Private Mortgage Insurance v. Second Mortgages

One of the important developments that made the collapse of the housing price bubble relevant to banks and the national and international economy was the rise in low equity loans.

Many new homeowners had a down payment of 0% or 3% or 5%, rather than the traditional 20% of the purchase price. Home equity loans to finance other consumption were common, and were structured either as a refinancing of the entire mortgage with cash out, or as a second mortgage at a higher interest rate.

Banks attempted to cope with these risks in basically two ways. If there was a single loan with low equity, banks generally required private mortgage insurance. Alternately, two loans were made, one conventional with 20% equity, and the other a second mortgage (or even a third or fourth mortgage) secured by the remaining equity.

The tax code encouraged structuring the deal as a second mortgage rather than as private mortgage insurance, because, until recently, interest on a second mortgage was deductible, while private mortgage insurance was not.  (The mortgage insurance deduction has IIRC now expired.)

Second mortgages, which have higher interest rates than first mortgages, have the virtue for the lender having interest payments that automatically fall as the principal amount on the second mortgage declines, unlike private mortgage insurance, which is typically a constant premium until it is terminated, even though the risk declines over time. But, depending upon the amortization period of the second loan, one could be paying some debt at the higher second loan interest rate long after the amount outstanding on the two loans combined was less than 80% of the original purchase price. In contrast, laws enacted in the late 1990s to make it easier to cancel private mortgage insurance policies to be cancelled once a buyer has paid enough of the principal of a mortgage to have an outstanding loan balance of 80% or less of the purchase price.

Footnote: Bad Loans Go Bad Early

Loans that go bad are overwhelmingly new ones. A New York Times review of metro New York City foreclosures in 2007 found that:

[O]n Long Island, the average age of a loan in foreclosure last month was a little more than two years, according to Long Island Profiles. In January 2004, the average length of time it took borrowers struggling with loans was five years before losing their property through foreclosure.

Early foreclosures are particularly common in subprime loans, but quite early in most mortgages. A January 2005 study of subprime foreclosures, citing a 2001 study by HUD of the same subject found that:

[L]oans by subprime lenders in Baltimore were on average 1.8 years old at the start of foreclosure, compared to 3.2 years for prime and FHA loans.

Similarly, a December 2003 study of foreclosures in a set of 12,000 new mortgages by a federal government agency found that: "the average age of the loans at foreclosure in the data set was approximately 30 months or 2½ years."

The stereotypical mortgage has a thirty year term, very few have less than a fifteen year term, and increasingly before the subprime mortgage collapse, one was seeing even longer amortizations, with forty year terms or "interest only" payments for extended time periods.

Fifteen year loan terms tend not to be subprime.