by Calculated Risk on 3/30/2007 11:42:00 AM

From the Census Bureau: February 2007 Construction Spending at $1,170.8 Billion Annual Rate

The U.S. Census Bureau of the Department of Commerce announced today that construction spending during February 2007 was estimated at a seasonally adjusted annual rate of $1,170.8 billion, 0.3 percent above the revised January estimate of $1,167.7 billion. The February figure is 2.4 percent (±2.2%) below the February 2006 estimate of $1,199.9 billion.

During the first 2 months of this year, construction spending amounted to $159.9 billion, 2.4 percent (±2.2%) below the $163.8 billion for the same period in 2006.

Click on graph for larger image.

This graph shows the YoY change for the three major components of construction spending: Private Residential, Private Non-Residential, and Public.

While private residential spending has declined significantly, spending for both private non-residential and public construction have been strong. This will probably be one of the keys for the economy going forward: Will nonresidential construction spending follow residential "off the cliff" (the normal historical pattern)? Or will nonresidential spending stay strong. I'll revisit this discussion soon.

by Tanta on 3/30/2007 09:31:00 AM

I have described borrower-initiated fraud-for-housing loans as “self-underwritten.” The idea is that the borrower knows what the lender’s qualification standards are, knows he doesn’t meet those standards, and knows he cannot negotiate those standards away. His choices are, then, to accept the denial of credit, buy a cheaper house, or to lie or misrepresent the facts of his income, assets, employment, occupancy, and so on such that he appears to meet the standards. I call this “self-underwritten” because it rests on the borrower’s belief that he is a better judge of his prospects for carrying the loan successfully than the lender is; this belief allows him to justify his behavior as something other than criminal.

The fact that there have been so many “self-underwritten” loans in an environment of exceptionally lax standards for lender-underwritten loans is the key to puncturing this self-justification. It isn’t like we’ve had a credit crunch for years perpetuated by extremely risk-averse lenders, so that only perfect borrowers—or those who misrepresent themselves as perfect—can get a mortgage loan. Another way of putting this is that given how ugly so many of the lender-underwritten loans have been lately, there’s reason to think the self-underwritten ones are mostly butt-ugly. I take data on EPD rates for stated-income and zero-down loans, for instance, as some confirmation of this view.

Such a simple-minded perspective on things does, however, beg for additional complexity, and where else would you go for such additional analytic firepower than the New York Times? I offer you a third option: economist-underwritten loans.

Dr. Austan Goolsbee, a Real Economist™, presents the third way of understanding the issue in “’Irresponsible’ Mortgages Have Opened Doors to Many of the Excluded":

A study conducted by Kristopher Gerardi and Paul S. Willen from the Federal Reserve Bank of Boston and Harvey S. Rosen of Princeton, "Do Households Benefit from Financial Deregulation and Innovation? The Case of the Mortgage Market" (National Bureau of Economic Research Working Paper 12967), shows that the three decades from 1970 to 2000 witnessed an incredible flowering of new types of home loans. These innovations mainly served to give people power to make their own decisions about housing, and they ended up being quite sensible with their newfound access to capital.

These economists followed thousands of people over their lives and examined the evidence for whether mortgage markets have become more efficient over time. Lost in the current discussion about borrowers’ income levels in the subprime market is the fact that someone with a low income now but who stands to earn much more in the future would, in a perfect market, be able to borrow from a bank to buy a house. That is how economists view the efficiency of a capital market: people’s decisions unrestricted by the amount of money they have right now.

And this study shows that measured this way, the mortgage market has become more perfect, not more irresponsible. People tend to make good decisions about their own economic prospects. As Professor Rosen said in an interview, “Our findings suggest that people make sensible housing decisions in that the size of house they buy today relates to their future income, not just their current income and that the innovations in mortgages over 30 years gave many people the opportunity to own a home that they would not have otherwise had, just because they didn’t have enough assets in the bank at the moment they needed the house.”

The first time I read this I was, actually, so speechless that I could only respond with a quotation from our wise commenter mp: toad bones. Also, dog balls.

