The Heterogeneous Effects of Household Debt Relief
(with Manuel Adelino and Miguel Ferreira)
Large-scale debt forbearance is a key policy tool during crises, yet targeting is challenging due to information asymmetries. Using transaction-level data from a Portuguese bank during COVID-19, we find financially fragile households are more likely to enter forbearance, irrespective of income shock. Mortgage payment suspension increases consumption and savings, but effects differ across households. Low-wealth and low-income households exhibit higher consumption sensitivity to forbearance, while high-wealth and high-income households show greater saving sensitivity. Additionally, ineligible households accessing forbearance show a higher consumption sensitivity than eligible ones. Our findings suggest considering observable household characteristics could enhance the effectiveness of debt relief policies.

Do Specialized Distress Investors Undermine Firms Ex Ante? A Simple Model of Private Bailouts
(with Fernando Anjos and Irem Demirci)
We extend the canonical moral hazard model in corporate finance to include specialized distress investors (SDIs) who privately bail out financially-distressed firms. First, we show that SDIs can negatively impact credit rationing ex ante, even when SDIs create value and even though initial lenders are required to agree to SDI participation. Second, we show that SDI presence lowers project NPV when initial lenders can exploit a value-destroying SDI’s technology to expropriate equityholders ex post. These negative results notwithstanding, we also show that firms can sometimes mitigate these effects by reducing leverage and/or by focusing on projects with certain characteristics.

Capital Structure and Employee Consumption
I show that employer capital structure can affect employee consumption and saving decisions using new matched employer-employee data from Portugal. Specifically, I find that employees of highly leveraged firms exhibit lower marginal propensities to consume, particularly in non-essential goods and services. This suggests that financial distress costs can spill over to other, potentially unrelated firms. Employees in these firms do not receive higher wages to compensate for the increased risk. The effects are identified by exploiting negative industry-wide shocks. I rationalize the findings using a Diamond-Mortensen-Pissarides matching model, where heterogeneous risk-averse employees bargain with heterogeneous employers to determine wages. Consistent with the model, low-wealth individuals are most affected due to their relatively higher unemployment costs. My results suggest that financial distress costs are partially shifted to employees.