Several articles over the past few months have suggested that since measures of house price affordability remain near historic norms, the chance of a house price correction in Canada is slim. Certainly this is the strongest argument for continued stability in the Canadian housing market, but like any measure of real estate fundamentals (which admittedly includes real house prices and the price/rent and price/income ratios, which I often reference), affordability measures have their shortcomings. For today, let's set aside the two fundamental risks in relying on an affordability measure to forecast market strength:
1) The reality that interest rates remain near historic lows with only one long-term direction in which to move. As they do increase, barring very strong income growth, affordability will erode as a function of some pretty simple math.
2) The reality that the current housing boom has been a disproportionate driver of our economic and labour market growth. For example, residential investment remains near all-time highs as does GDP output and employment in housing-related industries. At the same time, home equity withdrawal through mortgage refinancing and home equity line of credit growth is running at levels consistent with what the U.S. experienced at its peak, artificially supporting consumption and aggregate demand. So employment and economic growth, and by extension incomes and hence affordability, would look very different without the current boom in real estate and credit-driven consumption.
While lots could be said about those two issues alone, today I want to instead explain why the two most commonly referenced affordability measures tend to portray housing as more affordable than it actually is.
Overstating typical down payment
RBC produces a fabulous quarterly report detailing affordability trends over time. It is an interesting and informative read and provides regional data, which the other commonly-referenced index does not. But it is not without its limitations.
The first is to assume that down payments have been stable over time, despite the fact that we know that is not the case. The household savings rate has fallen from the high teens in the 80s to roughly 4 per cent today. Even the Bank of Canada chimed in on this over the summer, suggesting that new buyers who take out insured mortgages tend to make close to the minimum required down payment. The reality is that 5 per cent down and cash-back mortgages that essentially allow for 100-per-cent financing have not always been a part of the mortgage landscape. That being the case, assuming that buyers make a 25-per-cent down payment would skew the reading.
The counterargument is that the RBC index also assumes a 25-year amortization, which is smaller than the allowable 30 years. However, the difference is not offsetting: The average resale price in Canada in April was just over $375,000. With a 25-per-cent down payment and the remainder amortized over 25 years at 3.5 per cent, the monthly payment is just over $1,400. If we adjust to a 90-per-cent down payment and amortize over 30 years, the payment jumps to $1,513.
Despite the fact that RBC assumes a stable 25-per-cent down payment and interest rates are near all-time lows, we find that affordability in most Canadian cities is worse than the average of the past 20 years.
Is the average house really only $260,000?
One would think that the Bank of Canada would be a pretty authoritative source on financial matters, yet they have a very peculiar way of calculating their affordability index. You can read about it here and see the index itself here.
Unlike RBC, the Bank of Canada assumes a 5-per-cent down payment. This is certainly conservative.
Where things get weird is in how they calculate the average house price. What they've done is set the price at $144,600 in 1990, which is simply the average resale price at the time as calculated by the Canadian Real Estate Association (CREA). From that point, the Bank of Canada estimates the change in the average house price by averaging out gains as reported by Statistics Canada's New House Price index (NHPI), and by Royal Lepage (i.e. CREA).
The issue here is that the NHPI is a quality-adjusted index, which means it seeks to measure the change in a comparable dwelling over time. This is a very important concept when it comes to tracking true inflation over time as you need an apples-to-apples comparison.
Its use in calculating the average house price in Canada for the sake of creating an affordability index is far more suspect.
We know that the average size of new dwellings has risen in Canada, with the average new house being just under 2,000 square feet, according to data from the Canadian Home Builders Association. This is up markedly since the 1970s when the average house size in Canada was under 1,100 square feet. So as a society, we've changed our expectations for what constitutes a "normal sized" house. The problem is that as we've demanded larger and larger homes with better amenities and have been willing to stretch the household budget further to get those, the NHPI has been busy factoring out these changes.
The end result is that by pegging 50 per cent of the growth in house prices to the NHPI, the average house price used in the Bank of Canada affordability index has significantly underperformed other measures of house price appreciation. The chart above illustrates the change in value of the average house used in the Bank of Canada affordability index. If you believe that the average house in Canada is $260,000, I can see how you'd think that there was no problem with affordability in Canada.