After thinking about it overnight, I have come to the conclusion that that’s still the wisest response, but you don’t get a good blog post out of simple incantations. In the “permanent income hypothesis” on which the economist-underwritten loan is based, the borrower’s belief that he will always be able to earn more money in the future, which justifies over-consumption of housing in the present into which he will grow, renders mortgage market “efficient” to the extent that it does away with such artificial constraints as down payment and DTI requirements—which are based on “the amount of money they have right now,” and adopts innovative standards depending on an individual borrower’s confidence in the amount of money he might have in a couple of years.

The evidence for this view is that economist-underwritten loans in the period 1970-2000 didn’t do so badly. Sure, a few of them went down, but it’s important to understand why:

Of course, basing loans on future earnings expectations is riskier than lending money to prime borrowers at 30-year fixed interest rates. That is why interest rates are higher for subprime borrowers and for big mortgages that require little money down. Sometimes the risks flop. Sometimes people even have to sell their properties because they cannot make the numbers work.

The traditional causes of foreclosure, even before there was subprime lending, were job loss, divorce and major medical expenses. And the national foreclosure data seem to suggest that these issues remain paramount. The latest numbers show that foreclosures have been concentrated not in places where real estate bubbles have supposedly been popping, but rather in places whose economies have stagnated — the hurricane-torn communities on the Gulf of Mexico and the industrial Midwest states like Ohio, Michigan and Indiana, where the domestic auto industry has suffered. These do not automatically point to subprime lending as the leading cause of foreclosure problems.

So in this period of happily performing economist-underwritten loans, there were some losers. Apparently the causes, job loss, divorce, and major medical expenses, which could be understood to mean situations in which current expenses are substantially greater than current income—have nothing to say about the idea that it is wise to take a loan that ignores one’s current income and expenses. Lenders, it appears, may consider future income; servicers, it appears, still keep refusing to accept aspirations rather than negotiable instruments to apply to a past-due balance.

Of course it’s not surprising that Goolsbee ignores the evidence of a house-price bubble, since there can apparently be no bubbles in perfect markets. Theories do that to you. But I don’t think theory can really explain the revolting disingenuousness at the end of his op-ed:

The Center for Responsible Lending estimated that in 2005, a majority of home loans to African-Americans and 40 percent of home loans to Hispanics were subprime loans. The existence and spread of subprime lending helps explain the drastic growth of homeownership for these same groups.

“Drastic”?

This is actually what CRL has to say on this topic:

According to the Fed report, even after adjusting for differences in the borrower characteristics contained in the HMDA data, African-American and Latino borrowers were more likely to receive higher-rate loans. Furthermore, a recent study released by CRL shows that disparities tend to persist even after additional adjustments were made for differences in credit scores, equity, and other risk factors not available in HMDA data. The Fed authors also adjust for originating lender. Though this adjustment reduces the disparities substantially, significant differences remain. . . .

The CRL study found that, even after controlling for legitimate risk factors, African-American and Latino borrowers were still more likely to receive higher-rate subprime loans than similarly-situated non-Latino white borrowers. With raw disparities in higher-rate loans between groups basically unchanged from 2004 to 2005, there is little reason to believe that legitimate risk factors would account for all of the disparity evident in the 2005 data.

In other words, CRL is suggesting that a pattern of finding subprime loans given to minority borrowers with similar credit, income, and equity profiles to non-Latino whites who get prime loans may imply a certain “inefficiency” in the mortgage market somewhere. For Goolsbee to use this data to buttress an unregulated free-for-all by claiming that it helps out the traditionally disadvantaged is, well, dishonest.

If you look hard at the data compiled by folks like CRL, you do have to face the problems inherent in the lender-underwritten mortgage market: when lenders are allowed to apply standards without public review, they certainly can end up applying those standards in a discriminatory fashion. When “reputable lenders” are allowed to exit entirely certain minority markets, leaving them to the tender embrace of the loan sharks, the “innovative” subprime market can quickly become mere predation. I have no beef with anyone who wants to see regulation of lenders to prevent these social evils.

However, if I had to choose between lender-underwritten and economist-underwritten loans? No contest.

by Calculated Risk on 3/29/2007 05:00:00 PM

Evan Koenig at the Dallas Fed writes: Vive la Différence (hat tip Mark Thoma)

"There are several disturbing similarities between the U.S. economy's recent behavior and its behavior in 2000–01, but also some reassuring differences."

See the above links for the details.

I'd like to expand the periods for Koenig's chart 4 and 5.

Click on graph for larger image.

