The Non-guarantee

Fannie and Freddie are required by their government charters to inform investors that their securities are not backed by the government. Investors, however, have generally treated their debts and other obligations as implicitly government-guaranteed, allowing them to borrow at rates only slightly higher than Treasury bond yields.

In recent years, the Treasury Department and the Federal Reserve have sought to undermine that confidence. For example, in a speech in 2004, then-Secretary John Snow said:

"GSEs are able to borrow at rates that are lower than their other financial competitors, other financial institutions, because they are perceived by many in the marketplace to have a relationship with the government, they are perceived to have a government guarantee."...

"We don't believe in a 'too big to fail' doctrine, but the reality is that the market treats the paper as if the government is backing it. We strongly resist that notion.

"You know that there is that perception. And it's not a healthy perception and we need to disabuse people of that perception. Investments in Fannie and Freddie are uninsured investments."

As recently as March of this year, Fed Chairman Bernanke said, "Only if GSE debt holders are persuaded that the failure of a GSE will subject them to losses will they have an incentive to exert market discipline."

No protection may be available for the GSEs' shareholders, but it will probably now be demonstrated that their creditors were right to ignore the Treasury and Fed trash talk. While there has been a debate about whether Bear Stearns should have been allowed to default, the danger to the financial system and the economy is vastly greater in this instance. Indeed, the interest rate premium on GSE debt over Treasury bonds actually narrowed last week, when it was recognized how unthinkable it would be to not honor the GSE obligations.

Risks

The GSEs provide funds for mortgage lending in two ways. They buy mortgages or mortgage-backed securities (MBS) to hold as portfolio investments, financed by issuing corporate bonds, and they provide guarantees against default for MBS that are sold to other investors (or to themselves). Much of the criticism of the GSEs by the Fed, Treasury, and others related to their increased reliance on portfolio purchases rather than guarantees. It was argued, with considerable justification, that holding mortgage assets in portfolio exposed the GSEs to greater risks. It was further argued that they did so to increase profits without lowering borrowing costs for home buyers. In the course of those arguments it was invariably stated that interest rate risk (from portfolio activity) was severe, while credit risk (present with both purchases and guarantees) was manageable and "not a serious problem." The current difficulties, however, are primarily the result of credit risk.

Some of those credit risk problems reflect regulatory requirements. Fannie and Freddie are regulated by the Office of Federal Housing Enterprise Oversight (OFHEO) with regard to safety and soundness (although that regulatory responsibility may be shifted to a new agency under pending legislation). They are also subject to regulation by HUD under mandates to serve low- and moderate-income households and neighborhoods. As originators and investors with more energy than brains expanded (subprime) lending to those borrowers and neighborhoods, it was difficult for Fannie and Freddie to increase their shares. The GSEs didn't want to buy or guarantee subprime loans, correctly perceiving them to be insanely risky. Instead they purchased securities created by subprime lenders, taking only the supposedly-safe tranches. Those portfolio purchases were counted toward their obligations to lend to lower-income home buyers, but are now part of the write-downs.

Raising Capital

On February 13, President Bush signed the Economic Stimulus Act of 2008, which included a temporary increase in the maximum loan eligible for sale to Fannie and Freddie. With that new authority, and with other sources of mortgage credit sidelined, the housing industry, Congress, and the Administration expected Fannie and Freddie to expand their activities. To do so, and to offset losses from write-downs, they would have to increase their capital bases. The regulatory capital requirements to support MBS guarantees are less than those to support portfolio assets, so they can facilitate more mortgage lending by shifting to that line of business, but they still require more capital to continue to fill the void left by defunct or hobbled private conduits.

Freddie Mac CEO Richard Syron, however, was under the impression that he was running a private-sector firm on behalf of shareholders. On March 12, he told a meeting of Wall Street analysts that Freddie didn't plan to raise additional capital and thereby dilute the value of the stock, which had already fallen by more than half. That was not well-received in Washington. A week later, Syron joined CEO Mudd of Fannie Mae and OFHEO Director Lockhart to announce a planned expansion.

