Privately-owned housing starts in September were at a seasonally adjusted annual rate (SAAR) of 872,000. This is 15.0 per cent (±12.1%) above the revised August estimate of 758,000 and is 34.8 per cent (±18.2%) above the September 2011 rate of 647,000, according to the US Census Bureau. That was way above the median consensus economists’ expectations for a a SAAR of 768,000 as surveyed by Briefing.com.
This is another example of economists underestimating the strength of the economy in September. Their consensus guesses have been below where the data was ultimately reported, in virtually every major economic report for September except industrial production. We should note however that the range of error in the SAAR number is extremely large, so the economists’ guesses and the SAAR data are all really just a crap-shoot. That being said, we still have to look at the numbers and discern as best we can what’s likely to be actually going on. For that I rely on the actual not seasonally adjusted data for a clearer picture.
The actual, not seasonally adjusted number for September housing starts was 79,000 units, up 8,900 from August and up 20,200 units or 34.4% from September 2011. That is a huge percentage gain, but considering the extremely low historical level from which this gain is measured, perhaps the number of units is a better comparative measure. By that standard, this was still the best year to year gain since near the top of the bubble in November 2005, with a gain of 22,600 at that time.
There’s no “usual” seasonal pattern from August to September. Sometimes starts are up and sometimes down. Hurricanes can have an impact if they make landfall in populated areas with a lot of housing construction, especially Florida, where there’s always plenty of housing construction in near coastal areas. September 2011 was up 4,300 units from August. In 2010, September was down 3,300 units. The average September change over the past 10 years was a decline of 2,800 units.
By any standard, the current number was very good. However the majority of the gain was in multifamily, which accounted for nearly 1/3 of all starts. The gain in single family starts was more muted.
Photo: Wall Street Examiner
Housing Starts and Employment
The actual, not seasonally adjusted number of September single family unit starts was 52,500 units, up 2,600 from August and up 15,200 units or 40.8% from September 2011. The year to year gain of 15,200 units was was the best annual gain since the top of the bubble in November 2005, when the year to year gain was over 22,000. The difference between now and then is that the market was oversupplied then. Today existing single family house inventory on the market is extremely low, and the pace of new house construction is still near historic lows in spite of the recent gain.
On a month to month basis, the gain, while small, was better than average. There’s no typical seasonal pattern in September. September 2011 was down 2,100 units from August. In 2010 September was virtually unchanged from August. The average September change over the past 10 years was a decline of 2,500 units. In those terms this September’s performance was very good.
When the market was oversupplied in 2005, housing starts began to decline a full year before both stock prices and full time employment topped out. But at the stock market bottom and ever since, the pattern of stock prices and the pattern of the annual change in housing starts has been strongly correlated. The growth rate in full time employment followed the rally in housing starts and the stock market coming off the low in 2009. Employment lagged housing starts, but led the stock market at the top in 2006 and 2007. The moral of the story may be that stock investors are the last to get the news, or that belief in stock bubbles dies hard. With housing starts still trending higher, there’s probably little likelihood of a stock market top before there’s evidence of the gains in housing topping out.
Housing starts are a measure of supply. Single family housing demand is best indicated by the level of new house sales. Builders have been known to get ahead of the market (as we are all too well aware) building more units than the market can absorb within a reasonable time. On that score, there may be cause for concern.
The top graph of the chart below shows single family starts versus sales. While starts have edged to new 42 month highs, sales are lagging, with the gap between starts and sales beginning to widen. That could be a sign that the single family market may be starting to overheat. At this point, it’s just an early caution sign that bears watching. The September new house sales figures to be released on October 24, will be an important one to watch. Strong sales would support the growth in starts, but sales weaker than the gains in starts could be a sign that trouble is brewing.
Photo: Wall Street Examiner
Housing Starts vs. Sales
The gain in multifamily unit starts has been disproportionate to the gain in single family starts. Most multifamily starts are rentals. The latest available government data from the census bureau pegged the national vacancy rate for multifamily near 9%, but declining, in the second quarter. Reis Reports does a private survey of apartment rental vacancy in the largest US metros, and it reported vacancy declining to 4.6% in September in those markets. However, Reis feels that the demand fundamentals have deteriorated in the past couple of months, with the decline in vacancy slowing and some slowing in absorption, while starts rise.
In my experience as a commercial real estate appraiser and market analyst from the late 1980s through 2002, a 5% vacancy rate was considered a tight market. Since then vacancy rates have typically been higher, with 8-10% often considered “norma.l” A vacancy rate that low today suggests good demand for multifamily housing. However with multifamily starts surging, the trends may be approaching equilibrium, which would result in slower growth going forward. Where the market could get in trouble would be if starts outpace demand sufficiently to build an oversupply. We’re not there yet by a long shot, but that is how all cycles have ended. They never simply remain at a healthy equilibrium.
Read more at Wall Street Examiner.