Debating the Mortgage Stress Test: To Loosen or Not To Loosen?
By: Diana Petramala
July 29, 2019
There has been considerable debate around the Canadian mortgage stress test (implemented by OSFI1 on January 1, 2018). Since these new rules have come into effect, the Greater Toronto Area (GTA) housing market experienced the biggest and most sustained drop in resale home activity in over a decade (see Figure 1). These weakened market conditions have led many to believe that the need for these guidelines have passed, resulting in calls for the government to loosen mortgage stress test rules.
Top policy watchdogs (such as the IMF and CMHC) are on the other side of the debate, warning that risks associated with hot housing markets are far too great and that the Federal Government agency should hold firm on the mortgage stress test rules.
CUR would like to weigh in with the top three reasons why we feel maintaining the mortgage stress test, in its current form, is a prudent policy stance at this time. The primary purpose of the mortgage stress test was to ensure households and financial institutions were not taking on too much financial risk in an over-heated market. An unintended, but welcomed, outcome was a cooling in the housing market. However, there is still the risk the market could be tipped back into hot conditions despite the stress test, a risk that could rise if the government were to loosen demand-side policies. Regulators should avoid encouraging policy actions that increase demand until limitations on housing supply are loosened.
Reason #1: The housing market is teetering on the edge of a re-acceleration
The housing market in the GTA has cooled, but it is not that cool.
Six quarters out from the implementation of the mortgage stress test, existing home sales in the GTA are down 19% from the levels reached in 2018. Even with this drop, home sales have just fallen back to levels that are more normal, following a long run of frothy market conditions (see Figure 1). Indeed, home sales are now hovering around their 20-year long-run average.
Meanwhile, houses are still really expensive, and require that the average homebuyer take on a heavy debt load for purchase. The average MLS home price in the GTA took a tumble with the introduction of the Ontario Foreign Homebuyers Tax in 2017, but barely budged following the implementation of the mortgage stress test. The MLS average home price in the GTA grew by 2.0% year-to-date between December 2018 and June 2019, meaning there has been very little relief to home seekers (see Figure 2).
In fact, breaking down home price performance by housing type shows there are still considerable price pressures in some segments of the market (see Figure 3). While the prices for single-family homes were up just 0.8% year-over-year in 2019, resale prices for condos were still rising at an almost 7% annual pace.
Meanwhile, pent-up demand for housing is likely bubbling under the surface. Millennials are quickly aging into their 30s (prime home-buying years), and population growth is booming across the region due to a pick-up in immigration between 2016 and 2018. Statistics Canada estimates that population growth in the GTA accelerated between 2016 and 2018, growing by an annual average of 110,000 people per year, compared to an average annual of 75,000 per year in the prior five years.
Lower mortgage interest rates further raise the risk of a re-acceleration in the GTA market. The mortgage stress test cannot be blamed for all of the decline in housing activity since early 2018. Higher mortgage interest rates likely also played a role. The Bank of Canada overnight rate rose 125 basis points between July 2017 and October 2018, resulting in a 70 to 110 basis point increase in mortgage rates over that period. Buying a house in the GTA was the most expensive it had been in over a quarter century in 2018, even when excluding the impact of the mortgage stress test (see Figure 4).
However, 5-year mortgage rates in Canada have since come back down, raising the risk of a re-acceleration in housing demand (see Figure 5). Even the posted rate that sets the bar for the income-check under the stress test came down.
Yet, not much has changed on the supply side, with active listings still below their long-run average as of June of this year (see Figure 6).
The recent stabilization in home prices may provide Millennial home seekers the opportunity for their incomes to catch up to the expensive level of home prices. With robust labour markets, household incomes are finally growing more than home prices. Meanwhile, income growth should help offset the impact of the stress test by 2020.
However, the mismatch between supply and demand in the GTA housing market raises the risk of a re-acceleration in home prices in the event of any policy measure that could stoke demand (such as a loosening of the mortgage stress test rules), but not supply. Should home prices rise by another 10 percent, the actual required monthly mortgage payment on a 25-year conventional mortgage on an average priced home would go up by $400 assuming interest rates stay at their current level. This is in addition to the extra $15,000 a homebuyer will have to save for a down payment.