OFHEO increased the amount that the GSE could invest, relative to their capital bases, in part based on the promises to raise additional capital. Subsequently, Fannie--the more politically-attuned sibling--sold some common and preferred shares (less than OFHEO wanted), but Freddie only has said it will raise capital "in the near future."

The quasi-public, quasi-private housing GSEs had come to be regarded by many as arrogant, greedy, opaque, reckless, and unnecesary. In current circumstances, "unnecessary", at least, must be dropped from that list.

Analyses by Joe Cortright (CEOs for Cities) and David Stiff (Fiserve) find that house prices have weakened more in outer suburbs than in more central locations in metropolitan areas.  That may be attributable to gas prices, although those are also the areas where it was easier to build—and overbuild.  Data from the Census Bureau's Housing Vacancy Survey showing that the homeowner vacancy rate for units built since 2000 is over 10 percent, while the rate for older units is only 2.2 percent, are consistent with the notion that areas with newer housing are more troubled.

Demographics should favor downtown and urbanized suburban areas over the coming decade.  For the past 40 years, the demands of baby boomers have dictated the character of new construction, leading to apartment construction in the early 1970s, followed by starter single-family homes in the 1980s and by larger, trade-up homes in the 1990s and the first half of this decade.  Now the children of the baby boomers, born in the 1980s and 1990s (and known as "the Echo", "Generation Y", or "Millenials") will increasingly influence new construction.  This will initially enhance demand for apartments and townhouses and communities with more urban character.  (There has also been conjecture that boomers will trade in their McMansions and move into cities, but that's less likely.)

Even without the recent rise in gas prices, commuting has become more frustrating.  This is reflected in the surveys of prospective home buyers conducted periodically by the National Association of Home Builders.  One of the questions regularly included in those surveys has been "What would you be prepared to accept in the home to make it more affordable for you?"  Twenty years ago, 35 percent of respondents said that they would accept a longer commute, but in the latest survey, only 21 percent gave that answer.  On the other hand, the share willing to accept a smaller house grew from 18 percent to 27 percent.  Over the past 20 years, the median new home size increased from about 1,800 square feet, to 2,300, and the share of commutes exceeding 10 miles increased from 42 percent to 53 percent, perhaps tipping the balance.

Central cities have seen some growth in recent years, with 18 of the largest 25 gaining population between 2000 and 2006, but large cities have not grown as fast as their suburbs, and the Texas and Arizona cities with the highest growth have political boundaries that essentially include their suburbs, with land area of 240 square miles or more. Most of the large cities that grew did so because natural increase and immigration offset net domestic outmigration.  Growth in households exceeded population growth, however, as single-person households replaced families with children.

In general, a shift toward more urbanized development likely won't mean a revival of older cities, beyond a few gentrified neighborhoods.  Rather, satellite and suburban areas will become more like cities.   Especially for older, inner suburbs that means more city problems as well as city amenities.

A key factor in the relative growth of cities and suburbs has been the movement of jobs to the suburbs.  That contributes to demand for apartments etc., in the suburbs, and frustrates the development and use of mass transit.

I created some tables from Census Bureau data showing information about cities and suburbs, mainly using the designations of principal/central cities and metropolitan areas.  Most of the information only goes through 2006, but I don't expect that later data will show radical changes yet.

Foreclosures are commonly not initiated until loans are about 5 months in arrears.  Foreclosures that are started do not always end with the lender owning the property, but where that occurs the process takes two months to more than a year, depending on state (and, sometimes, local) laws and procedures, with 5 months being about average.  A table from the "Hope Now" group of lenders shows estimates of the range among states.  An alternative, more detailed, analysis may be found in a Freddie Mac report by Cutts and Merrill.

