Canada’s New Home Data Problem
Complex statistical models are only as good as the data in them, and the industry must demand better transparency.
When starting any new company these days, an entrepreneur must consider whether technology will put them out of business in the future. That was certainly a consideration for me when starting a consulting firm that produces residential market studies and values new home projects. Will a computer be able to take available new home data and produce a reliable absorption and pricing valuation and forecast over the next five to 10 years? The answer was no.
Scott Page, an author and Professor of Complex Systems at the University of Michigan was a guest on The Knowledge Project podcast earlier this year and discussed the need for data analysts to be bilingual. What he meant was, they not only need to understand how data works and its limitations, they must fully understand the industry in which data is being derived: two very different skills. First off, understanding the residential real estate market and the multitude of factors that impact value is extremely difficult; I’m personally updating my mental models daily. Complex statistical models are only as good as the data in them, and Canada has a new home data problem.
The CMHC tracks things like starts, completions, standing inventory and single-family house prices for every major market in Canada. However, the house prices are recorded at closing, not at the time of the sale, and starts and completions are also lagging indicators, reflecting market conditions a year to four years earlier. CMHC doesn’t track new condo prices at all. This leaves most of the heavy lifting to for-profit private data collections firms.
During my career I’ve worked for three such firms, one in Dallas and two in the GTA. The last firm I worked at produced tremendous analysis of the data, as did our main competitor. However, the biggest issue was the collection. The accuracy of the data was occasionally called into question, and I vigorously defended the developers and brokers that provided it. Why wouldn’t they tell me the truth, the market was booming? Perhaps I was a little too trusting.
In 2009, market conditions changed, and sales were extremely slow in the first half of the year. Many developers, including some of the GTA’s most well-known, refused to give us their sales numbers. Luckily the market roared back to life, and those firms gave us their figures again. A couple years later I ran into a developer and congratulated him on selling 325 units during the opening month, having just received the figures from their brokerage. He told me that wasn’t accurate; they’d only sold about 250. When I questioned the brokerage, they admitted they’d fibbed because the other big launch at the time had sold well, and they wanted to make themselves look better. I like to believe that these are isolated instances, and that our industry understands that good data can help everyone, including the lenders that are essential to getting their land, construction and inventory financed.
Vancouver and Calgary have a couple firms that cover new housing activity, and there is a small boutique shop in Ottawa, but outside of those three major markets, there is no data at all on new home sales and pricing in most other Canadian markets. A local housing analyst out west told me developers in B.C. and Alberta can be particularly uncooperative during a slow market, and many sales packages include only the starting price, which makes it impossible to know the overall pricing and the value proposition of any individual location. These “starting from” sites can (and undoubtedly have) resulted in a developer charging significantly different prices to different buyers for very similar suites. The industry and new home buyers must demand better transparency.
Statistics Canada has recently began tracking new condo prices, but the indexes are really limited in what they provide. If we add on that the local real estate boards across Canada are very protective of their data, going as far as actively pursuing legal action against Realtors that publish detailed pricing on their websites or newsletters. There are other resale property and land portals that exist in various provinces as well, most of which make it nearly impossible to download large data sets for analysis.
Perhaps I shouldn’t be complaining, this lack of available data and piecemeal new home sales and pricing information in Canada makes someone with my skillset more valuable. You need someone to fill-in the missing puzzle pieces. However, the housing analysis would be better, the development risk would be lower, the consumer would be more informed and more likely to buy, and the entire industry would benefit if there was much better new home data available.
Several of my clients will probably not be happy with me, but measures need to be put in place to improve new home data tracking and reporting in Canada. To the industry, please provide accurate information to your local data firm and Statistics Canadas, and to consumers, support transparent developers.