This paper studies how lender-borrower relationships impact loan performance, pricing, and competition in U.S. mortgage markets. I establish that lenders learn private information about borrowers during the lending relationship, creating adverse selection among borrowers who switch lenders. Switching borrowers are 34% more likely to miss payments than borrowers who stay with their original lender, and lenders respond by charging switchers higher interest rates of up to 11 basis points. To identify the role of lender private information in generating this adverse selection, I use lender market exits as a quasi-experiment and show that borrowers forced to switch lenders perform better than other borrowers who switch lenders. I then estimate a model of relationship lending with endogenous adverse selection to quantify how policies which eliminate incumbent lenders' information advantage would reduce their incentives to offer lower initial rates, increasing mortgage rates by 5 basis points.

This paper studies the effect of housing affordability on fertility using a large pro-natalist housing policy. In 2019, the Hungarian government expanded housing subsidies and mortgage support for families purchasing homes in small municipalities, with eligibility determined by a population cutoff of 5,000 residents. A distinctive feature of the policy is that households could access housing support not only based on existing children, but also by committing ex-ante to having children in the future. Exploiting the population threshold, we implement a difference-in-differences design using linked census and administrative birth records covering the universe of women of reproductive age. We find that eligibility for the program increased fertility by 3.0 percent in the five years after implementation. Dynamic estimates show no differential pre-trends and suggest that the effects persist beyond short-term timing shifts. The results provide causal evidence that relaxing housing finance constraints prior to childbearing increases fertility.

Analyzing the costs and benefits of liquidity regulation in the modern context is challenging due to the presence of other interventions whose effects commingle with the impact of liquidity regulation. This paper investigates the impact of liquidity regulation in a historical context in which liquidity regulation's effects can be more effectively isolated from those of other policy interventions: the National Banking Era in the late 19th century United States. Our quasi-experimental design uses population-based cutoffs that determined the stringency of reserve requirements to identify the causal impact of liquidity regulation on bank lending. We present evidence that tighter liquidity requirements crowd out bank lending, illustrating a cost of liquidity regulation.