Inman

Too-good-to-be-true loans end an era

The 10-year T-note improved this week, back down from the spooky-healthy job-market surprise two weeks ago, mortgages about 6.25 percent. However, the 10-year is now smack in the middle of a three-week trading range with no reason to move until the economy declares itself in new data.

We’ll get housing news next week, and then nothing until the first week of May. I think the data suggests a steadily weakening economy, but that general perception is built into today’s rates. New claims for unemployment insurance are suspiciously high; March retail sales were stronger than expected, up .7 percent, but robbing from April; and industrial production rose, but going nowhere. New housing starts and permits were up a hair, but this is a false signal: the only way for builders to unload excess land is to build houses and give them away, further depressing local markets.

The key items: corporate capital spending is sinking, down to 1.5 percent growth from forecast 6-7 percent, and “free cash flow” has dipped to deep negative — by 5 percent of GDP, as stock-buyback tail-chasing nears its endgame. Core CPI was under control in March, up .1 percent, but overall up .6 percent, freezing the Fed no matter what weakness lies ahead.

The media have for the moment worn themselves out on the subprime meltdown but continue to miss the subtle changes affecting millions of buyers and owners.

Markets in trouble are distinguished by illiquidity and widening spreads.

Normally, the flow of mortgage paper is greased all the way from retail to the derivative mill. As that great machine throws gear teeth and intermittently gnashes to a halt, intermediaries up-stream are worried about getting stuck with paper — they would become “lenders,” which everyone knows is insanely dangerous.

Two bulletins to us in the last weeks: an excellent wholesaler/securitizer of Alt-A loans (junk, not trash), suddenly announced that the maximum lock-length would be 15 days, and only for loans fully approved. The second: a major warehouse bank (these provide credit to mortgage banks to fund loans at closing, the credit line paid down when the downstream securitizer buys/pays for the loan) announced that it would no longer warehouse second mortgages. At all. Then a week later reversed, but only for seconds underwritten by the downstream buyer.

These policy changes are on or over the edge of panic, in fear that the system will lock up altogether, in terminal absence of liquidity.

The system is already in terminal illiquidity, but progressive, from poorest product headed up to some yet-unknown middle zone, pricing spreads moving from daylight to tomb. Citibank, the definitive piggyback wholesaler, changed its whole underwriting matrix by 5 percent of loan-to-value; whatever it would do at 100 percent it will now approve only at 95 percent and so on down the scale.

Alt-A is completely misunderstood outside the trade. At its weak end, it is trash, but at its better end, quality is as good as any Fannie ever patted. Alt-A just means “not Fannie.” Example: We have a client moving to the area without a job, but with a half-million dollars in savings, a first-time buyer, a $450,000 condo with 25 percent down. Fine career in IT, will have a job as soon as he wants one; FICO score in the 800s.

This loan four months ago would have been priced perhaps a half-percent above today’s 6.25 percent Fannies. Today, the talking starts at 8 percent. This guy may decide to get a job rather quicker, but there are many other no-risk, “no-doc” candidates who won’t be buying soon — not until the true risk of default is understood. We’re years away from that resolution, and in the meantime, underwriting and pricing of marginal product is going to be herky-jerky, here-today, gone-tomorrow, back-again, but not the same.

Never ever again will it be as it was from 2001 to 2006.

Lou Barnes is a mortgage broker and nationally syndicated columnist based in Boulder, Colo. He can be reached at lbarnes@boulderwest.com.