By Michael Englund
Housing starts -- a measure of the number of new homes on which builders have started construction -- fell 5.6%, to a 2.014 million pace in October. This is from an upwardly revised rate of 2.134 million in September (2.108 million previously). Single-family housing starts dropped 3.7%, after a 3% gain in September, while multifamily starts were down 14.8%.
Permits fell 6.7%, to 2.071 million. Housing completions were flat at 1.964 million. Despite the larger-than-expected decline in starts, the pace remains very healthy.
The report revealed declines in starts for all four geographic regions, but with notably lofty levels of activity in the South for starts, permits, and housing under construction. More generally, permits and housing under construction were stronger than starts -- reflecting the counterintuitive observation that the weather was actually worse in October across the country than in both August and September, despite the hurricanes. We assume that starts are still oscillating in the 2 million to 2.1 million area, given the trajectory of the permit figures.
We expect a solid 0.8% rise in construction spending in October that will likely be followed by monthly gains in the 1% area through the winter months, led both by new home construction and repair work. For October, the drop in the multifamily "under-construction" data suggests that the number could be weaker. But the single-family figures were robust, and home improvement is likely to be strong as well.
The figures are consistent with residential construction growth in the same 10% area in the fourth quarter as will be reported for the third quarter, once the previous 4.8% real-growth clip from the last third-quarter gross domestic product (GDP) report is revised higher.
Overall, low interest rates, a solid economy, and good growth in income and profits are continuing to drive strength in housing. This momentum is despite recent jitters in the markets about a "bubble pop," which we view thus far as little more than a typical market reaction to normal seasonal weakness in real estate (see BW Online, 11/09/05, "Don't Let Housing's Seasons Scare You").
In our view, the housing sector's peak will likely not be seen until early 2006. There is powerful housing-sector momentum, and the "conundrum" for long-term yields will still leave mortgage rates at historically low levels of 7% or less in 2006, despite Fed tightenings.
Englund is chief economist for Action Economics