5 year fixed is the way to go any advice rate guy?

  • Tbahz says:

    The net difference being 30 bps is not enough to rally the market or new buyers to the market. The change in using a 2% spread over contract as a stress is reasonable and easy enough to calculate from lender to lender. If you want to really calculate costs of home ownership, you’ll use 100% of condo fees, 100% of the average national heating costs and 100% of the average national homeowners insurance all of which are under estimated in calculating TDS and GDS. Homeowner insurance is mandatory for any bank mortgage but it’s not factored into the ratios.

    The constant “rates are going to increase” rhetoric has been going on for 10+ years and it’s not happened. The economy is too fragile to larger increases that if they were 2% over five years would cripple the Economy and spur a ton of bankruptcies. Now with Canada’s economy being even more fragile with a weak Liberal government that has ruined international investment in Canada, Canadians are increasingly more at risk to losing jobs etc. Regardless, inflation will be impacted by decreasing amounts of discretionary incomes and higher home prices. The recent Coronavirus outbreak on the other side of the planet is an example of stifling economic factor that was not anticipated and it had an impact on rates. Further, the liberal government has done enough damage to impact job losses across the country which is another factor depressing rate increases.

    My point is, we may never see 6% discounted rates again. And if we do, it will take over ten years for us to get there and while that is happening people will adjust. But I just don’t see it.

Leave a Reply

Your email address will not be published.