If the foreclosed property is sold at auction, the lender is typically the only bidder, and whether there is an auction or not, the lender ends up with REO (real-estate owned).  There may be a delay of several more months before the REO home is put up for sale, frequently because of the need to repair damage caused by vandalism or neglect.

In recent years, it appears that less than half of foreclosures resulted in the lender taking title to the home.  Despite various widely heralded plans to reduce foreclosures and evictions, the proportion of foreclosures resulting in forfeiture of the property is already rising and likely to increase further.

Thus, the problems shown by the recent rise in delinquencies and foreclosures still haven't been fully reflected in the housing market.  Even without additional increases in delinquencies and initial foreclosure actions, more homes and residents will be dumped on the market over the next year.

Foreclosures are costly, for communities as well as for borrowers and lenders.  Tenants in rental housing are also typically displaced when a foreclosure occurs.  It would be nice if all this pain could be avoided, but often there is no feasible alternative, and dragging out the process may end up making the situation worse.  The foreclosure process is only one factor in the housing market disarray, but recovery in states where foreclosure is longest and messiest may be delayed even more than elsewhere.

]]> http://michaelcarliner.com/blog/2008/06/15/foreclosure-tsunami/feed/ http://michaelcarliner.com/blog/2008/05/21/fha-to-the-rescue/ http://michaelcarliner.com/blog/2008/05/21/fha-to-the-rescue/#comments Wed, 21 May 2008 21:25:35 +0000 Michael Carliner http://michaelcarliner.com/blog/2008/05/21/fha-to-the-rescue/ As the housing and mortgage markets have foundered, the FHA (Federal Housing Administration) has suddenly become the focus of a variety of efforts to stabilize the market. It has already filled much of the breach left by vanishing financing sources, but now it is being eyed for more dubious heroics.

FHA doesn't provide mortgages, but it guarantees that the mortgages will be repaid, if not by the borrower, then by the government. Although in theory taxpayers are ultimately responsible, the main FHA program is a mutual insurance arrangement under which borrowers pay premiums to insure against their own defaults, potentially receiving rebates of money left over after claims are paid. Unlike conventional (not-government-insured) mortgages, FHA loans may be, and usually are, packaged into Ginnie Mae mortgage-backed securities, which are fully and unambiguously backed by the federal government (on top of the federal insurance on the underlying mortgages).

When FHA was created in 1934, it was intended for economic stimulus and mostly insured loans on new homes in segregated communities (see my 1998 article). A few decades later, however, it evolved into a program mainly for insuring mortgages for lower-income, minority, and/or first-time home buyers. FHA-insured loans were available to borrowers who lacked the downpayments and credit ratings required for conventional prime loans.

With the rise in subprime lending, especially after 2001, the FHA share shriveled (as documented and analyzed in a June 2007 GAO report). Subprime lenders were willing to accept borrowers that even FHA, with its relatively-relaxed standards, rejected. Subprime loans also, in some cases, offered lower initial payments, but had terms that became more expensive over time.

Now subprime lending has dried up, and FHA lending has soared--more than doubling in the past year, despite fewer home sales. The growth is likely to accelerate, moreover, as a result of a sharp increase in the maximum loan size, provided for in the February 2008 economic stimulus package. The "FHA Secure" program provided somewhat more lenient underwriting rules for refinances of conventional ARMs, starting in September 2007, and those rules will be further liberalized in July 2008, promising additional volume, and possibly greater risk.

Legislation now under consideration, and apparently headed toward adoption, would create a much more ambitious role for FHA. New loans would be made available to seriously delinquent borrowers facing foreclosure whose homes are worth less than the outstanding balances on their mortgages. The holder of the previous mortgage (aka mortgagee--the lender, not the borrower) must be willing to accept less than the current value of the home as payment in full. The rescued borrower would have to agree to split any subsequent gains from resale of the house with FHA.

