Deficiency judgments as a mortgage pricing factor

Pruszkowski, Tomasz Article Deficiency judgments as a mortgage pricing factor International Journal of Management and Economics Provided in Cooperation with: SGH Warsaw School of Economics, Warsaw Suggested Citation: Pruszkowski, Tomasz (2017) : Deficiency judgments as a mortgage pricing factor, International Journal of Management and Economics, ISSN 2543-5361, De Gruyter Open, Warsaw, Vol. 53, Iss. 2, pp. 57-68, https://doi.org/10.1515/ijme-2017-0012 This Version is available at: https://hdl.handle.net/10419/309637 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by-nc-nd/3.0/ DOI: 10.1515/ijme-2017-0012 International Journal of Management and Economics Volume 53, Issue 2, April–June 2017, pp.57–68; http://www.sgh.waw.pl/ijme/ Tomasz Pruszkowski1 Department of Capital Markets, Warsaw School of Economics, Warszawa, Poland Deficiency Judgments as aMortgage Pricing Factor Abstract The subject of the deficiency judgments has been poorly examined due toalack of relevant data and the complexity of the issue. Some comprehensive studies have explored whether allowing deficiency judgments decreases the likelihood of strategic defaults inthe U. S. mortgage market. Little, however, has been done todetermine whether there is any direct correlation between legal standing allowing recourse and loan pricing. Hence, additional work regarding this subject is needed. This study seeks tofill this gap by exploring the impact of allowing deficiency judgments on mortgage pricing policy invarious U. S. states. Seven distinctive mortgage types intwo groups of states were compared. We conclude that there is nostatistically significant difference between recourse and non-recourse states interms of mortgage pricing, regardless of mortgage type. Keywords: deficiency judgment, recourse and non-recourse mortgages, mortgage pricing policies JEL:G21, G28, K25, K35 Introduction Aset number of factors determining the likelihood of default and foreclosure emerge from empirical research and theoretical analysis. Primary among them are deficiency © 2017 Tomasz Pruszkowski. This is an open access article distributed under the Creative Commons Attribution-NonCommercial-NoDerivs license (http://creativecommons.org/licenses/by-nc-nd/3.0/). Tomasz Pruszkowski 58 judgments, equity of redemption, the statutory right of redemption and judicial/ non-judicial foreclosures [Clauretie, Herzog, 1989, pp.221–233], as well as bankruptcy laws. In this study we differentiated between states inwhich the mortgage holders can file for adeficiency judgment if the proceeds from aforeclosure sale are insufficient tocover their mortgage debt, and states inwhich lenders are prohibited or severely restricted from pursuing the borrower. Based on the foregoing dichotomy, we analyzed whether there are risk assessment differences between these two groups of states and, consequently, more (less) stringent pricing policies instates considered non-recourse (recourse). Mortgage interest rates stem from interbank offer rates and assessments of default risk. Lenders have little direct influence on the former. Therefore, the latter should pre-eminently differentiate pricing policies among lenders. Since mortgages are long-term obligations, interest rate surges increase clients’ likelihood of default. And property price declines reduce the ability of borrowers’ torefinance. This is especially true inthe U. S. market where, unlike inEurope, interest rates are usually fixed either for the whole term of the mortgage or for the first few years (FRM, 5/1 ARM, 7/1 ARM). The mortgage market inthe U. S. –as opposed toEurope –is often perceived as recourse, because most states allow deficiency judgments. Among recourse states, 21 have nolimits and 27 impose restrictions. Adeficiency judgment usually refers toalender’s claim for the difference between aproperty’s value at the time of foreclosure and the outstanding loan balance, accrued interest, and legal and other costs acquired inthe course of the procedure following default. It is believed that the single most important reason for default is the borrower’s financial situation deterioration. This reason has been the most common explanation for rising foreclosures inthe Great Recession. In that financial shock, negative equity has been viewed as acollateral consequence of severe real estate depreciation, rather than apotential risk factor for mortgage default. The literature often portrays amortgage as atype of financial asset [Kau etal., 1993, pp.288–299]. As