Three Triggers? Negative Equity, Income Shocks and Institutions as Determinants of Mortgage Default

Abstract

In understanding the determinants of mortgage default, the consensus has moved from an 'option theory' model to the 'double trigger' hypothesis. Nonetheless, that consensus is based on within-country studies of default. This paper examines the determinants of mortgage default across five European countries, using a large dataset of over 2.3 million active mortgage loans originated between 1991 and 2013 across over 150 banks. The analysis finds support for both elements of the double trigger: while negative equity itself is a relatively small contributor to default, the effect of unemployment, and other variables such as the interest rate, is stronger for those in negative equity. The double trigger, however, varies by country: country-specific factors are found to have a large effect on default rates. For any given level of a loan's Loan to Value ('LTV') ratio, and as LTV changes, borrowers were more sensitive to the interest rate and unemployment in Ireland and Portugal than in the UK or the Netherlands.

Housing loan default in Malaysia: an analytical insight and policy implications

“…1. Campbell and Dietrich (1983), Quercia and Stegman (1992), Hakim and Haddad (1999), Deng et al (2000), Noordewier et al (2001), Archer et al (2002), Feldman and Gross (2005), Kelly (2008), Goodman and Smith (2010), Elul et al (2010), Goodman and Smith (2010), Kau et al (2011), Quercia et al (2011), Archer and Smith (2013), Park and Bang (2014), Campbell and Cocco (2015), Jones and Sirmans (2015), Tian et al (2016), Badarinza et al (2018), Laufer (2018), Gupta (2019), Cloyne et al (2019), Defusco et al (2020), Adzis et al (2020), Agarwal et al (2020), de Haan and Mastrogiacomo (2020), Ross and Shibut (2020), Linn and Lyons (2020), Ganong and Noel (2020), Nakatani (2020). …”

Malaysian residential mortgage loan default: a micro-level analysis

“…Loan characteristic factors are factors related to loan attributes such as initial LTV ratio (Deng et al, 1996;Hakim and Haddad, 1999;Elul et al, 2010;Quercia et al, 2012), Malaysian residential mortgage loan default amount of initial loan (Yang et al, 1998;Hakim and Haddad, 1999;Kelly, 2008;Soyer andXu, 2010 andQuercia et al, 2012), mortgage age (Campbell and Dietrich, 1983;Quercia and Stegman, 1992;Springer and Waller, 1993;LaCour-Little, 2004;Ghent and Kudlyak, 2011), negative equity (Deng et al, 2000;Linn and Lyons, 2019) and call option value (LaCour-Little, 2004;Quercia et al, 2012). Trigger events are defined as loss of employment (Quercia and Stegman, 1992;Deng et al, 1996Deng et al, , 2000Capozza et al, 1997;Elmer and Seelig, 1999;Elul et al, 2010;Quercia et al, 2012;Linn and Lyons, 2019) and divorce (Capozza et al, 1997;Deng et al, 2000). Borrower characteristics factors are factors related to borrower attributes such as payment-to-income ratio (Stansell and Millar, 1976;Vandell, 1978;Springer and Waller, 1993;Yang et al, 1998;LaCour-Little and Malpezzi, 2003;Bajari et al, 2008;Kelly, 2008 andArcher andSmith, 2013), borrower income…”

“…Trigger events are defined as loss of employment (Quercia and Stegman, 1992;Deng et al, 1996Deng et al, , 2000Capozza et al, 1997;Elmer and Seelig, 1999;Elul et al, 2010;Quercia et al, 2012;Linn and Lyons, 2019) and divorce (Capozza et al, 1997;Deng et al, 2000). Borrower characteristics factors are factors related to borrower attributes such as payment-to-income ratio (Stansell and Millar, 1976;Vandell, 1978;Springer and Waller, 1993;Yang et al, 1998;LaCour-Little and Malpezzi, 2003;Bajari et al, 2008;Kelly, 2008 andArcher andSmith, 2013), borrower income (Yang et al, 1998;Hakim and Haddad, 1999;Jones, 1993;LaCour-Little and Malpezzi, 2003;Feldman and Gross, 2005;Quercia et al, 2012;Linn and Lyons, 2019) and other borrower characteristics such as age, job position and number of dependents (Jones, 1993). Local housing market and macroeconomic factors that might contribute to mortgage loan default as reported in the existing studies are house price volatility (Capozza et al, 1997;Yang et al, 1998), interest rate spread (Springer and Waller, 1993;Capozza et al, 1997) and interest rate volatility (Quigley and Van Order, 1995).…”

Social traits and credit card default: a two-stage prediction framework

“…The primary source of our data is the European Data Warehouse (EDW), which is a repository of the universe of loan and bond level data under the European Central Bank's loan level initiative. 2 EDW has recently started gaining traction, used in studies that examine various issues like the impact of collateral eligibility on credit supply (Van Bekkum et al, 2018), the association between asset transparency and credit supply (Balakrishnan & Ertan, 2019), the riskiness of securitised loans granted to small and medium enterprises (Bedin et al, 2019), the impact of lending standards on the default rates of residential loans (Gaudêncio et al, 2019), the determinants of mortgage default (Linn & Lyons, 2020), and the association between the energy efficiency of buildings and mortgage defaults (Billio et al, 2021). When it comes to credit card default data, there is only one case study by Licari et al (2021) that uses the Markov decision process model to generate a dynamic credit limit policy.…”