Housing and the Financial Crisis
July 24, 2012
Adam Guren and Timothy McQuade, Harvard University
The ongoing housing bust precipitated a wave of mortgage defaults, with over 7 percent of the owner-occupied housing stock experiencing a foreclosure. Guren and McQuade present a model that shows how foreclosures can exacerbate a housing bust and delay the housing market recovery. By raising the ratio of sellers to buyers, by making buyers more selective, and by changing the composition of houses that sell, foreclosures freeze up the market for retail (non-foreclosure) sales and reduce both price and volume. Because negative equity is necessary for default, these general equilibrium effects on prices can create price-default spirals that amplify an initial shock. To assess the magnitude of these channels, the model is calibrated to simulate the downturn. The amplification channel is significant. The model successfully explains aggregate and retail price declines, the foreclosure share of volume, and the number of foreclosures both nationwide and across MSAs. While the model can explain variation in sales across MSAs, it cannot account for the aggregate level of the volume decline, suggesting that other forces have reduced sales nationwide. The quantitative analysis implies that in the last several years, foreclosures exacerbated aggregate price declines by approximately 50 percent and declines in the prices of retail homes by approximately 30 percent.
Alexander Chinco, New York University, and Christopher Mayer, Columbia University and NBER
Chinco and Mayer investigate the role that out of town second house buyers ("distant speculators") played in bubble formation in the U.S. residential housing market. Distant speculators are likely to have an excessive reliance on capital gains for financial returns and to be less informed about local market conditions. Using transactions-level data that allow them to identify the address of both the purchased property and the primary residence of the buyer, the authors show that increases in purchases by distant speculators (but not local speculators) are strongly correlated with appreciation in both high house prices and implied-to-actual rent (IAR) ratios -- a proxy for mispricing in the housing market. They develop a simple model that helps them to address the issue of reverse causality. Consistent with this model, they show that the size of the MSA that out of town second house buyers come from is positively related to the impact of distant speculators on house price and IAR ratio appreciation rates in the target MSA. This suggests that out of town second house buyers are not simply responding to unobserved changes in home values in the target MSA. The authors conclude by demonstrating the large impact that distant speculators have on the local economy, with out of town second house purchases equaling as much as 5 percent of total output in Las Vegas during the boom.
Johannes Stroebel, Stanford University
Stroebel empirically analyzes the sources and magnitude of asymmetric information between competing lenders in residential mortgage lending. Large property developers often cooperate with vertically integrated mortgage lenders to provide financing offers to buyers of their newly constructed homes. Stroebel shows that these integrated lenders have superior information about the construction quality of individual homes and exploit this information to lend against higher quality collateral. To compensate for the resulting adverse selection, non-integrated lenders charge higher interest rates when competing against an integrated lender.
Sumit Agarwal, National University of Singapore; Gene Amromin, Federal Reserve Bank of Chicago; Itzhak Ben-David, Ohio State University; Souphala Chomsisengphet, Office of the Comptroller of the Currency; Tomasz Piskorski, Columbia University; and Amit Seru, University of Chicago and NBER
The main rationale for policy intervention in debt renegotiation is to enhance such activity when foreclosures are perceived to be inefficiently high. Agarwal, Amromin, Ben-David, Chomsisengphet, Piskorski, and Seru examine the ability of the government to influence debt renegotiation by empirically evaluating the effects of the 2009 Home Affordable Modification Program that provided intermediaries (servicers) with sizeable financial incentives to renegotiate mortgages. A difference-in-difference strategy that exploits variation in program eligibility criteria reveals that the program generated an increase in the intensity of renegotiations while adversely affecting effectiveness of renegotiations performed outside the program. Renegotiations induced by the program resulted in a modest reduction in rate of foreclosures but did not alter the rate of house price decline, durable consumption, or employment in regions with higher exposure to the program. The overall impact of the program will be substantially limited, because it will induce renegotiations that will reach just one-third of its targeted 3 to 4 million indebted households. This shortfall is due in large part to low renegotiation intensity on the part of a few large servicers who responded at half the rate of others. This muted response cannot be explained by differences in contract, borrower, or regional characteristics of mortgages across servicers. Instead, the low renegotiation activity which is also observed before the program reflects servicer-specific factors that appear to be related to preexisting organizational capabilities. These findings reveal that the ability of government to quickly induce changes in behavior of large intermediaries through financial incentives is quite limited, underscoring significant barriers to the effectiveness of such policies.