such, it is atool determining credit risk. Option-pricing modeling is usually applied toassess the likelihood that aparticular borrower will default, and includes deficiency judgments as afactor inexercising that option [Johnes, 1993, pp.115–138]. This approach implies that aborrower will default only when the option is in-the-money, meaning that discounted peculiar proceeds and non-peculiar gains combined exceed the costs incurred and discounted potential loss. In other words, intrinsic value should notbe less than zero totrigger default on any given mortgage loan. Regardless of how controversial this line of reasoning is, it is more appropriate inanon-recourse state where future cash flows resulting from default are more predictable for aborrower who considers whether tostop loan payments. Applying this model inarecourse state –where neither the scope northe timing of amortgagee action can be accurately foreseen –requires amore elaborate methodology. Though mortgage defaults usually emanate directly from borrowers’ financial problems, some people default even when they are capable of paying monthly installments. These Deficiency Judgments as aMortgage Pricing Factor 59 borrowers are referred toas strategic defaulters. Several researchers believe that there is acause and effect relationship between deficiency judgment availability and strategic default prevalence inany given state. That is, this type of default is higher innon-recourse states than inrecourse states. There is little consensus about whether deficiency judgment availability has any direct impact on strategic default rates. This is partially due toashortage of empirical evidence driven by lack of data availability which, when it exists at all is usually proprietary and therefore unavailable toresearchers. In addition, classification factors are ambiguous; there is nosingle, generally accepted definition of strategic default due tothe lack of standardized databases. Rather, inconsistent, erratic definitions complicate matching the data that is available toresearchers. Consequently, areliable comparison between studies is challenging. Conflicting results are also attributable toavariety of factors relevant tothe default risk assessment process, such as vintage and geographical location [Experian and Oliver Wyman, 2010, pp.4–6], volume [Ghent, Kudlyak, 2011, pp.3139–3186], whether the mortgage is afirst or second lien [Jagtiani, Lang, 2011, pp.7–23], and morality [Guiso etal., 2013, pp.1473–1515 and Guiso etal., 2009, pp.1–33] and [White, 2010a, pp.1–12] and [White, 2010b, p.971] and [Bridgeman, 2011, pp.123–153]. Afew studies referring tobehavioral theories have tacked these problems [Wilkinson-Ryan, 2011, pp.1547–1583 and Riddiough, Wyatt, 1994, pp.299–318]. Most of these studies compare U. S. states where, by law or legal custom, different remedies are available topursue aborrower for the debt owed. No static model can explain the examined question for the reasons mentioned above. However we attempt todetermine the impact of particular legal regulations on economics through behavioral channels. This paper consists of literature review concerning deficiency judgments as afactor affecting strategic default, followed by astatistical analysis of whether lack of recourse is priced into mortgage loans. Adiscussion of our findings is then provided and followed by final remarks. Literature Review The findings presented inthe literature are far from conclusive. Researchers are divided evenly between those convinced that recourse has asignificant role inpreventing strategic default and those that see noevidence confirming any causal relationship. Data drive the disagreement among researchers, as there is still nocomprehensive data source or unambiguous definition of astrategic default. One of the more comprehensive studies is based on proprietary data from LPS Applied Analytics, Ghent etal. [Ghent, Kudlyak, 2011, 2011, pp.3139–3186], which found that recourse has nodirect implication, inabsolute terms, on default rates inrecourse versus non-recourse states, but it does help tolower borrower sensitivity tonegative equity. Moreover, that study suggests that this relationship is strengthened for properties appraised at $ 500,000 to$ 750,000, Tomasz Pruszkowski 60 for which owner, borrowers are twice as likely todefault inrecourse states. Astudy by Market Intelligence Report supports this observation [Experian, Wyman, 2010, pp.8–18]. The advantage of this report is an unrestrained access tomortgage and non-mortgage data obligations