Low delinquency rates prove mortgage lending’s too tight

Underwriting is not rocket science -- we had it right for 55 years

Two things this week: The aftershocks of last week’s jobs surprise, and astounding word that delinquencies on new mortgages are too low.

The weak jobs report knocked down long-term rates, and although everyone is suspicious of the report, the improvement has held. The all-defining 10-year T-note is trading near 2.85 percent, 30-year mortgages sliding down close to 4.5 percent.

Vault image via Shutterstock.
Vault image via Shutterstock.

After any wild-card report, everyone scours new data looking for confirmation or contradiction.

The poor “nonfarm payrolls” data last Friday came from the Bureau of Labor Statistics’ “establishment survey” of big business, in the last few years the healthiest segment of the economy.

This week we got the National Federation of Independent Business survey of small business: no change in overall conditions, just pacing back and forth in the same trench since late 2009. However, the NFIB employment component did improve ever-so-slightly at midyear last year and held in December. No acceleration, but no deterioration.

Industrial production has no particular correlation with jobs, but has risen in five straight months. This run may be another temporary inventory-overbuild, but GDP running over 3 percent may leak into hiring. We need rising wages even more than jobs.

Retail sales were OK in December, up from November, but year-over-year 2013 finished with a gain of 4.2 percent before subtraction for inflation, well off the 5.4 percent gain in 2012. The CPI core rate of inflation at 0.1 percent is stuck at less than half the Fed’s target.

Meet enough borrowers and review enough files, it's not hard to tell who can be trusted with a small down payment, and which condo project is safe to lend in."

Then, mortgages performing too well.

For the last dozen years we have quarreled about how to calibrate underwriting of mortgages. A hysterical mob still imagines bubbles around every corner, and another nasty bunch of thugs wants to shut down anything involving government, well-calibrated or not.

How to know, too tight, or too loose?

Abraham Lincoln was asked, how long should a man’s legs be? Abe replied, “Long enough to reach the ground.”

Proper measurement of underwriting begins at the end, with the default rate. Too many defaults means that underwriting was too easy (although a flow of really bad credit will defer the reckoning by pushing up prices — not today). Too few defaults is the signature of too tough.

From the back of the room, a hardhead: “We don’t want any defaults at all.”

But as you tighten toward no defaults, you reach a diminishing return, turning down acceptable credits in mania for perfection. Doing so will intercept recovery from recession (as it has),  impair any normal market, and make loans available only to the rich.

Specifics. In normal times 1945-2000, Fannie-style underwriting produced delinquency rates just above 4 percent and foreclosure rates near 0.7 percent. Today the overall U.S. delinquency rate is still 6.25 percent, some 4.5 million loans.

However, new loans made since 2010 have had the lowest rates of delinquency ever measured. One of the best metrics is early default: Performance in the first six months of a new loan reveals underwriting “misses” — those who had dreamed of owning a home but were not ready and quickly were late on a payment.

From a 2007 peak at 0.6 percent of Fannie loans late-paying at six months, now 0.05 percent. Even FHAs, are down to 0.09 percent, from more than 1 percent.

That squeeze was not easy to accomplish. Weak markets create high rates of delinquency, and housing is still in trouble in a lot of places.

The newest poster children are New York and New Jersey, with delinquency rates of 12.4 percent and 14.6 percent, joining Florida still at 14.6 percent, all three boosted by self-inflicted local legal mire and flinching from foreclosure.

California and Arizona at 5.6 percent and 5.5 percent are strong evidence that foreclosure medicine is worth the pain.

The Rust Belt — Illinois, Indiana, Ohio, Pennsylvania — is still 10 percent-plus, as is the old Confederacy and abandoned Nevada.

The previous conservator of Fannie and Freddie, Edward DeMarco, booby-trapped their operations. Ending bubble foolishness was easy; zealous trip-wiring has hurt all of us. His replacement, Mel Watt, knows defusing is necessary but needs political support.

Economic recovery aside, overtight credit punishes all young and recovering households.

Meet enough borrowers and review enough files, it’s not hard to tell who can be trusted with a small down payment, and which condo project is safe to lend in.

It’s not rocket science: We had it right for 55 years.

Lou Barnes is a mortgage broker based in Boulder, Colo. He can be reached at lbarnes@pmglending.com.


Related Articles

Comments