A few years ago the Canadian Real Estate Association decided that it was tired of the variability in average and median prices. When average prices shot up the regulators got antsy and starting talking about how they could cool the market. When prices dropped, they desperately tried to explain it away by talking about changes in sales mix and seasonality.
The alternative is a repeat sales index like the long standing Teranet HPI which tracks how the prices of homes are changing by tracking homes that sell twice (a sales pair) and taking the difference in price (and doing a bunch of statistical magic to try to remove outliers like heavily renovated flips). This works well, but Teranet only tracked the market as a whole rather than individual property types or regions.
So the CREA came out with their own repeat sales index, the MLS HPI. Unlike the Teranet, you can slice and dice this index into arbitrarily small sub markets. They do this by coming up with a benchmark or typical home in each region and property type. So they will have a set of characteristics of the typical 1 storey single family home in Gordon Head, and in Oak Bay, and in the westshore, or the core. Same for a typical condo in downtown and so on. Once they have defined the typical home, they try to track the price of that home over time.
For example, do you want to know how much the price of the benchmark townhouse in North Saanich increased between July 2016 and April 2017? No problem. Click some buttons in the handy dandy MLS Dashboard (you plebes only get the national one but I get the detailed one for Victoria) and you can pull up a chart to show exactly that.
And there you have it. The typical townhouse in North Saanich increased from a value of $602,300 to $690,000 in that span of 9 months. Pay no attention to the fact that there was not a single sale of any townhouse in North Saanich in that entire period. None at all. So maybe it’s a stretch to call it a repeat sale index and you should take that graph with one of these.
However I don’t want to be too harsh on the MLS HPI since I really don’t mind it overall. I’m sure that when there are no sales in a certain region of a certain property type they use sales in neighbouring regions where there were some, or in different property types and then adjust them based on historical relationships between the two. In the end you get a number that is about as good a guess as to market value for the benchmark house as you’re going to get. I do wish that the VREB would indicate when the data quality goes down the tubes for a region in their monthly reports but in the end it’s a minor sin.
Couple weeks ago they updated their definitions of what constitutes a typical home in each region so I thought I’d pull out a few municipalities and fields to see what this magical benchmark home is.
| Muni | Storeys | Lot Size | Living Area | Beds | Baths | Built |
|---|---|---|---|---|---|---|
| Colwood | 1 | 8405 | 1430 | 3 | 2 | 1979 |
| Colwood | 2 | 8058 | 1841 | 3 | 2 | 1991 |
| Esquimalt | 1 | 6221 | 1276 | 3 | 2 | 1958 |
| Esquimalt | 2 | 6098 | 1547 | 3 | 2 | 1953 |
| Langford | 1 | 7500 | 1377 | 3 | 2 | 1992 |
| Langford | 2 | 5227 | 1660 | 3 | 2.5 | 2006 |
| North Saanich | 1 | 22215 | 1897 | 3 | 2 | 1981 |
| North Saanich | 2 | 22215 | 2464 | 3 | 2.5 | 1989 |
| Saanich East | 1 | 8373 | 1484 | 3 | 2 | 1967 |
| Saanich East | 2 | 9147 | 2099 | 3 | 2 | 1980 |
| Saanich West | 1 | 7405 | 1323 | 3 | 2 | 1958 |
| Saanich West | 2 | 7173 | 1784 | 3 | 2 | 1980 |
| Victoria | 1 | 5662 | 1216 | 2 | 2 | 1949 |
| Victoria | 2 | 5662 | 1771 | 3 | 2 | 1920 |
| Vic West | 1 | 4791 | 1243 | 3 | 2 | 1967 |
| Vic West | 2 | 5227 | 1500 | 3 | 2 | 1912 |
The “living area” is only above ground, so not very accurate given how many places have finished basements. Still, kind of neat to see the differences in character reflected in the data, from the cramped lots in Vic West, to the estates in North Saanich, to the character homes downtown, and 70s boxes in Saanich. In the downtown we see the 2 storey buildings are the older character homes, while in the westshore they tend to be the newer builds.
From looking at some areas with largely homogeneous housing stock like Gordon Head, I believe the benchmark is more of an average rather than the most common property, but still you can see the broad differences in what is common by region.
