A weighted average price to better measure price variation

When measuring the aggregate price of all residential properties in a given area and time period, it is important to remember that the price is always influenced by the distribution of transactions in terms of both geography and property category. Real estate market analysts are more interested in the rate of change of average prices between two periods of time than average prices in themselves. Thus, when we measure the evolution of property prices from one period to another, we cannot be sure that we’re correctly measuring price variations if the sales breakdown is not identical between the two periods. If this is not the case, price variations may be attributable to a change in the sales breakdown. The QFREB has therefore developed a simple calculation that can prevent such changes from distorting variations in price.

Regional Disparities in Property Prices

The geographic breakdown of sales is particularly important as prices can differ significantly in various geographic areas. For example, with property prices being much higher in British Columbia than in Québec, it is evident that total sales concluded in each of these two provinces will influence the national average. For example, a decrease in sales in British Columbia and an increase in sale in Québec would have a downward influence on the average price of Canadian properties.

Let’s now consider an example specific to Québec by calculating the price variation between January 2016 and January 2017.

If 500 transactions were concluded in January 2016 in the Montréal Census Metropolitan Area (CMA) at an average price of $350,000 and 500 were concluded in the rest of the province at an average price of $200,000, the provincial average price would be $275,000.

Now suppose that in January 2017, the average price does not change, neither in Montréal nor in the rest of the province. However, this time, 600 transactions were concluded in the Montréal area and 400 in the rest of the province. The provincial average price then rises to $290,000.

The price variation would therefore be 5%, which could lead us to incorrectly conclude that prices have increased in the province. However, the increase is due solely to the fact that the Montréal share of total provincial sales rose from 50% to 60%.

To get around this problem, the QFREB calculates a weighted average price by using the share of sales concluded in each CMA (Gatineau, Montréal, Québec City, Saguenay, Sherbrooke and Trois-Rivières). In other words, each CMA always carries the same weight when the provincial average price is calculated. Each region’s weight was determined according to their share of the provincial total of Centris® transactions in a reference period which runs from 2008 to 2012 inclusively.

Three Property Categories, Three Different Weights

Within each CMA, the average property price is also influenced by the breakdown of sales by single-family homes, condominiums and plexes (2-5 units). When these proportions change between two periods, the variation in prices are affected. The evolving condominium market is a good example of this phenomenon. In the Québec City metropolitan area, the share of condominium sales increased from just under 15% in 2000 to 24% in 2012. Since condominiums generally sell for less than single-family homes and plexes, the increase in the share of condominiums has pulled the average price downward in the Québec City area over the years.

To correct this effect, we applied a fixed weight for each property category in each of the six metropolitan areas. These reflect each property category’s share of Centris® transactions of the total residential sales in its area from 2008 to 2012. These weighted values are applied to the entire series, making it easier to compare the different periods.

Comparing Apples with Apples

In summary, for each metropolitan area, we obtain a weighted average price by using fixed shares for each property category. We then calculate a weighted average price for the whole province by assigning a fixed weighted value to each metropolitan area.

Consequently, changes in average prices between two periods better reflect reality. The benefit of using a weighted average price is that it ensures that price variations are not simply the result of categorical or geographic variables.