January 27, 2008

Spinning the Stats (Nov'07)

(c) View from Silicon Valley, 2007.  All rights reserved.



The paper published this months' (Nov'07) statistics last Sunday.

As usual, the data was headlined by touts and spin, but also some useful data points we feel demand further attention.  For example:

"The value of newly approved residential construction in Santa Clara County sank to $69.4 million in November, less than half of what it was in November, 2006 and the lowest level in six years..."

This paragraph also included a tout on non-residential construction.  Unfortunately, it will be another few months (or maybe more?) before the bubble in office and retail construction is acknowledged.

For example, just in Sunnyvale the bridge to nowhere is going up across Highway 237.  A mile or two south, Sunnyvale's new downtown project is spewing dust and dirt into the air.  The city's press release touting this project apparently saw no irony in their simultaneous claim of "a green and sustainable community."

Just a bit north of the bridge to nowhere is "the largest building project in the South Bay."  Moffett Towers will add 1.7M square feet of class “A” office space.  Unfortunately, the city doesn't talk much about plans to get all the extra vehicles in or out of this office complex through the notorious US-101 / Highway237 / Mathilda /Moffett Park Drive disaster, err --five-way intersection. 

The Juniper and Yahoo traffic already clogging that area apparently wasn't enough to deter a developer from assuming he could flip the buildings to a pension fund and move his own business away from that mess.  Or maybe the anticipated Yahoo layoffs will be even bigger than rumors claim.

Sorry, we got a little carried away.  Back to the stats:

"The amount of office and research and development space available for leasing fell to 36M square feet, as of December 1, the lowest since October, 2001."

Fortunately, we have our seventh grade algebra and a little history data against which we can show it off.  The table accompanying the paper's tout says 16.7% of total office and R&D space is available. (For now, we'll skip the argument over the definition of "available.")  Divide the reported 36,030,904 sqaure feet by 16.7% and you deduce the total amount of space available (TAM) is 215,753.916.1 square feet.

We started tracking this figure back in October, 2002.  At that time, the report was 21.7% of space available, which was reported as 47.0M square feet, implying a 215,589,861.7 TAM.

Since October, 2002, the same table reports a total of $5,356M , or $5.4B with a B, in new non-residential construction was approved.  It is not clear if this includes government projects such San Jose's new city hall or Sunnyvale's new downtown complex, so the actual figure might be even higher.

Even so, $5.4B is new non-residential construction over a period of five-plus years yielded a net change in the office/R&D TAM of only 164,044.4 square feet, or  0.08% or 0.000761.  (Just for fun, these figures could be used to imply builders paid $32,650 per square foot??  Even divided by 10, this figure is ridiculous. Divide by 50 and the cost is only ~5% more than Brocade paid for their new buildings at the height of the boom.)

Dang it, we wandered off-topic again.  We haven't even touched on our favorite topic of all:

"The number of jobs in Silicon Valley rose to 922,100 in November, the highest since July, 2002.  The number of residents in the region with jobs rose to it highest level since July (2007?!?), while the percentage of those available to work who can't find a job inched back up to 5 percent from 4.9 percent the month before."

We believe this mis-direction needs to be pointed out as often as it is touted. 

Right away we notice this table usually touts Santa Clara County "residents employed."  Today they switched to "number of jobs" and compounded it by saying "in Silicon Valley."  The verbiage says 922.1K but the table below it read 922.9K.

For the moment, let's accept the argument on their terms.  Let's also rise above the snark and assume "in Silicon Valley" and 922.1K were just more clerical errors.

A total of 922.9K jobs is indeed the highest figure in 2007.  Howeer, June, 2007 was 921.2, implying 340 new jobs per month in the last five months?

We strongly prefer Santa Clara County "residents employed."  After all, we're a resident. 

Perhaps more relevant to our readers, it seems logical to believe housing prices depend on residents employed.  If folks outside the county become the primary drivers of employment growth, a prime pillar of price growth for local houses is undermined.

"Residents employed" reported for past Novembers includes:
2007:   833.3
2006:   817.1
2004:   819.0
2002:   852.7
2000: 1,057.3
1998:   961.3
1996:   899.0
1995:   857.1
1994:   809.7
1992:   798.1

Yes, 2007 was up from 2006.  It was below Novembers in 2002 and 2000 but those months are widely dismissed as "the bubble years" and therefore not relevant to current conditions.

We insist current residents employed below 2002 and 2000 is actually relevant.  After all, house prices are way above 2002 and 2000 prices.  House pirces around here are supposedly reasonable due to all the high-paying jobs.  Except there are 200K+ fewer of those jobs than in 2000, when house prices were quite a bit cheaper.

Arrgh!  Back to the topic at hand:
Current jobs are up, marginally, from 2006 but only re-visiting levels seen in 1995 and 1996.   In what rational universe is no growth in 12 years is defined as "strength?"

The paper goes on to tout hotels "managed to hold on to a 9.1 percent increase in the average room rate year over year."  Rooms sold were down -19.7% in one month and -6.3% year-over-year (y-o-y) but prices managed to increase enough to offset this drop.  Does that sound like strength to you?

Likewise, house prices were up 9.5% y-o-y even though sales were down -9.1% in one month and -41.7%(!!) y-o-y.   We'll stand back and let you readers fill out the list of issues we all have with these numbers.

In the end, the MN150 was reported down -6.7% in one month.

Conclusion:
Despite the spin pointed out above, the paper seemed to admit there are signs of weakness embedded in these statistics.  Or maybe we're the only ones who see it that way?

 ...to see all the stats, please visit:  http://www.viewfromsiliconvalley.com/id66.html