by Calculated Risk on 5/07/2008 06:59:00 PM

The Fed's Senior Loan Officer Opinion Survey is qualitative, not quantitative, and there has been some discussion on the predictive ability of the survey.

Luckily there was a paper written in 2000 that examined 'the value of the Senior Loan Officer Opinion Survey in predicting both lending and output'. See: Listening to Loan Officers: The Impact of Commercial Credit Standards on Lending and Output by New York Fed researchers Cara S. Lown, Donald P. Morgan, and Sonali Rohatgi.

From their conclusion:

Off and on since 1967, the Federal Reserve has surveyed loan officers at a small sample of large banks about their commercial credit standards. The idea behind the survey is that the availability of bank credit depends not just on interest rates, but on credit standards as well. Notwithstanding the small and changing sample, the checkered pattern of questions, and the sometimes curious responses of lenders, the reports are informative. The changes in standards that they report help to predict both commercial bank lending and GDP, even after controlling for past economic conditions and interest rates. Standards matter even in the 1990s, when capital markets were supposed to have eclipsed the role of banks in the economy. Changes in standards also help to predict narrower measures of business activity, where commercial credit availability from banks seems most crucial. The connection between bank standards and inventories is especially promising, because inventory investment is notoriously unpredictable and heavily bank dependent.

A shock to credit standards and its aftermath very much resemble a “credit crunch.” Loan officers tighten standards very sharply for a few quarters, but ease up only gradually: two to three years pass before standards are back to their initial level. Commercial loans at banks plummet immediately after the tightening in standards and continue to fall until lenders ease up. Output falls as well, and the federal funds rate, which we identify with the stance of monetary policy, is lowered. All in all, listening to loan officers tells us quite a lot.
emphasis added

Click on graph for larger image.

The authors provide these graphs that show the response of GDP, and in the amount of commercial and industrial loans, following a credit tightening shock. The impact on GDP is mostly within the first year, and peaks about 3 quarters after the shock.

The impact on lending lasts for a few years, and peaks about 2 years after the shock.

In the most recent tightening cycle (see graph here), there have been two tightening shocks: the first started in late 2006, and the 2nd was at the end of 2007. If the current cycle follows the normal pattern, the impact from the significant tightening at the end of 2007 should hit GDP later this year, and impact commercial loans for the next 2 to 3 years.

by Calculated Risk on 5/07/2008 06:27:00 PM

Last week I presented a chart of home improvement investment in real terms. I argued that home improvement spending could fall 15% to 20% in real terms based on previous home improvement slumps.

Here is the chart:

Click on graph for larger image.

The BEA reports that real spending on home improvement fell 2% in Q1 2008 (from Q4 2007), and has fallen about 4% in real terms from the peak. This is probably just the beginning of the home improvement slump; if this housing bust is similar to the early '80s or '90s, real home improvement investment will slump 15% to 20%.

Note: This graph shows real home improvement investment (2000 dollars) since 1959. Recessions are in light blue with the current recession "probable". (source: BEA)

Here is a 2nd graph in nominal terms comparing residential investment in single family structures (left scale), with investment in home improvement (right scale).

This graph shows how sharp the decline in single family structure investment has been - and how little impact (so far) the housing bust has had on home improvement spending. I don't expect the improvement investment slump to be anywhere near as severe as the single family structure collapse, but this does show there is potentially a significant downside.

by Anonymous on 5/07/2008 04:42:00 PM

Just yesterday Fannie Mae mentioned in its Q1 2008 Earnings Release that, as part of its "Keys to Recovery" initiatives, it would offer "a new refinancing option for up-to-date but 'underwater' borrowers with loans owned by Fannie Mae that will allow for refinancing up to 120 percent of a property's current value." That, so far, is all the information I have directly from Fannie Mae on this subject.

Unfortunately it got Dean Baker worked up. I respect Dr. Baker a great deal--he was calling the housing bubble long before it was cool--but I think he's got the wrong end of this:

This is a difficult move to justify from the standpoint of either taxpayers or homeowners.

The basic point is that homeowners will start out in these mortgages hugely underwater. Fannie’s policy means that it is prepared to lend $360,000 on a home that is appraised at $300,000. This gap implies that the homeowner can effectively put $60,000 in their pocket by turning the house back to the bank the day after the loan is issued. If the price drops another 10 percent in a year (prices are currently falling at an annual rate of more than 20 percent in the Case-Shiller 20 City Index), then this homeowner will be $90,000 underwater next May. If a seller would face 6 percent transactions costs, then in this example, walking away would provide a $106,000 premium compared with the option of a short sale.

This gap provides an enormous incentive for homeowners to default on their mortgage. Many homeowners will undoubtedly choose this option rather than make excessive mortgage payments on a house that is worth far less than the mortgage. A high default rate will of course lead to large losses for Fannie Mae and increase the likelihood that it will need a taxpayer bailout.

