Transcripts of the monetary policymaking body of the Federal Reserve from 2002–2008.

  • Good afternoon, everyone. I think you all are aware that this is the last meeting Governor Gramlich plans to attend. There will be a luncheon in his honor after the meeting ends tomorrow. It’s also the last meeting for Ed Ettin, who will retire at the end of July, and for Marvin Goodfriend from the Richmond Bank, who will move to Carnegie Mellon at the end of the summer. We’ll miss all of them. Ed, as some of us know, has been attending FOMC meetings since the late 1970s, and Marvin has been attending since 1993. As for Ned, there are certain special things that I’m going to say about him tomorrow, which I won’t repeat here. We’ll let him be a little nervous. [Laughter]

  • I’ve got a few things to say, too.

  • Ah, but I will have the gavel! The first item on the agenda is the selection of Richard T. Freeman as an Associate Economist of the FOMC to serve until the election of a successor at the first regularly scheduled meeting of the Committee in 2006. Would somebody like to move that nomination?

  • I’ll move that nomination.

  • Without objection, that is approved. We will be discussing a special topic today at your request, housing valuations and monetary policy. You’ve obviously received the documents the staff has distributed. Mine are buried in this stack here; I have them. Let’s turn now to Dave Stockton.

  • Thank you, Mr. Chairman. We have five briefings this afternoon and the issues covered in those briefings are so intertwined that we thought it would be best if we just presented all five and then opened the floor for discussion and questions afterward. We recognize that a long presentation, coming right after lunch and while a number of you are still probably a bit jet-lagged, could test your powers of concentration, but I think you’ll find the presenters to be both well informed and entertaining. [Laughter] So with that we’ll begin.

  • Thank you. My presentation begins on the third page of the handout you received.

    It seems that everybody is talking about house prices, and the upper panel of your first exhibit shows why: House prices, adjusted for general inflation, have risen at a rapid pace in recent years and did not even pause during the last recession. Indeed, the real rate of appreciation has increased, and the most recent readings have been at annual rates greater than 7 percent. By comparison, the average annual increase in real house prices during the past 30 years is only about 1¾ percent.

    The next two panels illustrate some of the eye-popping gains that have been recorded in selected metropolitan areas. For example, as shown in the middle left panel, real house prices increased about 16 percent in San Francisco and 30 percent in Las Vegas during the four-quarter period ending in the first quarter; as shown to the right, the most recent gain was 13 percent in New York and 20 percent in Miami.

    Rapid price appreciation has sparked debate about whether housing has become overvalued, and the popular press is filled with stories suggesting that it has. As summarized in the lower left panel, anecdotes suggesting that the housing market is overheated include those about increased speculation, purchase decisions that are perhaps too dependent on rosy assessments of future appreciation, and increased reliance on novel forms of financing without full recognition of the associated risks.

    Although these anecdotes are suggestive, they do not provide a benchmark for valuing housing. Two approaches that do provide a benchmark are listed to the right. One is to ask if housing is affordable for a typical family. Some analysts have argued that prices are too high relative to incomes, while others say low interest rates have kept required monthly mortgage payments affordable. Another approach is to ask if house prices are properly aligned with rents. I have pursued both approaches in my research, and have concluded that rents provide a preferable benchmark for valuing housing. I will therefore focus my prepared remarks on this approach.

    As summarized in the upper left panel of your next exhibit, a strength of this approach is that it employs a standard asset pricing framework, such as that often used in studies of stock market valuation. In this framework, rental payments in the housing market are analogous to dividends in the stock market. Seen this way, the price of a house should reflect the appropriately discounted stream of expected rents; high prices could be justified by high rents or by low carrying costs, which include interest payments, net taxes, and depreciation. But if prices appear unusually high relative to rents and carrying costs, one might conclude that housing is overvalued.

    As highlighted to the right, I have implemented the framework using repeat- transactions price indexes from the Office of Federal Housing Enterprise Oversight [OFHEO] and Freddie Mac and the tenants’ rent index from the consumer price index. I made several adjustments to these series to address some of the shortcomings of the published data. As you will see in Dick Peach’s presentation, he and I disagree about the best way to measure house prices. I would be happy to discuss the issue during the question period.

    That said, the red line in the middle panel shows the estimated price-rent ratio for the stock of housing, and the black line shows the estimated real carrying cost of housing. The first point to note is that the measured price-rent ratio is currently higher than at any earlier time for which we have data. Moreover, the run-up in prices appears to be far greater than can be explained by carrying costs, at least if we use the historical relationship between the two series as our guide. Although theory suggests a tight link between carrying costs and the price-rent ratio, the data suggest that the actual link is more tenuous. At the simplest level, while the price-rent ratio is at a historical high, the carrying cost is not at a historical low. More formally, regression analysis suggests that prices are about 20 percent too high given rents and carrying costs.

    One might reasonably ask what this potential 20 percent overvaluation portends for house prices. The lower panel summarizes the historical experience on this question. The panel shows a scatter plot of the price-rent ratio (on the horizontal axis) and changes in real house prices over the subsequent three years (on the vertical axis); the panel also includes a fitted regression line. As I mentioned a moment ago, carrying costs are only slightly correlated with the price-rent ratio. Thus, their effects can be excluded from the chart without materially affecting the results.

    The downward slope of the regression line summarizes a key finding: When house prices have been high relative to rents, subsequent price changes have typically been smaller than usual. For example, on the right side of the panel we can see that in the second quarter of 1979, the price-rent ratio stood above 24, and real prices decreased 7 percent during the subsequent three years. Near the other end of the scale, in the third quarter of 1970 the price-rent ratio stood just above 19, and real prices increased more than 5 percent during the subsequent three years. However, as is readily apparent from the chart, the relationship is loose. Most notably, in the first quarter of 2002, the last observation for which we have a reading for the subsequent three-year change in house prices, the price-rent ratio stood at 22. Although the regression suggests that real prices should have been about flat since then, real prices actually increased more than 20 percent, and the price-rent ratio rose to about 27—literally off the chart.

    To give an impression of what has happened in local markets, the upper panel of your next chart summarizes housing-market conditions for four metropolitan areas: San Francisco, New York, Chicago, and Miami. The panel displays the deviation of each city’s price-rent ratio from its long-run level in the second quarter of 1979 (the red bars), the fourth quarter of 1989 (the black bars), and the first quarter of 2005 (the green bars). The numbers above the bars indicate their height. The first two episodes represent previous peaks of the price-rent ratio at the national level. The third episode is where we are now.

    The numbers below the red and black bars show the performance of real house prices in each city in the three years following each episode. For instance, as shown by the red bar, the price-rent ratio in San Francisco was 7 percent above its long-run level in the middle of 1979. During the subsequent three years, real house prices there fell 5 percent. At the end of 1989 (the black bar) the price-rent ratio in San Francisco was 15 percent above its long-run level, and real prices there fell 12 percent during the subsequent three years. The most recent reading for San Francisco indicates that the price-rent ratio is now further above its long-run level than in either 1979 or 1989.

    The price-rent ratios suggest that housing is overvalued in the other three cities as well, but to differing degrees. Although the price-rent ratio in New York is elevated, housing does not look much more overvalued there than it did in the late 1980s. House prices appear elevated relative to rents in Chicago as well, but it is Miami that stands out as the most overheated of the four markets shown here.

    The evidence that I have shown you so far suggests that rents provide a loose tether for house prices; prices deviate from their long-run relationship with rents for extended periods, but not indefinitely. However, to provide a forecast of house prices, we must consider both the long-run relationship between prices and rents and the short-run dynamics among prices and other economic factors. The middle left panel of your exhibit lists several factors that affect real house-price dynamics in the short run. In particular, the Board’s staff has found that a basic model should include lagged changes in house prices, as well as changes in real income, real carrying costs, and the unemployment rate. This basic model does not take a stand on whether housing is overvalued. Alternatively, our error-correction model adds to the basic model the lagged level of the price-rent ratio and the lagged level of carrying costs. It therefore provides a formal way to model short-run dynamics around a long-run equilibrium level. The estimated coefficient on the lagged price- rent ratio is negative, capturing the idea that there is a long-run relationship between house prices and rents and that price changes are restrained when prices are high relative to rents.

    The middle right panel shows the implications of these two models for the four- quarter percent change in real house prices at the national level during the Greenbook forecast period. The basic model—in green—suggests that house prices will decelerate in the coming two years, but that they will still rise faster than their long-run average of about 1¾ percent per year. Thus, even a model that ignores the long-run relationship between the levels of prices and rents predicts a slowdown in house-price appreciation. Still, housing would become slightly more overvalued relative to rents at the end of the forecast period.

    The error-correction model—in red—calls for prices to decelerate more in the coming two years, with real prices actually edging down a bit in 2006. In this scenario, housing would be about 15 percent overvalued at the end of the forecast period. Note that the June Greenbook forecast—the black line—lies between the two model simulations.

    To sum up, the statistical evidence that I have presented today provides support for the view that the price-rent ratio can be a useful tool for summarizing housing valuations. The price-rent ratio is currently very high by historical standards, suggesting that housing might be overvalued by as much as 20 percent. Historical experience suggests that the change in real house prices going forward will be slower than in recent years. Unfortunately, the evidence cannot rule out either further rapid gains in house prices for a time or a more rapid correction back to fundamentals. Just as the price-dividend ratio is an imperfect tool for forecasting stock prices, the price-rent ratio should be considered only a rough guide to valuations in the housing market. Thank you.

  • As Josh noted, everyone’s talking about house prices; but they also seem to be talking a great deal about mortgages. The popular consensus appears to be that homebuyers, especially in hot housing markets, now make token down payments and can just scrape into their homes by resorting to interest-only mortgages; in this view, borrowers and lenders alike are vulnerable to any fall in house prices. In my prepared remarks I will address each of these issues.

    An obvious measure of the vulnerability of a borrower to a decline in his house price is the ratio of the current mortgage balance to the current value of his home; this ratio is known as the current loan-to-value, or LTV, ratio. The top panel of your first exhibit compares the distribution of the estimated current LTV as of September 2003 (the black bars) with the distribution as of March 2005 (the red bars).

    As shown by the leftmost red bar, 64 percent of borrowers currently have LTVs below 70 percent. Moving to the right, a further 18 percent currently have LTVs between 70 and 79 percent; 14 percent have LTVs between 80 and 89 percent; and only 4 percent have LTVs of 90 percent and above. Moreover, the recent rapid appreciation in house prices has actually outstripped mortgage debt growth, so that the average LTV has fallen over the past 18 months.

    The middle left panel uses the same data to concentrate on the most leveraged borrowers. For a given drop in nominal house prices (the horizontal axis), the vertical axis gives the percent of borrowers who would have negative equity. As before, the black line reflects data from September 2003 while the red line reflects the most recent data. As shown by the red line, following a house-price drop of 10 percent, an estimated 4 percent of borrowers would have negative equity, while a drop of 20 percent would leave about 18 percent of borrowers in negative territory.

    The middle right panel summarizes average LTVs at origination for homes purchased in 2004, by state, relative to price appreciation in that state over the previous four years. States with lower-than-average appreciation, such as Utah, Texas, and Oklahoma, are at the left, while states with faster-than-average appreciation, such as California, Massachusetts, and New York are at the right. As shown, LTV at origination in 2004 was actually lower in states with more appreciation. Thus, homebuyers in California and other high-appreciation states made larger down payments relative to the price of their house than homebuyers in low-appreciation states such as Oklahoma.

    Increasing home equity, mainly driven by rising house prices, has supported mortgage credit quality in recent quarters. The bottom left panel plots delinquency rates for loans held on banks’ books (the black line) as well as for the broader MBA [Mortgage Bankers Association] measure (the red line). Both have fallen significantly in recent years.

    The bottom right panel summarizes the vulnerability of borrowers to house- price shocks. The average LTV on mortgages has declined over the past 18 months, and most households currently have substantial equity in their homes. In the past few years, borrowers have benefited from rapidly rising house prices, which have kept mortgage delinquencies at extremely low levels. However, an estimated 4 percent of borrowers are highly leveraged and could lose all of their home equity if house prices were to fall 10 percent.

    As I noted earlier, recent anecdotal reports have highlighted a second potential risk lurking within mortgage markets—the sudden popularity of interest-only mortgages. These are the subject of your next exhibit.

    The statistics presented in exhibit 2 are taken exclusively from data on private- label—that is, non-GSE [government-sponsored enterprise]—residential mortgage- backed securities, or RMBS, pools. The overwhelming advantages of these pools for my purposes are that they are very transparent; that is, a great deal of information is available on the underlying mortgages and that they contain many heterodox mortgages, including interest-only mortgages. Many of the recent press articles citing the growth of novel mortgages are based on RMBS pool data and thus, because such mortgages are overrepresented in RMBS pools, have exaggerated, in my view, the role of heterodox mortgages.

    Turning to the data, line 1 of the top left panel shows that the dollar value of RMBS pools has nearly doubled over the past two years. Moreover, as shown in line 2, RMBS pools backed by interest-only, or IO, mortgages have increased almost sixfold, and now amount to nearly $300 billion. However, total home mortgage debt, line 3, has also increased over the past two years. As shown in the memo line, interest-only RMBS pools now account for 3.6 percent of all home mortgage debt, up from less than 1 percent two years ago, but still a small share of all mortgages.

    As their name implies, interest-only mortgages do not require the borrower to make principal payments, at least during an initial period. Borrowers willing to use IO mortgages could qualify for a larger mortgage and thus be able to buy a more expensive house. The top right panel gives some idea of the relationship between price appreciation and the popularity of IO mortgages. The vertical axis shows the fraction of IO mortgages used to purchase houses, by state, in 2004. Again, I should emphasize that the IO shares are calculated within the RMBS world, and so are probably overstated. The horizontal axis gives state-level house-price appreciation from 1999 through 2003. As you can see, IO mortgages are somewhat more popular in states that saw more appreciation, although, as seen in the inset box, the correlation is not particularly strong.

    While the principal value of an IO loan doesn’t decline, if the initial down payment is large enough, the borrower may have a substantial equity cushion against price shocks. The middle panel reports the LTV at origination for IO loans made over the past three years. As shown, most IO mortgages had LTVs below 80 percent, although the trend over time has been away from the very lowest LTVs. That said, lenders have continued to make relatively few IO loans with LTVs above 80 percent. Those higher LTV loans now account for about 15 percent of all outstanding IOs.