The previous lender would presumably only agree to such a deal if borrower and the house appeared to be pretty worthless. With the homeowner obligated to share any gains with FHA, there wouldn't be much incentive to improve, or even to maintain, the house. Compounding the adverse selection and moral hazard issues, this would seem to be a pretty difficult program to administer.

Desperate times may call for desperate measures, but the congressional plans look a little too desperate.

Even though the contraction in residential construction has extended to virtually all areas and market segments, there are some differences revealed by the data. Multifamily starts, for example, haven't fallen by much over the past couple of years, despite the 25 percent decline in March from a random uptick in February. For the first quarter of 2008, multifamily starts averaged 313,000 at an annual rate, exceeding the annual total for 2007 and not much below the annual total of 352,500 in 2005. This partly shows a failure of condo construction to adequately respond to the severely-glutted condo supply, but largely reflects increased construction of, and relatively-healthy demand for, rental apartments.

Among the four Census Bureau regions, the Northeast has experienced the least severe single-family decline, with new home sales down by 20 percent from 2005 to 2007, compared to 40 percent nationally. But with the Northeast accounting for only about 8 percent of new home sales, that doesn't provide much support for the national economy. The steepest regional declines have been in the Midwest and West, where new home sales and single family starts fell by nearly half from 2005 to 2007 and have fallen further in the early months of 2008.

As is typical during housing downturns, construction of single-family homes on customers' land has been more resilient than "for sale" construction on builder-owned land. About 20 percent of single-family starts occurred on customer-owned land in 2005, but that share was up to 32 percent by the 4th quarter of 2007.

The inventory of new homes for sale has declined somewhat in recent months, down to 471,000 at the end of February 2008, compared to 544,000 a year earlier. That still represents 9.5 months supply at the latest sales rate. Moreover, although there were fewer homes for sale in the construction pipeline, the number of completed unsold homes was up from February 2007. Among those completed homes for sale, the median time since completion was 7.2 months, compared to an already-high 4.2 months a year earlier. And that doesn't include homes with sales contracts that had been cancelled.

For aspiring first-time homebuyers, perhaps all this represents possible bargains, although home prices are still much higher than they were five years ago. For existing home owners, the housing and financial industries, and the national economy, the data paint a pretty grim picture. Current production is below the rate needed to accommodate longer-term demand, and eventually that will eliminate the excess supply. The process will take a while, however, especially with a recession likely to constrain new household formations, and with foreclosures dumping more supply on the market.

Employment growth in Nevada exceeded the national average every year from 1984 to 2006. In 13 of those 23 years, Nevada's growth rate was the highest among all states. Arizona and Florida also consistently exceeded the national employment growth rate. All three states have now fallen below the national average.

I calculated estimates of state-by-state employment involved in building, selling, and financing housing. Using data from 2005 County Business Patterns (the latest year available for that measure of detailed employment by industry), as well as some other sources, plus some crude but reasonable estimates of the share of employment in various industries dependent on housing, I figured that about 5.9 percent of jobs nationwide were related to housing growth and turnover. Nevada had the largest share among the states, with 9.0 percent, even though the building materials and other inputs used there largely came from other states. New York had the smallest share, at 4.4 percent, according to this calculation.

In growth-addicted states such as Nevada, Arizona, and Florida, reduced housing demand means significantly fewer jobs. And with so many recent migrants in the population, the likely impact on outmigration is greater than in states where there are stronger ties to the community. The unwinding of speculation and overbuilding thus translates into further reductions in housing demand. The demand-construction-employment-population-demand feedback loop supercharged home building during their long growth periods, but now it is operating in reverse. Nevada, Arizona, and Florida are not going to become chronic laggards like Pennsylvania or Michigan, but it may take a couple more years before they return to their growth trajectories.

Not all the over-achieving states of the past decade with large housing production sectors have been clobbered by the recent unpleasantness. Utah and Idaho, for example, seem to be doing all right. It is not entirely clear why they have been less affected. Maybe people who might have gone to Nevada or Arizona are moving there instead.

