Asset bubbles and supply failures: where are the qualified sellers?

AuthorGouldey, Bruce K.
PositionEssay

From their peak during the third quarter of 2007, to their trough during the first quarter of 2009, stock prices as measured by the S&P 500 fell 48 percent (see Figure 1). Through the third quarter of 2009, stock prices subsequently increased more than 40 percent. In contrast, housing prices, as measured by the Case-Shiller index, which fell 31 percent from their peak in the first quarter of 2006 to the first quarter of 2009, had not yet shown any sign of recovery.

In both markets, we observed the bursting of an asset bubble, but, while one market quickly began the process of recovery, the other did not. As of the writing of this article, the housing market is still characterized by large unsold inventories, rising foreclosures, and a damaged construction industry. Why did the stock market, which fell farther and faster and--with the exception of the financial services and auto industries--was not supported by the government, recover when the housing market did not?

We argue that, when a housing bubble bursts, the combination of high loan-to-value mortgages and costly foreclosures can inhibit house prices from quickly falling to their new equilibrium levels. The adjustment problem manifests itself, among other ways, in homeowners being unable to complete the sale of their houses at current market values. For some time, the resulting supply failure distorts the supply of houses offered for sale, inventories of houses listed for sale, and the relative prices of different quality houses. Public policies have been targeted mostly at maintaining house prices by propping up demand. Those policies exacerbate the problems associated with this supply failure and can result in substantial reduction of social welfare due to misallocation of resources.

[FIGURE 1 OMITTED]

The model we develop suggests instead that public policy should focus on unwinding uneconomic contracts in order to enable house prices to fall quickly to their equilibrium levels. Such policies would enable more homeowners to sell their houses at then current market prices, restoring the normal turnover of the housing stock and housing mobility to families. Indeed, by returning a measure of liquidity such policies would, in the long term, contribute to the demand for housing.

Asset Bubbles

An asset bubble can be said to exist when the price of the asset exceeds its fundamental or intrinsic value. (1) Krainer (2003), for example, says that a housing bubble occurs when the ratio of house prices to rent trends above its long-run average. Asset bubbles are usually if not always associated with an expansion of money and credit. In the recent stock and housing bubbles, financial innovations were involved during the run-up in prices, such as risk-shifting through collateralized debt obligations and credit default swaps. Those innovations facilitated an increase in the flow of credit to the housing market, but also built up a mismatching of assets and their claims. Eventually, the stock and housing bubbles burst because of these mismatches, mortgage defaults, and the loss of confidence in asset-backed commercial paper. Brunnermeier (2009) estimates that in the year following the stock market peak in October 2007, stocks lost $8 trillion. The fall in stock prices was brutal both in terms of the extent and quickness of the loss. The fall in housing values was nowhere near as brutal. As measured by the biennial American Housing Survey, the median value of houses fell $21,500 from 2007 to 2009, for a total loss of $1.9 trillion, about one quarter of the stock market loss. But, whereas stock values began to recover in the second quarter of 2009, housing values continued to deteriorate through 2009.

Previous examinations of the phenomenon of bubbles have focused mainly on the conditions for the existence of bubbles (e.g., Tirole 1982, 1985a, 1985b; Grossman and Yanagawa 1993; Allen et al. 1993; Kunieda 2008). Abreau (2003) examines the persistence and bursting of bubbles. Barlevy (2007) provides a good review of this literature. Allen and Gale (2000) show that financial leverage, through asymmetric information and risk-shifting, not only can cause bubbles, but also can exacerbate crises following the bursting of bubbles, particularly in assets with a supply that is relatively fixed. Glaeser et al. (2008) show that less elastic housing markets have longer and larger bubbles. They measure a bubble by a rising price-to-cost ratio in 79 metropolitan areas, where their cost index involves the sum of building cost per square foot and the cost of land.

Recent changes in mortgage financing have made the U.S. housing market less elastic. Government intervention in mortgage finance (e.g., through the lowering of FHA and VA down payments first to 3 percent and then to zero and increases in guaranteed loan maximums through Fannie Mac and Freddie Mac) resulted in a higher proportion of fixed rate (versus variable rate) mortgages, higher loan-to-value ratios, and lower interest rates. International banking regulations encouraged the acquisition of government-guaranteed mortgages by treating them as nearly risk-free. Initially financial engineering seemed to better spread default risk, interest rate risk, and liquidity across a wider spectrum of investors. Instead it may have concentrated the risks involved in mortgage lending so as to replace conventional risks with systemic risk and compromised the objectivity of regulatory bodies. In addition, financial engineering degraded transparency in mortgage finance. These institutional arrangements attenuated the housing bubble by increasing demand for residential houses. Green and Wachter (2005) and Coleman et al. (2008) show that the increase in subprime mortgage originations did not cause the housing price bubble in the late 2000s, but rather was a joint product, as new entrants displaced the government-sponsored enterprises (GSEs)--that is, Fannie Mae and Freddie Mac--in the packaging and sale of mortgage loans. Coleman et al. segment the mortgage market by purchase transaction size and show that, near the beginning of 2002, the percentage increase in price in the lowest tier started to increase at a faster rate than mid- and high-tier homes. This increase in the price in the lowest tier also was accompanied by an increase in the percentage of subprime mortgage originations and, later, by an increase in the percentage of alt-A mortgages. (2)

A significant aspect of the bubble-bursting in 2007 is that many financial markets froze. For a time even though bid and ask prices for certain financial assets were posted, no one was actually willing to trade at the posted prices. Easley and O'Hara (2010) show that markets will freeze when traders' portfolio preferences are incomplete in the presence of Knightian uncertainty. (3) When traders' preferences are incomplete and beliefs are uncertain, inertia can rule and no trades occur. The absence of trades makes unbiased valuation difficult. As indicated by time on the market, the housing market also froze. Yet, houses have real utility, and the bursting of the housing bubble should have had no impact on the house preferences of actual and potential homeowners.

In the next section, we argue that the combination of binding mortgage constraints and costly foreclosure distort the housing market upon the bursting of a housing bubble by inhibiting the fall of house prices to the levels necessary to clear the market. This distortion occurs because of a contraction in the number of qualified sellers. Homeowners cannot complete the sale because the current market value of their house is "underwater" or less than the balance on their mortgage and it would be costly to the homeowner to cover the deficiency.

Learner (2007) observes that housing follows a volume cycle, not a price cycle. That is, home prices are always sticky downward when demand declines. Instead of selling at depressed prices, homeowners choose not to make an elective move when prices in the short term are below intrinsic values, anticipating that prices will eventually return to their long-run equilibrium levels. This is a well-known example of time flexibility as a real option (e.g., see Brealey et al. 2006). However, when a housing bubble bursts, homeowners often face a different scenario. In this case, a falling price might still be greater than the new equilibrium price, but less than both the price paid for the house and the outstanding mortgage loan balance.

Below we develop a model describing first the generation of a housing bubble and then the bursting of the bubble. With regard to the bursting of the bubble, we utilize an agent model. (4) Our primary conclusion is that the combination of high loan-to-value mortgages and costly foreclosure have constrained the number of qualified sellers, causing the housing market to freeze, distorting relative prices, and inhibiting housing prices from quickly falling to market-clearing levels. Our analysis suggests that public policy should focus not on increasing housing demand but rather on removing constraints on the supply of houses being offered for sale so as to allow prices to quickly reach their equilibrium levels.

Formation of a Housing Bubble

For the present purpose, houses are not usefully characterized as commodities; rather, the facts that houses are discrete, durable, and differentiated goods must be taken into account. A house's price, accordingly, depends on its desirability relative to that of other houses. A particular house's desirability is affected by such factors as location, size, age, construction quality, lot, and finishes. In a perfect market all houses with the same desirability should sell at the same price and 'all houses with greater desirability should sell at a higher price than houses with lesser desirability. Otherwise, potential buyers of lesser desirable houses instead would purchase houses that are both more...

To continue reading

Request your trial

VLEX uses login cookies to provide you with a better browsing experience. If you click on 'Accept' or continue browsing this site we consider that you accept our cookie policy. ACCEPT