The first graph (compare to Koenig's chart 4) shows Koenig's method for projecting changes in residential investment is reasonable. There is no question that residential investment will fall much further, and will decline in every quarter in 2007.


The second graph (compare to Koenig's graph 5) compares changes in new orders for durable goods with changes in manufacturing employment. Once again, Koenig's approach is reasonable.

Based on this analysis, Koenig is expecting the net loss of 53K manufacturing jobs per month over the next 6 months. This is in addition to the residential construction job losses that I'm projecting will be in 75K+ per month range.

No wonder Koenig concludes:

"The big question is whether the drags from housing and manufacturing will let up before weakness there begins spilling over to the rest of the economy."

by Calculated Risk on 3/29/2007 03:01:00 PM

From Indymac: Additional Credit Loss Analysis on Alt-A and Subprime Lending

“Based on an objective analysis of the facts, talk of the ‘subprime contagion’ spreading to the Alt-A sector of the mortgage market is, in our view, overblown. Cumulative mortgage industry Alt-A loan losses over the last five years are 1/17th the loan losses for subprime loans based on the FALP data. In addition, as of Dec. 31, 2006, the 30+day delinquency percentage for Alt-A loans in the mortgage industry was 5.0 percent as compared to 21.7 percent for subprime loans, suggesting that the differential in loan loss performance for Alt-A versus subprime will continue into the future.”
Michael W. Perry, Indymac’s Chairman and CEO, March 29, 2007

I don't think anyone was suggesting that Alt-A delinquency rates would be as high as subprime. So Mr. Perry is creating a strawman here.

The suggestion is that Alt-A would see a substantial increase in delinquency rates, and, as a result, the sector specific credit crunch (with tighter lending standards) would also impact Alt-A. This is already happening. Yesterday I posted the analysis from ResCap of their new lending standards on 2006 originations. ResCap estimated that their new guidelines would have eliminated 20% of Alt-A loans in 2006.

The subprime contagion is already here.

by Tanta on 3/29/2007 01:15:00 PM

The AP reports on delinquencies for consumer credit:

The American Bankers Association, in its quarterly survey of consumer loans, reported Thursday that late payments on home equity loans rose to 1.92 percent in the October-December period. That was up sharply from 1.79 percent in the prior quarter and the highest since the first quarter of 2006. . . .

A separate survey released earlier this month by the Mortgage Bankers Association showed a big jump in late mortgage payments in the final quarter of last year, news that caused stocks on Wall Street to swoon.

The American Bankers Association's survey, meanwhile, also showed that the delinquency rate on “indirect” auto loans, which are arranged through dealerships, shot up to 2.57 percent in the fourth quarter, the highest since the second quarter of 2001, when the economy was in a recession. Late payments on other auto loans, however, dipped slightly.

The survey also showed that the percentage of credit card payments past due was 4.56 percent in the fourth quarter, down slightly from 4.57 percent in the third quarter.

Late payments on boat loans rose, while delinquent payment on mobile homes went down.

by Calculated Risk on 3/29/2007 01:08:00 PM

From the San Diego Union: Home loan defaults skyrocket in county

Homeowners throughout San Diego County are defaulting on their loans and losing their properties to foreclosure at an increasingly rapid pace, a shattering event that nevertheless remains far less extensive here than the mortgage problems arising elsewhere in the nation.

In the first two months of the year, there were four times more default notices issued in the county than in the first two months of last year, while foreclosures tripled over the same period, according to an analysis of data provided by locally based DataQuick Information Systems.

by Tanta on 3/29/2007 06:57:00 AM

My post yesterday on Sandra Braunstein's remarks on subprime lending brought up the issue of disclosures to consumers. As you may or may not know, there is an existing Regulation Z that requires lenders to provide ARM loan program disclosures to consumers no later than the time of application for an ARM loan. These disclosures are supposed to help the borrower understand the mechanics of the loan. Reg Z does not require specific text, although it does supply example text. It does spell out in some detail what issues must be addressed in the disclosures. Lenders usually choose to have different disclosure documents for different ARM types; a lender using a very long "multipurpose" disclosure risks having its good faith challenged.

Reg Z has been around forever; that's important to remember given the current situation we have of borrowers (not to mention brokers) who clearly don't understand the terms of these loans. The item on the agenda of someone like Braunstein is whether and how Reg Z may need to be supplemented or superceded with new disclosure requirements. Politically speaking, lenders generally lobby against any such changes to the regs, if for no other reason than that they don't want to go to the additional effort and expense of changing all the disclosures they currently use. You may conclude that at least some lenders have other reasons to object, namely that--theoretically, at least--a clearer disclosure of loan terms might lead some borrowers to refuse to accept some of these products.

I will be the first to admit that it's a struggle for someone like me to judge the adequacy of a Reg Z disclosure. Obviously, I can and do review them for accuracy. But there's a big difference between a statement being true and a statement being comprehensible or useful to a non-expert. And experts can quickly lose their ability to appreciate the difficulties of non-experts. In my view, that's one of the biggest problems with regulatory disclosure requirements: the "final sign off" is from some lawyer, not from the sort of non-lawyer non-expert for whom the document is designed.

Furthermore, these documents need to be viewed not just in terms of their intended readers' expertise, but also in the context in which they are provided. Over the years so many different disclosures and documents have accumulated in the "application package" that you cannot really judge the accessibility of a single document without bearing in mind all the other documents for the borrower to absorb.

And, of course, that borrower has only some period of time in which to absorb things. Those who are all in favor of the 12-second approval using high-speed technology are always those who counsel the borrower to take his time, take the docs home over the weekend, consult a friend about them, do some web-surfing, there's no hurry here, make sure you understand this before you cough up a nonrefundable application fee . . . right? I'm not anti-technology or anti-AUS, but I continue to have huge problems with using it at point of application, and selling the speed as an advantage. The assumption that any borrower should be in that much of a hurry to borrow that much money at that complicated a set of loan terms makes my skin crawl.

But perhaps I underestimate the borrower? Well, you tell me. Here's Thornburg Mortgage's 1-Month LIBOR Option ARM disclosure, and here's the corresponding Note. I chose these not just because they're freely available on the web, but also because in my opinion Thornburg's disclosure is as well and clearly written as they tend to come. So this is not a test of the worst disclosures out there.

I would be truly interested in what you all make of these documents. If you read my UberNerd post on neg am, or you've had one of these before, or you're in some part of the mortgage business, please bear in mind that you have more background information than the intended reader of the disclosure. If it only makes sense to an expert, or even just to a more than usually informed non-expert, then it isn't doing what Reg Z intended it to do.

by Calculated Risk on 3/28/2007 05:30:00 PM

Here is the presentation from the ResCap / GMAC Financial Services 2007 Investor Forum.

ResCap evaulated the impact of the their new underwriting guidelines on their 2006 vintage portfolio (hat tip Brian). The new guidelines would have eliminated 58% of subprime loans, 20% of Alt-A loans and 35% of all 2nds. Wow!

by Tanta on 3/28/2007 12:00:00 PM

There was some outrage in the comments recently about Sandra Braunstein’s recent testimony on subprime lending. Braunstein is the Director of the Division of Consumer and Community Affairs for the Federal Reserve.

There are many things one can say about Braunstein’s testimony. I want to drive a few little conceptual wedges into her nice smooth characterization of the current mortgage market in general and subprime lending in particular.

It starts early, in the fourth paragraph of the written statement:

Today, the mortgage lending business has changed dramatically with the development of national markets for mortgages, technological changes, and the advent of securitization. The traditional book-and-hold model of mortgage lending has shifted to an originate-to-distribute model.

So far, so good, except something important seems to be missing from the “national, technological, and securitized” holy trinity of major market forces. Braunstein continues:

While commercial banks still have a significant role in the mortgage origination and distribution process, they are no longer the leading originators or holders of residential mortgages. Securitization has allowed many financial institutions to use increasingly sophisticated strategies to package and resell home mortgages to investors. This has resulted in increased competition and a wide variety of mortgage products and choices for consumers, in a market in which mortgage brokers and mortgage finance companies compete aggressively with traditional banks to offer new products to would-be homeowners. [My emphasis]

In what strange world are mortgage brokers “competitors” of banks? Let us remember what a broker is: not a lender. A broker brings a borrower and a lender together, and pockets a commission of some sort for so doing. The broker has no capital to lend. The “competitors” of brokers are direct retail lenders, meaning those lenders with loan officer employees who offer loans directly to consumers without requiring the services of brokers.

