What Is The Micro-elasticity Of Mortgage Demand To Interest Rates?

What is the micro-elasticity of mortgage demand tointerest rates?Stephanie Lo1December 2, 20161Part of this work has been performed at the Federal Reserve Bank of Boston. Theviews expressed in this presentation are solely those of the author and not necessarilythose of the Federal Reserve Bank of Boston or the Federal Reserve System.Stephanie LoElasticity of mortgage demandDecember 2, 20161 / 25

OverviewWhat is the elasticity of mortgage demand to interest rates?Hard to measure: interest rate movements and house demand bothtied to macroeconomyImportant for understanding magnitude of monetary policy transmissionI formulate a regression discontinuity (RD) design to measure theimpact of interest rates on intensive and extensive margins ofmortgage demandI use a novel identification method, which features regulatory-inducedbreaks in mortgage rates across certain credit score thresholdsRare administrative dataset on daily o ered mortgage rates acrossborrower characteristicsValidity of RD shown using proprietary detailed mortgage data linkedon the individual-level with credit bureau dataStephanie LoElasticity of mortgage demandDecember 2, 20162 / 25

Economic Motivation: Inequality in New MortgagesIncreased mortgage share to high-FICO and high-income individualsLeft: Post-crisis, low FICO individuals saw decreasing mortgage shareRight: Red line indicates that the median income index for purchasemortgages has risen drasticallyResearch question: how much of this is due to increases in mortgagespreads?Stephanie LoElasticity of mortgage demandDecember 2, 20163 / 25

Loan Level Price Adjustments and Policy DebateLLPAs are mortgage fees that vary by FICO at originationLLPAs were instituted in 2008 to help Fannie and Freddie “manageour credit risks, mitigate losses, and ensure an adequate capitalposition”LLPAs have come under fire:“[Mortgage fees are] a potentially debilitating one-two punch for manyborrowers and it raises serious questions as to the unintendedconsequences of these ambitious fee hikes . The only known is thatvirtually all borrowers will soon face higher costs and rates, and afragile housing recovery will deal with yet another majorchallenge.“ – Mortgage News Daily, December 2013“”Eight years after the financial crisis, mortgage credit quality hasimproved dramatically and regulations have improved the industry’srisk management practices. We believe these changes justifyeliminating LLPAs“ – Letter to FHFA director in 2016, co-signed by25 groupsPolicy implications: Did LLPAs pose a risk to the fragile housingrecovery?Stephanie LoElasticity of mortgage demandDecember 2, 20164 / 25

LiteratureElasticity of “housing” to interest ratesGlaeser Shapiro (2003): housing demand to interest rates on state-levelhome mortgage interest deduction; no significant responseGlaeser Gottlieb Gyourko (2012): house prices are less responsive tointerest rates than standard model would predictFuster Zafar (2014): survey to measure sensitivity of housing demandto interest rates; low sensitivityElasticity of mortgage demandBest et al (2015): remortgagors bunch right below LTV “notches” ininterest rate scheduleDeFusco Paciorek (2014): jumbo-conforming bunchingFind relatively small demand response (100 bp decr. in rates ! 2-3prct incr. in demand)My measure has benefit of being more precise and allows for measureof extensive margin of demandMortgage credit supply during the crisisGoodman Li (2014); Anenberg et al (2015)Stephanie LoElasticity of mortgage demandDecember 2, 20165 / 25

Preview of resultsI find the elasticity of mortgage demand to interest rates is largeA 25 bp decrease in mortgage rates is associated with:Intensive: an increase in loan size of 15k (approx. 10% of averageorigination volume)Extensive: a 50% increase in likelihood of potential borrower todemand a loanI establish that these estimates are driven by demand-, not supply-,side factorsI estimate that, had mortgage rates been the same across FICO680-719 borrowers as FICO 720 borrowers, there would have beenmuch more mortgage borrowing (about 15B more than actual 43.5B)Stephanie LoElasticity of mortgage demandDecember 2, 20166 / 25