assigned toparticular borrowers (provided by Experian), which permits calculation and statistical analysis based on aspecific, arbitrary definition of astrategic defaulter. Specifically, mortgage and non-mortgage obligations were coupled and strategic defaulters defined as acustomers with 60 to180 or more days past due on their mortgage, with noarrears regarding other types of financial commitments (e.g., less than sixty days past due on auto loans, less than ninety days past due on bank cards, retail cards, and personal loans, and being current on other financial commitments –as verified for the six month period after the first sixty days past due on the mortgage). The report concludes that borrowers with one mortgage are the most likely towalk away from their home despite having financial resources tocontinue monthly payments. The authors also find that strategic default was the prevailing form of default among sub-prime borrowers. Geographical patterns also emerge from the report. Between 2005 and July 2009 defaults increased eighty and fifty-three times inCalifornia and Florida respectively. The timing between borrowing and defaults is also noteworthy: strategic defaulters inthe first half of 2009 were six times more likely tohave obtained their mortgages in2006 as compared tothose with mortgages originating in2004. These claims have been repeatedly challenged. Due tothe fact that other states have been consistently keeping the status of recourse/non-recourse policy, Li and Oswald focused on Nevada, which ended recourse judgments in2009 (mortgage loans made after October 2009 were collateralized by primary single family homes) [Li, Oswald, 2014, pp.2–23]. Using unique mortgage loan data, the authors claim that this change did notdecrease defaults or foreclosures, and that mortgage demand remained stable despite stricter underwriting processes (lenders reduced approval rates and loan size). Adifferent perspective emerges from amore general study by Kanis Saengchote [Saengchote, 2014, pp.2–28] pertaining toall U. S. states. The author uses the BAPCPA (Bankruptcy Abuse Prevention and Protection Act of 2005) as aproxy for states that permits recourse judgment. Although both papers have asimilar starting point, their conclusions differ. Saengchote finds that stronger recourse laws may deter strategic defaulters, and significantly increase credit supply inrecourse states after BAPCPA. This is considered aconsequence of perceiving mortgage collateral as afactor reducing risk as bankruptcy law become strengthened, and therefore the number of eligible borrowers is reduced. He also finds arelationship with the underwriting process. Hatchondo, Martines and Sanchez constructed amodel totry todetermine how recourse mortgages and LTV limit regulations can be used tomitigate default risk [Hatchondo etal., 2014, pp.10–48]. Their study concluded that regulation being too lax or too harsh may inflict harm tothe property market. Lenders deprived of primary recourse remedies may seek more defaults. Overly stringent regulations may decrease housing demand Deficiency Judgments as aMortgage Pricing Factor 61 and diminish ahousehold’s ability toprovide self-insurance for themselves. The authors claim that these adverse results may be mitigated by combining relatively mild recourse laws with Loan-to-Value limits, which could lower default rates and strengthen demand without impairing ahome buyers ability tocover the risk using self-insurance. Another study focusing on the relationship between recourse and non-recourse states [Bhutta etal., 2010, pp.14–29] seeks toinvestigate the underlying cause of mortgage defaults inArizona, California, Florida and Nevada (covering non-prime mortgages inthe period between 2006 and September 2009). Their findings suggest that –depending on the negative equity threshold –both the “double trigger” and “strategic default” theories are correct. The “double trigger” theory states that life events and liquidity constraints, along with “ruthless calculation” may trigger adecision todefault on an “underwater” mortgage. “Strategic default” theory states that rational calculation drives borrowers todefault when negative equity is involved. The authors attempt toquantify this threshold, finding that the median borrower refrains from default until negative equity exceeds 62%. On the other hand, when negative equity is 10% or less, the combination of life events and negative equity can trigger delinquent payments. Literature on what implications lender remedies have on voluntary defaults, and the price implication resulting from different legal rights, are inshort supply. The few that have been conducted concentrate more on coinsurance and risk transfer between mortgage holders and insurers, rather than the general costs imposed on abank or borrower. Based on quantitative analysis, Ambrose etal. advise private mortgage insurers to increase transaction costs associated with mortgage put option for borrowers. To do so, they recommend shortening the period between default and foreclosure and increasing “inthe money” deficiency judgments todecrease strategic defaults, and cut costs. The authors also recommend that governmental insurers (who provide de facto PMI) actively seek deficiency judgments. This recommendation rests on two study findings: (a) default probability increases with the expected delay between default and actual foreclosure and (b) areverse correlation between likelihood of deficiency judgment and default [Ambrose etal., 1997, pp.314–325]. These findings are generally corroborated by Jagtiani, Lang [2011, pp.7–23]. Their study concentrated on factors determining aborrower’s propensity todefault on afirst lien while staying current on asecond lien mortgage. The authors revealed several statistically relevant regularities. Among them is that negative equity was aprerequisite for default, but insufficient totrigger it. This explains why there is asignificant proportion of borrowers defaulted on their first lien mortgage while remaining current on their second lien mortgage. Borrowers were especially likely todo sowhen the second lien was aHELOC (home equity line of credit) rather than HELOAN (home equity loan), as the former provided borrowers with acredit line. Surprisingly, though, researchers found noevidence that mortgage quality (prime, alt–Aor subprime) played asignificant role indetermining this behavior. Tomasz Pruszkowski 62 Perhaps the most perplexing results concerning pricing specifically emerge from astudy conducted by [Ghent, Kudlyak, 2011, pp.3174–3177], indicating that interest rates are actually higher inrecourse, versus non-recourse states (excluding privately held mortgages). They also found the strongest deterrent effect for strategic defaulters tobe properties appraised at $ 750.000 to$ 1.000.000, which may suggest astrong correlation between recourse and pricing within that price category. Data and Methodology One would expect that the correlation between legal standing inagiven state and probability of default would impact risk assessment. Therefore that risk, being acrucial pricing element, would be reflected ininterest rates. Since most studies ignore the differences between recourse and nonrecourse states, we investigate the issue from apractical perspective. We consider interest rates tobe the most adequate determinant of potential price differentiation between states, since interest rates reflect the cost of capital and of risk involved indifferent types of loans. To determine whether there is aclear relationship, we juxtapose average interest rates across recourse and non-recourse states as determined by Ghent and Kudlyak [Ghent, Kudlyak, 2011, pp.3143–3146]. We then split the data into seven groups toreflect distinctive types of mortgages (instead of pooling these data together). While this may increase the likelihood of having tochoose non-parametric calculation methods due toamore scattered distribution, it enhanced the practical results. Seven types of mortgages were taken into consideration, namely: thirty year –term mortgages with afixed interest rate (30 year fixed), fifteen year-term mortgages with afixed interest rate (15 year fixed), hybrid adjustable-rate mortgages (5/1 ARM), thirty year-term refinanced mortgages (30 yr fixed mtg refi), fifteen years-term refinanced mortgages (15yr fixed mtg refi), hybrid adjustable-rate refinanced mortgage (7/1 ARM refi), and fifteen year-term jumbo refinanced mortgages with fixed interest rates (15 yr jumbo fix mtg refi). The hybrid adjustable-rate mortgage (5/1 ARM) has afixed interest rate for 5 years, that is then adjusted annually. From year six onwards, the interest rate is based on an index factor and apredetermined margin. The hybrid adjustable-rate refinanced mortgage (7/1ARM refi) has the same structure, but it lasts seven years. These mortgage types can be used when the property will likely be sold before the initial fixed payment period. Adjustable-rate mortgages or ARM are much less popular inthe U. S. than incontinental Europe. According to2010 studies by Bankrate, less than 10% of would-be mortgage holders chose this type of mortgage [available at: http://www.bankrate.com/ accessed: November 8, 2016]. The fifteen year-term jumbo refinanced mortgage with fixed interest rate (15 yr jumbo fix mtg refi) refers toamortgage exceeding conforming limits imposed by Fannie Mae and Freddie Mac regulations. Cut-off points differ throughout the U. S., but inmost states mortgages over $ 417.000 are qualified as jumbo. Deficiency Judgments as aMortgage Pricing Factor 63 TABLE 1. States classification of various mortgage types with their average interest rates State Recourse/ non-recourse state 30 year fixed 15 year fixed 5/1 ARM 30 yr fixed mtg refi 15 yr fixed mtg refi 7/1 ARM refi 15 yr jumbo fixed mtg refi Alabama recourse 3.45 2.74 3.05 3.49 2.77 3.17 3.14 Alaska non-recourse 3.44 2.75 3.11 3.56 2.8 3.12 3.18 Arizona non-recourse 3.5 2.78 3.02 3.53 2.8 3.11 3.91 Arkansas recourse 3.47 2.75 3.38 3.52 2.79 3.23 3.21 California non-recourse 3.49 2.76 3.04 3.49 2.76 3.16 4.06 Colorado recourse 3.44 2.72 3 3.49 2.76 3.12 4.07 Connecticut recourse 3.47 2.76 2.99 3.48 2.76 3.15 3.14 Delawere recourse 3.48 2.76 3.04 3.52 2.79 3.17 3.19 Florida recourse 3.57 2.74 2.99 3.6 2.77 3.16 4.06 Georgia recourse 3.44 2.73 3.07 3.48 2.76 3.15 3.9 Hawaii recourse 3.44 2.72 2.99 3.52 2.79 3.22 3.17 Idaho recourse 3.44 2.71 2.99 3.52 2.74 3.12 3.58 Illinois recourse 3.46 2.74 3.03 3.5 2.78 3.16 3.16 Indiana recourse 3.48 2.77 2.91 3.52 2.8 3.19 3.12 Iowa non-recourse 3.45 2.75 3.11 3.5 2.79 3.19 3.14 Kansas recourse 3.46 2.75 3.06 3.51 2.79 3.18 3.14 Kentucky recourse 3.47 2.76 3.13 3.47 2.76 3.09 3.2 Louisiana recourse 3.46 2.76 3.06 3.51 2.79 3.18 3.13 Maine recourse 3.46 2.74 3.06 3.5 2.77 3.17 3.17 Maryland recourse 3.47 2.74 2.95 3.49 2.76 3.13 3.58 Massachusetts recourse 3.45 2.72 2.86 3.5 2.76 3.16 3.18 Michigan recourse 3.47 2.75 3.05 3.5 2.78 3.16 3.85 Minnesota non-recourse 3.46 2.75 3.09 3.51 2.79 3.19 3.2 Mississippi recourse 3.46 2.76 3.08 3.5 2.79 3.18 3.14 Missouri recourse 3.46 2.75 3.29 3.51 2.79 3.21 3.18 Montana non-recourse 3.47 2.76 3.11 3.58 2.8 3.13 3.18 Nebraska recourse 3.94 2.94 2.94 4.1 3.07 3.39 4.04 Nevada recourse 3.45 2.74 3.08 3.51 2.78 3.13 3.19 New_Hampshire recourse 3.46 2.73 3.06 3.51 2.78 3.11 3.18 New_Jersey recourse 3.45 2.72 2.85 3.48 2.74 3.1 3.33 New_Mexico recourse 3.45 2.73 2.99 3.5 2.77 3.14 3.38 New_York recourse 3.5 2.76 2.85 3.55 2.8 3.14 3.61 North_Carolina 3.46 2.76 3.01 3.5 2.79 3.16 3.39 Tomasz Pruszkowski 64 State Recourse/ non-recourse state 30 year fixed 15 year fixed 5/1 ARM 30 yr fixed mtg refi 15 yr fixed mtg refi 7/1 ARM refi 15 yr jumbo fixed mtg refi North_Dacota non-recourse 3.46 2.76 3.1 3.52 2.81 3.13 3.2 Ohio recourse 3.46 2.75 3.04 3.5 2.78 3.18 3.22 Oklahoma recourse 3.45 2.75 3.08 3.5 2.79 3.2 3.15 Oregon non-recourse 3.44 2.71 3.06 3.49 2.75 3.3 4.37 Pennsylvania recourse 3.43 2.71 2.95 3.46 2.74 3.11 4.09 Rhodejsland recourse 3.46 2.74 3.03 3.51 2.78 3.1 3.15 South_Carolina recourse 3.45 2.75 3.02 3.5 2.78 3.16 3.15 South_Dacota recourse 3.47 2.76 3.14 3.53 2.81 3.13 3.18 Tennessee recourse 3.44 2.72 2.97 3.48 2.75 3.13 3.2 Texas recourse 3.46 2.74 2.97 3.49 2.76 3.14 3.92 Utah recourse 3.45 2.75 3.05 3.49 2.78 3.11 3.89 Vermont recourse 3.45 2.74 3.1 3.5 2.78 3.18 3.16 Virginia recourse 3.47 2.75 3.01 3.51 2.78 3.15 3.73 Washington non-recourse 3.48 2.75 3.02 3.48 2.76 3.12 3.63 West_Wirginia recourse 3.45 2.74 3.05 3.5 2.78 3.18 3.2 Wisconsin non-recourse 3.45 2.73 2.84 3.48 2.75 3.11 3.6 Wyoming recourse 3.47 2.76 3.1 3.54 2.81 3.13 3.31 S o u r c e : own elaboration based on classification determined by Ghent, A. C. and Kudlyak, M., inRecourse and Residential Mortgage Default: Evidence from US States. The Review of Financial Studies, 24 (9) and average interest rates values provided by Bankrate.com (availablae at: http://www.bankrate.com/national-mortgage-rates/, accessed: November 7, 2016). Results This statistical elaboration compares the means of each type of mortgage interest rate inrecourse and non-recourse states. An analysis of histograms revealed that the data under scrutiny did notfulfill the assumptions necessary toapply aparametric test. The assumption was that the data are notnormally distributed. Both skewness and kurtosis statistics corroborated this assumption. Based on histogram inspections, skewness, and kurtosis statistics corroborated by the Kolmogorov-Smirnov test presented below, we determined that assumptions for at-test are notmet and decided toproceed with anon-parametric equivalent.