by Calculated Risk on 5/05/2008 12:38:00 PM

If the imminent slowdown in non-residential structure investment is similar to the percentage declines during the '90/'91 and '01 recessions, then non-residential investment could decline as much as 15% to 20% over the next four quarters, from the $501 billion seasonally adjusted annual rate (SAAR) in Q4 2007, to about $400 billion to $425 billion in Q4 2008. (see: CRE Bust: How Deep, How Fast?)

Most of that possible decline will probably come from three key categories: office buildings, multimerchandise shopping, and lodging.

Click on graph for larger image.

This graph shows the investment in these three categories over the last ten years (as a percent of GDP). Note: data from the BEA. The BEA started breaking out office and multimerchandise shopping in 1997.

Lodging and multimerchandise shopping saw the largest booms, while office space was less than the office boom in the late '90s. If all three categories decline to the recent cycle lows (as a percent of GDP), this will be a decline of about $60 billion in non-residential investment (SAAR). This breaks down to a $22 billion decline for office investment, $13 billion for multimerchandise shopping, and $25 billion for lodging.

Multimerchandise shopping tends to be closely associated with residential investment (developers add strip malls and shopping centers as new communities are built), so the bust in shopping center and strip mall investment is the most predictable (strip mall vacancy rates have risen sharply). Also, a sharp decline in lodging investment also seems very likely given the significantly tighter lending standards for hotels. And office investment will probably slump too based on the recent Grubb & Ellis forecast: Big rise seen in unoccupied office space

Other areas of non-residential structure investment might hold up: such as hospitals, manufacturing (because of exports), and power and mining investment. But overall the decline in non-residential structure investment will probably be significant.