Welcome!
I am a PhD Candidate in Finance at the Simon Business School, University of Rochester. My research interests are in mortgages and securitization, corporate finance, and real estate.
I am on the 2026-2027 job market.
My job market paper studies how CARES Act mortgage forbearance unintentionally reduced new mortgage lending by imposing liquidity and operational burdens on mortgage servicers.
Email: zliu88@simon.rochester.edu
Presentation: FMA (2026), SFA (2026), University of Rochester (2026)
Abstract: I document an unintended intermediary-side consequence of the Coronavirus Aid, Relief, and Economic Security (CARES) Act mortgage forbearance program. The policy allowed borrowers to pause payments but imposed liquidity and operational burdens on mortgage servicers, many of which also originate mortgages. To address the joint determination of forbearance and lending policies, I construct a shift-share exposure measure combining lenders' pre-pandemic geographic footprints with county forbearance rates and compare origination outcomes of differentially exposed lenders within the same county-month. A one-standard-deviation increase in forbearance exposure reduces monthly mortgage originations by 20.6% during March-December 2020. The lending contraction is concentrated among lenders that retain servicing or hold larger mortgage servicing asset positions, consistent with a servicing channel. High-exposure lenders approve fewer otherwise similar applications and charge higher rate spreads, while borrower characteristics remain stable, suggesting a lender-driven contraction. My results reveal that relief policies for existing borrowers can constrain credit to new borrowers through intermediary frictions.
with Yang Yi
Presentations: AEA (2026, Poster), FMA (2025), SFA (2025), EFA (2025), CICF (2025), Boulder Summer Conference on Consumer Financial Decision Making (2025, Poster), University of Rochester (2025), University of Alberta (2025). (includes presentations by coauthors)
Abstract: Do political shifts shape household leverage cycles? Comparing otherwise similar mortgage borrowers in consistently Democratic- and Republican-leaning areas, we find that those politically aligned with the U.S. president request larger loans and use higher leverage. Lender screening only partially offsets this surge in demand, yielding a 10% relative increase in originations in aligned regions. The effects intensify in politically homogeneous areas, during periods of heightened partisan conflict, and among financially unconstrained borrowers. Survey evidence shows that aligned individuals hold more optimistic housing market expectations. Yet these beliefs do not materialize in stronger fundamentals, leading to higher delinquency in aligned regions.
Abstract: Lenders' decisions about which loans to keep on their balance sheets and which to sell through securitization shape the allocation of credit risk. I examine how banks' retention decisions depend on their local branch presence. Using U.S. conforming mortgages originated from 2012 to 2021, I find that banks are more likely to retain mortgages near their branches and sell more distant loans. Comparing loans originated by the same bank and accounting for local market conditions, I find that the retention probability is 3.9 percentage points lower at ten miles than within one mile of the bank's nearest branch, a difference equivalent to 13% of the unconditional retention rate. Banks' strategic retention has real impacts on loan performance: Among mortgages acquired by Fannie Mae, loans in areas where the originating bank has a stronger local presence are less likely to become delinquent. Overall, the results suggest an informational advantage of local branches.