Plan1Motivation2Brief Methodology3Results4RD Validity5Economic Implications6ConclusionStephanie LoElasticity of mortgage demandDecember 2, 20167 / 25

Credit score cuto s l Amount680700Stephanie 0660LTV Ratio250200appraisal amount (k)50403020106408000760300Orig. Propensity600020001406604000number of mortgages200180160origination amount (k)6.565.5mortgage rate5640mortgage propensityNo. of Originations10000Orig. Amount220Mortgage Rates640660680700720740Elasticity of mortgage demand760640660680700720December 2, 20168 / 25

LLPAs drive mortgage rate di erentials across borrowersIn December 2007, GSEs imposedupfront-fees called LLPAs01LLPA23Can think of as additional insurancepremiaVary by discrete credit score bins (i.e.680-699, 700-719, etc.)600650FICO score700750While technically paid by banks atclosing, I show evidence that these arefully passed on to consumersServe as an exogenous wedge in interestrate across borrowers (even thoughlenders see same rate)Stephanie LoElasticity of mortgage demandDecember 2, 20169 / 25

Mortgage rate di erences across timeLenders price the upfront LLPAthey pay into mortgage rateMagnitude of rate impactdepends on MBS spreads,prepayment forecasts, etc.Mortgage rates di erences arefairly small on average (27 bpfor FICO 680-700, 90thpercentile 47 bp)Stephanie LoElasticity of mortgage demandDecember 2, 201610 / 25

Example of RDMortgage rates discretely jump up at 720 but are constant within-binfor 700-719 and 720-739Driven by LLPAsMortgage propensity also has break in trend at this cuto I define elasticity as this jump, @N, over the change in rate, @rStephanie LoElasticity of mortgage demandDecember 2, 201611 / 25

Plan1Motivation2Brief Methodology3Results4RD Validity5Economic Implications6ConclusionStephanie LoElasticity of mortgage demandDecember 2, 201612 / 25

Base RD resultsOrigination amountBase origination amountAppraisal amountBase appraisal amountLoan-to-value ratioBase loan-to-value ratioMortgage propensityBase mortgage ***[19.63,38.88]28.3295% confidence intervals in brackets p 0.10, p 0.05, p 0.01Interpretation is change across threshold per 100bp change in interestratesFICO 700 likely most reliable estimateStephanie LoElasticity of mortgage demandDecember 2, 201613 / 25

Sanity check / back-of-envelope on payment implicationsSuppose borrower starts with 5% 30-yr FRM, 200k originationamountMonthly payment is 1073Suppose interest rates fall to 4%Holding monthly payment constant:Can now have origination amount of 226kHolding origination amount constant:Monthly payment falls to 954The magnitude of my estimates implies borrowers spend more thanjust the constant-payment strategy would imply 52- 80k incr in origination amount, vs. constant-payment-implied 26ksubstitution toward debtStephanie LoElasticity of mortgage demandDecember 2, 201614 / 25

Intensive elasticity estimates, by size of rate spreadPer basis point of rate spreadFICO 700.4.4lwald.1lwald20bp-40bp5bp-20bp40bp-60bpFICO p40bp-60bpAggregate e ectFICO 700FICO .2.3.3.3.4FICO 720.5.5.5FICO p20bp-40bpDeclining marginal responsiveness per basis point rate spreadStephanie LoElasticity of mortgage demandDecember 2, 201615 / 25

Plan1Motivation2Brief Methodology3Results4RD Validity5Economic Implications6ConclusionStephanie LoElasticity of mortgage demandDecember 2, 201616 / 25

Supply side does not seem to be driving the resultsforeclosure rates.01.0200default rates.02.04.03Foreclosure by FICO.06Default by FICO650700750FICO at origination800650700750FICO at origination800One might worry that lenders also use rule-of-thumb for rejectingmortgages, so that credit supply is looser above threshold, thereforeinducing more and larger mortgages above cuto I argue this is not a concern for high FICO scores:There is no incentive for this ! securitization rates do not varydiscretely across cuto If this is the case, you would expect higher default rates right abovethe cuto Stephanie LoElasticity of mortgage demandDecember 2, 201617 / 25

