## _wp14212 ## Source details **Canonical URL:** [_wp14212](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp14212.pdf) ## Other formats - [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp14212.pdf.md) - [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp14212.pdf.json) --- ### Research question and motivation - Investigates the impact of foreclosure laws on lending decisions by banks, focusing on the trade-off between borrower protection and credit access. - Main focus: judicial foreclosure laws (whether states require the foreclosure process be handled by the court) versus non-judicial (power-of-sale) alternatives. - Rationale: judicial procedures are more time consuming and require legal professionals, implying higher costs to creditors (see Clauretie and Herzog, 1990; Schill, 1991). - Foreclosure losses noted: foregone interest, attorney fees, court costs, property taxes, repairs, hazard insurance and other indirect costs. ### Empirical strategy and identification - Quasi-experimental design exploiting two sources of exogenous variation: - Variation in foreclosure laws across state borders. - Discontinuity in loans’ eligibility to guarantees from Government Sponsored Enterprises (GSEs) at the regulatory loan-limit cutoff (conforming vs. non-conforming/jumbo loans). - Rationale for GSE cutoff: GSE guarantees shift the financial burden of foreclosure costs onto the GSEs; guarantee fees are uniform across states. - Uses loan-level data from urban areas straddling state borders and distinguishes conforming (GSE-eligible) from non-conforming (jumbo) loans. - Controls include bank fixed effects to study bank lending decisions directly and address supply-side variation. - Introduces a continuous measure of foreclosure timelines obtained from the U.S. Foreclosure Network (USFN), estimating number of days to foreclose based solely on state laws, complementing the binary judicial/non-judicial indicator. ### Data and scope - Sample: loans from urban areas straddling state borders (cross-border urban sample). - Unit of analysis: loan-level lending decisions (probability to reject a loan application). - Data sources and measures: - HMDA loan-level data reported to the Federal Reserve. - RealtyTrac for judicial/non-judicial classification. - USFN timelines (USFN measure corresponds to year 2004; constructed purely from state regulations). - Freddie Mac processing timelines (correlation of 0.8 with USFN). - Sample restrictions and period: - Loan type: conventional (excluding FHA, VA, Farm Service Agency, Rural Housing Service). - Property type: single family; loan purpose: home purchase; occupancy: owner-occupied principal dwelling. - Geographic focus: metropolitan counties on state borders; final geographic sample: 52 urban areas straddling state borders (exploits changes across 49 state borders). - Time period: 2001-2006. - Bank activity filter: banks receiving more than 30 applications per year and county. - Sample size and coverage: - Around 1.3 million loan applications. - 699 banks. - 198 counties. - 42 states. - Exploits 49 state borders. ### Key findings (overview) - Judicial foreclosure laws are associated with a significant increase in the relative rejection rate on jumbo loans. - Impact on conforming loans is weak and overall not statistically significant. - Pattern consistent with foreclosure costs affecting lending when costs are borne by banks (jumbos) but not when costs are borne by GSEs (conforming loans). - Demand-side checks show the relative number, volume, and characteristics of jumbo loans compared to conforming loans do not correlate with judicial foreclosure laws across borders. - Results robust to: - Binary judicial/non-judicial measure; and - Continuous foreclosure time-frame measures from USFN. - Results subjected to robustness, falsification, and sensitivity tests (including Online Appendix). ### Methodological contributions - Uses continuous measures of foreclosure time frames (USFN days estimates) instead of relying solely on a binary indicator. - Leverages exogenous GSE conforming loan-limit cutoff to sharpen identification and address concerns about unobservable confounders. - Baseline regression (LPM) specification: - R_{i,k,u,s,t} = 1 if loan application i at bank k in urban area u, state s, time t was rejected. - Controls include bank (α_k), urban area (β_u), year (ρ_t) fixed effects, applicant and loan characteristics X_{i,k,u,s,t}, and regulatory variables Reg_s. - Main interaction: Reg_s × J where J is jumbo dummy: - R_{i,k,u,s,t} = α_k + β_u + ρ_t + θX_{i,k,u,s,t} + γReg_s + δJ + φReg_s J + ε_{i,k,u,s,t} - LPM preferred for √N consistency with N→∞ and fixed T. ### State foreclosure laws and measurement - Foreclosure process: - Judicial foreclosure: lender initiates foreclosure