by Calculated Risk on 2/29/2008 03:47:00 PM

Last week, the NY Times mentioned the recent estimate from Dr. Mark Zandi, chief economist at Economy.com, of 8.8 million homeowers currently with zero or negative equity.

Here is an overview of the calculation from Zandi's Plan to Take Bolder Action on Mortgages (excerpted with permission from Mark Zandi):

By the end of March, an estimated 8.8 million homeowners, equal to 10.3% of the single-family housing stock, are expected to have zero or negative equity. This calculation is based on estimates of homeowner equity across more than 40,000 Zip codes nationwide. The amount of first and second mortgage debt outstanding, including both home equity lines of credit and closed-end second loans, is derived from credit file data. The estimates of housing values are based on repeat-sale house price indices from Fiserv Case Shiller, which are benchmarked to house prices available from the 2000 Census. In those Zip codes where repeat sales price data are not available, Moody’s Economy.com's estimates of county house prices are used. The current distribution of homeowner equity across Zip codes also highlights the risks posed by further price declines. If prices fall 20% from their peak to their expected trough next spring, some 13.8 million households will be under water by then, equal to 15.9% of the stock.

The following graph shows my estimates from last December of the number of homeowners with zero or negative equity based on several different price declines. This was based on data from First American.

At the end of 2006, there were approximately 3.5 million U.S. homeowners with no or negative equity. (approximately 7% of the 51 million household with mortgages).

By the end of 2007, I calculated the number would have risen to about 5.6 million. However, prices fell further than I expected in 2008, and this actually is very close to Zandi's estimate for the end of March 2008.

With a 20% price decline (peak to trough), I calculated 13.6 million homeowners with zero or negative equity; Zandi's updated numbers put that at 13.8 million.

Zandi didn't provide an estimate for a 30% peak to trough price decline, but my 20 million estimate is probably pretty close. Note: I calculated the percent of homeowners with mortgages with no or negative equity; Zandi calculated the percent of the total single-family housing stock.