MARCH 21, 2012
It’s been frequently suggested that Vancouver’s lack of available land has driven house prices to extreme levels. That same thought has also been used to justify Toronto house prices, though to a lesser degree.
I want to share a couple thoughts on that front:
1) The number of dwellings in every major Canadian city has increased in line with growing population.
There should be no net effect of growing population when enough dwellings have been built to meet demographic demand (though speculative demand is another story altogether).
Below is a chart showing the cumulative growth in population and number of dwellings in Canada’s three largest cities since 2001.
While Vancouver’s population has grown 17% since 2001, the total number of residential dwellings has grown 22%. Toronto has grown by 17% but has grown its total residential dwelling stock by 24% in that time.
In terms of housing starts, we note that construction of single family homes in Vancouver is running more or less in line with decade averages while multiple starts (condos apartments, rows, etc) are well above decade trend. If there really is a lack of buildable land, it certainly doesn’t seem to be stopping developers from bringing supply to market.
The same can not be said of Toronto and Montreal where single family housing starts are down substantially from levels of 10 years ago while multiple starts are well above historic norms.
2) Population density: Should Vancouver command such a premium?
Below is a chart I’ve compiled that compares the population density of North American cities to the median resale house price. Yes, Vancouver is densely populated, but it also commands a near 100% price premium over the median resale price of homes in New York and San Francisco. Note also that the resale price for Vancouver is actually for the entire metro area, meaning the city-specific number will be significantly higher.
Vancouver house prices are not just expensive in nominal terms, they are also very stretched relative to incomes. In the chart below, population density is plotted against the median multiple (median house price divided by median household income) as calculated by Demographia.
3) Rents not indicating supply pressure
Finally, consider the price/rent ratio in Vancouver relative to US cities. If indeed there is a land constraint issue, one would expect a premium on house prices but also on rentals as any population is made up of renters and owners. Even in densely-populated cities, house prices and rents historically move more or less in tandem.
Of course, a credit bubble can distort that ratio as houses are bought on credit while rents are paid out of income. Credit spigots can easily be opened, allowing marginal buyers to enter the ownership pool, but incomes are much more “sticky”. Coupled with a change in mass psychology that embraces greater ‘ownership premium’ and the price/rent ratio can, for a time, become distorted. But among fundamental measures of house price sustainability, there is none more important and none that displays a greater mean-reverting tendency than the price/rent ratio.
The IMF recently calculated Vancouver residential dwellings to have an average price/rent ratio of nearly 60. Toronto was estimated at nearly 40. Compare that with estimated price/rent ratios of US cities in 2005, courtesy of CNN:
The bottom line is that land constraints and population density simply cannot account for prices at these levels. Both Vancouver and Toronto house prices are being driven and sustained by a credit boom of historic magnitude in Canada.