Fannie’s policy does have the effect of aiding banks that made bad mortgages. The new mortgages will allow these mortgages to be paid off. If matters were left to the market, the banks would almost certainly suffer large losses.

Baker is assuming that Fannie Mae will allow cash-out refinances in this program; although the mention in the earnings release doesn't specify that, I certainly assumed when I read it that we were talking about no-cash out refinances. (Fannie Mae's term for those, by the way, is "limited cash out" refinances. By this they mean that the loan balance can increase, but only to pay closing costs or pay off subordinate liens. That is what the rest of world means by "no cash out"--no cash disbursed to the borrower or to pay off non-mortgage debts.)

Fannie does make it clear that we are talking about Fannie Mae-owned loans. That is significant for two reasons. First, if the loans are upside down, it's already Fannie Mae's problem. To use Baker's example, if the borrower already owes $360,000 on a $300,000 home, the situation isn't made worse by refinancing it into a new loan with a lower payment. For Fannie to purchase a refinance of loans it already owns--presumably at a lower rate or payment, which improves the borrower's position and thus the strength of the loan--is not to take on risk you didn't already have. Second, of course, this isn't bailing out banks or anyone else.

That is why I assumed--and will confirm as soon as I find the Announcement from Fannie Mae--that these are no-cash-out refis. It would, indeed, worsen the risk of an existing underwater loan to let the borrower take more cash out.

Finally, it is specifically limited to performing loans. These are borrowers who are not, generally, eligible for a "workout" because they're not delinquent. But if they have hybrid ARMs coming up on a reset, or fixed rates that are higher than current market rates, this gives them the opportunity to get into a lower rate and payment while other costs--gas, anyone?--are taking more out of their pocketbooks. It isn't clear to me why this would increase any incentive to default, or increase Fannie Mae's losses if the borrowers did subsequently default. The loans are already underwater; even putting them 5% more underwater by rolling in closing costs seems to me, under the circumstances, to be less frightening than letting performing underwater ARMs get to a reset that will be hard for the borrower to bear. Not every borrower who is upside down will default, but every borrower with an unaffordable payment will in the current environment.

So there's a whole lot wrong with a whole lot of pressure to make the GSEs bail out the problems of the mortgage and housing markets, but so far this one sounds to me like Fannie Mae "bailing out" Fannie Mae, and, well, they ought to do that if it makes sense. Fannie Mae certainly does need to get a press release out clarifying the cash-out issue right away, before more nothingburgers get supersized.

by Calculated Risk on 5/07/2008 03:37:00 PM

From Bloomberg: U.S. Consumer Debt Rises More Than Forecast in March

Consumer credit increased by $15.3 billion for the month to $2.56 trillion, the biggest monthly rise since November, the Federal Reserve said today in Washington. In February, credit rose by $6.5 billion, previously reported as an increase of $5.2 billion.
...
``Consumers are strapped as incomes are not keeping up with inflation and this is leading them to rely increasingly on credit to see them through the worst housing downturn since the Great Depression,'' said Chris Rupkey, chief financial economist at Bank of Tokyo-Mitsubishi in New York. ``The days of extracting cash from one's home to spend on goods and services are long gone.''

Say goodbye to the Home ATM, and hello to credit cards.

by Calculated Risk on 5/07/2008 12:09:00 PM

Brett Arends at the WSJ asks: Is Housing Slump at a Bottom?

[The following chart] from Wellesley College Prof. Karl E. Case, one of the leading experts on the housing market in the country. And it suggests we may be at, or near, the bottom of the housing crash.

... new housing starts have at last slumped below the seemingly magical one million mark. That happened in March. Every time that has happened in the last 50 years, it proved to be the bottom of a recession.

"It is really remarkable how much where we are today looks like the bottom we've had in the last three cycles," Mr. Case says. "Every time we've gone below a million starts, the market has cleared at that moment."

First, I think any article discussing the housing "bottom" should start by defining what they mean by "bottom". Are they talking about starts? New home sales? Residential investment? Housing prices? Or some other metric?

Most people think of the bottom in terms of price, and in most housing busts, starts, residential investment, and new home sales all bottom long before housing prices bottom. The linked article seems to confuse a bottom for housing starts with a bottom for housing prices, and that is incorrect.

Second, we can write the supply side of the equation as:

Supply = new units added - units sold + existing units for sale. Looking at just housing starts provides only one portion of the equation (this leaves out rental units too - a substitute product).

Here is a graph of inventory (new and existing) for sale at year end (March for 2008):

Click on graph for larger image.

In the supply-demand equation for housing, prices will be under pressure until the total supply declines significantly. So even if housing starts are near a bottom, there will no quick recovery for starts, and prices will continue to decline, until the total inventory is reduced.

by Calculated Risk on 5/07/2008 10:31:00 AM

From the NAR: Soft Existing-Home Sales Expected Near-Term But to Rise Midsummer

The Pending Home Sales Index, a forward-looking indicator based on contracts signed in March, edged down 1.0 percent to 83.0 from a downwardly revised level of 83.8 in February, and was 20.1 percent lower than the March 2007 index of 103.9.

