, , , ,

This post is a case study into the information on the state of the market for residential housing and, more specifically, about some of the differences between good information and what is seen in the popular media.

Consider the market for homes in the Toronto area. The Toronto Real Estate Board (TREB) offers very good information, with lots of details and insights based on de-averaging.  (The data described below, come from the March 2015 report.)

Those details would bore most people, which is why there are researchers whose job is to pick it apart more carefully (and why students should be required to attend courses on “Research Methods” or “Market Analysis”.) For other people, the media fills the space with more interesting stories. You have probably read stories about how there is a price bubble in Canada. The size of the bubble is up for debate; I have heard reputable sources conclude that prices are 35% too high or 10- 30% or whatever is in tomorrow’s paper. (Understanding the nuances of this debate is why you should attend those courses.)

Many people find such stories unconvincing. Research on personal judgment indicates that more colourful personal examples are easier to remember. That is what the media provide. The problem is that the most colourful examples are rarely representative.  That difference can be dangerous if you want to make a reasoned decision.

Consider the information published in the Real Estate section of the Globe and Mail (GM) every Friday. (In many ways, this newspaper is a better source of information than some others.) They pick a small number of recent sales and describe them; since the last Friday in February, 42 properties have been profiled. The selection is informative for its selectivity and what careful research would see as omissions.

First, consider the selling price: are the profiled homes typical? No. They tend to be single detached houses, with prices higher than the prices of single detached houses reported by TREB. This figure shows the TREB distribution in red and the GM’s distribution in blue.

GM SP distributionThis difference should not be surprising: more expensive homes are probably better looking and sell more papers.

The problem facing anybody trying to learn about market conditions is these houses differ according to other indicators of market conditions: how fast these houses sell and the relationship between the list price and the selling price.

The houses profiled in the GM sell extraordinarily quickly. According to TREB, the average days on market is about 20 days. If you exclude 3 profiles, including one of more than 250 days, then the average according to GM was less than 8 days. It is dangerous to think that you can sell your home equally fast. Further, according to TREB, the number of sales is consistently less than the number of new listings, by a sizable fraction.

The discount from list price is another widely-used indicator of market conditions. When I started work in this area, a normal discount below list price was 3- 4% and the selling price rarely exceeded the list price. Now the selling price was 1% more than the list price on average in the TREB data. The profiled houses sold for 6% more than the list price, on average. This figure shows that many of the profiled houses sold for a lot more than 6%. (Note that the list price is not a “market price”, since it is controlled by the seller only. Several researchers are trying to understand why a bargaining-type  price mechanism has changed to an auction-type mechanism and what the change reveals.)

GM Discount There is a lot of excitement and it is easy to lose sight of what really matters. Examples are interesting but may not be representative. A housing boom and a housing crash differ in that the prices are going in different directions but differences will also be evident in other indicators. These changes can be hard to decipher based on what the media select to report.