    Finally, anecdotes often emphasize that IO loans are extended to borrowers with lower credit quality. The bottom panel shows the distribution of credit quality, measured by FICO scores, among IO borrowers. As a rough approximation, most lenders define prime quality borrowers as those with FICO scores of 660 or above. The leftmost set of bars thus represents subprime borrowers. As shown, between 8 and 10 percent of IO loans have been extended to these borrowers. Moving to the right, the next two sets of bars show that the great majority of IO borrowers had solid credit scores between 660 and 779. As shown by the rightmost set of bars, about 10 percent of IO borrowers had credit scores above 780. On the whole, therefore, the credit quality of borrowers using interest-only loans does not appear particularly risky.

    One might wonder if financial institutions and investors have, in the face of the continuing housing boom, dropped their defenses against the mortgage losses that would accompany a house-price bust. The top left panel of your next exhibit lists the main institutions exposed to residential mortgage credit risk and the main types of mortgages held by these institutions. The housing GSEs, line 1, almost exclusively hold or guarantee conforming mortgages with fixed rates. Private mortgage insurers, line 2, insure the component of mortgage principal that exceeds 80 percent of the property’s value and so are effectively exposed to the credit risk associated with high-LTV loans. RMBS pools (line 3), as well as banks and thrifts (line 4), hold a wide variety of different mortgage types, including traditional fixed- rate mortgages as well as variable-rate and junior liens.

    The top right panel emphasizes that the housing GSEs hold very little credit risk. As shown on line 1, the average LTV at origination of GSE-guaranteed mortgages was 70 percent; based on regional house-price appreciation, the estimated current LTV of these mortgages (line 2) has fallen to 57 percent. The average credit score of the underlying borrowers (line 3) is also solidly in the “prime” category. Obviously, these average values mask some variation in the borrower population, which no doubt contains some higher-risk borrowers. However, as shown on line 4, 19 percent of the mortgages guaranteed by the GSEs carry some form of credit enhancement. If one of these mortgages defaults, the GSE receives a payment from the insurer, usually a private mortgage insurance, or PMI, company.

    The middle left panel examines the health of the PMI industry. As shown by the black line (left axis), the ratio of total insured mortgages to capital, the risk- capital ratio, has declined steadily over the past 10 years, indicating that PMI companies have historically high capital on hand relative to the risks they insure. The red line (right axis) shows net underwriting income—that is, income from premiums less losses and expenses, relative to capital. After suffering large losses in the late 1980s, PMI companies have consistently recorded positive underwriting income. In sum, PMI companies appear to have built up a historically large cushion to absorb the losses that might be associated with a widespread drop in real estate values.

    The middle right panel analyzes the risks posed to investors in RMBS pools. These pools contain some of the riskier outstanding mortgages. However, they are structured so that investors can choose their risk exposure. Further, RMBS pools are exceptionally transparent, so investors have extensive information on each mortgage in the pool. In principle, investors should have understood, and appropriately priced, the risks inherent in these mortgages. In practice, however, investors price these mortgages using loss models, which are estimated using relatively little data from major house-price declines.

    The bottom two panels discuss the exposure of the 8,900 banks and thrifts in the United States to residential mortgage credit risk. The panel to the left divides institutions into quartiles by the fraction of their portfolios accounted for by residential mortgage assets, defined as whole loans, home equity lines of credit, non- GSE RMBS pools, and residual tranches on securitized mortgages. As shown, for institutions in the bottom quartile, mortgages account for less than 5 percent of total assets. This fraction rises by quartile until, for institutions at the top, mortgages account for more than 40 percent of assets. Residential mortgage credit risk is more concentrated at these institutions than at the institutions in the lower quartiles.

    The panel to the right shows the average size and capital-to-asset ratio of institutions in each of the quartiles. Reading down the first column, which gives average institution size, one can see that smaller institutions are concentrated in the first two quartiles, which have relatively little mortgage exposure. Reading down the second column, which gives average tier 1 capital ratios, institutions in all quartiles are extremely well capitalized. Thus, institutions with large amounts of mortgage credit risk on their portfolios are well positioned to handle severe losses.

    To sum up, neither borrowers nor lenders appear particularly shaky. Indeed, the evidence points in the opposite direction: borrowers have large equity cushions, interest-only mortgages are not an especially sinister development, and financial institutions are quite healthy. Nonetheless, even the most sanguine analyst quails when contemplating a historically unprecedented drop in nationwide nominal house prices. Such a drop will obviously hurt both borrowers and lenders and will also no doubt expose weaknesses that will only be obvious in hindsight. Thus, perhaps it would be best simply to venture the judgment that the national mortgage system might bend, but will likely not break, in the face of a large drop in house prices. That concludes my prepared remarks.

  • Hardly a day goes by without another anecdote-laden article in the press claiming that the U.S. is experiencing a housing bubble that will soon burst, with disastrous consequences for the economy. Indeed, housing market activity has been quite robust for some time now, with starts and sales of single- family homes reaching all time highs in recent months and home prices rising rapidly, particularly along the east and west coasts of the country. But such activity could be the result of solid fundamentals underlying the housing market. After all, both nominal and real long-term interest rates have declined substantially over the last decade. Productivity growth has been surprisingly strong since the mid-1990s, producing rapid real income growth primarily for those in the upper half of the income distribution. And the large baby-boom generation has entered its peak earning years and appears to have strong preferences for large homes loaded with amenities.

    One of the conditions of an asset bubble is that the price of the asset has risen well above what is consistent with underlying fundamentals. In the current debate, two measures of relative value have been applied to single-family home prices— price relative to income and rent relative to price. In the comments that follow I will concentrate on the price-to-income measure, but my conclusions apply equally to the rent-to-price measure.

    In most analyses of the ratio of home price to income, the measure of home price that is used is the repeat-sales home price index published by the Office of Federal Housing Enterprise Oversight (exhibit 1). This index tracks changes in the average price of homes purchased (or mortgages refinanced) with loans purchased by Fannie Mae and Freddie Mac, or conforming conventional loans. Therefore, it excludes cash sales as well as purchases or refinancings financed with FHA [Federal Housing Administration], VA [Veterans Affairs], and jumbo conventional mortgages. It is called a repeat-sales index because it is derived by observing the sales prices—or appraised values, in the case of refinancings—of properties at specific addresses at two or more points in time. Finally, it is a transactions-based index in that it reflects the prices of homes that are sold (or refinanced) rather than the entire universe of single-family homes.

    A lesser known home price index is the constant-quality new home price index published by the Bureau of the Census (exhibit 2). This index is based on a sample of new homes sold, regardless of how the sale was financed. Hedonic methods are employed to hold the physical and locational characteristics constant over time. This index is part of the Census Bureau statistical program through which the single- family residential investment deflator of the national income and product accounts is derived. As shown in exhibit 3, the increase in prices indicated by these two indexes is quite different. For example, over the four years from 2000 to 2004, the OFHEO index increased at a compound annual rate of 8.2 percent, while the constant-quality index increased at a 5.4 percent annual rate. As shown in exhibit 4, the current ratio of price over median family income derived from these two indexes is vastly different. If the OFHEO index is giving an accurate picture of what is happening to home prices, I think one could say with some confidence that prices have been bid up to unsustainable levels. However, if the constant-quality index is a better reflection of reality, home prices actually look somewhat low relative to median family income, particularly compared to the late 1970s. I believe the constant- quality index provides a more accurate indication of what is happening to the price of a typical single-family home. In contrast, the OFHEO index is subject to upward biases that accumulate over time and distort ratios such as price-to-income and income-to-rents.

    To help us understand the biases in the OFHEO index, exhibit 5 presents the distribution by value of all single-family homes in the U.S. in 2003, with the specific values at the 25th, 50th, 75th, and 80th percentiles. 4 The median value in 2003 was $150,000 with the distribution skewed toward the right. The value at the 25th percentile was $90,000 while the value at the 75th percentile was $250,000. We do not know with certainty where the OFHEO index falls on this distribution, as it is an index rather than a series of values. But we can be reasonably certain that it lies somewhere between the average price of all existing single-family homes sold and the average price of homes purchased with conventional loans. That means the OFHEO index is a closer reflection of what is happening at the 75th percentile rather than the 50th percentile. Moreover, it is very likely that over time the point on that distribution represented by the OFHEO index has been drifting to the right. One cause of this rightward drift is what I call transaction bias. As shown in exhibit 6, the American Housing Survey (AHS) data suggest that both appreciation rates and turnover rates increase as one moves out the home value distribution. For example, from 1997 to 2003 the compound annual rate of appreciation at the 25th percentile was 4.5 percent, increasing to 8.7 percent at the 80th percentile. Corresponding average turnover rates for the period from 1997 to 2003 were 5.9 percent and 7.4 percent. That means, of course, that the average rate of appreciation of the units that turn over is higher than the average rate of appreciation of the entire distribution. While the amount of bias in any one year is likely to be small, it does cumulate over time and becomes quite important when one is comparing levels versus income or rents.

    Another potential upward bias in the OFHEO index is that while it is a repeat- sales index, there is evidence to suggest that it is not a constant-quality index. In addition to the strong pace of new housing starts, another aspect of the housing boom of the past decade has been a significant increase additions and alterations to the existing housing stock. Exhibit 7 presents in the top panel the ratio of the OFHEO index over the constant-quality index plotted over the period from 1977 to the present. In the lower panel are plotted real improvements per unit of housing stock per year over the same period. Over the past decade real improvements per unit have increased about 25 percent, which appears to be associated with a further increase in the ratio of the OFHEO index over the constant-quality home price index. Research suggests that higher-income households have a higher income elasticity of demand for improvements to their primary residences, suggesting that this source of upward bias is likely to be more pronounced in the right half of the distribution of all single-family homes. 5

    Another way of looking at the issue of home prices over income is to go back to the AHS data and see what is happening at various points on the distribution of all single-family homes. This information is presented in exhibit 8. At the 25th percentile the ratio of home price to income has been relatively stable, while it has increased sharply at the 75th and 80th percentiles, reminiscent of the price-to­ income ratio computed with the OFHEO index. Let me pause just a moment and emphasize that I am comparing home prices at a percentile with the incomes of the people who live in those homes at the same percentile. This chart is likely the equivalent of the finding that the increase of home prices has been most pronounced in areas of the country where home prices were already relatively high due in part to relatively inelastic supply. It is likely that in these areas of the country, where land values are high, the inclination to make substantial improvements to existing properties is the greatest.

    Clearly, not everyone agrees that the constant-quality new home price index provides an accurate indication of what is happening to the price of a typical single- family home. For example, it has been argued that most new construction takes place at the fringe of metropolitan areas where land prices may not be rising as fast as is intra-marginal land. There are several counter arguments. First, the theory leading to the conclusion that intra-marginal land values increase at a faster rate than land at the fringe is based on a theory of the development of a metropolitan area with fairly restrictive assumptions. Modern metropolitan areas have multiple commercial/employment centers. Many households have preferences for rural or suburban residences over urban residences. Restrictions on development to counter suburban sprawl have reportedly resulted in sharp increases in prices of parcels of land suitable for home building. Finally, I would like to note that the increases in land prices implicit in the constant-quality home price index, shown in exhibit 9, are substantial, particularly in the Northeast and the West. These estimates were derived assuming that land represents 50 percent of the value of the total property and that the prices of the other inputs increase at the same rate in all regions of the country.

    In closing, I would like to comment on one other aspect of the housing bubble issue that has received substantial attention. Earlier this year a major real estate related trade association released the results of a survey indicating that a significant percentage of single-family home sales were of investment properties and second homes. This was widely interpreted as evidence of speculative buying of rental properties, another feature of a housing bubble. Again, I believe that such reports should be viewed skeptically. According to the AHS, in 2003 single-family investment properties, defined as homes renter-occupied or for rent, represented 14.2 percent of all single-family homes while second homes represented another 4.7 percent. Therefore, we should not be surprised that such properties represent about 20 percent of the sales that take place at any point in time. Moreover, the American Housing Survey data indicate that single-family investment properties have been declining as a share of all single-family homes for some time and declined in absolute numbers from 2001 to 2003. A principal reason the rental vacancy rate for single-family homes has risen in recent years is that the number of renter-occupied single-family homes has declined as people switch from renting to owning. So if a lot of people are buying rental properties, it must also be the case that a lot are selling as well. That concludes my report. Thank you.

  • I will review some general issues related to monetary policy and asset prices. Let me start—at the top of page 1—by assuming that an asset price can, in theory at least, be separated into a component determined by underlying economic fundamentals and a non-fundamental or bubble component. An asset price may be in line with its fundamentals, so the bubble component is zero, or bubbles could be positive or negative—perhaps representing irrational euphoria or pessimism.

    Two types of monetary policy responses to movements in an asset price have been proposed. I refer to the first type as “Standard Policy” because there is widespread agreement that it represents the minimum appropriate policy response. The Standard Policy responds to an asset price to the extent it conveys information to the central bank about the future path of output and inflation—the goal variables of monetary policy. For example, a booming stock market is usually followed by higher demand and increased inflationary pressures, so tighter policy would be needed to offset these consequences. Even for the Standard Policy response, it will likely be useful to identify, if possible, the separate components of the asset price. In particular, the bubble component may exhibit more volatile dynamics and be a pernicious source of macroeconomic risk, so optimal policy would likely react more to bubbles than to movements in the fundamental component.

    The second type of response, the “Bubble Policy,” follows the Standard Policy as a base case, but, in certain circumstances, it also takes steps to contain or reduce the asset price bubble. Proponents of a Bubble Policy argue that movements in the bubble component can have serious adverse consequences for macroeconomic performance that monetary policy cannot readily offset after the fact, so it is preferable to try to eliminate this source of macroeconomic fluctuations directly. Furthermore, because bubbles often seem to display a self-reinforcing behavior, a little preemption and prevention early on can avoid later excesses.

    A best-case scenario for these two policies is illustrated in the lower half of the first page. Under ideal circumstances, the policymaker knows the fundamental and bubble components, and as history unfolds, the Standard Policy would likely recommend higher interest rates to offset any economic stimulus generated by the bubble before the crash and lower rates afterward. A Bubble Policy would go further and try to mitigate the fluctuations in the bubble and achieve an asset price path like AP′t. This would likely require higher interest rates than the Standard Policy before the crash and lower rates afterward, and it will likely trade off near- term deviations from the central bank’s macroeconomic goals for better overall macroeconomic performance later on. The fundamental difference between the two policies is that the Standard Policy takes the bubble component essentially as given or exogenous, while the Bubble Policy takes into account the endogenous nature of the bubble component—specifically, a linkage between the policy instrument and the bubble.

