RAMs and Refinancing
Saturday, Jan. 6, 2018 10:15 AM - 12:15 PM
- Chair: Vicki Bogan, Cornell University
How Home Equity Extraction and Reverse Mortgages Affect the Credit Outcomes of Senior Households
AbstractThis paper examines how the extraction of home equity, including but not limited to equity extracted through reverse mortgages, affects credit outcomes of senior households. We use data from the Federal Reserve Bank of New York/Equifax Consumer Credit Panel, supplemented with our unique credit panel dataset of reverse mortgage borrowers. We track credit outcomes for seniors who extracted equity through cash-out refinancing, home equity lines of credit or home equity loans between 2008 and 2011, and a random sample of non-extractors. We estimate differences-indifferences by extraction channel using individual fixed effects panel regression. We find that seniors extracting equity through reverse mortgages have greater reductions in consumer debt, and are less likely to become delinquent or foreclose three years post origination relative to other extractors and non-extractors. These effects are greater among households who experienced a credit shock within the two years prior to loan origination. To help isolate the effect of the extraction channel on credit outcomes, we re-estimate our models with a matched sample of consumers at the time of extraction. We find that otherwise similar HECM borrowers have larger reductions in credit card debt post extraction than other equity borrowers and non-borrowers, with no significant difference in the rates of delinquency on non-housing debt post extraction. For HECM borrowers, we find that increased initial withdrawal and increased monthly cash flow contribute to the reduction in credit card debt.
An Empirical Study of Termination Behavior of Reverse Mortgages
AbstractReverse mortgages generally have open maturity dates. The variability of the exact termination time of a mortgage is one of the most important risks faced by the lenders and mortgage insurers. This paper analyzes the termination experience of the reverse mortgages in the US. We identify that a reverse mortgage can be terminated by three distinct events: refinance, mortality, and mobility. Using FHA insured Home Equity Conversion Mortgage loan level data, we estimated the probabilities of the termination by individual events. The results show that the three termination events are driven by different factors. Refinances are mainly driven by the appreciation of the house value, especially during the first three years. Mortality terminations follow closely the US mortality tables, which are governed by age and gender. Mobility caused terminations are sensitive to borrower’s age-gender characteristics as well as housing market conditions. In addition, the initial cash draw pattern has significant but different impacts on all three termination types.
Mortgage Default with Positive Equity
AbstractFrictionless models of mortgage default predict that defaulters are underwater.
Available evidence suggests otherwise, but there are well-known concerns with this evidence,
and so it is unclear how many defaulters have positive equity. This paper provides the first
formal estimates of the Loan-to-Value ratios (LTVs) of defaulters, using rich microdata and
a robust Bayesian estimation procedure. It finds that 27-47% of foreclosed homeowners had
positive equity from 2011-2013, which implies roughly 81-87% had positive equity in more
"normal" times. Motivated by this evidence, the paper then develops a quantitative lifecycle
model of mortgage default with standard search frictions. In the model, homeowners who miss
a mortgage payment may make it up the next period. As a result, abovewater homeowners
sometimes choose to miss a payment rather than sell after an income shock. If their income
recovers, they make up their payment and keep their home. If it does not, they may sell their
home in a frictional market. Otherwise, they lose their home to foreclosure. The estimated
model matches key untargeted moments, including the foreclosure rate and the proportion of
defaulters with positive equity. In the model, a policy called "lender recourse" - which allows
lenders to seize the assets of underwater defaulters - is generally ineffective at reducing default
rates, consistent with the evidence.
University of Southern California
California State University-Fullerton
Federal Reserve Bank of Philadelphia
- D1 - Household Behavior and Family Economics
- R2 - Household Analysis