in court; generally slower and imposes costs on lenders. - Non-judicial foreclosure: liquidation via power-of-sale handled by a trustee; relatively quicker. - Classification: - Following Mian et al. (2011), states classified as judicial/non-judicial using RealtyTrac. - 20 states identified as requiring a judicial process. - Judicial foreclosure states concentrated in the eastern United States. - Statutory variations: - Recourse/deficiency judgments: 8 states allow deficiency judgments. - Right of redemption: 9 states give borrowers a statutory right of redemption (typically between three months and a year). - Historical exogeneity: many state mortgage-law differences date from the nineteenth century; recourse laws influenced by farm foreclosure experience in the early 1930s. - Improved measurement: - USFN estimates of days to foreclose (constructed purely from regulations) and Freddie Mac timelines used; USFN rescaled to [0,1] for some analyses. ### Empirical results: interaction of jumbo status with state foreclosure laws (section III(C)) - Main interaction findings: - Jumbo × judicial foreclosure: jumbo loan rejection rate in judicial states is higher by 2.3 percentage points (statistically significant at the 1% level). - Jumbo × deficiency prohibition: positive and weakly significant, suggesting an additional increase in rejection rate on jumbo loans by around 3 percentage points in states with no recourse. - Continuous timeline measures: - Regulatory processing timelines do not have a significant effect on rejection rate in the full sample. - Interaction with jumbo dummy: - USFN measure: relative increase in rejection rate on jumbo loans varies by around 3.4 percentage points between a state with zero delays (Texas scores 0.06 on USFN rescaled measure) and New York (USFN rescaled measure = 1). - Freddie Mac timelines show similar results. - Interpretation: judicial foreclosure matters for jumbos but is almost irrelevant for conforming loans due to GSE guarantees. ### Band samples and falsification (C.2) - Band regressions within 20%, 10%, and 5% of the jumbo cutoff; control for county fixed effects. - Descriptive: Figure 5 (20% band) shows in non-judicial states conforming and jumbo rejection rates are very close; in judicial states the difference is larger by 2.5 percentage points for the median county (mean rejection rate higher by around 1 percentage point). - Regression evidence: - Jumbo × judicial coefficient consistently positive and significant across band samples; magnitude smaller in 20% and 10% bands, larger in 5% band. - Interactions of jumbo with redemption or deficiency prohibition not significant. - Falsification: imaginary jumbo cutoffs at 80% and 120% of actual cutoff produce interaction coefficients far from significant in all sub-samples. ### Demand around the jumbo cutoff (C.3) - Concern: effects might be demand-driven rather than supply-driven. - Approach: within 20% band compute ratios of number and dollar volume of jumbo to conforming applications; ratio of average applicant income; ratio of average census tract income. - Regression (3): J_NJ_{c,u,s,t} = α_u + ρ_t + γ Reg_s + ε_{c,u,s,t} - Results (Table 6): - Columns (1)-(4) judicial dummy; (5)-(8) USFN timelines; (9)-(12) Freddie Mac timelines. - All coefficients on foreclosure laws are far from significant. - Interpretation: no significant correlation between foreclosure requirements and demand around the jumbo cutoff; supports supply-response interpretation. ### Economic magnitude (C.4) - Empirical magnitudes: - Judicial foreclosure associated with around a 2 percentage point relative increase in the rejection rate on jumbo loans. - Processing delays associated with around a 3 percentage point increase in rejection rate going from minimum delays (Texas) to very long delays (New York). - Back-of-the-envelope aggregate illustration (mainland U.S., 2001–2006): - Around 300 thousand applications to jumbo loans in judicial states between 2001 and 2006, with around $150 billion in 2006 U.S. dollars. - A 2 percentage point increase in the rejection rate corresponds to around 6 thousand loans with a value of around $3 billion. - Applying the 2 percentage point rejection rate to total loans (conforming and jumbo) during that period yields around 60 thousand loans with a total volume of around $10 billion. - Implied difference (illustrative impact of GSE cross-subsidization) is $ 7 billion over that period (computed as the difference between the two scenarios). - Policy note: FHFA planned an increase in guarantee fees on states with the longest processing delays (all requiring judicial foreclosure) in December 2013; the decision was later put on hold for further study following the appointment of a new director. ### Bank heterogeneity (C.5) - Bank characteristics examined: size, geographic diversification, reliance on private securitization. - Correlations among characteristics: Size and diversification ρ = 0.6; diversification and private securitization ρ = 0.15. - Measures: - Size = log of total assets. - Diversification = 1 - sum of squared shares of a bank’s loans by county (ranges from 0 to near 1). - Private securitization = share of loans sold to non-GSE private buyers (from HMDA loan-sale entry). - Dummy variables mark banks in the top third for each characteristic. - Regression (4) on jumbo loan subsample with bank and county fixed effects: - USFN timelines * Securitization: positive and significant; one standard deviation (0.26) increase in securitization associated with around a 4 percentage point increase in the rejection rate moving from 0 delays to maximum delays. - Freddie Mac timelines * Securitization: positive and significant. - Dummy-variable specifications: top-tier Diversification and Securitization interactions with timelines show positive and significant coefficients. - Interpretation: Banks relying more on private securitization and banks in top tiers of diversification/securitization are more likely to curtail lending in areas with long foreclosure delays. ### Robustness, falsification, and sensitivity checks - Robustness checks include bandwidth/sample choices around the jumbo cutoff, falsification tests with imaginary cutoffs, and tests for relative demand. - Results robust to alternative measures (binary judicial dummy, USFN continuous timelines, Freddie Mac timelines). - Bank fixed effects and cross-border design address supply-side identification concerns. - Standard errors clustered at urban area and bank levels. ### Interpretation and policy implications - Borrower-friendly foreclosure laws, particularly those producing long judicial foreclosure delays, materially reduce credit supply for loans where foreclosure costs are retained by banks (jumbos). - Uniform GSE guarantees shielded the conforming market from these adverse effects; ignoring GSE cross-subsidization would understate borrower-friendly laws’ impact on credit supply. - Heterogeneity across banks and across borders within urban areas is important for understanding how legal environments translate into credit supply outcomes. - Policy relevance: findings inform cost-benefit assessments of borrower-protection regulations and highlight how GSE cross-subsidization mitigated overall impacts on credit supply. *Source: _wp14212 - 1. Summary Statistics; Section II; section III(C), C.2–C.5; Conclusion.* ### 1. Summary Statistics ............................................................................................ ### _wp14212 - 1. Summary Statistics ............................................................................................ ### Research question and motivation - Investigates the impact of foreclosure laws on lending decisions by banks, focusing on the trade-off between borrower protection and credit access. - Main focus: judicial foreclosure laws (whether states require the foreclosure process be handled by the court) versus non-judicial (power-of-sale) alternatives. - Rationale: judicial procedures are more time consuming and require legal professionals, implying higher costs to creditors (see Clauretie and Herzog, 1990; Schill, 1991). - Foreclosure losses noted in the source include: foregone interest, attorney fees, court costs, property taxes, repairs, hazard insurance and other indirect costs. ### Empirical strategy and identification - Uses a quasi-experimental design exploiting two sources of exogenous variation: - Variation in foreclosure laws across state borders. - Discontinuity in loans’ eligibility to guarantees from Government Sponsored Enterprises (GSEs) at the regulatory loan-limit cutoff (conforming vs. non-conforming/jumbo loans). - Rationale for exploiting the GSE cutoff: GSE guarantees shift the financial burden of foreclosure costs onto the GSEs, and guarantee fees are uniform across states. - Uses loan-level data from urban areas straddling state borders and distinguishes conforming (GSE-eligible) from non-conforming (jumbo) loans. - Controls include bank fixed effects to study bank lending decisions directly and address supply-side variation. - Introduces a continuous measure of foreclosure timelines obtained from the U.S. Foreclosure Network (USFN), which estimates the number of days it takes to foreclose on a property solely based on state laws, complementing the standard binary judicial/non-judicial indicator. ### Data and scope - Sample: loans from urban areas straddling state borders (cross-border urban sample). - Unit of analysis: loan-level lending decisions (probability to reject a loan application). - Additional design features: the paper examines heterogeneity in the banking sector across borders even within the same urban area. ### Key findings - Judicial foreclosure laws are associated with a significant increase in the relative rejection rate on jumbo loans. - The impact of judicial foreclosure laws on rejection rates for conforming loans is weak and overall not statistically significant. - The observed pattern is consistent with the hypothesis that foreclosure costs matter for banks’ lending when those costs are borne by the banks (as with jumbo loans), but not when costs are borne by GSEs (as with conforming loans). - Demand-side checks: - The relative number, volume, and characteristics of jumbo loans compared to conforming loans do not correlate with judicial foreclosure laws across borders, supporting a supply-side interpretation. - Results are robust to using either: - The standard binary judicial/non-judicial measure used in the literature; or - Continuous foreclosure time-frame measures from authoritative sources (USFN). - Results are subjected to a battery of robustness and falsification tests (including an Online Appendix) that strengthen the findings. ### Methodological contributions - Uses more direct and continuous measures of foreclosure time frames (USFN days estimates) rather than relying solely on a binary judicial/non-judicial indicator. - Leverages the exogenous GSE conforming loan-limit cutoff to sharpen identification and address concerns about unobservable confounders. ### Robustness, falsification, and sensitivity checks (overview) - Performs robustness checks including bandwidth/sample choices around the jumbo cutoff. - Conducts falsification tests. - Tests for relative demand to ensure observed cross-border differences are not driven by demand-side forces. - Examines bank characteristics and how they affect sensitivity to foreclosure laws (details presented in the full paper and Online Appendix). ### Interpretation and implications - Evidence indicates foreclosure laws materially affect credit supply for loans where foreclosure costs are retained by banks (jumbos), but not for conforming loans whose foreclosure costs are borne by GSEs. - Heterogeneity across banks and across borders within urban areas matters for understanding how legal environments translate into credit supply outcomes. *Source: _wp14212 - 1. Summary Statistics ............................................................................................* ### Section II) related to deficiency judgments and right of redemption, but we do not highlight them due to ### _wp14212 - Section II) related to deficiency judgments and right of redemption, but we do not highlight them due to ### Key findings and contributions - Banks that rely more on private securitization (as opposed to balance sheet lending) are "significantly most responsive" to variation in foreclosure timelines. - Some evidence that credit supply by banks that are more geographically diversified is more negatively affected by longer foreclosure timelines. - The paper isolates supply-side factors and exploits an exogenous jumbo cutoff to distinguish loans’ eligibility to GSE guarantees. - The paper uses new data that provide a more direct measure of costs associated with judicial foreclosures. - On the average loan, foreclosure laws do not have a significant impact on banks’ decision (benchmark regressions show judicial and other foreclosure-law dummies are not significant for the average loan). ### State foreclosure laws (analysis of legal variation) - Foreclosure process summary: - Judicial foreclosure: lender initiates foreclosure in court; court issues decree, stipulates notices, and oversees procedure; generally slower and imposes costs on lenders. - Non-judicial foreclosure: liquidation via power-of-sale handled by a trustee; relatively quicker. - Classification and heterogeneity: - Following Mian et al. (2011), states are classified as judicial or non-judicial using RealtyTrac. - 20 states are identified as requiring a judicial process (shaded in gray in the referenced Figure 1). - Judicial foreclosure states are concentrated in the eastern United States. - Substantial variation exists in foreclosure processing timelines within judicial and non-judicial groups. - Additional statutory variations: - Recourse/deficiency judgments: 8 states allow lenders to pursue a deficiency judgment against the borrower’s other assets (i.e., allow recourse). - Right of redemption: 9 states give borrowers a statutory right of redemption, typically ranging between three months and a year. - Historical exogeneity: - Ghent (2013) finds that, with the exception of anti-deficiency statutes, most differences in state mortgage law date from the nineteenth century and many from before the US civil War; judicial vs. non-judicial distinctions can be traced to idiosyncratic factors and individual judges. - Recourse/deficiency judgment laws can be partly traced to states’ experience with farm foreclosure in the early 1930s. - Improved measurement: - The paper exploits data from the U.S. Foreclosure Network (USFN) that estimate the number of days to foreclose based solely on regulations (abstracting from operational delays). - An index of timelines (scaled by the longest timelines) shows strong correlation between judicial foreclosure and USFN timelines, but also substantial within-group variation. - Freddie Mac processing timelines are used as a supplementary measure (correlation of 0.8 with USFN). ### Empirical strategy (identification and methodology) - Objective: determine whether borrower-friendly foreclosure laws affect bank lending decisions at origination, focusing on supply-side effects while controlling for demand and bank heterogeneity. - Main approaches and controls: - Use individual loan-level HMDA data to control for borrower characteristics, regional fixed effects, and bank fixed effects. - Restrict sample to urban areas straddling state borders to minimize unobservable regional differences. - Exploit exogenous discontinuity in GSE coverage at the jumbo cutoff to compare jumbo and conforming loans. - Baseline regression (linear probability model, LPM): - Dependent variable R_{i,k,u,s,t} = 1 if loan application i at bank k in urban area u, state s, time t was rejected. - Controls: bank (α_k), urban area (β_u), year (ρ_t) fixed effects; applicant and loan characteristics X_{i,k,u,s,t} (income, gender, race, census-tract income, loan-to-income ratio). - Regulatory variables Reg_s include: (i) judicial foreclosure requirement dummy, (ii) statutory right of redemption dummy, (iii) prohibition of deficiency judgment dummy. Judicial dummy also replaced by continuous foreclosure timelines. - LPM preferred because with N→∞ and T fixed, LPM estimates are √N consistent whereas Probit/Logit can be inconsistent in this setting. - Exploiting GSE segmentation: - GSEs (Fannie Mae and Freddie Mac) provide guarantees that compensate creditors and servicers for foreclosure costs; guarantee fees do not vary across states and eligibility is transparent and non-discriminatory. - GSE market share context: - By 2003, around 45 percent of all outstanding mortgages had GSE guarantees. - The GSE eligible market is roughly around 75% of the overall market; the share of the GSEs in their market was closer to 60 percent. - The GSE share declined to some extent during 2004-06 when the private label market boomed. - Loans above the legal GSE cutoff are "jumbo" and are less liquid and often held by banks or sold to private securitization pools. - Key regression includes interaction of foreclosure-law variables with a jumbo dummy J to identify differential effects on jumbo loans: - R_{i,k,u,s,t} = α_k + β_u + ρ_t + θX_{i,k,u,s,t} + γReg_s + δJ + φReg_s J + ε_{i,k,u,s,t} - Identification strengthened by estimating the model in the full sample and in a subsample of applications within a small interval around the jumbo cutoff, controlling for similar loan sizes and testing for demand-side confounding. ### Data and sample construction - Data sources: - HMDA (Home Mortgage Disclosure Act) loan-level data reported to the Federal Reserve. - RealtyTrac for judicial/non-judicial classification and USFN timelines (USFN measure corresponds to year 2004). - USFN timeline data are constructed purely from state regulations. - Freddie Mac provides servicer processing timelines (permissible processing delays remain constant between 2001 and 2006). - Sample restrictions and period: - Loan type: conventional (excluding FHA, VA, Farm Service Agency, Rural Housing Service). - Property type: single family; loan purpose: home purchase (excluding refinancing or home improvement); occupancy: owner-occupied principal dwelling. - Geographic focus: metropolitan counties lying on state borders, defined as contiguous counties that belong to an MSA. - Final geographic sample: 52 urban areas straddling state borders (sample exploits changes in regulations across 49 state borders). - Time period: 2001-2006 (exclude years before 2001 due to lack of data on foreclosure processing days; do not extend beyond 2006 to avoid crisis-specific factors). - Bank activity filter: from each year and county include banks that received more than 30 applications. - Sample size and coverage: - Around 1.3 million loan applications. - 699 banks. - 198 counties. - 42 states. - Analysis exploits 49 state borders. - Summary statistics and diagnostics: - Table 1 (referenced) shows: - The jumbo market makes up around 10% of total loans, with this share higher at the peak of the boom. - The average rejection rate in the sample decreases over most of the boom period. - Loan-to-income (LTI) increased during the boom period. - Cross-border balance checks: - Panel A of Table 2 compares county characteristics (income, employment, minority, construction, mortgage delinquency) between judicial and non-judicial subsamples; mean t-tests show no significant differences. - Panel B of Table 2 regresses these characteristics on urban area fixed effects and foreclosure measures (judicial dummy, USFN, Freddie Mac); no significant relation between foreclosure duration and county characteristics is found. - Banking landscape heterogeneity: - In 2005, considering banks operating in county-pairs: - There are 801 banks operating in 157 county pairs in the extended sample used for this exercise. - Among bank and county-pair combinations where the bank operates in at least one county, only about 40 percent of cases show the bank operating in both counties. - Market shares vary widely across borders; numerous banks have market shares above 40 percent in one county but modest presence in neighboring counties. - Conclusion: substantial heterogeneity in the banking landscape across state borders motivates use of bank fixed effects. ### Empirical results (overview) - Estimation details: - Regressions include urban area, bank, and time fixed effects. - Standard errors clustered at both the urban area and bank levels. - Benchmark regression findings (Table 3 referenced): - Applicant and loan characteristic effects: higher rejection probability for applicants in lower-income census tracts, with lower income, and with higher loan-to-income ratios. - Jumbo dummy coefficient is positive. - Judicial foreclosure dummy coefficient is small and far from significant. - Coefficients on statutory right of redemption and prohibition of deficiency judgment are also not significant for the average loan. - Interpretation: - On average, foreclosure laws do not appear to significantly affect banks’ lending decisions at origination for the average loan. - The paper’s identification strategy (jumbo interaction) is designed to detect effects concentrated in segments not shielded by uniform GSE guarantees; results for φ (Reg_s × J) are the key test for whether borrower-friendly laws discourage lending in the jumbo segment (results discussion and detailed estimates are provided in subsequent tables referenced in the paper). *Source: _wp14212 - Section II) related to deficiency judgments and right of redemption, but we do not highlight them due to (PDF).* ### section III(C), the absence of a significant effect of state laws could be simply a result of ### _wp14212 - section III(C), the absence of a significant effect of state laws could be simply a result of ### Interaction of jumbo status with state foreclosure laws and GSE guarantees - Estimated regression (2) interacts the jumbo dummy with state laws to account for GSE guarantees shielding conforming loans. - Findings: - Positive and strongly significant coefficient on jumbo dummy * judicial foreclosure: jumbo loan rejection rate in judicial states is higher by 2.3 percentage points. This difference is statistically significant at the 1% level. - Interaction between jumbo dummy and deficiency prohibition dummy is positive and weakly significant, suggesting an additional increase in the rejection rate on jumbo loans by around 3 percentage points in states with no recourse. - Continuous measures of foreclosure timelines (USFN rescaled to [0,1] and Freddie Mac timelines): - Regulatory processing timelines do not have a significant effect on the rejection rate in the full sample. - Interaction with the jumbo dummy shows significance for jumbo loans: - USFN measure: relative increase in rejection rate on jumbo loans varies by around 3.4 percentage points between a state with zero delays (Texas scores 0.06 on the USFN rescaled measure) and New York (USFN rescaled measure = 1). - Freddie Mac timelines show similar results. - Interpretation: - Whether states required judicial foreclosure was almost irrelevant to the supply of conforming loans but meaningful for jumbo loans. - The increase in the rejection rate above the jumbo cutoff (by around 3 percentage points) was almost twice as large in states with long judicial foreclosure timelines. ### C.2 Band samples and falsification - Strategy: - Repeat regression (2) on subsamples of loans within 20%, 10%, and 5% of the jumbo cutoff to focus on loans more similar in size and borrower income. - Control for county fixed effects in band regressions. - Descriptive illustration: - Figure 5 (20% band): in non-judicial states the rejection rates for conforming and jumbo loans are very close; in judicial states the difference is larger by 2.5 percentage