    A decision tree for choosing between the Standard and Bubble Policies is shown on page 2. In brief, it asks three questions: (1) Can policymakers identify a bubble? (2) Will fallout from a bubble be significant and hard to rectify ex post? and (3) Is monetary policy the right method to use to deflate the bubble?

    The answer to the first question—can policymakers identify a bubble?—is “no” if the particular asset price appears aligned with fundamentals. Some have argued that this is nearly always the case because estimates of fundamentals are so imprecise and because asset prices reflect the collective information and wisdom of professional traders in organized markets. If policymakers cannot discern a bubble, then the Standard Policy is the only feasible response.

    But suppose an asset price bubble is identified. Then the second hurdle is whether bubble fluctuations have significant macroeconomic fallout that monetary policy cannot readily offset after the fact. A negative answer to this question is appropriate in two situations. First, if the bubble is in an asset market that is small in domestic economic terms—for example, a localized real estate market—then a central banker should avoid attempts at asset price realignment. Second, even when there are significant macroeconomic consequences from an asset price bubble boom and bust, if they occur with a sufficient lag so the policymaker can adopt a wait-and­ see attitude, then the Standard Policy is again appropriate. This second case seems relevant if fluctuations in the bubble component have only conventional effects on aggregate demand and supply through changes in wealth, the cost of capital, and balance sheets. Then, to a first approximation, the lags involved in these channels are about as long as the lags in the monetary transmission mechanism; therefore, the Standard Policy should suffice. For example, fluctuations in equity prices will affect wealth and consumer demand, but a nimble central banker can essentially offset these consequences by changing interest rates in reaction to—that is, after—the equity price movements.

    Now to the case where asset price movements have significant macroeconomic consequences and those consequences are hard to clean up after the fact through monetary policy. The most often mentioned possibility is that a bursting asset price bubble will lead to a broad financial crisis and credit crunch. Such financial instability is likely to be transmitted to the economy much more quickly than can be offset by interest rate policy. This may set the stage for invoking a Bubble Policy. Another example is when the asset price misalignment results in significant misallocations of resources, which distort aggregate demand and supply across sectors and over time and impede the achievement of the highest possible long-run economic growth. For example, the dot-com bubble spurred overinvestment in fiber optic cable and decimated the provision of venture capital for new technology startups for years. Of course, after the fact, it is difficult to unwind these problems with the blunt instrument of monetary policy, and, depending on the specifics, it is possible to conceive of a situation in which reducing the bubble in advance is a preferred policy strategy.

    The final hurdle before invoking a Bubble Policy involves assessing whether monetary policy is the right way to deflate the asset price bubble. Ideally, for the Bubble Policy, a moderate adjustment of interest rates could constrain the bubble and greatly reduce the risk of severe future macroeconomic dislocations. However, bubbles, even if identified, often do not appear influenced by monetary policy actions in a predictable way. Furthermore, even if changing interest rates could alter the bubble path, such a strategy may involve substantial costs, including near-term deviations from the macroeconomic goals of the central bank as well as potential political or moral hazard complications. Finally, even if monetary policy can affect the bubble, alternative strategies to deflate it, such as changes in financial regulations or supervision, may be more targeted and have a lower cost.

    This decision tree is a daunting triple jump, but not a total prohibition against bubble reduction. Indeed, page 3 suggests different real-time answers to these three questions for two historical episodes. The first episode is the run-up in the stock market during the late 1990s. In 1999 and 2000, one could have made the case that there was an equity price bubble in the high-technology sector and perhaps in the overall market as well. Also, during that time, the possible capital misallocation from the run-up in prices and the possible financial instability that might have followed a bursting of the bubble may have appeared difficult to rectify. However, it was also unlikely that monetary policy could have deflated the equity price bubble without substantial costs to the economy. In the event, of course, a Bubble Policy was not followed, but arguably, the consequences from any bubble boom and bust have been manageable.

    A different example is provided by the bond market collapse in 1994. One could argue that this “inflation scare,” which pushed up yields on 30-year bonds by over 2 percentage points, resulted in an asset price misalignment that was fairly apparent to the FOMC during the second half of 1994. If this bond bubble had persisted, the widespread propagation of the associated fears of higher inflation could have had severe consequences that would have been costly to unwind with monetary policy later on. Finally, with regard to deflatability, it did appear likely that monetary policy could guide bond prices back to fundamentals. Indeed, one interpretation of the FOMC’s actions in 1994 is that it purposefully and successfully contained a bond market bubble with sizable increases in the funds rate. It is an open question which of these two episodes is the more relevant one today. That concludes my remarks.

  • I’ll be referring to the exhibits beginning on page 26. In my presentation today, I’ll lay out a few scenarios that illustrate the potential macroeconomic fallout resulting from a significant decline in house prices, and I will examine policy responses that minimize it.

    I’ll start by describing the possible size of the current problem—assuming there is one. As pointed out in Dick Peach’s presentation, there are serious difficulties in accurately measuring both actual and “fundamental” house prices. But, for the purposes of my presentation, I will take as a working hypothesis that house prices are high relative to fundamentals—or, in terms of the decision tree that Glenn just laid out, I assume that the answer to his question 1 is “Yes, the asset price appears misaligned.” As Josh Gallin indicated, it would take up to a 20 percent decline in house prices to bring the price-to-rent ratio back in line with fundamentals. With housing wealth standing at around $18 trillion today, such a drop in house prices would extinguish $3.6 trillion of household wealth. That’s equal to about 30 percent of GDP. Based on a marginal propensity to consume from housing wealth of 3½ cents on the dollar, this decline in wealth would entail a nearly 1½ percentage point increase in the personal saving rate. And, according to estimates from the FRB/US model, it implies a 40 basis point reduction in the long-run neutral real funds rate.

    It may be useful to put these figures into context by comparing them to those associated with the stock market overvaluation reached in early 2000. Stock prices at that time were arguably some 50 to 70 percent overvalued. Correction of prices to fundamentals at that time would have implied a reduction in household wealth of $6.7 trillion, equal to about 70 percent of contemporaneous GDP. In the event, stock market wealth fell by $4.6 trillion between March 2000 and March 2001, and at its lowest point was down $8.5 trillion. There is considerable uncertainty regarding the magnitude of the effects of changes in stock market and housing wealth on household spending; nonetheless, it seems clear the magnitude of the current potential problem is much smaller than, and perhaps only half as large as, that of the stock market bubble. Of course, if house prices continue to soar—and in the San Francisco Bay Area, at least, they show no signs of slowing—the magnitude of the housing overvaluation problem will rise as well.

    A cautionary note worth emphasizing is that the monetary policy cushion available today, as measured by the prevailing federal funds rate, is noticeably smaller than it was in early 2000 at the peak of the stock market.

    The first question that comes to mind is: What should monetary policy do, if anything, about the apparent overvaluation in house prices? The answer to that depends crucially on the answer to Glenn’s second question: “Do bubble fluctuations result in large macroeconomic consequences that monetary policy cannot readily offset?” Therefore, I now explore the effects of a bubble collapse and the ability of policy to respond effectively to them.

    I consider three scenarios in which a housing bubble deflates relatively quickly. I use the FRB/US model to quantify these effects. Note that for the materials posted last Wednesday, I based my simulations on the April Greenbook projection. I have since updated the simulations, and the ones that I will be showing today are based on the extended June Greenbook projection. In each scenario, I assume that house prices fall by 20 percent relative to the baseline over the next 2½ years.

    In the first two scenarios, I consider the effects of exogenous declines in house prices that apparently come “out of the blue;” the third scenario considers the possibility of an interrelated decline in both bond and house prices. For each scenario, I consider two types of monetary policy response. The first is the optimal perfect foresight policy that is very similar to those reported in past Bluebooks. The path of policy is chosen to minimize the sum of the squared deviations of the unemployment rate and the core PCE [personal consumption expenditures] price inflation rate from their respective targets and the squared changes in the funds rate, with the latter receiving relatively little weight. Perfect foresight means that the policymaker today has advance knowledge of all future shocks to the economy. I assume that the inflation objective is 1½ percent and that the unemployment rate target is 5 percent, equal to the staff’s estimate of the NAIRU [non-accelerating inflation rate of unemployment]. The second policy is one that responds only to events as they occur and does not respond directly to anticipated future developments related to house prices. For this purpose, I use a modified version of the Taylor rule, with a coefficient of 1 on the output gap and of ½ on the inflation gap, and with the long-run natural rate of interest that appears in the rule varying in accord with sustained changes in housing wealth and bond premiums.

    The second page of my exhibit shows the results from model simulations of a 20 percent decline in house prices, where only the standard channels included in the FRB/US model are in play. For comparison, I have also plotted the baseline paths, based on the June extended Greenbook projection but modified under the assumption that monetary policy is set optimally in the way I just described. Because a decline in house prices primarily influences demand, not supply, it does not pose a difficult tradeoff between policy goals. In addition, according to the model, the macroeconomic effects play out gradually and are moderate in magnitude, giving policy time to respond. The optimal policy calls for a path for the funds rate that averages about 3¼ percent during 2007 and 2008—about 40 basis points below the baseline path. Under this policy, the unemployment and inflation rates are nearly the same as in the baseline. The very small rise in inflation reflects the effects of the depreciation of the dollar resulting from the reduction in domestic interest rates.

    The modified Taylor rule is able to mimic the outcomes under the optimal policy reasonably well, indicating that policy need not fully anticipate future house- price declines to be effective, but can simply respond to events as they unfold. Note that if the house-price decline were larger (or the marginal propensity to consume out of housing wealth bigger), then the policy implication would simply be to cut rates by proportionally more.

    In summary, assuming that the FRB/US model does a good job of capturing the macroeconomic implications of declining house prices, such an event does not pose a particularly difficult challenge for monetary policy. One lingering concern, however, is that the model may be missing other important avenues by which large movements in house prices affect the economy. I now consider a scenario that entertains that possibility.

    In the second scenario, I incorporate extra-model spillovers from falling house prices, presumably reflecting a decline in confidence and an extinguishing of household spending that had been fueled by equity extraction from mortgage refinancing, as described in the “heightened spending response” simulation reported in the June Greenbook. Importantly, the demand spillovers are assumed to have a very different dynamic from the model’s standard wealth effect: They kick in much more rapidly and eventually dissipate so that they have no long-run effects. As shown in the next page of the exhibit, owing to the extra drag from these spillovers, the optimal policy calls for a series of rate cuts, bringing the funds rate to 2¼ percent by the middle of next year. The funds rate does not again reach 3 percent until the spring of 2007. The optimal policy is able to stem the rise in unemployment that would otherwise occur at the cost of a modest and short-lived uptick in inflation.

    The Taylor rule, however, is not as successful. It fails to anticipate the spillover effects and responds too timidly once they occur. Still, it contains the rise in unemployment to only about ½ percentage point above baseline and moves inflation slightly more rapidly toward the assumed inflation objective. I should note that, given the uncertainty regarding the size and timing of such spillovers, the ideal outcome in the optimal policy simulation exaggerates the real-world ability of monetary policy to offset the effects of such shocks.

    As I noted before, the thought experiment behind these first two sets of simulations is that house prices fall in a kind of vacuum, without any relationship to other events. Some commentators have argued that the current high level of house prices is the outcome of a history of very low interest rates and past house-price appreciation that has given rise to irrationally optimistic expectations of future appreciation. Indeed, a simple estimated equation relating the current price-to-rent ratio to the user cost of housing and past house-price appreciation does a reasonably good job of explaining much of the run-up in house prices over the past several years. If this explanation holds water, a potential risk to housing prices and the outlook in general lies in the path of longer-term interest rates—which have been surprisingly low, given prevailing economic conditions—and the usual behavior of term premiums. Bond yields could return to more normal levels and in so doing contribute to a downward trajectory in house prices.

    I explore such a possibility in the final scenario, which builds on Scenario 2, and is the subject of the next page of my exhibit. I now add the assumption that the term premiums on long-term bonds rise by 75 basis points, relative to baseline, over the second half of this year and remain at these levels. This aspect of the scenario is similar to an alternative scenario reported in the June Greenbook. This shock to bond premiums by itself reduces the long-term equilibrium real funds rate by about 70 basis points. Optimal policy calls for the funds rate to fall below 1 percent by the middle of next year and for the funds rate to remain below 3 percent through 2008. The optimal policy is just able to contain the rise in unemployment without confronting the zero bound. The Taylor rule, on the other hand, responds too gradually to events, and, as a result, the unemployment rate reaches 6 percent in early 2007.

    This scenario presents a difficult challenge for monetary policy, especially in light of the looming zero lower bound on interest rates under the optimal policy. More generally, it highlights that the risks posed by a house-price decline are magnified if they occur in tandem with other events that damp economic activity.

    In addition, as Glenn mentioned, a house-price misalignment may misallocate resources to housing-related activities. These conditions suggest considering Glenn’s question 3, whether an alternative policy is needed that aims to deflate a house-price overvaluation before it grows too large—the subject of the next page of my exhibit. The potential effectiveness of such a preemptive policy, however, is less than clear. The basic empirical relationship between house prices, interest rates, and other factors, upon which such a policy necessarily would rely, is imperfectly understood and may have changed over time. Moreover, as seen in the earlier presentations, there remains considerable uncertainty over the degree of overvaluation. Thus, the successful use of monetary policy to reduce the magnitude of a misalignment of house prices would be a daunting task, even assuming that such a goal were deemed desirable. This concludes our prepared remarks.

  • We’re here to answer your questions and to listen to your discussion.

  • Well, I believe there is a very critical statistical difference here which, before we get very far down the road, I think requires resolution. You three gentlemen are talking about different sets of reality, and you’re using the same basic data system. You all use the American Housing Survey, you use OFHEO, and you use the Census price index. I’ve heard comments from a few of you periodically on such matters.

