This is the first part of a Blogpost that responds to research findings reported through a recent Globe and Mail article that states condo rents are declining.
A recent report by an equity research firm came to the conclusion that condo rents are on the decline, which, in turn, could trigger a sell-off of investor units that would cause condo prices to “drop dramatically.” They found that on a per square foot basis, rents declined by 1.6% from last year. The basis of their analysis came from researching ‘for rent’ postings on Craigslist in 47 projects, and comparing them to the results from last year’s exercise. They added that mortgage pre-approvals are not needed to buy new condos and buyers may have trouble closing, which could lead to “significant losses” for developers and lenders.
Urbanation does not have access to the actual report, so we are basing our response on the information published in the Globe and Mail.
The first part to this Blogpost will address the more glaring issues related to the quality of this research. The second to follow will address claims of impending rent and price declines and closing risk.
Using Craigslist as the basis for determining changes in market rents poses at least a few major problems.
Craigslist doesn’t report transacted rents
It’s important to recognize that Craigslist is a listings website – it shows how much owners are asking for their rentals, not what they are getting. While it may be true that listing values can generally be used as a close approximation for transaction values, it is less likely to be the case on Craigslist. These listings can often be posted by individuals with very little knowledge of the market, and are more likely to under- or over-price their units, relative to units on the MLS system, which are posted by realtors who work in the market every day. If there is a difference in the degree of over- or under-pricing units between two periods, the change in average rents calculated becomes unreliable.
The sample size is too small
The listings vs. transaction value issue may not be that problematic if one were using a large enough sample that could mask differences in the degree of over or under pricing units between two periods. But 47 buildings isn’t enough. It represents only 3% of the total number of condo buildings in the Toronto CMA. It’s also unclear if the same buildings were monitored at both points in time. Given that they report, 553 postings in October of this year versus 148 last year – this isn’t likely the case. Rents can be very project-specific, and even slight changes in the sample can impact results.
Inconsistent reporting of unit sizes
It was stated in the article that the research ‘controlled for unit sizes where possible’ to measure changes in rents per square foot. When measuring changes on a per square foot basis in a small sample, precision is paramount – you can’t look at rents per square foot for some units and not for others. And the unit sizes must be accurate, or at least consistently reported across sample periods. Is it a stretch to think that some people posting on Craigslist may exaggerate? Again, this may not be that big of a problem if everyone exaggerated by the same degree. But if one person posts a 650 square foot unit for $1,600 in 2012 and the same unit gets posted in 2013 for the same rent but it’s now advertised as 675 feet – it reduces the rent per square foot by almost 4%. Similar issues may arise using unit size information taking from MLS listings, which is why Urbanation goes through the arduous process of maintaining a complete database on unit sizes taken from survey measurements on record at the land registry.
Who's research should be taken with a “grain of salt”?
We recognize that transactions through the MLS system don’t represent all activity in the condo rental market, but we do believe it represents the strong majority. Rentals through MLS will reach 20,000 units this year, which equals an annual turnover rate of roughly 8% of the entire stock. Adding in units sold yields a total annual turnover rate of about 15%. The ‘true’ annual turnover rate of all condos is likely closer to 20%, which means total rental transactions through other means such as Craigslist probably add about 1/3 to the total.
This means Urbanation captures about 2/3 of the entire condo rental market. The activity within the 47 buildings monitored through Craigslist would capture no more than 5%. Nonetheless, the authors of this research cautioned that claims such as Urbanation’s that rents per square foot are rising by 4% year over year should be taken “with a grain of salt”. We wonder, what would be appropriate to take with their claims?