Back in mid-November1, TD Economics introduced its national Home Price Index to help provide a clearer sense of price trends in the midst of a swift downturn in sales. The issue with using the regular average price in such a context - or, for that matter, when sales are booming in pricier markets - is that it provides a distorted picture of the evolution of individual home values.
The national monthly average price is a basket of prices in which each local market is weighted by the number of sales during that month. In typical times, sales volumes tend to be reflective of the market size, as measured by the number of households living within any given market. However, if sales drop by half in a market like say, Vancouver, which is Canada's priciest market, the national average price is being pulled down by the sudden drop in the weight applied to the most expensive market. The weights applied to the markets with cheaper housing prices correspondingly increase. This can happen even if all individual item (home) prices in the basket are steady or increasing. The accompanying table provides a simple example using real data to illustrate this point.
Our TD HPI helps control for price volatility solely related to sales across
Canada's major metropolitan markets.
The latest snapshot, as at March, still shows that the standard national average price decline overstates individual market declines, but to a lesser extent than in months prior. While the standard (CREA) average price recorded a 9.5% year-over-year drop, the TD HPI was 8.1% lower than a year prior. This difference of 1.4 percentage points between the two measures had been as large as 8 percentage points in November 2008, but has been narrowing steadily since then. This is consistent with the evolution of sales, which dropped off sharply late last year in Canada's priciest markets. While sales volumes were still 25% lower than a year prior in the first quarter, they have stabilized close to the level established in the previous quarter (Q4.08). Furthermore, the year-over-year decline in sales is very broadly-based across metropolitan markets, with magnitudes inside a narrower range, typically 25 to 30 percent lower for most markets. The larger 35 and 40 percent sales drops (Q1.09/Q1.08) recorded respectively in Calgary and Vancouver help account for the fact the TD HPI and standard average CREA price are not yet on par.
From their respective peaks, which occured in December 2007 for both measures, the standard average price stood 13.3% lower as of March 2009, while the TD HPI had registered a 10.1% drop. We will continue to report on the TD HPI on a quarterly basis, and compare it to the standard national average price, with our next update on second quarter (Apr-Jun) data slated for late July. Our expectation is that sales will have firmed up slightly from those in the first quarter, and that Canada's major metropolitan markets will continue to move largely in tandem in the near-term. As a result, the national average price and the TD HPI should increasingly provide a similar and accurate picture of underlying residential price trends.