Subprime

The issues connected with subprime loans are only a part of the broader problems in the mortgage market, and mortgage finance problems are only part of the distress in the housing market. But mortgage lending was a big factor in speculative activity, housing oversupply, and inflated prices. And subprime loans accounted for a disproportionate share of recent wayward lending and current loan defaults.

The basic dynamic that led to the subprime debacle has been exposed pretty broadly already, but for a (literally) graphic exposition, you might refer to this description. Thanks to Bernie Markstein of NAHB for forwarding the link.

During the past year, particular concern has focused on the risks from imminent resets of hybrid subprime ARMs originated in 2006. A large share of subprime loans originated in 2005 and 2006 were 2/28 and 3/27 hybrids, meaning that they carried fixed rates for the first 2 or 3 years, after which the interest rate is reset every 6 months, based on a specified index and margin. The most common index is the 6-month LIBOR (London Interbank Offered Rate) – no doubt chosen because it was the rate least likely to be comprehensible to a subprime borrower. The margin is typically a whopping 600 basis points.

Although the initial fixed rates on these subprime ARMs were below the specified fully-indexed rates, they were not cheap. In some cases, the rate for the first couple of months was set very low, but during the rest of the fixed-rate period, the rates were well above prime fixed-rate loans. FDIC Chair Sheila Bair (one of the few officials to recognize the dangers before it all blew up) reported that the average for hybrid subprime ARMs in 2006 was 8.23 percent.

In early December, a program orchestrated by the Treasury Department was announced with great fanfare to defer resets for subprime borrowers who (a) were current and able to continue to make payments at the existing fixed rate and (b) would not be able to make payments after resets. It is still not clear whether the promised modifications were forthcoming, but since the announcement the 6-month LIBOR has fallen from about 5 percent to about 3 percent, implying that resets would only produce changes from 8.2 percent to about 9 percent, rather than to about 11 percent (subject to caps on per-period adjustment). Therefore, relatively few subprime borrowers can be characterized as able to pay before reset but unable to do so afterward. The reductions of the Fed funds rate and other Federal Reserve (and foreign central bank) actions have thus done much more to forestall reset payment shock than any modifications or forbearance actions.

While those factors have reduced the threat from reset payment shocks, that's not the central problem anyway. Very high shares of these loans were already in trouble before facing resets, as the sketchy underwriting that characterized subprime lending earlier in the decade became much more reckless in 2006.

Alt-A

Although many subprime loans had small downpayments and some had interest-only payments during the initial years, it is among the "near-prime" or "Alt-A" category that really exotic loans were most common. That's where there was the greatest concentration of loans to investors, downpayments using piggy-back loans, low-doc loans, low teaser rates, and non-fully-amortizing (interest only, option ARM, and balloon) loans. Average FICO (credit) scores of Alt-A borrowers were much closer to those of prime borrowers than to subprime borrowers, but the creative features represent risks that are hard to evaluate but (especially in combination) quite frightening.

Prime

While prime conforming loans, largely sold to Fannie and Freddie, and prime jumbo loans did not exhibit craziness comparable to the alt-A and subprime loans, even there the characteristics of loans became riskier, at the same time that inflation in house prices was adding to the potential risk in standard loans. For example, the interest-only share of loans in GSE pools jumped from about 8 percent in 2005 to about 16 percent in 2006.

The response

Two recent surveys of servicers, by the Mortgage Bankers Association and the Hope Now Alliance (with considerable overlap) attempted to measure the handling of delinquent subprime (and other) mortgages. They indicate that a substantial and growing share of lender responses involved loan modifications (new loan terms) or repayment plans (allowing more time to pay without changing terms). That was before any of the government-organized initiatives. Thus, it is unclear how much the involvement of the Treasury Department and other government entities changed things. Treasury argues anyway that it only encouraged a wholly-private effort, but encouragement from government officials with regulatory authority over your business can seem like more than casual advice.