Reason #2: Market reaction to mortgage regulation changes have tended to be temporary
The implementation of the mortgage stress test marks the seventh attempt by Federal Government agencies in the last decade to safeguard housing and financial markets by changing mortgage qualification rules (see Figure 1).
The trends shown in Figure 1 suggests that the impact of each regulation was temporary. New regulations would cool the market for two to three quarters, but activity would re-accelerate shortly thereafter, as borrowers and lenders figure out how to skirt around them.
It should come as no surprise though, that the impact of the 2018 mortgage stress test was more sustained than that of past regulatory changes, given it had more reach. Past regulation changes only targeted insured mortgages. And, many of these regulations could be circumvented by switching from an insured mortgage to an uninsured one, provided the buyer could accumulate enough of a down payment.
The 2018 mortgage stress test applies to all mortgages issued by federally regulated institutions, which the Bank of Canada estimates accounts for 75% of all mortgage lending.
Insured mortgages accounted for almost 50% of all new mortgages when the Federal Government started tightening mortgage qualification rules in 2008, though that share fell to 25% by 2018.
While borrowers have very few options to escape these new guidelines, alternatives do exist. The share of mortgage lending at non-regulated mortgage lenders is rising according to CMHC, although it remains small. The housing market will begin to shake off the new rules as borrowers find ways to avoid or to adjust to the stress test. At the same time, the risks of a return to frothy conditions in the market remain high.
Reason #3: Don’t let low delinquency rates fool you!
The biggest argument for reversing course on the mortgage stress test is that mortgage quality has gotten a lot better since their implementation.
Policy makers should be wary of this argument, however, as delinquency rates are not a leading indicator of budding risks.
Figure 7 shows the percent of mortgages in Ontario that are late on their payments by more than 90 days (i.e. delinquent) going back to 1990. The last time delinquency rates were as low as they are now was right before the great 1990s housing market crash. Delinquency rates did not rise during the 1990s until the market crashed and unemployment rose.
These delinquency rates are also a function of the amount of a household’s income that must be devoted to paying the interest cost on a mortgage loan (see Figure 7). The higher the debt servicing costs, the harder it is to repay your loans. Therefore, both the amount of debt and the interest rates paid on that debt are important determinants of future delinquency rates.
It is therefore prudent for governments to monitor risks associated with excessive real estate gains and the impact that has on debt loads, rather than to concentrate on delinquency rates.
Mortgage delinquency rates are low in Ontario, because interest rates have been so low for so long. Longer-term mortgage interest rates are still 200 to 300 basis points below what they were before the financial crisis. However, if interest rates were to rise by 200 basis points, a household taking out a 25-year conventional mortgage today to buy an average priced house in Toronto at $796,000 could be faced with a $300 increase in monthly mortgage payments when renewing in five years’ time. Household incomes are growing quickly in Toronto, but not that quickly. A rise in interest rates to more normal levels could push many households over the edge.
Conclusion
Now is not the time to change the mortgage stress test, for the risks of driving a re-acceleration in market activity in the GTA housing market are still far too great. While a crisis has been adverted and a soft landing has been achieved, there are still significant risks for an acceleration in housing market conditions. It is prudent for Canadian banking regulators to maintain stricter borrowing guidelines for now. There could be a role for more flexible qualification guidelines once interest rates return to more normal levels, however.
_____________________________________________________________________
Sources:
[1] The Office of the Superintendent of Financial Services (OSFI) is an independent federal government agency that regulates and supervises more than 400 federally regulated financial institutions and 1,200 pension plans. Federally regulated entities include all banks, federally incorporated or registered trust and loan companies, insurance companies, cooperative credit associations, fraternal benefit societies and private pension plans in Canada. OSFI’s scope of regulation does not include consumer or consumer-related issues or the securities sector, which are the responsibility of other federal and provincial agencies.
Diana Petramala is Senior Researcher at Toronto Metropolitan University's Centre for Urban Research and Land Development (CUR) in Toronto.
_________________________________________________________________________________________________________