D

Sensationalizing Math: Could Underwater Mortgage Reports Be Flawed?

This is going to be a long one, and I’ll be late for lunch, as usual. I just got off the phone with Gwen Moritz, Editor of Arkansas Business in Little Rock. Gwen took a look at a scintillating report in Saturday’s Arkansas Democrat-Gazette that said 25% of Arkansas mortgages are “under water”. Steve Brown ran the same report Friday and I parroted, our number in Dallas being 30% underwater. Every media outlet in America ran the piece since it was wisely sent out localized and sensational. The study was made by a California company called First American CoreLogic. Gwen read the story and it nagged her — 25% of Arkansas mortgages under water? How can that be, she asked.

Now when I read the story, I too had my doubts, for this reason: if the reports CoreLogic used included HELOCS — home equity lines of credit — why would 30% of Dallas mortgages be underwater? In Texas we are limited to 80% loan to value home equity ratios. That would mean that if someone maxed out their HELOC in Texas, they’d have to have lost 20% of the value of their BRAND NEW 100% FINANCED home and then some to be underwater. Or maybe they’d be flat. Whatever, I thought, as I usually do, who am I to question the brainpower and computers of these geniuses in California? God, we are all so bad!

Well, Gwen questioned them. She did what I should have done and picked up the phone and said, you know, this just doesn’t look right. Artkansas never had a housing bubble. Neither did Texas. She did some research on CoreLogic’s methodology and found out — hold the presses!!! — the company’s numbers may be biased on the HIGH SIDE . In other words, giving us facts that are worse than they really are.

Most of the mortgages used in the study are less than six years old — really? Did the whole world re-finance in the last six years? When she called CoreLogic, Gwen got them to admit actually 85% are less than six years old.

The study assumes that the borrower owes the balance of the mortgage when they took out the loan — in other words, it does not take into account monthly mortgage payments made against that balance. Why? Because of the way CoreLogic gets it’s information — it aggregates information from lenders, matching total mortgaged amounts in geographic chunks, like how much was loaned in a specific area or zip code. Then the computers whiz out numbers against property values. (Obviously I over-simplify.) They get this proprietary info from the lenders and use original loan amounts from public data records. But they base it on the original loan amount — they never get to come into our file cabinets and see how much money we’ve actually paid down on that mortgage or home equity LOC.

See where I’m going with this?

Gwen also found out the study includes the HELOCs but again, uses only the maximum credit figures since that is all the information that is available. So it looks like every one’s HELOC is at the max. Don’t know about you, but I have to pay down on mine every month and in fact, ours is almost paid off.

So if you are using figures that assume homeowners still owe every penny in debt they ever took out on a loan, figures that give no credit to what homeowners have in fact paid, then Gwen asks, maybe that study could over-estimate the number of folks whose mortgages are underwater.

“It’s biased to the high side,” says Gwen. “We need to be a little more skeptical of these studies.”

No kidding.

(Note: Thanks to Gwen Moritz for not only questioning the CoreLogic report but notifying the Alliance of Area Business Publishers of her query via email, which Glenn Hunter, editor of award-winning DCEO, then forwarded to me.)