Existing home sales will probably decline all year - the opposite of the NAR forecast.

by Calculated Risk on 5/07/2008 09:34:00 AM

From Bloomberg: Vallejo, California City Officials Vote to File for Bankruptcy

Vallejo, California's city council voted to go into bankruptcy, saying the city doesn't have enough money to pay its bills ... because it ran out of money amid the worst housing slump in the U.S. in 26 years.

The city of 117,000 is facing ballooning labor costs and declining housing-related tax revenue that have left it near insolvency. ... The area has been one of the hardest hit in Northern California by the housing market slump. Home prices in Solano County, where the town resides, dropped 19 percent in January from the year before, according to DataQuick ...

by Calculated Risk on 5/07/2008 12:14:00 AM

From Bloomberg: Fed's Hoenig Says `Serious' Inflation Risk May Prompt Rate Hike

``There is a significant risk that higher inflation will become embedded in the economy and require significant monetary policy tightening to reduce it,'' Hoenig said in the prepared text of a speech in Denver. Consumers are gaining an ``inflation psychology to an extent that I have not seen since the 1970s and early 1980s.''
...
``A sharp slowdown in growth has put the economy at the brink of a recession while, at the same time, rising commodity prices have caused inflation pressures to rise considerably,'' Hoenig said to the Economic Club of Colorado.
...
The combination of slowing growth and inflation is ``troublesome,'' Hoenig said. Rising global commodity prices and higher prices of imported goods from China and other markets are pushing up prices.

``Some would dismiss these rising inflationary pressures as temporary,'' he said. ``I believe they are more serious.''

Here is Hoenig's speech.

Update: here is a graph of the University of Michigan Inflation Expectations from the St Louis Fed.



Expectations are not well anchored, and inflation can become embedded in the system, even with rising unemployment. This is probably a huge concern for the Fed.

by Calculated Risk on 5/06/2008 06:45:00 PM

Here are a couple of interviews with analysts that were wrong about housing (emphasis added) ...

From a Newsweek interview with ex-NAR economist David Lereah: It’s Going to Get Worse

"[I] just didn't realize the scope, the extent, the magnitude of the loose underwriting—not looking at incomes and wages, just providing so many mortgage loans based on [expected] future price appreciation rather than the creditworthiness of the borrower," Lereah says. "That got so out of hand, and none of us realized the magnitude of it until it was too late."

And from Jon Lansner at the O.C. Register: Insider Q&A learns how a housing expert goofed

Note: this is the "expert" that was mentioned on the front page of the O.C. Register in 2005 - see Lansner's Emblem of O.C.’s $600,000 home market, now a short sale)

Us: Where did you go wrong?

Walter: Along with almost everyone else, I didn’t recognize that the 2003 thru mid-2006 housing market boom was caused almost exclusively by the introduction (and pushing) of low-teaser-rate loans.
...
Us: What was it you didn’t foresee? Where did the demand go?

Walter: Shame on me! After 40 years of analyzing my way through four economic and housing market booms and busts, I knew that the fuel for every boom also eventually burns it up and causes it to go bust. I had my head in the sand and wasn’t paying attention.If I had been paying attention I would have known that a high percentage of those low-teaser-rate loans would go into foreclosure and bring the whole house of greed tumbling down. But I wasn’t paying attention.

Hoocoodanode?

by Calculated Risk on 5/06/2008 03:44:00 PM

Inventory numbers from the Census Bureau do not include cancellations - and although cancellation rates are still above normal, the rate has declined from the record levels of last year.

As examples, D.R. Horton reported a cancellation rate of 33% for the most recent quarter, down from 48% in the Fall of 2007. And Centex reported a cancellation rate of 29%, below the mid-30s of 2007.

Each builder has their own downpayment and cancellation policies. Some builders require much higher downpayments and therefore have lower cancellation rates. For Centex, a cancellation rate in the low 20s was normal during good times. For D.R. Horton, a cancellation rate below 20% was common during the boom.

Cancellation rates are important when analyzing the New Home data from the Census Bureau. What matters is the change in cancellation rates, not the absolute level. Falling cancellation rates mean the Census Bureau is probably underestimating sales, and underestimating the decline in inventory.

This graph shows my adjusted inventory estimate using cancellation rates from several of the large public builders. This suggests that inventory levels are now declining. Unfortunately this number can only be calculated on a quarterly basis, and not until several of the homebuilders file their quarterly reports with the SEC.

My inventory estimate is 560 thousand (as opposed the Census Bureau reported 468 thousand) as of the end of March. This is actually positive news, since the inventory level decreased by 42 thousand in Q1 by this method. This also suggests the Census Bureau understated sales slightly in Q1.

Also note that most condominiums are not included in new home inventory (or new home sales) from the Census Bureau. Areas with significant condominium developments probably have higher levels of inventory.