This article came out from the Federal Reserve Bank of New York. It’s a wonderful look at debt from an analytical point of view. I’m so glad they wrote it.
When Debts Compete, Which Wins?
When faced with financial hardship, borrowers might choose to repay some debts while falling behind on others—potentially going into default. Such choices provide insight into consumers’ spending priorities and can help us better understand the condition of borrowers under financial distress. In this post, we examine how consumers prioritize their default choices. Do consumers under financial stress default on their credit cards first? Or are they more likely to default on their mortgage?
Using data from the New York Fed Consumer Credit Panel / Equifax (CCP), we can begin to answer these questions. We empirically model this situation as a kind of competition between types of debt, with each category of debt vying for the limited funds of a particular borrower. When this borrower has two forms of debt and only pays one, we predict which type of debt is most likely to be repaid. The repaid debt is designated the “winner” of our pseudo-competition. For example, if an individual pays his or her mortgage while defaulting on credit cards, we describe mortgage debt as the winner and credit card debt as the loser.
Once we consider this situation as a competition, we can see that predicting which debt is kept current is similar to predicting victory probabilities for a football game or a chess match—a task for which “Elo ratings” are used. An Elo rating reflects the strength of a player, or in this case debt, to win a head-to-head competition for payment priority. The higher the rating, the more likely the debt is to be kept in good standing.
We similarly estimate ratings for each type of debt using a symmetric logistic regression. Our approach can accommodate multiple kinds of debt simultaneously, giving a single rating for each category of debt by predicting the likelihood of payment relative to that for all other debt types. These ratings can also be calculated by year, state, or other subsets to see if this relative prioritization of debt changes over time or varies across individuals. In addition, this approach can control for loan characteristics, such as the balances due on the different categories of loans.
Looking at the ratings by year in the chart below, we can see that auto and mortgage debt are strongly prioritized over credit card debt. In the year 2000, for example, the ratings imply that auto and mortgage loans were respectively 10.8 and 10.6 times more likely to be paid in a competition with credit card debt. This makes intuitive sense, as individuals put their cars or homes at risk by defaulting on this collateralized type of loan. Generally borrowers have less to lose when defaulting on credit card bills.
Interestingly, we observe a sharp drop in the prioritization of mortgage debt relative to auto and credit card debt during the 2007–09 recession. Prior to the recession, individuals were about equally likely to prioritize auto loans as they were mortgage loans when given the choice, showing a slightly greater tendency to pay the latter. At the onset of the recession, this pecking order reverses, and auto loans become more than twice as likely to be paid when competing with mortgage loans.
Calculating our prioritization ratings by state, we can observe which parts of the country experienced the largest drops in mortgage prioritization between the pre- and post-recession periods (1999–2007 and 2008–2015, respectively). In the map below, red shading indicates larger declines in mortgage prioritization. Mortgage prioritization fell across the United States in the aftermath of the recession, with California, Nevada, and Florida experiencing the largest drops in mortgage prioritization. These states also experienced large home price declines during the recession. One possible explanation for this phenomenon is that individuals in states where house prices declined more had less equity in their homes and therefore less to lose by defaulting on their mortgage payments.
Borrower types offer another dimension for assessing debt prioritization patterns. The chart below compares the debt prioritization profiles of low- and high-credit-score individuals (defined as being in the bottom or top half of the Equifax risk-score distribution in the previous year). We see that while both sets of individuals experienced a decline in mortgage prioritization, the trend was particularly pronounced among borrowers with previously high credit scores. In fact, there was a period when high-credit-score individuals prioritized credit card debt over mortgage debt when choosing between the two!
These novel debt prioritization metrics allow us to characterize which debts individuals choose to keep in good standing. In particular, we are able to observe a large drop in the relative prioritization of mortgages in the post-recession period, a trend that is particularly pronounced in states that experienced a large decline in housing prices and among individuals with historically high credit scores. Our findings are consistent with strategic default motives—that is, borrowers choosing to cease payment on collateral whose value falls below the borrowed amount. However, additional analysis is needed to better identify precisely why people choose to prioritize one form of debt over another. In future work, we will look to further explain these patterns by considering the presence of nonrecourse laws and judicial foreclosures, the average time to foreclosure, home equity levels, and other such factors.
The views expressed in this post are those of the authors and do not necessarily reflect the position of the Federal Reserve Bank of New York or the Federal Reserve System. Any errors or omissions are the responsibility of the authors.