This is the second in a two-part series exploring the impact the mortgage stress tests had on borrowers, as well as their relevance today. You can read part 1 here: A deep-dive into the mortgage stress tests: Did they serve their purpose?
In this part, I will review past recommendations I have made concerning the stress tests, along with some additional and unintended risks the stress tests have brought about.
Deciding on the qualifying interest rate
In the past, I argued that the qualifying rate should be the higher of 3.75% or the actual contract rate. Given what’s happening now, was I wrong? Actually, I think what’s happening now vindicates my position.
The argument for the 3.75% test rate had two major parts:
Firstly, the renewal will occur in the future. The stress testing has not considered that the borrowers’ incomes will have increased and the mortgage principals will have been reduced. Therefore, as I calculated earlier, most people who borrowed five years ago at 3.25% should now be able to afford a renewal at 5% or even higher.
Secondly, I have argued that if the future rates at renewal are higher than the test rate, then it will be because the economy is very strong: incomes will also have increased a lot, mitigating the impact of the higher rates. That has happened.
At this time, it looks to me like these arguments were correct.
The policy team at the Canadian Home Builders Association (“CHBA”) was aware of my analysis. They thought about it, and made a useful addition, that the qualifying rate should be increased for terms shorter than five years and for variable rates. I concluded that the increment should be a quarter point for each year that the initial term is shorter than five years.
In retrospect, I’m comfortable that my old recommendation (with the addition from the CHBA) is correct: for five-year fixed-rate mortgages, the qualifying interest rate should be the greater of 3.75% or the contracted interest rate. There should be increments for terms shorter than 5 years and for variable rates.
Today, a typical “special offer” rate for a 5-year fixed rate mortgage is 5.15%. Under my proposed policy, a borrower making that choice would be tested at that rate. If they take a 2-year fixed of 5.2%, they would be tested at 5.95% (the contract rate plus 0.25% times three years). And if they take a variable rate of 4.95%, the test would be at 6.2% (1.25% above the contract rate).
Instead, all of these borrowers will be tested at about 7%. This testing won’t take account of income growth, principal repayment, or where interest rates might be in the future.
Under this proposal, the variable rate mortgages issued during 2021 at rates of 1.3-1.4% would have been tested at 5% (the 3.75% minimum rate plus 1.25%).
An alternative would have been to take 5-year fixed rate mortgages, at contracted rates from 1.7% to 2.7%, but tested at 3.75%. That would have encouraged more borrowers to take the 5-year fixed rate options. In consequence, the current risk environment would be considerably better if that had been the case.
Stress testing has created some additional risks
In addition to affecting choices of variable versus fixed rates, the extremely high qualifying interest rates affected mortgage choices, causing movements to non-insured, non-traditional (non-federally-regulated) lenders, often at interest rates higher than were offered by regulated lenders. Already, this effect has had negative economic impacts, by raising the amounts of interest that are paid compared to what would be paid if the mortgages were via prime lenders. In addition,
At present, alternative-lender mortgages are subject to at least two forms of renewal risk: increased payments and secondly, given the increased sense of risk in financial markets, there may be less funds available for renewals by alternative lenders. Therefore, there may be situations where these mortgages can’t be renewed at all at a reasonable market rate, either at the same lender or via a transfer.
This risk for alternative mortgages is currently a major threat to the housing market and the economy. The federal government should discuss this, and what it can do. This should include lowering the barriers that prevent transfers of those mortgages to regulated lenders and/or insured mortgages.
Another negative event is that a succession of mortgage regulation tightening during the past decade has weighed not just on housing resales, but also on pre-construction sales of new homes.
This has caused housing starts to be less than they might have been. This is one of the multiple factors that have contributed to the housing supply crisis and therefore to the excessive price growth that has occurred.
Another issue: from time to time, there have been concerns about the requirement for stress testing of mortgages that are transferred to a federally-regulated lender. This regulation doesn’t reduce risk in the financial system because the mortgages already exist and therefore risk already exists: at best, all the policy does is influence where the risk is located.
But, it creates risk for the borrowers that they could be trapped at their current lender, and therefore be exploited by that lender. Current interest rates will result in qualifying rates in the area of 7%, so this risk has become elevated. The regulator (OSFI) should eliminate (or sharply lessen) that requirement. I see that the Canadian Real Estate Association (CREA) is also making this argument.
On balance, due to these issues, I’m not persuaded that we are better off as a result of the stress tests. Yes, stress testing has been necessary, but the design is flawed.
Looking forward
Obviously, there is a lot of uncertainty about future interest rates.
The most recent forecasts from the major banks suggest that rates are now at peak levels, but they expect only small reductions during the coming year (not more than a half-point for bonds and no change for the short end).
I have a different view (admittedly coming from a bearish opinion, that the sharply higher interest rates might depress the housing market so badly that Canada enters a severe economic recession).
I expect that rates will be considerably lower by the middle of next year (and if they aren’t, we will be in for a lot of pain that will last for a long time).
In the monthly Housing Market Digest reports that I publish on my website, I have argued that the inflation that is occurring now is mainly due to disruptions to supply, and high interest rates won’t fix that. In fact, they will make it worse by discouraging investment in new productive capacity. We are already seeing an example of that in the housing market, where new home sales have fallen precipitously, which will result in much-reduced housing starts in 2023 and 2024. It is possible that five years from now the housing supply crisis will have become even worse because of these high interest rates (and the further depressive effects of the stress tests).
What happens to inflation will depend largely on what happens to the supply issues. It is possible that those issues may ease, or they could conceivably get worse. If events are favourable, the Bank of Canada should seize that as an opportunity to declare victory and reduce interest rates.