"If we're building 500,000 [houses a year], we are eating up that [excess housing] inventory. And the faster we eat up that inventory, the better off we are." -- Warren Buffett

Call it a silver lining: Homebuilding is going nowhere fast compared with household creation. And as Buffett points out, that's good news, since it means we're chewing through excess inventory. But it also shows the housing market is still a ways from recovery mode.  

September housing starts came in at an annualized 590,000, below expectations. Over the past year, here's how monthly starts have fared:

Month

Housing Starts

Annualized, Seasonally Adjusted

September 2008

822,000

October 2008

763,000

November 2008

655,000

December 2008

556,000

January 2009

488,000

February 2009

574,000

March 2009

521,000

April 2009

479,000

May 2009

551,000

June 2009

590,000

July 2009

593,000

August 2009

587,000

September 2009

590,000

Source: U.S. Census.

This is starting to show a fairly clear trend: Housing starts have probably bottomed, but are painfully slow to rebound. Shares of homebuilders like Beazer Homes (NYSE:BZH), Hovnanian (NYSE:HOV), and Pulte Homes (NYSE:PHM) all sold off on the latest numbers.

You can blame that slow rebound on a few things: First, foreclosures are still a disaster. Plenty of subprime borrowers have been purged from the market, but now a wave of option-ARM loans are resetting and recasting, pushing armies of "prime" borrowers into foreclosure. More foreclosures equals added supply. More supply depresses the need to build.

Second, the $8,000 first-time homebuyer credit is scheduled to end next month, meaning demand from first-time homeowners will fade, as happened when cash-for-clunkers ended. In anticipation, homebuilders are likely hunkering down. This should surprise no one: When you pay people to buy homes, they buy more. Stop paying them, and they buy less. No magic tricks here.

The big issue is how much more supply needs to be eaten away. This is an incredibly difficult question, because the "shadow inventory" is anyone's guess. Shadow inventory is composed mainly of two parts:

  • Bank-owned foreclosed homes not yet on the market, presumably because banks hope prices will rebound and don't want to recognize losses.
  • Homeowners who want to sell, but are holding off for the same reason as banks: They're waiting for prices to rebound. 

No one knows how large this shadow inventory is, only that it's somewhere between huge and catastrophic. Any rebound in prices could explode new inventory, pushing prices back down.

What we do know, however, are the on-market inventory numbers:

Month

Months of Supply, Unsold Homes

August 2008

10.6

September 2008

10.1

October 2008

10.2

November 2008

11.0

December 2008

9.4

January 2009

9.7

February 2009

9.7

March 2009

9.6

April 2009

10.1

May 2009

9.8

June 2009

9.4

July 2009

9.3

August 2009

8.5

Source: U.S. Census and HUD.

With 8.5 months' supply, the housing market is as stable as it's been in a while. Problem is, 8.5 months is still high, with about six months' supply historically being a healthy norm. Plus, shadow inventory could make the real supply of unsold homes astronomically higher than it appears. In one candid estimate, Stan Humphries, chief economist at Zillow.com, predicted that more realistic numbers could "take about four years to run through." Wonderful. That's like being kicked in the groin while you're on the gurney.   

All of which leads to the unsurprising conclusion: Housing is still in the dumps, and will remain face down in the mud for a while. Not only is this an uncomfortable truth for homeowners, but also housing-heavy banks like Wells Fargo (NYSE:WFC), Bank of America (NYSE:BAC), and even JPMorgan Chase (NYSE:JPM). A lot of investor sentiment is riding on the hope that the worst is behind us. And while it may be, that alone doesn't mean what lies ahead will be anything close to good.

For related Foolishness: