by Calculated Risk on 8/07/2007 10:45:00 AM

From Greg Ip and Jon E. Hilsenrath at the WSJ: How Credit Got So Easy
And Why It's Tightening. This article tries to explain how we got here. The authors discuss why the Fed set rates so low in 2001 and quotes Greenspan today:

"We tried in 2004 to move long-term rates higher in order to get mortgage interest rates up and take some of the fizz out of the housing market. But we failed."

The authors then discuss the changes in the mortgage business:

"All of us felt the suction from Wall Street. One day you would get an email saying, 'We will buy no-doc loans at 95% loan-to-value,' and an old-timer like me had never seen one," says Mr. Barnes [co-owner of a small Colorado mortgage bank]. "It wasn't long before the no-doc emails said 100%."
...
At first, delinquencies were surprisingly low. As a result, the credit ratings for bonds backed by the mortgages assumed a modest default rate. Standards for getting a mortgage fell. ...

The delinquency rate was a mirage: It was low mainly because home prices were rising so much that borrowers who fell behind could easily refinance. When home prices stopped rising in 2006, and fell in some regions, that game ended.

And they also touch on the easy credit for LBOs.

... lenders began to ease borrowing requirements. They agreed, for instance, to "covenant-lite debt," which dropped once-standard performance requirements, and "PIK-toggle" notes, which allowed borrowers to toggle interest payments on and off like a faucet.

Bankers began marketing debt deals for companies that, unlike Yellow Pages, didn't have comfortable cash flow. There was Chrysler, burning cash rather than producing it. And there was First Data Corp., whose post-takeover cash flow would barely cover interest payments and capital spending, according to Standard & Poor's LCD, a unit of S&P which tracks the high-yield market.

Last month, investors began to balk. Now many banks find themselves having committed to lend about $200 billion that they had intended to turn over to investors, but can't.

These are just a few short excerpts from a very good summary article.

by Anonymous on 8/07/2007 09:27:00 AM

I was going to post something on this yesterday, but it blew my mind so badly that I had to take refuge in Star Trek reruns until my faith in technology was restored. Sorry about the delay and everything.

Fitch sent out a press release describing the modifications it is making to the software it uses to rate mortgage-backed securities. I was, personally, a little shocked to discover that Fitch just now got around to working with state-level adjustments for RE market volatility. They are still working on incorporating MSA (Metropolitan Statistical Area) level factors. In other words, their analysis of loss frequency and severity due to rapid changes in a local real estate market just went from a blunt instrument to a somewhat less blunt instrument; the "pinpoint laser" thing is in the works.

But these two items are another order of magnitude entirely:

3) New Documentation Category: ResiLogic initially provided three categories for documentation, reflecting the categorization of loans in the model development sample. However, after evaluating lender programs and the recent performance of subprime and Alt-A mortgages, Fitch's analysts felt that greater differentiation was called for. This has resulted in the introduction of a new 'Low' documentation category, which indicates default risk which is greater than the existing 'Reduced' category but less than the existing 'None' category. Fitch will publish a research report the week of Aug. 6, 2007 describing the mapping between lenders documentation programs and the new categories.

4) Change in DTI Ratio: Back-end DTI ratio data is generally more available than front-end data in the data files provided by mortgage issuers. Therefore Fitch has modified ResiLogic to utilize back-end DTI. However, Fitch continues to be concerned by the prevalence of missing DTI data. Fitch will use a default assumption for missing subprime DTI of 50%.

OK, so we'll have to wait for the new report to find out how bad the doc type problem is, but I'm going to guess that Fitch has been lumping things like an AUS-approved loan with a single paystub and a verbal verification of employment in the same bucket with a "stated income" program for a W-2 borrower in which there is no documentation of income whatsoever. That would tend to improve the performance history of true "stated income" loans.

I'm hoping I'm wrong about that, but the following item gives me zero confidence in the matter. Long-time readers of this site are pretty familiar with "DTI," which people like me use to refer to the borrower's total monthly obligations, including but not limited to the mortgage payment, divided by the borrower's gross monthly income. The reason most of you all have never heard of "HTI" is that it's so damned obsolete in most quarters that not even your Tanta cares about it.

Many moons ago, lenders used two ratios to qualify borrowers for mortgages: the so-called "front ratio" or HTI, which included just the mortgage payment, and the "back ratio" or DTI, which included total debt. After a number of years, industry consensus became that HTI really isn't very strongly predictive of default as a separate measure, and so nearly every lender dispensed with a benchmark for HTI. Lately we've obviously gone a little nuts with the DTIs we will allow, but the point is that every credit model out there--except, apparently, Fitch's--used DTI as the sole measure in the model, or at most the primary determinant of capacity evaluation with HTI being considerably less weighted in the analysis.

Fitch is saying one of two things, and I'm still not sure which: either they really did use HTI instead of DTI in their models, which means that the weighted average DTI information they've been publishing is junk, or they've really been using DTI all along but didn't realize it, because they didn't understand the difference. I'm inclined to the latter view only because I believe that DTI is the only data they're getting from issuers. I do not know how anyone could look at DTIs of 50% and assume that other debt had to be added to that to get the borrower's total picture and still rate these deals with a straight face.

Of course Fitch is saying that this factor is of only modest impact. Well, duh. My own view of the matter has always been that DTI is a critical measure of loan quality, but not if you don't verify the "I" part. If you let borrowers and originators just back into a "stated income" that produces the right ratio, it doesn't exactly matter what that number is, and it's better to ignore it entirely in your evaluation of credit than to use it, because you end up giving "credit" to loans that have low but entirely bogus DTIs. I don't exactly have huge confidence in Fitch's handling of this problem now that I know that they've also had some trouble identifying doc types with any precision.

This is not a little bitty deal, a quibbling over geeky details of underwriting archana. This is big bad-ass deal. Fitch just came out and admitted that "ResiLogic" is working with corrupt or missing data, that Fitch's analysts don't know how to "crack" a mortgage tape (that is, how to map an originator-specific data field into a standard one that can compare across deals), and that there is no reason for any of us to believe them when they say that certain loan characteristics are more or less predictive of default. Some of you may have suspected this before today. But I'll admit that I didn't think they were making mistakes at this elementary a level. It's hard to get the rocket science right. Defining the DTI field? Lord love a duck.

In other news, Fitch is having a conference call today at 11:00 ET to discuss this matter. I may attempt to listen in, although I'm telling you now I'm not sure I can stand any more of this.

by Anonymous on 8/07/2007 07:44:00 AM

Bank of America is with me.

Bloomberg:

Aug. 7 (Bloomberg) -- A California Bible school sued its bankers on July 31 for fraud in connection with swaps, and you have to wonder how many other municipalities can say what the school said in its lawsuit.

We did everything they told us to.

They negotiated all the terms on our behalf.

They assured us they were representing our best interests.

We relied on them to get everything right.

We didn't know what we were doing.

Sounds like a bunch of subprime deadbeats to me.*

*Warning: You are consuming irony. This early in the day, too.

by Calculated Risk on 8/06/2007 09:17:00 PM

From the WSJ: Mortgage Fears Drive Up Rates On Jumbo Loans

Turmoil in the U.S. home-mortgage market is starting to pinch even buyers of high-end homes with good credit records ...
...
Lenders ... now are jacking up rates on jumbo mortgages for prime borrowers. These mortgages exceed the $417,000 limit for loans eligible for purchase and guarantee by Fannie and Freddie. They account for about 16% of the total mortgage market, according to Inside Mortgage Finance, a trade publication, and are especially prevalent in California, New Jersey, New York City, Washington, D.C., and other locales with high home costs.

Lenders were charging an average 7.34% for prime 30-year fixed-rate jumbo loans yesterday, according to a survey by financial publisher HSH Associates. That is up from an average of about 7.1% last week and 6.5% in mid-May.

by Calculated Risk on 8/06/2007 08:42:00 PM

The cover story of BusinessWeek is the "Bonfire of the Builders".

And that brings up the The cover curse

We've taken grief for decades about the cover curse at BW. When you see a glowing cover story on a company, the common wisdom says, sell. Now a trio of financial analysts has carried out a study of this cover curse at BW, Forbes and Fortune. (ex Andreesen). And they conclude, I'm sorry to see, that there's something to the curse. Basically, a positive or negative cover story marks the end of a company's "extreme" behavior, either good or bad. After the cover story, they tend to regress, in their corporate way, toward the mean.

Does the "cover curse" mean the worst is over for the homebuilders?

As a reminder, this was the cover of Time Magazine in June 2005, marking the top of the housing boom.

The magazines tend to put "extreme" stories on their covers. But just because the story is extreme, doesn't mean it won't stay extreme.

And right now housing fundamentals of supply and demand suggests that the bottom for housing is still well into the future.

So I doubt this BusinessWeek cover will mark the bottom for housing.

by Anonymous on 8/06/2007 05:41:00 PM

Luminent, a mortgage REIT, halted trading this morning after a 30% drop in the share price on "news pending." Here's the release:

SAN FRANCISCO, CA., August 6, 2007 — Luminent Mortgage Capital, Inc. (NYSE: LUM) announced today that, since August 3, 2007, the mortgage industry, and the financing methods that the mortgage industry relies upon, have deteriorated significantly and in an unprecedented fashion. Effectively, the secondary market for mortgage loans and mortgage-backed securities has seized-up. As a result, Luminent is simultaneously experiencing a significant increase in margin calls on its highest quality assets and a decrease on the financing advance rates provided by its lenders.
In a Board of Directors meeting today, Luminent’s Board unanimously voted to take the following actions:
• The Board of Directors suspended payment of Luminent’s second quarter cash dividend of 32 cents per share on Luminent’s common stock.

• The Board of Directors extended the maturity of the outstanding commercial paper issued by Luminent Star Funding Trust I, a special purpose subsidiary of Luminent, by 110 days.

• The Board of Directors cancelled Luminent’s second quarter 2007 earnings release conference call, scheduled for Thursday, August 9, 2007, at 10:00 a.m. PDT, to discuss its second quarter of 2007 results of operations.

• The Board of Directors delayed the filing of Luminent’s quarterly report on form 10-Q for the second quarter of 2007. Luminent’s second quarter of 2007 unaudited condensed financial information is attached to this press release. Luminent’s independent registered public accounting firm has not completed a review of the financial information for the three and six months ended June 30, 2007.

• The Board of Directors authorized Luminent’s senior management to inform the New York Stock Exchange of these unfolding events and, as a result, trading was halted in Luminent’s common stock.

The Board of Directors currently is considering the full range of strategic alternatives to enhance Luminent’s liquidity and preserve shareholder value during this period of market volatility.

LUM issued a PR on July 30 "reaffirming" its dividend after the AHM fiasco.

by Calculated Risk on 8/06/2007 03:07:00 PM

From MarketWatch: National City Home Equity stops taking loan applications

National City Home Equity, a unit of National City Corp. stopped taking loan applications on Monday, according to an e-mail that was sent to mortgage brokers. MarketWatch obtained a copy of the e-mail. National City Home Equity also said that decisions on loans that are currently in its pipeline would be made later today.

I've also heard that Aegis Mortgage suspended all loan originations on Monday. From crispy&cole in the comments:

Aegis unable to fund mortgage loans already in pipeline 2:12 PM ET, Aug 06, 2007
Aegis Mortgage suspends all loan originations 2:12 PM ET, Aug 06, 2007

by Anonymous on 8/06/2007 01:04:00 PM

Our friends at the NYT have gotten straight on the difference between a CDO and an MBS. I am very happy. One of these days, the single-class pass-through will come back into fashion and we'll discover that not all MBS are tranched, but that's small beer. We now have a firm basis on which to build insights.

Anyway, here's a better picture than you get in any of my crappy posts.

by Anonymous on 8/06/2007 08:47:00 AM

The Washington Post discusses the death of 100% financing. There is this gem:

National City Home Equity, a division of National City Bank, one of the nation's big home lenders, stopped funding some types of zero-down loans this month, said Ken Carter, the division's executive vice president.

"When home prices were appreciating and interest rates were declining, that product made sense," Carter said. "Today, we're on the opposite side of that coin, and it's not prudent to be stretching."

This is why we lenders cannot 'fess up, acknowledge our part in the mess, and get down to problem-solving. Two years ago, being "prudent" got you laughed out of meetings, in which you were told to take your carping negative hand-wringing somewhere else, because if a deal "made sense" to a speculator, it made sense to everybody. Now that we're scraping the charred husks of the speculators out of the EasyBake Oven, we can use the word "prudent" again.

A kind reader sent me this link to Option One's most recent fit of prudence. Among other things--"no Florida condos" probably got everyone's attention--there was this little change to the underwriting guidelines:

Funds available for asset verification (down payment/reserves) from a 401K will be limited to 70% of vested balance

I honestly don't know what Option One's old rule on 401(k)s was--I don't actually want to think about it--but it appears that heretofore Option One was allowing borrowers to pretend like they could liquidate their retirement accounts to cover their mortgage payments without paying any tax penalty. Since that idea probably had sense-making issues even at the time, I'm going to guess that the idea was that the balance would grow so fast that by the time anyone might need to liquidate (at 70%), Mr. Market would have made up for the haircut. If it made sense to assume that house prices would appreciate forever and Fed Funds would stay at a buck-twenty five, it surely wasn't so hard to imagine 20% annual returns on everybody's 401(k) forevermore.

by Calculated Risk on 8/06/2007 12:54:00 AM

Just for fun - as I write this the S&P500 futures are flat - here is a somewhat younger Bill Griffith, along with Ed Hart and Diana Koricke reporting on Black Monday on FNN: