Collateral Misreporting in the Residential Mortgage-Backed Security Market

Abstract

<jats:p> Securitized mortgage appraisals routinely target pre-specified valuations, 45% of purchase loan appraisals exactly equal purchase prices, and appraisals virtually never fall below purchase prices. As a result, appraisals exceed automated valuation model (AVM) valuations 60% of the time and are 5% higher than AVM valuations on average. High appraisals and indicators of appraisal targeting predict loan delinquency and residential mortgage-backed security (RMBS) losses and are priced at the loan level through higher interest rates, but have essentially no impact on RMBS pricing. Selection bias simulations and unfunded loan application appraisals indicate that high appraisals are intentional. The extent to which appraisals exceed AVM valuations varies across loan officers, mortgage brokers, and appraisers, and high appraisals are associated with more repeat business for appraisers, potentially incentivizing appraisers to inflate their appraisals. </jats:p><jats:p> This paper was accepted by Tomasz Piskorski, finance. </jats:p>

Distressed comps

“…Finally, in contrast to most databases that only include appraisal information for originated loans, our data include appraisals associated with applications that did not result in funded loans as well. 12 Thus, we are able to circumvent a selection issue, discussed by Kruger and Maturana (2021), Conklin et al (2020), andEriksen et al (2020), which is common in the literature.…”

Appraisal overvaluation: Evidence of price adjustment bias in sales Comparisons

“…Researchers often attribute the appraisal's propensity to support the collateral underwriting to agency incentives (see, e.g., Agarwal et al., 2015; Calem et al., 2015; Cho & Megbolugbe, 1996; Eriksen et al., 2020; Ferguson, 1988; Kruger & Maturana, 2020; Lentz & Wong, 1998). According to the agency theory, the principal–agent relationship between lenders and appraisers offers an institutional framework to understand the inherent bias in the appraisal process.…”

“…With the central role of collateral in mortgage credit decisions, the last mortgage crisis has prompted research on the identification and measurement of appraisal overvaluation in both purchase money and refinance mortgage transactions (Agarwal et al., 2015; Calem et al., 2015; Griffin & Maturana, 2016; Kruger & Maturana, 2020). However, the underlying mechanism that leads to appraisal overvaluation remains unclear.…”

“…According to the 2015 Home Mortgage Disclosure Act (HMDA) data, collateral was the third most frequently cited reason for loan application denial or 13.7% on first-lien purchase mortgages.5 Systematic appraisal bias or over-valuation in refinancing transactions are discussed and documented inAgarwal et al (2015),Agarwal et al (2020),Shi and Zhang (2015),Griffin and Maturana (2016), andKruger and Maturana (2020). Evidence of potentially inflated appraisals in HUD-backed reverse mortgage lending, Home-Equity Conversion Mortgage (HECM) program, can be found inPark (2017) andMayer (2020).…”

Ignoring Spatial and Spatiotemporal Dependence in the Disturbances Can Make Black Swans Appear Grey

“…The authors also show that high pricing errors can be the cause of an apparent appraisal bias even when the AVMs are unbiased. Krugery and Maturanaz (2020) analyze appraisals and AVMs on a large sample of US non-agency securitized loans originated between 2001 and 2007. They highlighted that the potential sources of the errors for AVMs and appraisals are different.…”