But for any bank with a wholesale origination model (which could, you know, be related to its “securitization” model in some fashion), brokers are not competitors, they’re sources of business. The wholesaler’s competition is other wholesalers, some of which are not depositories. Why am I making such a big deal about this? Because brokers have no “new products” to “offer” to “would-be homeowners.” They go out and find products offered by wholesalers to offer to would-be homeowners. Is this an unimportant distinction? I don’t think so.

Two paragraphs later, the lurking problem gets even worse:

One of the products of this new mortgage market is subprime lending. Subprime lending has grown rapidly in recent years. In 1994, fewer than 5 percent of mortgage originations were subprime, but by 2005 about 20 percent of new mortgage loans were subprime. The expanded access to subprime mortgage credit has helped fuel growth in homeownership.

It’s certainly questionable that subprime has really directly fueled homeownership to the extent Braunstein here assumes. As commenters have pointed out, there is evidence that subprime is still mostly a refinance business, not a purchase money business. To say, for instance, that record numbers of new home purchases have been made with subprime loans is not to say that most subprime loans are for purchases. Subprime on the whole, as far as I can see, has been fueling the MEW machine, and only indirectly the ownership machine. It is hard to build a MEW market with a bunch of renters.

But that aside, I am puzzled by the phrase “expanded access to subprime mortgage credit.” This assumes that “access” is a question of borrowers having access to creditors, rather than, perhaps, a question of creditors having access to borrowers. The whole idea of “predatory lending,” which is a subset of subprime lending, is that there are lenders who want to lend going after borrowers who may not have supplied the “demand” until someone fast-talked them into it. Even in the more “respectable” parts of the subprime and Alt-A business, I would argue, the "disintermediation" of “national markets, technology, and securitization,” which rely to a large extent on the “intermediation” of brokers, can function as much as supply creating demand than the other way around.

Those who wish to throw the Econ 101 textbooks at me will have to explain to me just how, exactly, borrower demand for loans they obviously do not understand, and that are not anywhere close to being in their best interest, gets created. Are we talking about a demand for credit or a demand for income-substitutes? And those who want to say that it’s all a matter of borrowers substituting short-term interest for long-term interest need to explain this EPD epidemic to me. Either those EPD loans were 100% fraudulent—borrowers who never intended to own the property or make the payments—or some of them were borrowers who never stood a chance of receiving even short-term benefit from the loan. I’m not sure which case is more comforting, but I surely can’t see here unambiguous evidence for pent-up demand that simmered for years until the “new mortgage market” Braunstein discusses suddenly offered the product everyone had been waiting for. Next thing you know, someone is going to tell me about the invention of advertising.

Quite honestly, after the first couple of paragraphs I just lost interest in anything else Braunstein had to say. It isn’t even, in my view, that her other testimony is wrong or worthless. It’s simply that I no longer care to hear anyone, and especially regulators, continue to talk about the mortgage “market” without talking about the mortgage business. I’m not interested in people talking about securitization and other forms of “disintermediation” until they address wholesale and correspondent originations and telemarketers and other kinds of “intermediation.” At some point these regulators are going to have to quit with the executive summaries and get into the UberNerd weeds with the rest of us, or else we’re going to keep getting this "disclosure" regulation (you can do pretty much anything as long as you slip in a disclosure that says so in the mice type) that works nicely in Econ 101 and doesn’t do diddly when the Borg adapts. Until they’re willing to recognize that a demand “bubble” exists, and that intermediation is not "competition," nothing they come up with in the name of “consumer access” is going to help much.

by Calculated Risk on 3/28/2007 10:57:00 AM

Back in February, Bernanke suggested the following risks:

The risks to this outlook are significant. To the downside, the ultimate extent of the housing market correction is difficult to forecast and may prove greater than we anticipate. Similarly, spillover effects from developments in the housing market onto consumer spending and employment in housing-related industries may be more pronounced than expected.
Chairman Bernanke, Feb 14, 2007

Now compare to Bernanke's testimony today:

This forecast is subject to a number of risks. To the downside, the correction in the housing market could turn out to be more severe than we currently expect, perhaps exacerbated by problems in the subprime sector. Moreover, we could yet see greater spillover from the weakness in housing to employment and consumer spending than has occurred thus far. The possibility that the recent weakness in business investment will persist is an additional downside risk.

Yet the forecast remains the same: "the economy appears likely to continue to expand at a moderate pace".