    There are a couple of issues that I find striking here. One is how little we apparently know, at least as has been communicated to me, on the effect of the underlying land prices that are involved in all of these indexes. You can’t get the land price out of the American Housing Survey. I don’t believe it’s in the decennial housing surveys. We do have farmland prices by county, and one would presumably be able to pick up what was going on at the fringe of various residential operations—it sounds like a tortuous job—but that still doesn’t solve the underlying question of what the ratio of land is to other values. And it strikes me that we need to come to grips with some of these issues, if for no other reason than I think it’s probably our job—because no one else seems to be taking it on—to try to second-guess the data that we’re getting. Nobody is raising the question about the quality of the Loanperformance data and the fact that we’re dealing with private pools, which I presume are in the data that are being put together with respect to the proportion of IO loans. I assume that conforming mortgages do not include IOs. Or do they? Are there parts of these pools that do?

  • An interest-only mortgage can be purchased by a GSE if it’s below the conforming loan limit.

  • And how much of the GSE portfolio includes IO loans? Do we know?

  • At the moment only Freddie Mac, I believe, has a pilot program to buy these, and it’s very small.

  • That was my understanding. Does Fannie do any?

  • So, for all practical purposes, since private pools have gotten very large but are still somewhat less than the sum of Fannie and Freddie, the Loanperformance data at best are large exaggerations.

    There’s also an interesting question about the second house issue. You’re raising the data out of the American Housing Survey, and the last one was for 2003, but the actual acceleration of both turnover and presumed purchases occurs after that. We’ve been getting HMDA [Home Mortgage Disclosure Act] data on that for 2004, essentially doing it by month. In there, of course, are the mortgage originations for other than owner-occupied residences. One would expect that there’s a little fudging on the part of investors about what the homes are to be used for, but if the bias is presumed to be fairly standard, we’re going to learn something about the apparently uncontested assumption of the acceleration in turnover of existing homes and, indeed, the big increase in new home construction. All of that will be disaggregated into purchases for ownership, for investment, and for vacation homes.

    The data that we have seen—confidential data from individual banks—do show a rather significant pickup. The numbers don’t get as high as those from the National Association of Realtors—23 percent or something like that—but they indicate a fairly substantial pickup. And of the sum of second homes, a disproportionately large number are purchased by investors; the investor share very definitely has accelerated. This is one set of data—for the very large banks— that does show a marked acceleration into 2004 and 2005. So, I think we need to know much more about this.

    But let me just ask one simple question: Do we have any insight on the land component of those sales?

  • Sir, I can address that a bit. Actually, Morris Davis, who is an economist here at the Board, has written a paper on that very subject, and—

  • Was that a year or so ago?

  • Just about a year ago, exactly.

  • Okay. I can tell something from the tone of your voice. [Laughter] In any event, the idea there was to look at house prices and construction costs and to try to back out as a sort of mechanical operation—not an easy one—the value of the land. He shows that land prices have increased very rapidly and that the land share of the property values has gone up over time. As you mentioned, we know something about farm prices. But besides those two things, there really aren’t—

  • As an example, the Chicago Bank publishes quarterly data— fairly detailed data—on land prices in all the states in its District. Does anybody else?

  • Agricultural land, yes.

  • Yes. These are your data? You collect them?

  • Yes, we collect the data.

  • So you actually have data on each individual farm. You have a database from which one could actually construct the price changes over time to at least learn what is happening as the urban fringe moves out. So you have at least one dimension, which is this issue that you raise with respect to new homes and existing homes. Has anybody tried to look at that? I ask because they are the only data that I’m aware of which actually show a level of land prices over a significant period of time. And no one uses them on the grounds that agricultural land prices have nothing to do with residential land prices. But that can’t be true.

  • I agree, Mr. Chairman. Those data are being influenced in part by the expansion of residential areas out into what used to be farms or agricultural land. There are some interesting dynamics taking place because, in the Chicago metropolitan area, a lot of this land is now being used for residential purposes. However the sellers, for tax purposes, want to use the proceeds to buy agricultural land right away. So they then go downstate in Illinois and bid up the prices there. They are not as concerned about the price as long as they can make this tax-free exchange. So, that’s one dynamic. The other interesting dynamic, of course, is that as agricultural subsidies have increased so rapidly—they are now about 50 percent of net farm income—there’s less concern about the value of land coming down.

  • It’s capitalized into the land price.

  • But you still have to pay the agricultural subsidy to buy the land to build a house. There’s no way to get out from under that.

  • Mr. Chairman, we have, in fact, looked at farmland values and at the data that are collected by the Reserve Banks, and there has been an appreciable acceleration in farmland prices. It varies a considerable degree by District. For example, in areas like the San Francisco District or the Richmond District, where the interface of the urban expansion into agricultural areas probably is greater, we’ve seen a much more pronounced acceleration than, for example, in the Dallas District.

    As we put that all together, it is true that we don’t have a good series broken down by regional land prices per se. But it’s very difficult to believe that the acceleration we’ve seen in residential property prices doesn’t reflect to a very significant degree an increase in land prices. While it may be difficult to do Morris’s type of calculation, we just haven’t seen anything like the acceleration in basic construction costs that would suggest anything other than a substantial increase in land values. And when we looked at those farmland prices to see if there was some confirmation of that impression in the agricultural land prices, it seemed pretty clear that we were seeing that. Again, that doesn’t necessarily mean that the farmland is overvalued. I think the same questions are open on the agricultural land prices as they are in Dick’s and Josh’s debate about whether or not housing is overvalued, but there has been a substantial step-up in recent years.

  • We do obviously have BEA [Bureau of Economic Analysis] data on transaction costs and, one would presume, profit margins, so that in a way we have at least one element of the markup over cost. The differences that we’re looking at here are so large that it’s not a minor question. Indeed, if you take Dick’s bubble—or lack thereof, I should say—you get the impression that deflation is probably going on. [Laughter] I’d be curious, frankly, to have Andreas, who has looked at this to a great extent, comment on Dick’s paper.

  • I think Josh and Dick should have an opportunity to talk a little more about their differences of view in this area.

  • Not that I don’t have an opinion. [Laughter]

  • I thought you’d be the ideal intermediary. [Laughter]

  • Well, let me make just a couple of observations.

  • Let me stop you. You’ve made your point. I think you were responding to the data of your two colleagues. Let them respond, and then you respond to them.

  • Loosely speaking, what we’re trying to do is to get a sense of changes in the price of a typical home—what has been happening to that price over time—adjusting for quality. I guess I’ll start with my concerns about the constant-quality figure, and then I’ll address Dick’s concerns about the transactions price index that I used.

    One reason I worry about the constant-quality number, and Dick alluded to it, is that it’s a constant-quality price index for new homes, and new homes typically are not built where existing homes are already situated. You can see that when you drive out of any city, basically. The big developments are on the outskirts of the cities, and those outskirts are moving further out. I recognize that cities can be multi-centered and so forth, but new homes—while not always built on the cheapest available land—typically are built on the cheapest available land. And the constant- quality price index for new homes has only crude controls for location. So I don’t think that index is able to fully handle that issue. That’s why I like the repeat-transactions price index. As Dick mentioned, it basically matches up addresses over time and looks at the price changes.

  • Have you adjusted that for the modernization per housing unit?

  • How much does that take off?

  • Well, as Dick pointed out in his briefing, there are data on improvements. So I use those data and divide that by the stock of housing and subtract it off. But on the other side of that argument is the fact that while many homes get additions made to them, all homes are also deteriorating somewhat over time, which will bias things in the other direction. And we have data on that. So, improvements are affecting price appreciation one way, and depreciation is affecting it the other way. Once you take both into account, the net effect is only a few tenths a year. Now, these things can cumulate over time, but I use data on improvements and on depreciation to remove those biases from the published data. So, there are problems with the published data, but we don’t have to make do with the published data.

    Likewise, there’s a transactions bias. Homes that appreciate more tend to transact more, and that’s going to bias upward the published price index that I start with. But there are studies on the size of this transactions bias. Dick provided some data on it. I’ve taken estimates of the size of this transactions bias and removed it from the published data. So, I start with the published data, but I make various adjustments. One last point is that because the price index is based on mortgages purchased or securitized by the GSEs, it’s going to miss FHA and VA loans and also jumbo loans.

    There is a published private sector price index put out by Case and Shiller that actually goes to deed-level data, and they look at all home purchases. They do a repeat-transactions method, but they look at all purchases and throw out only refinancings and exchanges among families. That index actually shows faster appreciation since 1999. That would seem to suggest that the homes with jumbo loans may be appreciating faster and that by leaving out jumbos I am perhaps biasing things down.

    So, my final takeaway from this is that there’s no perfect way to measure house prices. All these measures have their various flaws. But I think that a properly adjusted repeat-transactions price index—one that uses not just the published data but makes adjustments for the shortcomings—is a better way to get a read on what’s going on in the housing market.

  • I have just three comments. First, I’m not disputing that land prices where houses are being built would be lower than land prices closer to the center of a metropolitan area. The issue is the rate of increase of those land prices. And we don’t really know what’s happening to the rate of increase of land prices at different points along the continuum.

  • It’s the weighting effect. As you move out, the price levels are lower, so there’s a downward bias, whereas for existing homes that bias can’t exist.

  • That’s true. That assumes that there is this nice gradual gradient of price change for land as one moves out from the center of a metropolitan area. I don’t know the answer, but my observation is that that is not what occurs. In the New York metropolitan area, for example, I think one could cite many locations where land prices closer to Manhattan, say in Newark, are not rising as rapidly as in Short Hills or in Hunterdon County. So, that relationship probably just doesn’t hold in reality.

    Another thing that we’ve observed, particularly in our metropolitan area, is that state and local governments are now aggressive bidders for parcels of land that are potential sites for residential development. So it’s conceivable that the price of land on which one could build a home now is rising extremely rapidly—and perhaps even more rapidly than land closer in that has already been developed.

  • Is there any agreement about what proportion the land is currently—some rough order of magnitude—relative to the total value of homes sold? For example, is land 10 percent of the sales price, on average, or 20 percent or 30 percent?

  • I think it varies substantially across locations. I think Morris found it to have become quite high; he estimated that the number had moved up above 50 percent.

  • The average lot size also has been declining rather rapidly, indicating that the price of land is going up rapidly.

  • I’m not asking for the change, I’m looking for the level. Governor Gramlich.

  • Well, one more question on the data. Josh, you need the share of land to do the price-rent ratio, don’t you? You adjust for improvements, but surely that’s only on the structure, not on the land.

  • So the share of land in the value is increasing. Is that extracted from your price-rent ratio in some way?

  • Yes. What you do is get the depreciation rate, and that’s relative to the structure. It actually doesn’t make too much of a difference exactly how you do it. But, again, I got land share numbers from Morris Davis, and one can basically tamp it down year by year.

  • Okay. Then it’s also a ratio. And the CPI tenant rent measure is for a constant unit over time. So what you try to do is to normalize on that?

  • Yes. My price series and my rent series are both indexes to start with, so they are unit-less. Again, leaning on my colleague, Morris, we get a base-year estimate of rents for a constant-quality unit. We get the base-year dollar price and then multiply by that.

  • Okay. John, I have one question for you. How do we interpret the optimal policy? We, the policymakers, know all the shocks. So when we move the funds rate on the red line, is that to be interpreted as acting in advance of the house-price change, simultaneously with the house-price change, or what?

  • In the optimal policy simulations, from the beginning of the simulations, which is the third quarter of this year, you know the entire future path of house prices and any other shocks that I add. So you are acting in anticipation of those future price changes.

  • So if we see house prices going up, we’re in effect cutting the funds rate while the house prices are going up?

  • That’s correct. That’s why I was trying to compare it with the Taylor rule—which doesn’t have that aspect of responding to developments as they unfold—to see how well that would do.

  • I wanted to make one other point with regard to Josh’s work, and that is that using the tenant rent component of the CPI and comparing it with the OFHEO index is, in my opinion, a real apples-to-oranges type of comparison. As I pointed out in my comments, the OFHEO index measures what is happening around the 75th percentile of the distribution of single- family homes where we know incomes are rising more rapidly. Using a rent component that on the one hand is looking at what is going on in the multi-family housing market to a large extent—

  • Actually, it’s based on single-family homes which are rented but have the characteristic that they could be sold.

  • Not in the tenant rent component he is using.

  • The rent component I presume we’re using is the owners’ equivalent rent, are we not?

  • Actually, it’s not. But it doesn’t make any difference. [Laughter] I could easily have done it with the owners’ equivalent rent, but I would have had to cut off all my pictures in 1983 because that’s when that series begins. I went with the tenant rent strictly for the length of the time series. But if you look at the relative levels of these two series over time, while the growth rates do differ a little bit, you’re talking about really small potatoes—on the order of 2 percent.

  • If you keep getting a significant rise in home ownership— meaning a shifting of existing families from rent only—you are going to bias that ratio.

  • Absolutely. As an aside, but regarding that particular point, I’ve been working one day a week at the Bureau of Labor Statistics, and I have access to the confidential underlying micro data that they use to build up the CPI. So, some of my work is trying to get at just these types of criticisms. But from the underlying data, I have been able to look at rent appreciation for different segments of the rental market, trying to see—with home ownership rates going up— where the homeowner market is stealing, so to speak, from the rental market and how big the biases are. And it does look as if the higher end of the rental market has been showing somewhat smaller rent increases over time, suggesting that that’s where it’s coming from, but not dramatically.

  • That raises an important point because, as we know, wages of nonsupervisory workers, which are essentially payroll data, are going up 3 percent, whereas implicitly that of supervisory workers is going up double or three times that. That’s changing; it has accelerated in the most recent period.

  • Let me note just one last thing. The pictures are going to look the same as those Dick showed, whether we look at house prices relative to income or to rent. So, even if the rent series might have this problem or that problem—and it does have problems—house prices have grown quite dramatically using the adjusted data that I used relative to incomes as well.

  • I guess my question is addressed to Glenn and John. Obviously, the United States is unique, but we’re not the only country where a central bank has been trying to deal with this issue. We’ve seen recently that monetary authorities in the United Kingdom, Australia, and perhaps the Netherlands are all talking about the possibility of using monetary policy to try to deal with the house-price issue. So, what, if any, lesson can we learn from overseas? I recognize that our particular situation is unique, but we’re not the only ones dealing with this problem.

  • That’s an interesting question. I think one of the points to be considered is that those countries are all smaller than the United States, and typically their residential property markets are centered in one or two cities—London or Sydney, for example. So perhaps that’s a factor that just looms larger in the policy calculation. And that may be the reason why the central bankers in those countries have been perhaps a little more proactive, although that issue is still being worked out, I think. They have done a bit more jawboning, so to speak. But Australia, in particular, has been a little more explicit about factoring in housing market developments in the policy calculation. Again, they don’t target house prices, but they certainly seem to take them into account a bit more than the Unites States does, in terms of setting monetary policy.