Andrew Haughwout, Donghoon Lee, Joseph Tracy, and Wilbert van der Klaauw, Federal Reserve Bank of New York
Haughwout, Lee, Tracy, and van der Klaauw explore a mostly undocumented but important dimension of the housing market crisis: the role played by real estate investors. Using unique credit-report data, they document large increases in the share of purchases, and subsequently delinquencies, by real estate investors. In states that experienced the largest housing booms and busts, at the peak of the market almost half of purchase mortgage originations were associated with investors. In part by apparently misreporting their intentions to occupy the property, investors took on more leverage, contributing to higher rates of default. These findings have important implications for policies designed to address the consequences and recurrence of housing market bubbles.
Kristopher Gerardi, Federal Reserve Bank of Atlanta; Eric Rosenblatt and Vincent Yao, Fannie Mae; and Paul Willen, Federal Reserve Bank of Boston and NBER
In a recent set of influential papers, researchers have argued that residential mortgage foreclosures reduce the sale prices of nearby properties. Gerardi, Rosenblatt, Yao and Willen revisit this issue using a more robust identification strategy combined with new data that contain information on the location of properties secured by seriously delinquent mortgages and information on the condition of foreclosed properties. They find that while properties in virtually all stages of distress have statistically significant, negative effects on nearby home values, the magnitudes are economically small, peak before the distressed properties complete the foreclosure process, and go to zero about a year after the bank sells the property to a new homeowner. These estimates are very sensitive to the condition of the distressed property, with a positive correlation existing between house price growth and foreclosed properties identified as being in "above average" condition. The researchers argue that the most plausible explanation for these results is an externality resulting from reduced investment by owners of distressed property. Their analysis shows that policies that slow the transition from delinquency to foreclosure likely exacerbate the negative effect of mortgage distress on house prices.
Patrick Bayer, Duke University and NBER; Fernando Ferreira; and Stephen Ross, University of Connecticut
Homeownership has long been viewed as an important avenue for wealth accumulation and proposed as a mechanism for reducing racial disparities in wealth. However, it carries considerable risk that may increase the vulnerability of households with limited wealth or liquidity to negative economic or health shocks, with serious long-term consequences for wealth accumulation and credit-worthiness. Bayer, Ferreira, and Ross find large racial and ethnic differences in delinquencies and foreclosures in the years following the recent financial crisis, even for homeowners with similar credit scores. A substantial portion of these differences persist with the inclusion of lender-and-tract fixed effects and with controls for receiving a rate-spread loan intended in part to capture the influence of subprime lending. Contemporaneous controls for negative equity and employment designed to capture differences in the magnitude of shocks do little to erode these differences. This leaves a substantial potential explanatory role for the greater vulnerability of these households to negative shocks, and thus suggests that homeownership may not be an appropriate mechanism for reducing racial disparities in wealth.
Chao He and Yu Zhu, University of Wisconsin, Madison, and Randall Wright, University of Wisconsin, Madison and NBER
Housing, in addition to providing direct utility, facilitates credit transactions when home equity serves as collateral. He, Wright, and Zhu document big increases in home-equity loans coinciding with the start of the house-price boom, and suggest one explanation. When it is used as collateral, housing can bear a liquidity premium. Because liquidity is endogenous, even when fundamentals are deterministic, time-invariant equilibrium house prices can display complicated patterns including cyclic, chaotic, and stochastic trajectories some of which resemble bubbles. This framework is tractable, with exogenous or with endogenous supply, and with exogenous or endogenous credit limits. Yet it captures several salient qualitative features of actual housing markets. Numerical work shows that the model can capture some, if not all, quantitative features, as well. The authors also discuss the effects of monetary policy.
Anthony DeFusco and Wenjie Ding, University of Pennsylvania; and Fernando Ferreira and Joseph Gyourko, University of Pennsylvania and NBER
DeFusco, Ferreira, Gyourko, and Ding investigate whether contagion, which is commonly defined as a correlation across space between different markets above and beyond what normally exists, was an important factor in the last housing cycle. They address three empirical challenges: 1) defining the timing of the housing boom in a non ad hoc way; 2) avoiding specification search bias when only one time series is used to define both the timing and magnitude of contagion, and 3) controlling for omitted variable bias that arises from common local shocks. Using housing transactions data from 99 metropolitan areas between 1993 and 2009, they find that contagion affected both the timing and magnitude of local booms. The probability of a given market booming is about 10 percent higher when there is a standard deviation increase in the number of close neighbors that boomed last quarter. The elasticity of focal market price growth with respect to lagged near neighbors' price growth is about 0.l on average. This impact only exists when the focal market itself booms, and it persists for many years. Thus, contagion appears to have played a role in the expansion of local market booms, not just in their inception. However, there is no evidence here of contagion during the housing bust.