![]() |
|
|
|
|
Special Reports A DIFFERENT LOOK AT CANADIAN HOME PRICES November 20, 2008 Canada’s housing market has come into focus lately. Homeowners need not look very far south for reasons of concern, as the U.S. housing market continues to tumble. To gain a clear perspective on housing market trends, however, one needs good price data. One measure often quoted is the simple average price in any given month and market, such as those provided by the National Association of Realtors (NAR) in the U.S. and the Canadian Real Estate Association (CREA). As of the last available data from CREA (Oct. 2008), the average Canadian home price was 10% lower than a year prior. These measure the average price for existing home transactions in a given month and market regardless of unit type, size, or quality. When aggregating across local markets to arrive at a national price, prices are simply weighted by the relative sales volumes in each city. In normal times, these regional sales-weighted price measures are reliable. However, the same cannot be said when markets are at a turning point and sales fluctuate significantly. Such is the case in Canada at the moment. As at Oct. 08, sales were down 50% in British Columbia for example. Since average prices in British Columbia are the highest in the nation, the drop in sales tends to overstate the extent of price declines when applied to a simple national average. Analysts of U.S. housing markets typically prefer to look at repeat sales price measures, such as those provided by OFHEO and S&P/Case-Shiller, rather than a simple average. Such price indices offer a better apples-to-apples comparison over time by focusing on houses that have sold more than once. Canada lacks such a repeat sales price measure. As of its latest press release (Nov. 14), CREA has started to derive national prices using a more accurate stock-weighted method on the regions. Regional markets are weighted by the outstanding stock of homes, rather than sales. Along this latter vein, TD Economics has computed a TD Home Price Index (HPI), which weights the regional markets by the outstanding stock of homes within each market. This will help control for price volatility related solely to shifts in sales volumes – which arguably distort national figures. A simple example is provided in the accompanying table to highlight this last point. The TD HPI does not and cannot control for unit-type or segment shifts (e.g. expensive single-detached versus cheaper condos sold in any given month) within major metropolitan markets, but it does control for sales volume shifts across these major markets(1). We weigh each major market by its share of housing stock (as per cent of total) using the number of dwellings from Census data, interpolating as needed between Census years. To remain agnostic about post-2006 Census developments, weights are fixed after the 2006 Census until we get the next Census from 2011. The monthly TD HPI starts in Jan. 2002, which is, for the time being, the longest available consistent series. The result is a more reliable, less volatile national home price measure. As of the latest available data (Oct. 2008), the CREA sales-weighted major market home price measure was down 10.9% Y/Y. The TD HPI stock-weighted measure shows a much smaller decline of 4.6%. As suspected, the difference arises predominantly because of large double-digit drops in sales in some of Canada’s most expensive markets in British Columbia, which our index controls for. Over the course of the entire series (since Jan. 2002), the CREA price measure showed an average annual gain of 8.6%. The TD HPI gain was similar, but slightly higher 9.1% in nominal (current dollar) terms, and 7.4% in real (constant dollar(2)) terms. Year-to-date (Jan.-Oct. 08 compared to the same period last year), the TD HPI shows Canadian home prices to be still in the black at +3.6%, compared to -0.4% by the CREA measure. We will continue to monitor and report on the TD HPI and how it compares to the sales-weighted price throughout this turning point in Canadian housing markets.Pascal Gauthier, Economist Grant Bishop, Economist Endnotes (1) The 24 major markets used in compiling the TD_HPI are Calgary, Edmonton, Halifax-Dartmouth, Hamilton-Burlington, Kitchener-Waterloo, London & St. Thomas, Trois-Rivières, Montréal, Ottawa, Gatineau, Québec City, Regina, Saguenay, Saint John, Saskatoon, St. Catharines, Newfoundland & Labrador, Sudbury, Thunder Bay, Toronto (incl. Brampton, Durham Region, Mississauga, Orangeville and York Region), Greater Vancouver, Victoria, Windsor-Essex, and Winnipeg. (2) We adjust for inflation using the CPI (excluding shelter). For the full report in PDF format - including all charts and tables click here.
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Privacy Policy | Internet Security | Legal | TD Group Financial Services Site - Copyright © TD |
||