Joseph Vavra, Associate Professor of Finance

In recent years, many countries have used macroprudential policies such as tightening LTV requirements to potentially reduce financial stability risks. While some research suggests these policies are effective in reducing lending and house price growth, less is known about potential negative effects. For example, tighter limits on home equity borrowing may lead constrained households to borrower through higher-priced sources of credit like credit cards. Studying these issues by looking across countries is difficult since countries differ along lots of other dimensions. Using local within country variation is potentially more promising for cleaner identification, and in this project we propose to use policy discontinuities at the border with Texas to explore these issues. In Texas, equity extraction is only allowed if the resulting LTV remains below 80%, while other states do not have such restrictions. Kumar (2017) finds large effects of this policy on default using a border discontinuity design, but cannot actually measure equity withdrawal after origination or the effects on other borrowing, so it’s fairly indirect evidence. We want to use the CRISM data to more directly explore the effects of this policy on the relevant outcome variables of interest after controlling for borrower and loan characteristics. In particular, how much less home equity are Texas borrowers extracting compared to borrowers in neighboring states? Do borrowers in Texas carry more non-mortgage debt and are they more likely to become delinquent on this debt than borrowers in other states? How do the answers to these questions vary with borrower characteristics like age and credit score? Are Texas borrowers more vulnerable to economic shocks? One potential way to address this last question is to look at effects of oil price declines since mid-2014 interacted with the fact that certain locations in Texas have employment which is particularly sensitive to oil prices.