In a modern economy, governments apply regulations to a vast range of activities in response to concerns about safety, relationships between companies and their customers, and protecting the environment, among other issues. There is always a fundamental question: What is the right balance between managing risks versus supporting consumer choice?
The federal government’s regulations about residential mortgages have become unbalanced, and the rules are causing more harm than good. They have prevented hundreds of thousands of Canadians from making housing choices that they believe would be in their best interests, added to pressures within the rental sector and aggravated Canada’s housing supply crisis.
One significant concern is about the mortgage stress tests required by the Office of the Superintendent of Financial Institutions and the federal housing agency, the Canada Mortgage and Housing Corp. Testing is necessary, but these tests are excessive. They require that a new mortgage borrower be able to afford a payment that is two percentage points above the actual contracted interest rate. Before a lender can issue a new mortgage today at a typical rate of 5.5 per cent, it has to be tested at 7.5 per cent.
But the testing fails to consider an important factor that will determine future outcomes for borrowers: that their incomes will increase. OSFI and CMHC are surely aware of this argument, but they never seem to acknowledge it or respond to it.
We are now seeing a real-time and stark – in a positive way – demonstration of the impact of income growth on mortgage outcomes.
Interest rates have increased sharply over the past two years. During 2018, a typical rate for the most-used mortgage type (five-year, fixed-rate) was 3.3 per cent. Five years later, in 2023, those mortgages were renewed at a typical rate of 5.5 per cent.
Yet the number of Canadians experiencing mortgage difficulties has barely moved. As of December, only a small number of Canadians were behind on their mortgage payments by three or more months. The Canadian Bankers Association puts that rate at 0.18 per cent – less than two out of every 1,000 borrowers.
Yes, this is a large proportional rise – 26 per cent – from an all-time low of 0.14 per cent in August, 2022. At first glance, that increase seems bad. But if an extremely low number is increased by a large amount, it is still an extremely low number. The arrears rate for December is still lower than in any other month from January, 1990 – when the data set starts – up to July, 2021.
Some proponents of stress tests argue that the continued low level of the arrears rate shows that the tests have reduced risk and prevented mortgage distress.
There is a simple alternative explanation for the continued low arrears rate: It is due to income growth.
The accompanying table illustrates this. Typically, a renewal results in a payment increase of 20.4 per cent, but the borrower’s income has increased by more – 22.3 per cent. As a result, the new payment is actually more affordable than it was in 2018.
Correspondingly, the current arrears rate (0.18 per cent) is lower than it was in 2018 (0.24 per cent). Among mortgage borrowers who renewed during 2023, a very, very large majority were able to afford the payment increases.
Part of the argument for including income growth in the stress tests is that if interest rates are significantly higher in future, that would happen because the economy had been quite strong, and therefore incomes will have grown more quickly than usual. This is happening now: Over the past five years, the average weekly wage in Canada increased by 4.1 per cent per year. During the prior two decades, the average was 2.8 per cent per year.
The continued low rate of mortgage arrears doesn’t prove that the mortgage stress tests are working. Instead, it demonstrates that the stress tests need to incorporate income growth.
There is an easy way to do that. When income growth is taken into account, a two-point rise in the mortgage rate over five years can be adequately simulated by using an increment of 0.75 points above the contracted rate. OSFI has not responded to that math.
Related to this, there is an argument that the mortgage stress tests are looking at the wrong risk. History and statistical analysis tell us that changes in mortgage rates have very little impact on arrears. The much more important cause of mortgage defaults is loss of the ability to pay, usually because of reduced income, which is typically caused by job loss, but can also be due to illness or divorce.
The arrears rate remains extremely low because the employment situation in Canada is still very strong, and because incomes continue to grow. The unemployment rate for”prime age” adults (25 to 54 years) has recently increased and an impact on the arrears rate might be starting.
OSFI has acknowledged that loss of income is a risk factor. For example, in its January, 2023, consultation paper it said “The MQR [Minimum Qualifying Rate] has proven a crucial risk mitigant against this increase in rates, elevated inflation and the potential loss or reduction of borrower income.” But, it takes very little thought to conclude that an interest rate test tells us nothing about the consequences of job losses.
Risk is two-sided: It’s normal to focus on negative risks (as OSFI does), but in the case of incomes, there is a lot of positive risk. Not incorporating that positive risk in OSFI’s regulations is a failure of analysis and policy, with consequences that are very real.
In 2023, OSFI held a consultation on its mortgage regulations. In its summary of the submissions, it did not mention – let alone debunk – any comments about incorporating income growth. OSFI did receive comments about incomes and is surely aware of the arguments. I mentioned this, as did at least one national association. I expect others did, too.
The design failures in the mortgage stress tests mean that large numbers of Canadians are being prevented from making housing choices that they believe are in their best interests. In this respect, the mortgage stress tests are making us collectively worse off.
Since September, 2008, there have been multiple changes to mortgage lending regulations that have reduced the sizes of mortgages that Canadians can obtain. As a result, home buying has been reduced. Those regulatory changes have varied in their impacts. Many have been trivial, but two have been very consequential: the elimination of 30-year amortization periods for insured mortgages (2012) and the OSFI stress test (2018).
I calculate that, in combination, all the changes to mortgage regulations since 2012 have lowered resale home buying by about 10 per cent (a total of 600,000 lost sales) from what would otherwise have been the case. There must also be a large number of people who did manage to buy, but had to compromise to a choice that fell short of their reasonable expectations. That adds up to a lot of disappointment for a lot of middle-class Canadians.
In addition to reduced sales of resale properties, there has undoubtedly been an impact on sales of new homes and new housing starts. Since 2012, the mortgage regulations have reduced starts by about 200,000 dwelling units. Canada is now experiencing a housing crisis due to inadequate supply. Federal mortgage regulations have contributed to that crisis.
Will Dunning is a consulting economist based in Toronto. He specializes in housing market analysis.
Are you a young Canadian with money on your mind? To set yourself up for success and steer clear of costly mistakes, listen to our award-winning Stress Test podcast.
