Mortgage Default, Foreclosures and Bankruptcy in the Context of the Financial Crisis

The financial crisis of 2008 and the current recession were triggered by the bursting of the housing bubble and the subprime mortgage crisis that began in late 2006/early 2007. In this paper, we investigate whether personal bankruptcy law also played a role in creating the mortgage crisis or in making it worse. In particular, we show that there is a strong and previously unnoticed relationship between homeowners' decisions to default on their mortgages and their decisions to file for bankruptcy. In theory, bankruptcy and default could be either substitutes or complements. They could be substitutes if homeowners use bankruptcy to save their homes, assuming that they succeed in avoiding default. Alternately, they could be complements if homeowners use bankruptcy to reduce the cost of default, in which case they would be observed to do both or neither. In fact we show that the bankruptcy/default is complementary for most homeowners. We also investigate the relationship between foreclosure and bankruptcy and show that they are also strongly and positively related. The paper argues that these relationships have important public policy implications. In particular, foreclosures have very high social costs, and some of these costs are external to both borrowers and lenders. As a result, there is a social gain from discouraging bankruptcy filings, since fewer bankruptcies would also mean fewer defaults and foreclosures. We show that these considerations shift optimal bankruptcy law in a pro-creditor direction and shift other policies related to bankruptcy in a pro-debtor direction. To give some background on why the bankruptcy/mortgage default relationship has not been recognized, figure 1 shows the bankruptcy filing and mortgage default rates from 1980 to 2008. Both are scaled to equal one in 1980. The diagram suggests that the mortgage default rate was fairly steady over the period—the fraction of mortgage debt that defaulted was around 5% in the 1980s, dropped to 4.5% in the 1990s and early 2000s, and did not begin rising until 2007. In contrast, the bankruptcy filing rate has risen steadily: between 1980 and 2004, it increased from 3.5 to 14 per 1,000 households. A major bankruptcy reform went into effect in late 2005 that made bankruptcy law more pro-creditor and the filing rate responded by jumping to 18 per 1,000 households in 2005, as debtors rushed to file under the old rules. It then dropped sharply to 5.2 in 2006, but since then has resumed its upward trend in 2007, rising to 9.2 per 1,000 households in 2008. 2 If we run a regression explaining the bankruptcy filing rate as a function of the mortgage default rate, the coefficient of the default rate is negative and not statistically significant. Thus aggregate data over a long time period suggest little relationship between bankruptcy and mortgage default. But the picture changes if we examine micro data over a more recent time period. We use the LPS data, a large dataset of mortgages that are followed every month. Our sample consists of prime and subprime mortgages that originated in 2004 or 2005 and are followed monthly from origin until they are paid off, go into foreclosure, or until the end of our sample period in October 2008. 3 Mortgage default is defined to occur when homeowners are delinquent on their payments by one month or more. We also know when homeowners file for bankruptcy and when lenders begin the foreclosure process. Figure 2 shows monthly default rates, foreclosure rates, and bankruptcy filing rates, with prime and subprime mortgages shown separately. The period covered includes both the October 2005 bankruptcy reform and the start of the mortgage crisis. Using this data, the correlation coefficients between bankruptcy and mortgage default and between bankruptcy and foreclosure are .62 and .87, respectively. For subprime mortgages, the correlations are .86 and .87, respectively. Thus recent data suggest that homeowners' default and bankruptcy decisions are closely related and that homeowners' bankruptcy decisions are closely related to the start of foreclosure,download.