points for the median county (mean rejection rate also higher by around 1 percentage point). - Regression results (Table 4): - Consistently positive and significant coefficient on jumbo dummy * judicial requirement across band samples. - Magnitude consistent with Table 3; coefficients smaller in 20% and 10% bands but larger in the 5% band. - Interactions between jumbo dummy and redemption or deficiency prohibition dummies are not significant. - Falsification test (Table 5): - Imaginary jumbo cutoffs at 80% and 120% of the actual cutoff produce interaction coefficients far from significant in all sub-samples. - Results support hypothesis that the impact of judicial requirement is more pronounced for jumbo loans due to the discontinuation of GSE guarantees at the jumbo cutoff rather than other confounding factors. ### C.3 Demand around the jumbo cutoff - Empirical concern addressed: whether observed effects are demand-driven (changes in number/size of applications) rather than supply-driven. - Approach: - Use HMDA to observe total number of applications and compute ratios within a 20% band: number and dollar volume of jumbo to conforming applications; ratio of average applicant income; ratio of average census tract income (HMDA reports median census tract income). - Estimate regression (3): J_NJ_{c,u,s,t} = α_u + ρ_t + γ Reg_s + ε_{c,u,s,t} with urban and time fixed effects and state regulations (Reg_s). - Results (Table 6): - Columns (1)-(4) using judicial dummy, columns (5)-(8) USFN timelines, columns (9)-(12) Freddie Mac timelines. - All coefficients on foreclosure laws, including the judicial foreclosure dummy, are far from significant levels across specifications. - Interpretation: - No significant correlation between foreclosure requirements and quantitative or qualitative differences in demand around the jumbo cutoff. - Evidence supports the identifying assumption that observed effects reflect supply responses to foreclosure costs. ### C.4 Economic magnitude - Summary of empirical magnitudes: - Judicial foreclosure associated with around a two percentage point relative increase in the rejection rate on jumbo loans. - Processing delays associated with around a 3 percentage point increase in the rejection rate going from a state with minimum delays like Texas to a state with very long delays like New York. - Back-of-the-envelope aggregate illustration (mainland U.S., 2001–2006): - Around 300 thousand applications to jumbo loans in judicial states between 2001 and 2006, with around $150 billion in 2006 U.S. dollars. - A 2 percentage point increase in the rejection rate corresponds to around 6 thousand loans with a value of around $3 billion. - Counterfactual applying the 2 percentage point rejection rate to the total number of loans (conforming and jumbo) during that period yields around 60 thousand loans with a total volume of around $10 billion. - Implied difference (illustrative impact of GSE cross-subsidization) is $ 7 billion over that period (computed as the difference between the two scenarios). - Policy note reported in the source: - FHFA planned an increase in guarantee fees on a number of states with the longest processing delays (all requiring judicial foreclosure) in December 2013; the decision was later put on hold for further study following the appointment of a new director. ### C.5 Bank heterogeneity - Motivation: - Examine whether bank size, geographic diversification, and reliance on private securitization affect sensitivity of lending to foreclosure laws. - Correlations among characteristics: Size and diversification ρ = 0.6; diversification and private securitization ρ = 0.15. - Measures: - Size = log of total assets. - Diversification = 1 - sum of squared shares of a bank’s loans by county (ranges from 0 to near 1). - Private securitization = share of loans sold to non-GSE private buyers (computed from HMDA loan-sale entry). - Create dummy variables marking banks in the top third of the distribution for each characteristic. - Regression (4) on jumbo loan subsample with bank and county fixed effects and HMDA controls. - Results (Table 7): - Continuous-variable specifications: - USFN timelines * Securitization: positive and significant; one standard deviation (0.26) increase in securitization associated with around a 4 percentage point increase in the rejection rate when moving from a state with 0 delays to a state with maximum delays. - USFN timelines * Diversification: coefficient negative in continuous specification but not robust across measures. - Freddie Mac timelines * Securitization: positive and significant. - Dummy-variable specifications (top third): - Interactions of bank characteristics with the judicial dummy are not significant. - Interactions of timelines with top-tier Diversification and Securitization dummies show positive and significant coefficients. - Interpretation: - Banks that rely more on private securitization and banks in the top tier of diversification and securitization are much more likely to curtail lending in areas with long foreclosure delays. ### Conclusion (summary from section IV) - Main results: - Borrower-friendly foreclosure laws, particularly those producing long judicial foreclosure delays, have a significant and meaningful negative impact on the supply of mortgage credit, primarily in the jumbo segment. - The uniform GSE guarantees shielded the conforming market from these adverse effects, and overlooking GSEs would underestimate borrower-friendly laws’ impact on credit supply. - Policy relevance: - Findings inform cost-benefit assessments of borrower-protection regulations and highlight how GSE cross-subsidization mitigated overall impacts on credit supply; reconsideration of such cross-subsidization may change credit outcomes. *Source: _wp14212 - section III(C), the absence of a significant effect of state laws could be simply a result of* ### References ### References ### Books and monographs - Acharya, V. V.,Richardson, M.,Van Nieuwerburgh, S.andWhite, L. J.(2011). Guaran- teed to fail: Fannie mae, freddie mac, and the debacle of mortgage finance.Princeton University Press. - Wooldridge, J. M.(2002).Econometric Analysis of Cross Section and Panel Data. MIT Press. ### Journal articles - Berkowitz, J.andWhite, M.(2004). Bankruptcy and small firms’ access to credit.RAND Journal of Economics,35(1), 69–84. - Clauretie, T. M.andHerzog, T.(1990). The effect of state foreclosure laws on loan losses: Ev- idence from the mortgage insurance industry.Journal of Monetary, Credit and Banking,22(2), 221–233. - Evans, D. S.andWright, J. D.(2009). Effect of the consumer financial protection agency act of 2009 on consumer credit.Loyola Consumer Law Review,22(3), 277–335. - Gerardi, K.,Lambie-Hanson, L.andWillen, P. S.(2013). Do borrower rights improve borrower outcomes? evidence from the foreclosure process.Journal of Urban Economics,73(1), 1–17. - Ghent, A. C.andKudlyak, M.(2011). Recourse and residential mortgage default: Evidence from U.S. states.Review of Financial Studies,27(6), 3139–3186. - Goodman, J.andLevitin, A.(2014). Bankruptcy law and the cost of credit: The impact of cramdown on mortgage interest rates.Journal of Law and Economics,57(1), 139–158. - Gropp, R.,Scholz, J.andWhite, M.(1997). Personal bankruptcy and credit supply and demand.The Quarterly Journal of Economics,112(1), 217–251. - Huang, R.(2008). Evaluating the real effect of bank branching deregulation: Comparing contigu- ous counties across U.S. state border.Journal of Financial Economics,87(3), 678–705. - Keys, B.,Mukherjee, T.,Seru, A.andVig, V.(2010). Did securitization lead to lax screening? Evidence from subprime loans.Quarterly Journal of Economics,125(1). - Loutskina, E.andStrahan, P. E.(2009). Securitization and the declining impact of bank finance on loan supply: Evidence from mortgage originations.Journal of Finance,64(2), 861– 889. - —and—(2011). Informed and uninformed investment in housing: The downside of diversification. Review of Financial Studies,24(5), 1447–1480. - Mian, A.,Sufi, A.andTrebbi, F.(2011). Foreclosures, house prices, and the real economy. NBER Working paper No. 16685. - Pence, K. M.(2006). Foreclosing on opportunity: State laws and mortgage credit.Review of Economics and Statistics,88(1), 177–182. - Schill, M. H.(1991). An economic analysis of mortgagor protection laws.Virginia Law Review, 77(3), 489–538. - Zhu, S.andPace, R. K.(2011). The influence of foreclosure delays on borrowers default behavior. 46th Annual AREUEA Conference Paper. ### Working papers, discussion papers, and conference papers - Calem, P.,Jagtiani, J.andLang, W. W.(2014). Foreclosure delay and consumer credit performance.Federal Reserve Bank of Philadelphia Working Paper No. 14-8. - Favara, G.andImbs, J. M.(2013). Credit supply and the price of housing.CEPR Discussion Paper No. DP8129. - Ghent, A.(2013). America’s mortgage laws in historical perspective.Arizona State University working paper. ### Reports, testimony, and committee publications - Bipartisan Policy Center(2013). The consumer financial protection bureau: Measuring the progress of a new agency.Financial Regulatory Reform Initiative. - Kupiec, P. H.(2014). Government financial policy and credit availability.Statement before the Subcommittee on Monetary Policy and Trade. - U.S. House of Representatives(2012). The consumer financial protection bureau’s threat to credit access in the united states.Committee on Oversight and Government Reform. *Source: _wp14212 - References* --- _Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp14212.pdf_