By: Larry MacDonald
Today’s release of the Teranet-National Bank National Composite House Price Index shows a vigorous gain of 2% in Canadian resale house prices from July to August. The national house price index (based on six cities) has now risen four straight months (and likely will show further increases in September and October).
Monthly rises in August were 2.7% in Toronto, 2.0% in Calgary, 1.7% in Vancouver, 1.5% in Ottawa, 1.2% in Montreal and 0.6% in Halifax. “For Toronto it was the fourth consecutive rise of 2% or more, taking the cumulative gain to 9.4% in just four months,” noted the monthly report from Teranet and National Bank.
Montreal, Halifax and Ottawa prices in August are now above their respective peaks attained during the housing boom. August house prices remain below boom-era peaks in Toronto (-3.0%), Vancouver (-7.7%) and Calgary (-12.9%).
This recent leap in house prices is putting housing back into overvalued territory at the national level, going by the IMF model. As for traditional valuation yardsticks (as mentioned in the IMF study) the existing state of overvaluation is becoming more substantial.
5 Responses to “ Housing Bubble Part Deux? ”
Larry, do you understand the gist of the IMF model? I started trying to wade through it but it’s pretty technical.
Are they comparing Canadian housing prices to our standard of living somehow and finding that the prices are too high compared with a similar comparison in the rest of the world?
By Patrick on Oct 28, 2009
Patrick
I took a quick look and I believe one way of explaining it would go as follows: they estimated (using quarterly data for each of five provinces) a statistical equation that relates growth in real house prices to i) growth in real per capita income, ii) growth in real mortgage credit, iii) past growth in real house prices, and iv) a mean reversion factor. Population growth was significant only in Sask. Values for the explanatory variables were plugged into the equations to impute expected price values, which were compared to actual values to assess whether they were overvalued or not. If I had more time I could dig a little deeper and provide a better explanation but the above may do for now, hopefully.
By Larry MacDonald on Oct 28, 2009
Ok thanks. I get the gist: they’ve identified some factors that seem to be correlated with housing prices, and derived a formula linking them all together; and right now actual prices are higher than those predicted by the formula.
It’s a bit egghead for me (and that’s saying a lot; my head is often criticized for its egginess). I don’t know how one would look at such a model and convince oneself that all the factors have been accounted for.
I’ve concluded that house prices are too high based on a comparison with renting, because after all, that is the alternative to buying a house. Only very recently has housing become competitive from that point of view, and I personally don’t believe it will last long.
I guess the one thing the IMF model could capture that I’m missing is the “house premium” that people are willing to pay for that feeling of ownership.
By Patrick on Nov 3, 2009
Patrick
It’s possible the IMF study may have a specification bias but presumably the IMF has some standards in its research (although if it was published in a peer-reviewed journal it would probaly have even more credibility). Interesting posts you have in your blog about renting vs. buying a house. Perhaps another way of highlighting the appeal of renting at this time is the house-price-to-rent ratio, which is about 60% above its historical average in Canada according to the IMF paper. Anyway, why don’t you do a post on the studies that have looked at the rent vs. buy a house decision? I’d be interested in seeing that (you could run it here as a guest post too, if you like).
By Larry MacDonald on Nov 3, 2009