Capozza and Thomson have described how subprime lenders expect high delinquencies and have a greater incentive for forbearance than prime lenders. Compared to prime loans, the interest rates on any further payments are higher and the recovery from foreclosure and sale is likely to be lower. Ultimately, despite repayment plans and modifications, troubled subprime loans are likely to lead to forced sales or forfeitures, but only after lengthy delays. Therefore, much of the impact on the already-glutted housing market still lies ahead.

One of the least documented aspects of default and foreclosure is the fate of the residents. Do they find housing elsewhere as renters or owners, double-up with others, or become homeless? In addition to the human cost, the answer to that question has significant implications for the market.

Among the limited audience who may read this are people who know much about this subject than I. Please contribute your thoughts and correct my errors.

]]> http://michaelcarliner.com/blog/2008/02/20/the-mortgage-muddle/feed/ http://michaelcarliner.com/blog/2008/02/01/let-the-recession-begin/ http://michaelcarliner.com/blog/2008/02/01/let-the-recession-begin/#comments Fri, 01 Feb 2008 14:05:36 +0000 Michael Carliner http://michaelcarliner.com/blog/2008/02/01/let-the-recession-begin/

Payroll employment fell by 17 thousand in January, the first month-to-month decline since 2003. The downtick was led by a decline of 27 thousand in construction employment, from 7,475,000 to 7,448,000. Residential construction employment, for builders and special trades (subcontractors), fell by 28 thousand, from 3,107,700 to 3,079,600.

A big drop in construction employment was slow in coming, but declines accelerated late in 2007. Even though housing starts, completions, sales, etc., had fallen sharply, employment in residential construction has declined much more modestly. Between March 2006 and December 2007, residential construction employment fell by less than 10 percent. Single family starts fell by 50 percent during that period, and the number of single family units under construction (a more relevant measure than starts) fell by 37 percent. Total units under construction fell by 25 percent.

The number of multifamily units under construction didn’t decline, but the average multifamily unit uses only about half as much labor and material as an average single family home. As a share of new residential construction value put in place, multifamily accounts for less than one-sixth of the total, even after the single-family collapse.

Revisions in the payroll data that were included in today's release showed weaker growth in total employment as well as construction employment during the past 2 years, but the overall picture didn't change. Various hypotheses have been suggested for the disconnect between construction activity and employment, such as a failure to count illegal immigrant workers, deficiencies in the payroll survey, and confusion in distinguishing between residential and nonresidential workers, but none of those explanations was really adequate. Just as there was initially a lack of apparent impact of housing on consumption, it was mainly a question of lags, rather than the absence of a relationship. Because employment in residential remodeling is included along with new construction, however, there is some cushion.

For some of the discussion and analysis that has addressed this, see the July 2007 comments by Richard Berner of Morgan Stanley, the October 2007 BLS analysis of the construction payroll data, and the discussion of the data in the January 2008 analysis of the effect of the housing slump on productivity from the Federal Reserve Bank of Chicago.

It's not official, but......

Although recessions are commonly thought of as consisting of 2 quarters of negative GDP change, the NBER committee that is effectively the arbiter of business cycles actually uses monthly statistics and judgment to determine when recessions begin and end. Often, however, recessions are well-underway, or even over, before they declare that one has started.

As explained in a memo from October 2003, the Business Cycle Dating Committee looks primarily at 4 measures: real personal income excluding transfers, payroll employment, industrial production, and real manufacturing and trade sales. I’m not completely sure I’ve used the same series as they do, but it appears that real personal income in December (the latest number available), was down slightly from a possible peak in September 2007. Industrial production may have peaked in July. Sales are only available through November, but were down from a possible peak in October. With the January employment decline, it looks like a duck, walks like a duck, swims like a duck, and now quacks like a duck. I think it’s a duck.