Securitization and default 9,0.000329]For FICO 700 and 720, 95% CIs of RD discontinuity in the probabilityof securitization and probability of default include 0Base securitization rate (within first 36 months) is around 50% –discontinuity magnitude is smallBase default rate is about 3%, and discontinuity statisticallyindistinguishable from 0Stephanie LoElasticity of mortgage demandDecember 2, 201618 / 25

FICO scores are virtually impossible to manipulate.01Density.02.03.04Credit scores often move /-20points over the course of 6monthsWhile some individuals may tryto manipulate FICO scores,0credit scoring is a “black box“there are a lot of moving parts600650700750FICO score, 6 mo. previousOrigination FICO 699Stephanie Lo800Origination FICO 700Imperfect control implies RDshould be valid a la Lee (2008)Elasticity of mortgage demandDecember 2, 201619 / 25

Borrowers just across FICO breakpoints are identicalBankcard balance, currentBase amount, bankcard balanceCar debt, conditional on having car debtBase amount, car debtCredit utilizationBase amount, credit 7[-0.122,0.437]0.5840.106[-0.0987,0.311]0.55195% confidence intervals in brackets p 0.10, p 0.05, p 0.01FICO cuto s for credit variables, such as bankcard debt, car debt, andcredit utilization (as percent of limit) are not statistically significantlydi erent from 0Estimates noisy due to noisy dataStephanie LoElasticity of mortgage demandDecember 2, 201620 / 25

Results not driven by di erential mortgage shopping600600fico score at origination650700750median credit score at origination650700750800B. Search vs. orig: median credit score800A. Search vs. orig: FICO600650700initial fico score750800650700750initial median credit score800D. Number of months searched2.52.53number of months searched3cumulative searches3.544.53.55C. Number of mortgage inquiries600600650700FICO at origination750800600650700FICO at origination750800No discrete breakpoints in mortgage shopping behavior or search vs.originated credit scoreStephanie LoElasticity of mortgage demandDecember 2, 201621 / 25

Plan1Motivation2Brief Methodology3Results4RD Validity5Economic Implications6ConclusionStephanie LoElasticity of mortgage demandDecember 2, 201622 / 25

Economic ImplicationsSuppose, over the sample period (Oct08 - Dec14), there were nodi erential LLPAs across 680 and 720 FICO scoresThen, FICO 680-719 borrowers (about 20 MM Americans) would:have faced interest rates that were on average 10-25 basis points lowerhave demanded approximately 25% more mortgages totalwould have originated mortgages about 11k larger eachPotential increase in total mortgage demand of 15 B (o actual newmortgage debt of 43.5 B)About 10 B from new mortgage debt and 5 B from increased size ofexisting mortgage debtOverall e ect could of course be larger, taking into account lowerFICO-score borrowersStephanie LoElasticity of mortgage demandDecember 2, 201623 / 25

Plan1Motivation2Brief Methodology3Results4RD Validity5Economic Implications6ConclusionStephanie LoElasticity of mortgage demandDecember 2, 201624 / 25

ConclusionNovel identification method to measure microelasticity of mortgagedemand to interest ratesHigh FICO borrowersRegulatory “wedge” in interest rates faced across thresholds !regression discontinuityDemand, not supply, drivenBorrowers’ demand for debt is sensitive to interest rates; 25 bpdecrease in interest rates associated with:Intensive: increase in loan size of 15k (approx. 10% of averageorigination volume)Extensive: 50% increase in likelihood of potential borrower to demanda loanImplications for “missing” mortgage demand post-crisis and efficacyof monetary policyStephanie LoElasticity of mortgage demandDecember 2, 201625 / 25

I find the elasticity of mortgage demand to interest rates is large A 25 bp decrease in mortgage rates is associated with: Intensive: an increase in loan size of 15k (approx. 10% of average origination volume) Extensive: a 50% increase in likelihood of potential borrower to demand a loan