Saturday, January 10, 2009

TD Economics: R.I.P. residential construction boom, 2002-2008.

Canadian housing starts largely unchanged at 177,300 units in December.

R.I.P. residential construction boom, 2002-2008.

After falling from 212,000 units to an upwardly revised 178,000 units (from a preliminary 172,000) in November, we got confirmation this morning that housing starts have indeed embarked on a cyclical downtrend. In other words, that the drop recorded in November was not a kink in the data or dismissible as purely weather-related or sheer volatility, which monthly housing starts figures can be prone to exhibit, especially in the multiple-family unit segment. The year 2008 will have marked the seventh consecutive year where total housing starts where higher than 200,000 units. Residential construction activity had already turned the corner a while ago, however, and today’s data help confirm this. After running at a pace well below the rate of formation of new household for an entire decade (1991-2001), housing starts ramped up significantly starting in 2002 to satisfy pent-up demand from that previous decade. It is our view that this pent-up demand has been absorbed. Consequently, housing starts will, over the long-term, have to come in line closer to the latest estimate of household formation rates, roughly 175,000. R.I.P construction boom, 2002-2008.

Is 175,000 units or so the new norm or level we should expect going forward? Not likely. In a recessionary context, we think housing starts in Canada will undershoot that benchmark for a while. On a national scale, our forecast calls for a bottom in housing starts near 140,000 units in the fourth quarter of this year – hence another 20% lower than the last recorded levels. On an annual average basis, starts will likely average around 150,000 units in both 2009 and 2010. From a regional perspective, the housing starts downtrend, while broadly-based, has been, and will continue to be, most severe in B.C. and Alberta. No province is immune from this, however, and we expect every province to record double-digit percentage dips in housing starts for 2009 when compared to the still lofty levels of 2008.

The employment data also line up well in confirming this long-awaited, and just as long in coming, downturn. During the first three quarters of 2008, the construction industry (as defined in the Labour Force Survey (LFS), which includes non-residential construction) was creating jobs at an average monthly pace of 11,200. A sharp U-turn occurred in the fourth quarter of 2008 to close out the year on a sour note, with jobs being shed at an average monthly pace of 15,300. Back-checking with the establishment payroll survey helps confirm that the residential segment, rather that the non-residential or engineering segments of construction, is indeed most likely responsible for the construction employment losses showing up in the LFS data. All said, it certainly seems as though an important chapter in which the residential construction industry was contributing to Canadian growth and employment in an outsized fashion, has come to an end. Lagging slightly behind the rest of the economy, residential construction has crossed over into the recessionary chapter, and looks unlikely to come out of it before 2010.

Pascal Gauthier, Economist, 416-944-5730

4 comments:

jesse said...

Actually that's a pretty upbeat report. Also their "pent-up demand" argument is weak. Where did all this "pent-up demand" live before this real estate boom? Guess what: it's the same place they will go now the boom is over. I'm sure Mr. Gauthier can research where this is in another press release.

mohican said...

I think the report is also quite upbeat. The ownership rate has risen dramatically over the past 7 years and Canada does not have high enough population growth to account for the level of starts that we've seen. The population has grown the past few years so there was a fundamental justification for some of the building but the real key is in the household ownership rate, which has risen significantly and does not have much room to rise further. In fact, I am looking for it to decline. This probably means that national housing starts will bottom in the 100,000 annual range in 2010 or 2011.

jesse said...

Yes but the ownership rate is not the problem. Those that didn't own used to rent so it's really about local in-migration and lower numbers of people per dwelling. Since the mix seems to be more smaller dwellings now than before, the high number of starts may be partially justified.

Unknown said...

The Canada Mortgage and Housing Corporation (CMHC) reported December housing starts declined to an annualized 177,300 from an upwardly revised November figure of 178,000. The annualized figures for the final two months of 2008 were both well below the actual total number of starts for the year: 212,366.Dovetailing with the CMHC housing starts report was news from Statistics Canada's Labour Force Survey that construction jobs declined in December by 44,300 positions. The statistical agency also released a report revealing that the value of building permits in November, residential and non-residential declined 11.8%. The number of residential units that municipalities issued construction permits for dropped 30% compared to November 2007 and the value of residential permits fell 9% in value compared to October.TD economist Pascal Gauthier said that for most the last six years, pent-up demand from the 1990s helped propel new home starts to levels above the formation rate of new Canadian households.see more in....Toronto Lofts