Quarterly data from the Census Bureau's Housing Vacancy Survey were released today, offering no encouragement to offset the dismal results for December new and existing home sales reported during the past week. The overall homeowner vacancy rate for the 4th quarter of 2007 was shown as 2.8 percent. That was up from 2.7 percent in the third quarter, and matched the record set in the first quarter of last year. The homeowner vacancy rate has been reported on a quarterly basis since 1956. Prior to 2006, it never exceeded 2.0 percent. That is perhaps the most graphic and disturbing measure of the extent of imbalance in the market.

The homeowner vacancy rate is defined as the number of vacant housing units for sale, divided by the sum of the vacant fior sale plus owner occupied plus sold but not yet occupied. The title is a bit of a misnomer, since many of the vacant units for sale were not previously owner-occupied and may be sold to investors rather than for owner-occupancy. Mobile homes and condos are included along with conventional single-family structures, and and the vacant-for-sale category includes both existing (previously-occupied) homes and completed new homes. Only about half of all new and existing homes being offered for sale are vacant, but they represent a greater burden than homes that are still occupied or that have not been completed.

About 15 percent of the 2.18 million vacant homes for sale were not previously occupied, according to the HVS data. That translates into a larger number than the 195 thousand completed unsold new homes reported in the new home sales report, because new condos and mobile homes are included, and because completed homes with cancelled contracts are, in theory, included in the HVS but not in the sales report. Of course, the two surveys--especially the HVS--are not sufficiently precise to justify any major inferences from the difference.

If we assume that a normal/equilibrium homeowner vacancy rate is, say, 1.7 percent (the average in 2004), then the excess number of vacant homes for sale is over 800 thousand. Even though current production is below long-term potential absorption, it would take a while to soak up that excess. Add in the possible effects of widespread foreclosures and the picture is even more dismal.

Myers and Ryu make estimates of the rate at which people in different age brackets bought and sold homes during 1995-2000, and report that among the population up to age 65 there were more purchases than sales, but beyond that the "sell rate" exceeded the "buy rate."  Given the impending shift in the age distribution, they extrapolate from that to a glut in housing supply.  Although the article in the Journal of the American Planning Association is more nuanced than the press reports drawn from it, it is still misleading.  It indicates that a major shift is imminent and that it will occur because boomers will "retire, relocate, and eventually withdraw from the housing market."

Sales by homeowners age 60 and over in 2000 were estimated by comparing their numbers to the numbers of owners from the same cohorts in 1990, using the difference as "a measure of all the home sales that were not followed by purchasing another home, but by renting, moving to a retirement home, or death."  It is that last component that undoubtedly accounts for much of their estimate. In a key chart with the vertical axis representing persons in each age group buying and selling, the scale ranges from zero to six percent, and the "sell rate" soars off the chart for those aged 80 and over. The caption says that "8.8 percent of persons 80 and older sold homes each year."  In that context, "sold" is a euphemism. In fact, according to data from the National Center for Health Statistics, more than 11 percent of persons 80 and older died each year during that period.

As I found when I looked at this subject in 1996, most older homeowners don't move, and if they do they are unlikely to become renters, move in with their children, etc.  Indeed, they were more likely than younger movers to buy newly-built homes. In connection with a presentation I gave in April 2006, I noted reasons to expect that the leading edge of the baby boom would be even less likely to opt out of home ownership in the near term.

Extensive and thorough analysis of housing demand among the older population may be found in a series of articles about housing wealth written over the course of two decades by Steven Venti and David Wise.  The most recent paper that I have was written in 2001.  They found that "in the absence of a precipitating shock--death of a spouse or entry of a family member into a nursing home--families are unlikely to discontinue home ownership.  And even when there is a precipitating shock, discontinuing ownership is the exception rather than the rule."

There will be a negative effect on housing demand from the passing of the 1946-1964 baby boom, but it won't be soon or sudden.