Research

(with Rodolfo Rigato) [submitted] [IMF WP version]

Abstract.  Leveraging project-level data from India and quasi-experimental variation in credit supply, we show firms expedite ongoing capital expenditure projects and start fewer projects when credit tightens. We rationalize our findings with a novel investment model featuring endogenous time to build. Tighter credit incentivizes firms to accelerate near-completion projects for faster cash flows while delaying early-stage projects, a mechanism borne out in the data. A quantitative model disciplined by our estimates shows endogenous time to build generates strong endogenous amplification and state-dependence of investment responses along the project maturity distribution. We illustrate implications for interest rate and tax policy.

Presentations.  Johns Hopkins SAIS, IMF-ER Summer Conference, SED Annual Meetings 2025, Western Finance Association Annual Meetings, IMF Macro-Finance Research Conference, IMF Research Department, STEG-CEPR, Sociedade Brasileira de Econometria*, SED Winter Meetings 2024, Boston College, European Central Bank*, NBER Summer Institute, IMF, Federal Reserve Bank of San Francisco, Federal Reserve Board, Cornerstone, McCombs School of Business of The University of Texas at Austin, Toulouse School of Economics, CREI-UPF, The Wharton School of the University of Pennsylvania, Analysis Group

(with Rodolfo Rigato) [submitted]

Online Appendix

Abstract.  We show that precautionary savings are too sensitive to consumption risk in standard heterogeneous agent (HA) models. First, we estimate the response of savings to variation in unemployment insurance (UI), a key cushion for consumption risk from job loss. We find small, statistically non-significant effects. Second, we show that such estimates reject HA models with standard time preferences; but that introducing present-biased households suffices to bridge model and data. Empirically discipling precautionary savings is policy-relevant, weakening the role of UI as an automatic stabilizer and discretionary stabilization tool: an estimated HANK model predicts a 45% smaller fiscal multiplier of temporary UI extensions; and a 40% weaker effect of UI in reducing aggregate consumption volatility.

Presentations.  SED Annual Meetings 2025*, Banco Central do Brasil*, IMF, University of California, Los Angeles*, Federal Reserve Bank of San Francisco*, FGV- EESP*, Insper*, PUC-Rio*, Bank of Canada*, European Central Bank*

(with Michael Blank)

Abstract.  How do firms shape the transmission of macroeconomic shocks and policy? Financial accelerator theories emphasize the role of firm-level financing frictions in amplifying the macroeconomic impact of aggregate shocks. While this literature generally focuses on capital investment, we consider how the effects of monetary shocks are amplified through links between financing frictions and labor demand. Under financial acceleration, firms reduce labor demand as financing constraints become more severe in response to adverse shocks, lowering labor income, and thereby aggregate demand. We empirically test this channel with a “micro-to-macro” approach based off the universe of US public firms. We first show that firms that ex ante appear to be relatively financially constrained contract employment more after a monetary tightening. We then assess the aggregate implications of this employment channel through a regional design. We construct measures of a given county’s exposure to public firms with differential financial constraints, and document that more exposed counties exhibit stronger employment declines following contractionary monetary shocks. Preliminary evidence suggests that within-county spillovers of constrained firms to the regional labor market are concentrated in non-tradable establishments, suggesting that interactions between aggregate demand and financial amplification through employment are operative.