OCTOBER 17, 2012
Canadian housing sales fall in September:
The Canadian Real Estate Association has released final September sales data. September sales totals came in at the lowest level since 2001. This comes as no surprise given the significant decline in sales seen in Canada’s four largest metros, as I outlined in several posts last week. Click on the links to see the ugly numbers from Montreal, Toronto, Ottawa, and Vancouver.
Despite this, CREA put on a brave face, noting in their press release that, “The number of home sales processed through the MLS® Systems of real estate Boards and Associations in Canada rose 2.5 per cent on a month-over-month basis in September 2012.”
Let’s chalk this up to the “miracle” of seasonal adjusting, as nominal sales fell 11% m/m, the second largest September m/m decline in a decade. The 14% upwards adjustment in sales was the largest seasonal adjustment ever for the month September.
Chart courtesy of Scotia. Hat tip to Mike Fotiou.
CREA attributed this record-high seasonal adjustment partially to the fact that September had only 20 weekdays, which is indeed rare, yet 1990, 2001, and 2007 also saw 20 weekdays in September but had much smaller seasonal adjustments in sales.
The CREA press release was also quick to point out that, “The national sales-to-new listings ratio, a measure of market balance, stood at 49 per cent in September 2012, remaining near the midpoint of a balanced market.”
Once again, the methodology used to seasonally adjust this data strikes me as suspect, as can be seen in the non seasonally-adjusted data below. The bottom line is that you have to go back to the early ‘90s to find a September sales-to-new list ratio this low:
The sales numbers were bad, even once accounting for fewer work days. This October will actually have two more week days than last year, so I expect the year-over-year decline to be more muted than we saw in September, likely in the 5-10% range. I do wonder if the fact that this October has more week days than last will be a prominent theme in CREA's analysis, as it was this past month. I doubt it.
Canadian debt-to-disposable income figures revised significantly higher:
Stats Canada has revised its methodology for calculating the household debt-to-disposable income ratio to align with IMF and UN data reporting practices.
The revision has resulted in a significantly higher debt-to-disposable income reading, with Q2 2012 figures adjusted from 152% to 164%. The chart below shows the new data plotted against a recent chart from the Bank of Canada, taken from the June 2012 Financial System Review publication. The dashed, red line represents the new data:
It’s worth noting that the US will be releasing their revised data in early 2013, but as I have frequently argued, the “cleanest” means of comparing debt burdens between two countries is to look at debt relative to GDP, as it removes the influence of differential tax regimes. The chart below shows household debt-to-GDP figures for Canada and the US:
Household debt-to-GDP patterns are very similar between Canada and the US. I’ve always questioned the significant gap that previously existed between Canadian and US debt-to-income figures in light of this. I suspect that even when the US revises their data next year, it will result in a negligible change.