In a recent report put out by the Canadian Real Estate Association (CREA), it was suggested that the Canadian Housing Market was starting to stabilize but that prices have decreased countrywide by 6.2% year over year.
RE/MAX has a long association with TD/Canada Trust and the mortgage specialist in our office, Ramy Ibrahim - keeps us up-to-date on the market from the bank’s point-of-view. Yesterday, I received an email from Ramy with a review of the CREA report by economists in the TD/Canada Trust office.
Here is a copy of that email:
A Second Look at Canadian Home Prices
Some quick number-crunching has validated our initial suspicion that the reported 6.2% Y/Y drop in the average home price in Canadian major markets last week may arguably exaggerate the true extent of the home price correction in September.
From the analysis we have conducted, it becomes evident that if we control for the weights of each city, the estimate for home price movement drops to -1.3% Y/Y, from -6.2% Y/Y.
We were initially surprised by the dramatic 6.2% Y/Y drop in the average home price in Canadian major markets reported by the Canadian Real Estate Agency (CREA) last week. This surprise turned to concern when we discovered that controls were not in place to deal with compositional shifts that may distort the actual behavior of the aggregated series. We indicated as much in our initial commentary on the data then. To assess our claim, we have now conducted a basic analysis on the data.
The Problem
The crux of the issue is that unless the underlying cities are properly weighted (that is, the weight is fixed over the periods in question), the results can be biased by dramatic changes in the volume of the sales in certain cities. In the case of this particular housing report, the dramatic 43.2% Y/Y fall in the number of homes sold in Greater Vancouver, and the fact that Vancouver prices are generally higher than average combined to make it look as though national home prices were falling sharply, when in fact it was driven by fewer expensive homes being sold in Vancouver, as a fraction of the whole. And that was clearly acknowledged by CREA.
The Solution
If one were to fix the weight assigned to each major city, this gets around the compositional distortions, and what’s more, the overall change in home prices would be less profound. To do this, we fixed the weight of the each city to their year-ago sales level (September 2007).
As a result of this calculation, the extent of the movement in home prices dropped from -6.2% Y/Y reported last week to -1.3% Y/Y, which suggests that Canadian existing home prices may be falling at a far more measured pace than indicated by the unadjusted statistic. And note that even this calculation does not totally eliminate all compositional effects, as we did not control for the type and quality of homes sold in each city.
Bottom Line
A more robust aggregation procedure indicates that the average existing Canadian home price dropped by 1.3% Y/Y in September, and not 6.2% Y/Y as reported by CREA. As a result, concerns that the Canadian housing sector may follow its U.S. counterpart into a major correction may be overblown.
Simultaneously, all of this emphasizes the great need in Canada for a higher level of precision in Canadian housing market data. We view a measure akin to the U.S. Case-Shiller as the gold standard.
This newsletter is provided for your information only. Conclusions and opinions given do not guarantee future events or performance. Facts and data provided are from sources we believe to be reliable, but we cannot guarantee they are complete or accurate. This newsletter is not intended to provide legal, accounting or tax advice and should not be relied on in that regard. COPYRIGHT 2008 by TD Securities Inc. “TD Securities” is a trademark of the TD Bank, representing TD Securities Inc., TD Securities (USA) LLC and certain investment and corporate banking activities of the TD Bank.
Recent Comments