  • But have the results, generally speaking, been positive or neutral? I know this policy is in its early days, but is there any judgment one could make even at this early stage on whether or not more jawboning—or, in the case of Australia, actually moving rates—has been helpful?

  • I think both Australia and the United Kingdom, to the extent they have conducted this type of policy, have felt it was fairly successful. They have been able to temper home price appreciation and restore some balance in the economy without any significant macroeconomic fallout. They have had lower home price appreciation but, again, their situations are perhaps very different from the situation for the United States.

  • I’d like to comment also on the policy discussion. It seems to me that to do this analysis correctly, the central bank really has to think about intervening more or less all the time because, if you have a one-off policy response and you promise that you’re never ever going to do it again, it perhaps would not be all that useful. So it seems to me that policymakers would need to think about routinely having asset prices be a consideration in policy analysis.

    I think that would be a terrible idea, and let me explain why. Let’s look at it from an optimal control point of view and suppose hypothetically that we have a policy instrument that is absolutely ideal. It has no side effects, is completely independent of the federal funds rate, or the macro policy instrument, and has perfectly predictable effects. That’s the ideal from a control- theoretic point of view. In that case, if the central bank were adjusting such an instrument and setting capital values, it seems to me that that would change the whole nature of the pricing mechanism in asset markets. And I think it would be a terrible idea in a market economy to have a government agency setting capital asset values. That’s why I would not start to go down this route; I believe it is really a dead-end.

  • I guess I’d make a distinction there, in that I think the proponents are talking not about setting asset prices in general but about trying perhaps to reduce the bubble component. Clearly, if asset prices are set equal to the fundamentals, that’s going to lead to the proper functioning of—

  • But the point is that there could be a little bubble component or a big bubble component. Obviously, policymakers wouldn’t apply the instrument when prices are regarded as reflecting the fundamentals, but the instrument would be continuously available. The whole point of such an instrument would be to keep the price right at the fundamentals. But then that would change the whole nature of asset pricing in a market economy. I think a proper policy analysis really can’t be done on a one-off basis in cases such as the stock market in the ’90s or house prices today.

  • I’m sympathetic to that view. It seems as if there might be a threshold. That is, policymakers might be more interested in certain asset markets or might be more interested in prices in asset markets in general at certain times. But I think you’re alluding to the moral hazard or political complications that could arise from this type of intervention or from trying to use this bubble component as a monetary transmission mechanism.

  • Essentially, it’s the same problem that arises with wage and price controls. Having that as a policy instrument that is sometimes used and sometimes not used completely changes the pricing mechanism in a market economy. I think that problem applies in spades to stock prices, equity prices, and bond prices. With house prices I think the problem is a little different, but I believe the same basic principle applies.

  • People also have made the argument that if policymakers are going to clean up the mess after the fact, there may be other problems. If you let equity prices rise on the upside but then essentially make some sort of insurance agreement on the downside, that again is a distortion—though perhaps an asymmetric one—that may also have complications.

  • There’s ample room, I think, for understanding the housing market within what you phrased as the standard policy. But I think going after asset prices directly is something very different. That’s the point that I’m trying to make.

  • Thank you, Mr. Chairman. I just wanted to make a couple of comments and also ask a quick question. My first comment relates to the run-up in price-rent ratios that we’ve seen. It seems to me that there might be a couple of factors that could explain at least some portion of the run-up, though probably not all of it, that weren’t mentioned in the presentations.

    First, it seems to me that financial innovations affecting housing could have improved the view of households regarding the desirability of housing as an asset to be held in portfolios and thus raised the equilibrium price-to-rent relationship for residential real estate. What I’m thinking of is the idea that equity held in residential real estate is a lot more accessible today than it has been in the past. Home equity credit at commercial banks is up fourfold since 1999, and many households obviously are now keenly aware that refinancing provides a low-cost avenue for tapping into the equity in their homes. So, in a sense, there might be less of a liquidity premium embodied in the return for housing. Also, if people feel that the liquidity constraints in holding housing as an asset are diminishing, that could explain a reduced need for precautionary saving in traditional liquid assets. It could even make people willing to put more of their wealth into down payments on houses and may have raised prices through that mechanism.

    The other thing that occurred to me is that there might be effects from tax changes. We’ve had changes in the rules for tax exemption and in 1997 on capital gains from the sale of primary residences that would make holding real estate assets more attractive. And the changes in capital gains taxes more generally in 1997 and then again in 2003 would have worked in the same direction.

    One of the things that we looked at that we thought was interesting was the behavior of price-rent ratios for residential housing and for commercial office space. Commercial office space price-rent ratios are highly cyclical—I guess they always have been—but it appears that the behavior of price-rent ratios in residential housing has closely mirrored what we’ve seen in commercial office space. The ratios for both have gone up about 30 to 35 percent since around 1998, though the dynamics are totally different. Commercial office space rents have been falling— it’s not that the prices have been rising—but the price-rent ratios have moved very similarly.

    A second comment I wanted to make concerns the relationship of creative finance to the housing market. One view that I think is very prevalent is that the use of credit in the form of piggyback loans, interest-only mortgages, option ARMs [adjustable-rate mortgages], and so forth, involves financial innovations that are feeding a kind of unsustainable bubble. But an alternative perspective on that is that high house prices, in fact, are curtailing effective demand for housing at this point and that house appreciation probably is poised to slow. So the increasing use of creative financing could be a sign of the final gasps of house-price appreciation at the pace we’ve seen and an indication that a slowing is at hand. Previously, lenders applied very rigid constraints on loan-to­ value ratios, but essentially those constraints are now being eased at the margin through these creative financing techniques. And that’s providing some elasticity to what was a firm roof. It may slightly diminish the price elasticity of the demand for housing, but the fact that it is blossoming now basically suggests that we really are at the ceiling where it’s binding and will ultimately constrain appreciation.

    Finally, with those two comments, a question. It concerns the presentation by Andreas and the numbers cited on loan-to-value ratios at origination. One of the things we’re seeing in California and elsewhere in our District—and maybe this is true nationwide—is a growing use of piggyback loans. Loan-to-value ratios of 90 to 95 percent are common in California, and we’ve even seen combination loan-to-value ratios and piggyback loans going up to 125 percent. I guess that means two things, one of which is that the traditional first mortgage looks utterly conventional. Those mortgages have an 80 percent loan-to-value ratio and I suppose they are being sold off to Fannie and Freddie. The other thing is that with such conventional mortgages being sold to Fannie and Freddie, there’s no need for private mortgage insurance. So Fannie’s and Freddie’s books may look better in some sense—less risky—than they really are because of all of the second mortgages going up to possibly 125 percent.

  • It sounds like a CDO [collateralized debt obligation]. That’s what it is, isn’t it?

  • Yes. So I wondered if that was something that you’re aware of and something that is included in the numbers.

  • Yes. The data that you see on my first exhibit in principle include second liens, closed-end second liens. And in particular in the graph that you cite I use the newly available 2004 data from HMDA which for the first time has collected data on whether or not the HMDA loan is a junior or a first lien. So the picture that you see there should reflect the growth of piggyback loans.

  • Could I just ask a quick question? Do we have data on price­ to-rent ratios of apartment dwellings? We do have data on prices of apartments, and we have rent. I think Dick raised a very interesting question about the numerator and the denominator. The way to square that is to get the same data for the same type of unit. Do we have that information and what does it show?

  • We do have those data. There’s an index that gives price per square foot and rent per square foot for apartments nationally and in various cities. And it shows a marked increase in the price-rent ratio.

  • Does it look like the ones that you were showing?

  • It doesn’t go back as far—it only goes back to 1986—but since 1996, yes.

  • It’s quite a steep increase.

  • Is the level the same, can you tell?

  • These apartment price data and rent data are by square footage. I guess I could make a calculation, but I don’t have those data here.

  • It’s worthwhile looking at that. Sorry about the interruption. Governor Olson.

  • I have a couple of questions. First of all, concerning the scenario of a 20 percent decline in housing prices: Is that weighted to any extent? Is it weighted so that it would reduce the froth or the increased valuation that had occurred up to a certain point? Or would that simply take the total amount of value and reduce that number by 20 percent?

  • All I’m doing is the latter—reducing the overall housing wealth by 20 percent relative to the base.

  • With Josh’s 20 percent, I assume that there is an implicit weighting, is there not? Would the 20 percent increase in value reflect to some extent the greater run-up in value in areas like California where there are large numbers of properties that have increased significantly more in value than others?

  • It’s basically a value-weighted measure.

  • Yours is. But is it the aggregate divided by the number of housing units? I would think that there would have to be an implicit weighting in your calculation. If, in fact, the decline in value—the Nasdaq effect, in other words—were to represent a return to normal from the increases that had occurred, would that have affected to any extent the risk exposure?

  • In terms of the FRB/US model, the only way that house prices and housing wealth enter is through the aggregate.

  • So, one of the shortcomings, even of a model with several hundred equations, is that it doesn’t have the distribution effects.

  • There are two parts to that. One is the overall impact on the model. The second is that, to the extent there is risk embedded in the underlying mortgages, it would tend to be where the run-up has been higher. I’m going to talk a little bit about that tomorrow.

    My second point is on land value. In the assessed valuation that I get from the county—I’ve been too frightened in recent years to look at it—doesn’t it break out the land value and the improvements? I assume that’s public knowledge, is it not? So, presumably, you could access that number. And my recollection is that for residential appraisals three different methodologies are used, one of which looks at the improvements vis-à-vis the land value. So I would guess that that information is available.

    My third point is just a visual observation, having to do with the incidence of teardowns in a market like this. When there is an asymmetry between the land value and the value of the improvements, you see an increase in the teardowns. Locally, that trend started in D.C. and it has now moved out into our neighborhood. So that is a critical component, it seems to me, of any buying decision. And I think those data would be available.

    I have one follow-up question on something the Chairman mentioned. Somebody suggested to me recently—I just heard a different figure on this—that the private pools of mortgage-backed loans are now larger than the GSE pools. Have you seen any recent data to that effect?

  • Not in the United States. I don’t know what country or planet— [laughter]

  • The planet was Earth. [Laughter] The country was the United States. And the person making the observation was talking about sales of the nonconforming product into what they see as a growing and undisciplined secondary market.

  • I’m sorry, I shouldn’t have said that. Do you mean the flow or the stock?

  • Well, that’s a good question. I think the statement was as to the flow.

  • Then I beg your pardon. Of course, the stock differences are enormous. As for the flow difference, on net, GSE guaranteed mortgages in both pools and to a small extent on the balance sheet hardly grew at all in 2004, while it was an explosive year for growth for the RMBS sector.

  • That’s true. That’s a good point. Thank you.

  • Following up on Governor Ferguson’s question, I would imagine that one of the differences here is that we have tremendous variations from market to market. And to the Chairman’s question about the proportion of housing prices attributable to the land value, I would assume that it has to do with local regulation and restrictions on land use. For example, if you look at the numbers for homes being considered or under permit for construction in Southern California—that includes Los Angeles, Ventura, and Orange County—in the ’80s it was in the 100,000 plus range. This year it’s 37,000, in part because of the imposition of highly restrictive regulations on land use. If you look at permits for current construction of single-family homes in North Carolina, where there are almost no restrictions whatsoever, it’s over 50,000.

    So, the question is: Can you assume that land price as a percentage of house price is a function of regulatory and other restrictions imposed on land use and on construction? Is there a way to measure this? I ask because it strikes me—and I’ve been thinking about other countries as well, and Spain is an interesting example—that monetary policy would be a very blunt general tool if, indeed, there are significant variations city by city, state by state, and market by market.

  • I’ll leave the monetary policy aspect to my colleagues. But let me make a couple of points about land use restriction. Certainly, land use restrictions, environmental restrictions, and those types of things have played a role in the rapid rise in house prices. That certainly is a factor. I mentioned in my briefing that I like to compare house prices to rent—and this is one of the reasons—as opposed to comparing them to incomes. If increased regulation boosts house prices, people are just going to have to pay more out of their income for housing because of the regulations put in place. But land use restrictions should affect rents as well. So, it should drive up the cost of buying a home, and it should also drive up the cost of renting a home. It can be a significant factor, but it should be in both of those elements.

    And this is somewhat of an aside, but I find it interesting. I talk to a lot of builders as part of my day-to-day work here, and I’ve asked them about this very issue. They say definitely that land use restriction is a big concern for them. But I also went back and looked at newspaper articles written in ’86, ’87, ’88, and ’89, and it’s truly remarkable how the view back then was that housing prices could never fall because, as people were saying left and right, there was no more buildable land left due to environmental regulations and land use restrictions. They were saying that right up to 1989 when house prices leveled out in nominal terms. I’m not saying these things aren’t real— the builders say they are real—but they’re certainly not new though they may be new in their effects. But just to get back to the point, I think they should be captured in the price and the rent data.

  • Just one follow-up comment. There’s a lot of evidence to suggest that a good part of the reason for the rapid rise in home prices in California, Washington, D.C., and along the East Coast, for example, is because of relatively inelastic supply—not that demand there is necessarily stronger. That’s pretty clear in the data. Whether it’s due to these land use restrictions or other factors, the fact is that these are already fairly densely populated areas and there’s definitely a more inelastic supply in these areas of the country.

  • I would imagine that it would be driven also by the kind of margins that the suppliers can capture. Just for verisimilitude, I talked to Hovnanian, who builds about 20,000 homes a year at roughly $300,000 apiece. His margins are being squeezed in California, but in these less restrictive areas he can capture a greater amount, so obviously he gravitates to where he can capture the greatest margin. Now that’s just one example, but common sense would suggest that it applies more broadly.

  • The argument for the rapid rise in land prices in 1837 was that land was fixed in quantity. [Laughter] So, new ideas are very rare. Vice Chair.

  • I remember that bubble! [Laughter] I have two questions. My first is for John Williams or perhaps for Dave Stockton. What is the right way to think about dealing with uncertainty in considering the policy question? Put aside Glenn’s question about whether you know anything ever about the relationship between prices and fundamentals at a given point. What if what you don’t know is simply the likely path of home prices going forward? You could take the group here around the table and assume some path, but there would be a fairly fat band of uncertainty around that path. What does your regime imply, John, for policy in that particular circumstance? Do you basically ignore housing prices and look at all the other things we look at?

  • I think that the Greenbook forecast, as I understand it—and maybe Dave should talk about this—is predicated on a particular assumption about the future path of housing prices and takes into account the kind of models that Josh was discussing. The staff looks at all the empirical evidence, just as they do for every equation on every aspect of the economy. So, monetary theory would tell you to come up with the best, most reasonable forecast and adopt a policy that is appropriate to that path but also consider, as you’re saying, all the risks and the distribution of the risks. It’s along the lines of some of the charts in the Greenbook, which show the distribution of risks and then contemplate the implications for the current set of policy options over that distribution of risks.

    I think the basic idea in the economic literature is that you first want to get a very reasonable path, and that would be more or less your baseline. It’s not a path, I should say, that just keeps housing prices constant or keeps any asset price constant. It should be the best forecast of these asset prices that you can come up with, but subject, of course, to the fact that these are very hard to predict.

  • And you wait and see and then adjust course based on developments as they unfold.

  • That would be the standard policy.

  • Just on that point, isn’t that the difference between the optimal policy and the Taylor rule? With the Taylor rule, you wait and see. In the optimal policy, you somehow know this so you can move earlier.

  • Right. The Taylor rule would just be one example, of course, of a policy that takes into account the information you have up to that point. It’s not optimal in some sense. It’s just a simple equation. It does represent this wait-and-see approach. I was using the optimal policy to illustrate what you would do if you knew everything. It’s kind of a comparison.

  • John can correct me if I’m wrong, but I think there is an important element to your question about what is the source of uncertainty. If the source of uncertainty is going to be the evolution of house prices, that basically is uncertainty of a linear type that’s affecting aggregate demand. So you take your best shot and, therefore, set policy accordingly. The other possibility is that you’re not certain how house prices will react as you change your instrument. That becomes multiplier uncertainty; and depending on the structure of the uncertainty in your model, you may either attenuate or not.

  • My second question is about policy, but not monetary policy. Glenn, in your note, you allude to other instruments if monetary policy doesn’t seem to be the appropriate tool to address a concern about lower value prices. What do we know about the history of the use of the supervisory tool in past periods of concern about real estate bubbles or imprudent lending? Do we have a rich history on the use of those instruments that tells us something about the efficacy or about our foresight in deploying them?

  • It’s called the real bills doctrine.

  • Since the bills doctrine. [Laughter]

  • Since you’re going to the British Embassy, I might note that in the United Kingdom they used to allow loan-to-value at origination of 120 percent. You could buy your property under water essentially by 20 percent. One of the recent restrictions they’ve imposed is a limit of 110 percent. So this time around they feel much better in the U.K about where they are with just 110 percent loan-to-value at origination. But the monetary authority there doesn’t have supervisory responsibility—

  • I meant in our history. In our history, have we used that tool to good or ill effect? Have we used it wisely and with foresight?

  • You’re biasing the answer. [Laughter]

  • The answer is obviously “yes.” [Laughter]

  • My understanding is that we’ve used it fairly often in the postwar period, in the ’60s, ’70s, and ’80s—in early 1980, for example. My understanding is that it hasn’t worked very well. There have been times when we’ve tried to jawbone the banking system’s allocation of credit.

  • Is the history one of using it too late, or of using it and its having no effect because there are other ways to get money?

  • May I say something here? I don’t have any quantitative studies on this, but based on talking to the folks who lived through it, I’d make a couple of observations. If we look at the 1980s—the most recent housing bubble that was nationwide. We saw the bubble bursting nationally as opposed to the pockets we’ve had with the California breakdown or the Boston breakdown, which really related to the local economies and local employment developments. I think the local economic employment situations were the drivers of delinquencies, and the regulators generally missed that because they missed the local economic impact. And I think people focus on the fact that those cycles were driven by local economic conditions.

    I think the period of the 1980s involved a broader failure on the part of supervisors. If we compare the 1980s experience with what is happening now, in the earlier period a lot of financial institutions were on an exam cycle that went five or six years. So, nobody on the supervisory side was in there looking at what was going on. And that period was before the time when securitization became a prevalent practice, so most of the risk was carried on the books of the banks. Also, many of the banks that were hit very badly were following developers—going out of their footprint and lending on properties that they didn’t know. And I think one of the risks we have today in our big banks is that while we say they are diversified geographically, many of their loans are to investors or purchasers of second-home resort properties. So, the lenders again are following their customers out of the local area, and in the location of the new property the local lenders aren’t looking at it; the customer’s lender is handling the transaction. So, we still may have that risk embedded within the financial institutions. It’s one of the things we’re focusing on.

    What is new about it this time, though, is that a lot of these nonconforming products are being securitized by the private sector. So the real question is: Where does the market discipline kick in? And as supervisors, can we fault an institution for responding to a market need when it is offloading the loans and the risk into these types of mortgage structures that Andreas has been describing? We clearly could if the financial institutions were buying the equity or mezzanine risk tranches and the risks were back on the institutions’ books. But in many cases that clearly isn’t what is happening. So, we have some different aspects this time around.

    Just to let everybody know, the OCC [Office of the Comptroller of the Currency] and the Fed currently are putting together a horizontal review to look at the fringe kind of lending activities where we do need to send some signals. We wrote the HELOC [home equity line of credit] guidance that came out last month. We’re working on this other one and hope to have it out in a couple of months. We don’t want to turn off safe loans or the traditional types of lending activities, but we need to figure out where to go on some of these practices that are on the fringes. But we haven’t done a sterling job. I think that’s why we’re trying to send out some guidance. We sent out the appraisal guidance a year or so ago. But some of the risky practices of the past are starting to be repeated, and it may be that the generation of lenders now didn’t live through the problems before.

  • Shall we break for coffee?

  • [Coffee break]

  • Thank you, Mr. Chairman. I wanted to make a few comments and then ask a question. First, I’d say that with all of the concerns about froth in housing markets, I found these presentations to be very informative, and I want to congratulate the people who spent a lot of time preparing them. I thought they were all very good presentations. But I also found the information comforting. We’ve all talked about the possibility of local housing bubbles and regional housing bubbles, and clearly there are some in the United States. But we never really looked at it on a national basis before. The net result for me was that I come away from the analysis not feeling any worse than I did before and probably a little better.

    First, I thought it was very helpful to see quantified—I think this was in Josh’s memo—the size of the potential bubble. He talked about a 20 percent drop in housing prices. But that was equal to only about 30 percent of GDP as compared to the drop in equity prices we had, which was more than twice that. Also, I had the feeling that appropriate monetary policy, as John said, could mitigate much of the distress that might occur. Moreover, the credit risk associated with home mortgages seems to be spread out across many institutions. Governor Bies said that a lot of analysis is being done now, and we’ll want to see the results of the analysis that the Board and the Comptroller are doing. But on the whole, the financial institutions seem to be in pretty good shape. The role of securitizing mortgages is to lay off risks to parties who are willing and able to bear the risks. Capital levels of the financial institutions are relatively high, so it appears that these markets are performing their roles well. And in the event of a sharp drop in housing prices, the odds of a spillover to financial institutions seem limited. And as I mentioned, it was helpful to hear the suggestion that monetary policy can be effective in responding to a sharp drop—if there is one—in housing prices. So I come away somewhat less concerned about the size and consequences of a housing bubble than I was before.

    The question I had relates to what Governor Yellen was asking about—financial innovation. I was going to make a similar point. The fact is that there has been a great deal of financial innovation in housing markets in the United States. The average person can borrow very easily on his home these days. And I was wondering if there have been—or if it is possible to do—any international comparisons on this. I wondered whether the price-rent ratios in other countries that may not have had the same degree of financial innovation we’ve had differ substantially from ours.

  • Well, I trust you did receive or are aware of the background document we circulated on the foreign experience.

  • We didn’t do price-rent ratios, and given what Josh has described in terms of the care he has to put in before he feels he has something close to the right number, I think we might feel a little hesitant to do so. I take it back; there are price-rent ratios—[laughter] in the back portion of that paper. What data we have, we used. But there are huge variations in the financing practices in housing markets across foreign countries. There is some degree of securitization but less than here. Practices with respect to the tax-favored treatment of borrowing to finance houses differ relative to those in the United States. There’s a host of different characteristics, in terms of variable-rate, fixed-rate, and other types of instruments used in various foreign countries.

    But one of the things our paper was intended to point out was the prevalence of really rapid rises in house prices. According to this chart—the first few pages of which are right here anyway— that shows through to an increase in the price-rent ratios as well. So, I refer you back to this paper. There is only a subset of the countries we covered for which we have price-rent charts, but there are some in the paper.

  • But you can’t really associate it with the degree of financial innovation that we’ve had in the United States because of this great variability? Is that what you’re saying, Karen?

  • I guess I wouldn’t necessarily associate it with, say, the development of the secondary market or the ease of equity extraction, and those sorts of things, but I’m not saying they’re irrelevant. For example, in this chart, Switzerland—which is a country that I think of as having made little progress in that regard—doesn’t have much of a price-rent ratio increase. And the U.K., which does have a more market-oriented approach, has a huge one.

    But it’s also true that that’s where the price differences have been. I think we’d be hard- pressed to link too tightly the increased liquidity that we ascribe to U.S. home equity as now having taken on the role of the single driving factor anyway. It might be relevant, but I don’t know that it distinguishes these countries.

  • Thank you, Mr. Chairman. I also want to thank the authors of the papers—the international paper as well as all of the papers that were talked about today—because I found them very helpful and reassuring, along the lines that Michael Moskow was discussing. I also thought that Janet’s comments on the financial innovations were insightful. Despite the fact that it’s hard to sort out the U.S. experience vis-à-vis other countries, the whole point of this—to me anyway—is that we’re seeing a phenomenon in the housing markets. So the question is: Do the fundamentals explain it or don’t they? And financial innovations are certainly part of the mix, in terms of changes in the fundamentals. So, to the extent that they’ve made many of the transactions in the housing market easier to do, that has to have had some impact on the underlying asset prices. So I thought that argument was a very interesting one, and it will be interesting to see if there is any way one can tease out the effects of that.

    My second point is that when you look at the relationship between rising prices on either the OFHEO or the constant-quality index against disposable income as opposed to median household income, you see an even more reassuring chart. And I would think there’s at least a little bit of logic to doing that, based on Dick Peach’s chart about the differences in house-price acceleration depending on income—with the value of higher-priced houses moving up faster than lower-priced houses. So I would think that there is some logic to looking at this with disposable income. I know that you have done those charts. I think we have every chart you could do! [Laughter] But it is interesting that one sees in that perspective somewhat less acceleration and a somewhat more reassuring picture.

    Finally, and this is more of a question, we talked about a 20 percent decline in house prices and what that would do in terms of the basic macroeconomic effects of it. But, of course, that wouldn’t happen overnight. Something would make it happen. And, as we consider our current policy stance, one of the conundrums—though I hate to use that word—is why haven’t the 10-year yield and, therefore, mortgage interest rates, taken the same upward path as rates at the short end of the curve. So a question I was asking myself—and I think it’s probably not so difficult to figure this out—was where the point is, as mortgage rates start to go up, when we look at that affordability curve and start to worry about households beginning to get into trouble. That would then have an impact on house prices. Obviously, higher interest rates would tend to level house prices off or take them down. They also would have an impact, other things equal, on people’s ability to afford the house that they’re in. How far do interest rates have to go up before that affordability curve starts to move in the direction that causes a problem in that regard?

  • If you look at the housing affordability picture—basically relating house prices and interest rates and income—it has been moving around in a fairly stable and favorable area for the last, say, 8 to 10 years.

  • As you know, mortgage rates have been moving up and down but on balance they have flattened out a bit over the past few years, if you compare, say, 2003 to now. So what has been happening over that time period is that affordability is edging down to the lower end of the range in which it has fluctuated for a while. There’s nothing scary about it, but affordability has been moving down. Certainly, we can take this argument a little further. If mortgage rates stay pretty stable and disposable income keeps rising at reasonable rates, affordability is going to start looking worse and worse if we keep getting the house-price gains that we’ve been seeing. That’s not necessarily an argument for overvaluation or undervaluation. But it is an argument for the view that the recent increases in house-price gains are not sustainable. If rates flatten out, affordability will start to deteriorate fairly significantly.

  • If rates flatten out or if they rise?

  • If rates flatten out; that’s because we would not be getting a kick any more from lower rates. House prices can go up and housing remains affordable if rates are lower. But if they flatten out and stay flat, one might think that it’s going to be hard for house-price gains to keep humming along the way they’ve been going. That’s going to start to eat into affordability.

  • Okay. I guess I was coming at it the other way.

  • I’m sorry, then.

  • No, no. Let me just think about that for a while. Go on.

  • Let me make one other point on the reasons why there may have been an increase in the equilibrium price-rent ratio, because I certainly agree with you and President Yellen that there are some good reasons. Those reasons could well include financial innovation, changes in capital gains taxation, and supply constraints. However, lots of asset price misalignments start out with situations where there are good reasons why those prices are rising rapidly. Productivity innovation and changes in business models, and so forth, were I think probably valid explanations for the increase in stock market prices, but that doesn’t mean—

  • While they were probably valid explanations for the stock market rise in the ’90s, they didn’t necessarily, in the end, go all the way to explaining how far those valuations had moved, even if they had somewhat of a solid economic impetus to begin with.

  • I was thinking primarily about the affordability issue. I know the affordability ratio involves at least three things moving simultaneously—the value of houses, disposable income, and interest rates. I was more or less looking just at the interest rate component.

  • I was wondering if interest rates were moving up instead of not moving up how that affects affordability.

  • It’s going to make housing significantly—

  • Well, yes, but where is the breaking point? If the 10-year rate goes up 1 percentage point, say, are we going to get worried about whether people can afford their houses? Or is there some range over which the rate can rise and we shouldn’t be concerned?

  • I certainly wouldn’t want to say the breaking point is here or there. But again, going back to some of the conversations I’ve had with homebuilders, for what it’s worth, they indicate that they’re not going to be too worried unless we see an increase—and one that occurs fast—on the order of 100 basis points.

  • So 1 percentage point, or 100 basis points, in a short amount of time—

  • Yes, but that’s based just on talking to some builders, as opposed to any firm statistical analysis. So I would be very cautious. I would certainly hesitate to draw a firm conclusion about a particular increase being a tipping point or a breaking point.

  • If I could just add a point: In the spring of ’87, we had a very rapid increase in rates—200 basis points—as we did again in 1994, and the housing market proved to be surprisingly resilient in the face of those fairly steep increases.

  • Thank you, Mr. Chairman. I have three questions. The first one is on the price-to-rent ratio. We’ve been treating it as if most of the adjustment has come on prices. And I wanted to ask Josh particularly whether, as you’ve been looking at the micro data and thinking about this, the dynamics of some of these innovations that have led to a shift from renting to home ownership might have artificially depressed rents relative to prices. And I wondered, after that shift is over, if rents will start rising faster and close the gap that way—use up some of that 20 percent. My question is what you thought of that. And my observation is that that would present a much more difficult situation for us sitting around this table. It would be kind of like a supply shock because prices would be rising, inflation would be higher—and that homeowners’ equivalent rent would be rising faster—and we’d face a more difficult situation. We pretty much know what to do if house prices fall. That’s a pure demand shock. But if rents start rising, that’s another matter. So I wondered if you’d comment on that.

    And then, while I have the floor, let me ask my questions of Glenn and John. Glenn, on the 1994 bubble analogy, I was surprised to see that classified as a bubble. I think there was some inflation scare then, but there was also a real rate adjustment at the same time. If you looked at any of the surveys, I think you wouldn’t have seen much of an increase in inflation expectations. I agree that we had to raise real rates in order to prevent that from happening. But that seems to me a very different animal than equity price changes or house-price changes because we are responsible for inflation. So if we see inflation moving, we’ve got to do something about that, whereas we’re not responsible for the relative prices of houses or equity and other things. So I wouldn’t have put 1994 on a list of situations we might think about as we’re looking at this issue of house-price gains. It seems to me very different. I’d like to hear your comment on that.

    And finally, my other question for John has to do with this point about the misallocation of resources. Doesn’t it matter what the state of the business cycle is? If we hadn’t had so many houses built and so much consumption over the last few years, we would have had more unemployment. So it’s not obvious that resources have been misallocated. The resources that went into building houses, furniture, and cars, and so forth might have been unemployed, especially if we had raised rates more in order to lean against the house-price increases. If we had, surely unemployment would be higher. So it seems to me that it’s one thing to talk about misallocating resources between two states of full employment, but it’s another thing to talk about a misallocation of resources where there would otherwise be slack in the economy. And the latter case I don’t think really is a misallocation of resources. There’s no opportunity cost.

  • Did you want to start on rents? That will give the others some time to think.

    I’ve looked at the rent side of the picture. As to this idea that perhaps prices are getting too high relative to rents, I’d argue that they’re going to come back in line. Now, maybe rents are going to be doing some of the correcting, as it were. The work that I’ve done gives just a little hint of evidence that maybe rents do a bit of the correcting. So, what we might see going forward is rent growth slightly higher than it otherwise would have been. But statistically speaking, it’s basically no different from zero. Statistically, it doesn’t look like rents do any of the correcting. What really seems to be happening is that rents go up at some rate determined by economic conditions and then prices move around them.

  • You don’t see anything in the current situation that would differentiate it from past history in that regard. I guess that was my question.

  • I do not, certainly not with respect to the rents. I would like to say also that I agree with the comments that financial innovations could very well be a justifiable reason for house prices to be higher than usual relative to other prices. I do think some things are different. But I don’t see anything to suggest that we have to be worried about rental inflation and, as a result, overall inflation.

  • With regard to equity prices and bond prices, the reason I included that comparison was to make a couple of points. For one, the comment is often made that we can’t second-guess financial markets or financial market participants. It seems as if that may be true for equity market participants but less true for bond market participants. There I take “inflation scare” or “credibility gap” or “conundrum” all to be another term for a situation in which we have an idea of where the fundamentals are and we’re not sure what bond market participants are thinking.

    Now, it’s true that there’s a difference in that you have a clear idea about the reaction function or where monetary policy is going and presumably can guide bond market participants with transparency regarding your notion of fundamentals. So there is a clear difference between the markets. The housing market is different from both the bond and the equity markets, and the question is: Where can we draw the lessons? The regional disparity is completely different from both markets, and that’s just a separate issue.

    In any event, one reason behind the housing price appreciation, perhaps, is that we have very low long rates. This is a bond rate conundrum. Perhaps the misalignment in bond prices is leading to this misalignment in housing prices. So, one could perhaps make the argument that it’s the bond price experience in 1994 that may be the relevant one for today. I think I’ll stop there and leave it at that.

  • In fact, the third scenario that I considered involved trying to emphasize that point—namely that, at least by some measures that people have come up with, there is a big difference in where bond rates are relative to standard estimates of fundamentals. It’s actually a much bigger problem for the economy than just the house-price effect directly, at least according to the FRB/US model. More importantly, it could be one of the factors driving a big part of the house- price appreciation. In terms of needing strong house prices to keep the economy moving, the way I view your third question is that if it weren’t for the house-price run-up, monetary policy would need to be easier, given current economic conditions. And I think that’s absolutely right.

    One way to think about my scenarios is just to reverse the signs, especially in scenarios 1 and 2, and think about it as this as the positive stimulus we’ve gotten from a 20 percent appreciation of housing prices and this is the positive effect we’ve gotten from some other factors. Especially scenario 1, I think you can see that way.

    The reason I mentioned the misallocation of resources toward housing-related activities in my presentation is that in quite a bit of the economic academic research about bubbles, the emphasis is that they actually lead to, as Glenn mentioned, a misallocation of resources. Therefore, these gaps between fundamental prices and actual prices should appear in the policymaker’s objective function, in addition to inflation, output and employment. So there is a notion here that that’s just another problem that you would want to balance off if you could.

    Now, I’d like to emphasize in my closing remarks that the assumption that you could affect the bubble is very problematic. As Josh himself mentioned, these relationships between housing prices and interest rates are just not as strong as one would think and not as strong as economic theory would suggest.

  • Just for the hell of it, I’d like to offer the hypothesis that property values are too low rather than too high. [Laughter] If you believe that Treasury indexed bonds are a good measure of the real rate of interest—and we were getting into this discussion—the real rate of interest has been cut in half in the past five years. For any asset with a perpetual stream of returns, like land, the capital value has doubled. Now, it may well be that the issue is with the indexed bonds and not with the housing prices. It may be that the right way to state this hypothesis is instead to say that the conundrum is that there’s a disconnect between the prices of some of these assets.

    In any event, if we are going to be in a world in which the real rate of interest is truly much lower than it was before—if that is going to be the situation for some time to come—then it seems to me that we could expect some long-continuing appreciation of land values in particular. Residential structures have a very long life, so they are almost priced that way. And let me link this analysis to the tear-down phenomenon that Mark Olson was talking about. I’ve been told—I don’t know the data here—that houses replacing tear-downs may account for about 300,000 units a year, which is a significant part of new building. Maybe someone knows the data on demolitions and replacements. I don’t.

  • That’s roughly the right number.

  • Okay. It’s a significant part of total new construction. And I know in the community where I live—I thought this was real estate hype when we bought our house—we were told that we had bought the land and got the house for free. But that’s just about the case, because the houses in my area—a subdivision built in the 1960s—are being torn down, and houses with 3,000 square feet are being replaced by houses with 6,000 to 8,000 square feet on nice one-acre lots. So the land value is significant.

    On this question about what is really happening to land values in urban areas, I don’t know whether the data are available to answer that. I suppose a detailed look at real estate records and tax records, if you had enough resources to do it, would enable you to look at properties where the houses are being torn down and replaced with new homes. And that would probably give you a pretty good measure of the underlying value of the land in those areas. In any event, when real rates of interest come down, one would expect to see increases in the value of the land relative to the structure. So one would expect, I think, to see more teardowns. I gather that, though it’s not a brand-new phenomenon, the scale of it is probably much increased in recent years from 10 or 20 years ago.

    At any rate, I offer those observations because, if we are in a world that is going to have much lower real rates of interest for some time to come, one would expect to see the price-to-rent ratio go up. Maybe this line in the chart has another 40 percent to go to get to equilibrium! [Laughter]

  • Mr. Chairman, I’d like to propose that he buy my house in Washington, [laughter] given that confidence.

  • If I’m right, you won’t need me to buy it.

  • I have an observation and a question about what has been happening to the mix of mortgage products in the last 12 to 18 months. In 2003, we saw refinancings peak; and then in 2004, total originations went from $3.8 trillion down to $2.8 trillion. So they dropped and are continuing to be soft this year. But what happened during that 12- to18-month period of time is that the number of ARMs shot up to about half of the originations. And based on anecdotal evidence, ARMs still seem to be running at least at that pace. Initially, with rates at record lows relative to those prevailing in the last 40 plus years, it made sense that rational consumers were locking in their mortgage rates and getting the houses they could afford.

    Now, we embarked a year ago today on a path of raising interest rates. We haven’t had a comparable movement at the long end, so long-term rates are basically about the same as they were. Why are so many more ARMs being generated? Is this an affordability issue? Is this a mortgage broker issue, where the brokers want to get a deal at any cost? What is happening? I’ve heard some folks say, “Well, the ARM is a way to do the loan now, and then we’ll turn around and do the fixed rate next year.”

  • Okay, let me try to answer that question. I can’t really comment on the hype mentality or the mortgage broker who uses a high pressure sales tactic. But in 2004, for example, I believe that, of the ARMs that were originated, less than 10 percent were traditional ARMs with a one-year reset period. The remaining 90 percent were hybrids that had a rate lock feature where the borrower is protected from interest rate movements for some period of time. The data on this are a little difficult to get, but it looks as if the median rate lock period was five years. If we think about people’s typical tenure in houses, five years isn’t a bad guess for how long people are going to stay in a particular house. A 30-year fixed-rate mortgage is, in effect, a hybrid mortgage that resets the day you move. So, I think of this to some degree as a financial innovation, and one that people are learning about through advertising or the press or whatever. And I believe this type of hybrid ARM probably makes sense for a lot of people who are making a decision on a mortgage loan.

    The other point that I’d make is that the fraction of originations that are ARMs often can overstate the role of ARMs in the total mortgage market, because people refinance out of ARMs into fixed-rate mortgages frequently. In fact, I’ve read in the Wall Street newsletters that some people seem to believe that the current rate of refinancing is elevated by this flow out of ARMs into fixed-rate loans.

  • Well, that was going to be my next question. The other part of this is that we have these older 3-, 5-, and 7-year ARMs, and 2005 is the initial wave of the 3-year ARMs having their first bump-up in rates or being refinanced out of existence.

  • And the numbers I’ve seen suggest that in the next three years 60 percent of these ARMs are going to have the rate bumped up or be refinanced. So, as we look at the cumulative impact of this—I’m going back to the affordability issue—what do we really see happening going forward in these mortgages?

  • Well, as you say, the rates are going to reset, and households will make decisions about whether to refinance to a different point on the yield curve. Or people can decide to move. I don’t think there’s going to be a catastrophe on the average reset date that—

  • No matter how fast rates move up.

  • Obviously, if rates went up sharply, that could be a problem.

  • You have to think in terms of the environment that we are forecasting to occur. It’s a pretty tame interest rate environment; so it’s not one that will necessarily generate significant concerns about the costs of those reset mortgages. Now, obviously, we could be wrong in both directions. One could imagine a situation where the rates rose faster. And that’s one of the channels by which monetary policy has traction on the economy; if households face higher rates, they would curtail their spending in response to that. But presumably you’d only be raising rates more rapidly because, in essence, you needed restraint on aggregate demand. So you would welcome that result, as long as it wasn’t precipitating some big nonlinear type of event in terms of people responding very, very sharply to that kind of change in the environment.

  • The second question I have relates to the nature of the price decline risk. If some of these ARMs can’t be refinanced and foreclosures actually start to occur—where financial institutions take over the property—are there any studies out there that indicate how much foreclosure volume could hit the market before it had a material impact on local house prices?

  • Not any recent studies that I’m aware of. We proposed doing a study jointly with a private data vendor company on precisely this topic but it didn’t really go anywhere. The question is whether foreclosure clusters can have a pernicious kind of spillover effect. I would say that lenders and loan servicers aren’t unaware of this problem. In my conversations with them, they’ve all said that their loss mitigation programs involve holding down the inventory of foreclosed properties that are on the market. They do not want to let them turn into this sort of “pocket of blight”—that is the term they used—as a result of the foreclosure process. That’s all the reassurance I can offer you.

  • We’ll just have to see how they really are going to engineer those processes. Thank you.

  • Yes. I have just a couple of comments and a question. Clearly, there are some perspectives from which housing prices seem to have drifted out of the usual relationship with indicators of fundamentals, and Joshua documented some of them. But it seems to me as if there are a lot of plausible stories one can tell about fundamentals that would explain or rationalize housing prices. Obviously, low interest rates have to top the list. Strong income growth among homeowning populations would be on the list, as would land use restrictions, which were mentioned earlier, and the recent surge in spending on home improvement. I found President Yellen’s suggestions intriguing. I’d like to offer my own, just in the spirit of adding potential explanations here. And it’s really a version of something Governor Kohn observed, which is that housing prices are relative prices.

    I’ve been struck by the fact that a collection of large metropolitan areas increasingly dominates the national housing figures and that house-price appreciation seems different across various urban regions. It suggests to me that housing values may be affected significantly by—I don’t know exactly how to phrase this—sort of the relative microeconomic value of agglomeration. By that I mean the value of the amenities in a city or the enhanced productivity associated with living in or near where one works. Now, in this age of telecommuting and the Internet, it’s easy to deduce that the value of living in a city has declined. But it seems plausible to me that the value of a thick labor market might be increasingly important for certain skill specialties. And it also seems plausible that the strong demand for urban amenities is evident in the recent vitality of many older urban cores. So I’d be interested if any of our housing data experts have any information relevant to that issue.

    While I have the floor, I’d like to make just an observation. It seems to me likely that a confluence of several fundamental factors might rationalize the current level of housing prices. So from that point of view, it’s hard for me to see how it would be reasonable to place a great deal of certainty on the notion that housing is significantly overvalued, or that there’s a bubble, or that it’s going to collapse really soon.

    I think these markets—this is echoing President Poole’s discussion—are too complex. I think our quantitative understanding of them is too limited to warrant second-guessing market forces. And beyond that, the models that we have of bubbles—Glenn wrote it down—are just some statistical noise added to an equation. I don’t think we have any models that give us any reason to hope that we can understand how interest rate changes would affect this little random statistical term added on to these equations.

    Having said that, housing prices pose a dilemma for us and are going to pose challenges for us soon, I think. Rapid appreciations in asset prices can make monetary policy more difficult. They tend to be associated with tightening labor markets. At the same time, there is a rise in the downside risk. So, even though I’m not very far down Glenn’s decision tree, these are still issues I’m paying attention to. It feels as if it could well occupy our attention here.

    But I’d be interested in your reaction to this agglomeration story.

  • Well, I think there’s something to it. There is the view that people, particularly higher-income people, place a high value on their time. If you don’t want to be caught in Washington, D.C., traffic, say, you might decide to sell your house in the suburbs and move into the city and in the process drive up land prices in D.C., which happens to be one of the metro areas with the highest rate of price increase.

    I’m not a regional specialist, but I know there are people at our Bank who are working on this issue of agglomeration benefits. And I think their conclusion is that there are a lot of reasons, particularly labor market associated reasons, why there may be more agglomeration benefits now than previously. That is exactly the opposite conclusion a lot of people thought would be the case as we move to more telecommuting and that sort of thing.

  • That certainly seems possible. When I look at Josh’s exhibit 3 on page 5, I see the Miami price-rent ratio at 64 percent above its trend. Now, it’s possible that everybody just woke up and decided, boy, there are people in Miami who are just really terrific to be around—it’s an exciting city and fascinating people live there. [Laughter] But it’s also possible that that statistic could be an indication that people have unrealistic expectations about the rate of increase in house prices expected in Miami. And there is certainly a lot of anecdotal evidence that in that particular city there is a lot of flipping of properties going on as well as other developments that might not be reflective of a purely equilibrium move in house prices or of agglomeration economies.

    So I’m still a little nervous about this. There are a lot of good reasons why prices ought to be high relative to rents and relative to incomes. And I think even President Poole’s suggestion that maybe housing is undervalued can’t be ruled out. We’ve done simple dividend discount-type calculations on rents and interest rates. And if you make a certain assumption about the growth rate of real rents going forward and the persistence of low interest rates, you can get figures showing further appreciation of maybe not 40 percent but rates that are pretty high. So we think that’s within the probability distribution. But we’re also worried that we’re seeing in many markets and for the nation as a whole a run-up in prices that certainly looks very unusual by historical standards. It could very well be that this time is different and it’s all being driven by fundamentals. But we don’t think you should rule out the possibility that you could be facing a period in which prices could be declining or just be softer.

    One point that has been made is that we obviously don’t know how the end will look—if there is an end. The end could come through a long period of just relatively subdued growth in nominal house prices. It wouldn’t have to be associated with a 20 percent decline. As we noted in the Greenbook, that is an extreme drop. In fact, to get to John’s scenario 3, a lot of extreme things have to happen. It takes an unusually large drop in house prices and a lot of spillover effects. As you recall, the first scenario was pretty tame; if you move down 50 basis points on interest rates, it offsets that. But if you layer on top of the decline in house prices a big drop in consumer confidence, a big equity extraction effect, or a much bigger wealth effect from housing—which looks to us to be pretty much on the edge—and throw in some covariance with a bond market event, the situation worsens substantially. But it takes a lot to get to a real disaster type of scenario.

    So even though we feel that house prices have moved out of alignment with the fundamentals, we don’t necessarily think the implications of that are that you’re going to be confronted immediately with some large problem. In fact, our best guess would be that the misalignment would unwind in ways that would be quite feasible for you to offset and insulate.

    There are questions on the supervisory side. There I don’t think the historical evidence suggests that supervisory policy has been used effectively to head off asset bubbles or to elegantly deflate them when they occur. What you might hope to do is to have in place policies that will prevent the kind of spillover effects of John’s scenario 2 so that you just have a wealth effect. In other words, house prices might go down or soften and that will show up on various agents’ balance sheets as a reduction in wealth. One might hope that they respond accordingly and that we won’t have the complications with intermediation or other kinds of things that would add to that effect. So our story basically is that we’re worried about valuations in the housing market, but we don’t necessarily see that as having profound consequences for your policy going forward.

  • Yes. I have a question for Karen about a couple of the charts. One of the things that struck me in looking at the charts for the United Kingdom—whether you look at price­ to-rent ratios or real house prices—is that they obviously have gone through several very sharp swings. If you look at the U.S. historical charts, they’re pretty benign. Dave was just talking about the possible fallout of sharp swings in those measures here, and I was curious whether we could learn anything about the macro consequences or the macro conditions associated with what has happened in the U.K.

  • Well, for one thing, the U.K.’s 1990 cycle was amplified by regulatory changes that preceded it, which led to mortgage lending that was excessive and not well supervised. It was a kind of blind-leading-the-blind situation: The regulator changed the rules and the financial institutions moved into the market and practices and norms changed. A great deal of lending took place. The supervisors were learning as much as the lenders were and—well, let me put it this way—it didn’t go well.

    In the past, there has certainly been a correspondence between household consumption and housing wealth in the U.K. So when they suffered the big drop in the 1990 rundown of housing prices, they also had a big change in household consumption out of disposable income, and so forth, and they had the makings of a recession as a result.

    Now, a couple of the characteristics of the U.K. market have always led us to think that it’s not a telling example. One is the prevalence of variable-rate mortgages, which causes the process of tightening monetary policy to contain the macroeconomy to have a bit of extra leverage over the discretionary income of households—to an extent that is not the case with fixed-rate mortgages Obviously, there’s a counterpart effect on the earnings of mortgage lenders, and so forth, in the U.S. economy that you need to take account of to fully understand that. But there is that characteristic. And there is the fact that house prices actually fell in the U.K., so they had big negative equity problems, which complicated the process of how to unwind the interaction of household behavior and financial intermediary behavior. To the extent people walked away from the houses, the financial intermediaries were getting collateral that perhaps no longer equaled the value of the loan. On the other hand, some households didn’t walk away; they just remained in a negative equity position for some time. And that had a long, dampening effect on their spending patterns. So there were complex reactions involved.

    But this time around, at least based on my conversations with U.K. officials, they think they’ve improved a lot of those things. So in that sense, there is something to be learned. Financial institutions now know better how to maintain their balance sheets and how to do this lending. The households know better, too; they’ve learned a bit. U.K. officials think they’ve seen a lessening to some degree of this tight link between housing wealth and consumption so that they’re not both on the run-up and then, when it stops, extrapolating what the consequences would be.

    On the other hand, I have to admit that I occasionally read my little machine while I sit here, and in the last hour it carried three statements from a member of the Bank of England’s Monetary Policy Committee who must have been making a speech. And all three statements the media chose to pick up pertained to whether or not housing prices ought to be relevant for monetary policy. [Laughter] And all of the statements were slightly two-handed, along the lines of: Monetary policy should not be driven by housing prices, but we must look at them closely. [Laughter]

    I think there are lessons to be learned; I’m not saying there aren’t. But certainly the 1990s episode had some characteristics that were far more extreme and that would never happen here because of institutional differences. And the present episode, which has remedied some of those earlier problems, I think is not such a bad deal. They’ve slowed the increase in housing prices. They aren’t having negative equity. They don’t have financial institutions that look like they’re going to become or technically are insolvent. I think in that sense they moved to improve their infrastructure, so to speak. They talk about the issue a lot but I think they feel the situation is okay at this time.

  • May I raise a question, which puzzles me, on chart 13? That has the new home price index and the OFHEO index. I don’t think there’s a big problem understanding the prices of new homes ex land. Since homes are largely customized, one would expect that, as indeed the data show, the growth in productivity—and presumably in unit labor costs in residential construction—is less than the average. That would lead you to conclude that at constant margins the average real price of homes is going up. And the number is, as I recall, somewhere between 0.5 and 0.8 or 0.9, or something like that. Now, after a house is constructed and priced, the price from there forward either is flat or goes down—I don’t think anybody presumes it goes up. This gets to the issue of whether the modernization or the depreciation overwhelms.

    You have a system in which you have initial prices of new homes, and let’s assume there’s a decay rate. So, as the homes age, if the decay rates are all constant, clearly any measure of constant- age existing homes will exactly parallel the new home price. Looking at this chart, it’s very evident that prices of existing homes on average are rising faster—generally going almost all the way back—than those of new homes. That could imply that there is a change in mix. In other words, if there’s a decay rate, then the presumption is that the average age of existing homes that are being sold is moving down. But none of the credible hypotheses—because that can’t go on very long— explains this difference. We know that the constant-quality new home price is a hedonic, reasonably well-constructed series. How do we reconcile these data? What’s happening?

  • Well, I think in part it may go back to one of the issues discussed earlier about the rate at which replacement is taking place. Now, I haven’t done this in the most scientific way, but if you just take periodic estimates of the stock of housing, whether from the Census or from the American Housing Survey, and try to line up changes in the stock with new production, it appears that we’re in a period when a lot of destruction of old housing is taking place. As you mentioned, the average estimate on tear-downs is 300,000, but I’ve seen estimates as high as 700,000 units.

  • Let me just say this. We have this series on vacancy rates and ratios, and that series actually has an implicit housing stock figure. If you take the figure for completions and match it up against that, for a good portion of the last 10 years you get negative replacement.

  • So it’s not evident what those replacement numbers are. We’re really at sea on what the actual total housing stock is, which is the problem we have with measuring population as well. So I’m not sure you can make the statement that you just made.

  • Well, I think for a shorter time horizon you can say that.

  • Do you mean like the period we’re in now?

  • Yes, over the past four or five years, I think it does hold true.

  • You mean as far as the completion rate being so high relative to household formation—if you believe the household formation data.

  • Right. Well, and also to the estimates of the stock.

  • Okay. So you’re suggesting that this is—but this pattern has gone on since 1980.

    SPEAKER(?). Page 14, Mr. Chairman.

  • It has been going on for a long time, but I think the separation has been most pronounced more recently.

  • That’s certainly true. But is it credible that we can have a consistently more rapid rise in prices of existing homes unless the value of the land is rising faster for those homes? Then there’s the issue we discussed earlier about moving out into the suburbs, which undercuts the land price value of the new homes. Do we have any insight into that at all?

  • That is my opinion on the matter. If you put a new home into place on a plot of land and follow that over time, the value of the structure might rise, reflecting an improvement to the house, or it might deteriorate. But as for the land, once the house is built, it falls into existing house land. That’s what it is from then on. And if the value of that package of land and structure is going up and you think the value of the structure is not, it’s coming in through the land.

    If you flip to the very last page of the handout, page 33, the lower panel shows levels of the repeat transactions prices in red, levels of the constant-quality price index for new homes in green, and an estimate of construction costs. That basically reflects the cost of a bundle of construction materials over time—a certain amount of lumber, a certain amount of—

  • That can’t be far wrong.

  • If you look at the pattern of the green line, the constant-quality and the cost line move roughly together. Obviously, the constant-quality price index has increased more rapidly recently, because that index does capture partially, I think, increases in land prices. But I think that the gaps between the three lines are telling us quite a bit about what’s going on with regard to the price of land sitting under the entire stock of housing.

  • You’re implying, however, an answer to the question I asked you before which you refused to answer [laughter], which is the ratio of the land value to total value.

  • There’s a difference between refusal and inability. [Laughter]

  • The ratio is—unless you were inferring it between the two prices, which begs the question—

  • Well, that’s exactly what Morris Davis—

  • Yes, I understand that. But it raises the question of whether the price data are accurate.

  • Absolutely. As I said, this is what I think is going on. In my view, that’s what is capturing land prices.

  • Did you ever say what adjustments you make in the repeat- transactions OFHEO numbers?

  • In my prepared text, no, I didn’t. But I can go over them briefly. Would you like me to—

  • Yes. I made an adjustment for improvement rate and depreciation rate. And there are actually two types of transaction biases I made adjustments for. One is the fact that homes that have greater appreciation tend to transact more and, therefore, they show up more in the repeat transaction database. So I made an adjustment for that. Also, two economists at OFHEO wrote a paper showing that there’s another kind of bias—basically the high-end homes tend to appreciate faster than the low-end homes—and that’s biasing the OFHEO as well. So I made an adjustment for that. Those are the three main adjustments that I made to the repeat transactions price data.

  • Governor Gramlich, did you want to make a comment?

  • I have three quick points. On Sue’s point earlier about foreclosures, I think foreclosure rates nationally are too low to do anything with analytically. But in my world of low-income housing, in certain urban neighborhoods they get up to 40 percent or so. One can’t help but think that what happens—it’s like the old redlining issue—is that first some foreclosures occur on a block, and pretty soon the values of the homes start going down, and then a lot of people stop making mortgage payments. I don’t think that has macro significance, but it is a big problem in certain neighborhoods. And that’s actually one thing I would like to work on when I leave here.

    On the discussion between Josh and Cathy earlier about affordability measures, I think that Cathy was trying to get you to say, Josh, what would happen if interest rates went up and nothing else happened. But—pardon?

  • That probably wouldn’t happen.

  • Well, that’s right. I understand that the link between interest rates and prices is weak. But really, there ought to be a full general equilibrium, if you will, calculation there. Affordability rates may tend to be a lot more stable than we would think just from interest rates alone, if it turns out that they do have an impact on prices.

    The third point: Janet raised what I thought was an interesting idea earlier that some people have picked up on, which is that maybe these credit innovations are endogenous. I recall a briefing that we had just a short time ago where the staff suggested that, in effect, what was happening—and I, frankly, was a little skeptical at the time—was that people were identified by how much they could pay for their house. And that, I think, makes the credit transaction automatically endogenous. That may be a better way to look at the whole situation, and I guess researchers at the Board have already started doing that. I don’t know how complete their research is, but given our myopic society, that may be a much better way to look at it—how much you can pay—and then nobody worries about what the implications are down the road. But that explains the initial buying decision.

  • Is that a statement or a question?

  • Yes. [Laughter] Do I have to say? If somebody wants to comment, they can.

  • Let’s choose to leave it as it stands and we’ll end this discussion with President Fisher. Do you want to make a comment?

  • No, sir. Mine was on a similar line, neither a question nor a statement. [Laughter]

  • Dino, how long are your remarks?

  • About 10 minutes.

  • Why don’t we leave them until first thing tomorrow morning? Is that okay? Let’s adjourn.

  • [Meeting recessed]