Publications
Input Sourcing under Climate Risk: Evidence from U.S. Manufacturing Firms with Joaquin Blaum and Sebastian Heise
Review of Economic Studies, May 2026
Abstract: We study the effect of risk on how firms organize their supply chains. We use transaction-level data on U.S. manufacturing imports to construct a novel measure of input sourcing risk based on the historical volatility of ocean shipping times. Our measure isolates the unexpected component of shipping times that is induced by weather conditions along more than 331,000 maritime routes. We first document that unexpected shipping delays significantly reduce importers' sales, profits, and employment. We then show that firms actively diversify weather risk by using more routes and foreign suppliers, although their import values decline. To rationalize these findings, we introduce shipping time risk into a general equilibrium model of importing with firm heterogeneity. Our quantitative analysis predicts substantial costs for the U.S. economy associated with supply chain risk.
General Equilibrium Effects in Space: Theory and Measurement with Costas Arkolakis and Rodrigo Adao
Forthcoming at American Economic Journal: Macroeconomics, May 2026
Abstract: We document that estimates of the differential effect of the China shock measured by Autor et al. (2013) are much larger than those implied by existing quantitative frameworks. This disconnect is more substantial when we account for regional exposure in consumption and import competition in other nearby regions. We then develop a reduced-form representation of spatial models in which changes in regional outcomes combine (i) each region's ``shift-share'' exposure to shocks in its excess labor demand, and (ii) the reduced-form effect that this exposure has directly on that region, and indirectly on other regions through spatial links. We use this representation to uncover the roots of the disconnect, and derive an empirical specification whose aggregation yields the shock’s general equilibrium impact. Finally, estimates of our reduced-form representation for the China shock yield large direct effects and reinforcing indirect effects, which arise from strong agglomeration forces, strong (weak) employment sensitivity to wages (prices), and regional production and trade links. As a result, we obtain employment losses that are larger and more dispersed than those implied by existing quantitative frameworks.
The Economics of Carbon Leakage Mitigation Policies with Stefan Ambec and Antonia Pacelli,
Journal of Environmental Economics and Management, May 2024
Abstract: In a trade model with endogenous emissions abatement, we investigate the impact of three policy instruments aimed at mitigating carbon leakage: free emission allowances, a Carbon Border Adjustment Mechanism (CBAM), and a CBAM with export rebates. We show that providing free allowances does not alter the incentives to abate carbon emissions, but, instead, fosters the entry of more carbon intensive producers. It "levels the playing field'' both domestically and internationally, and may even reverse the carbon leakage. In contrast, a CBAM only levels the playing field domestically, and may lead to an autarky equilibrium. To reverse the carbon leakage, a CBAM must be complemented with export rebates. We further show that a CBAM and export rebates improve welfare for any carbon price, and we identify the optimal share of free allowances with or without a CBAM. Finally, we perform a calibration exercise on cement and steel sectors to simulate the effects of the CBAM recently adopted by the European Union. Our model predicts a scenario with reverse carbon leakage and significant welfare gains for both sectors.
Demand Risk and Diversification through International Trade
Journal of International Economics, March 2022
Working paper version, Online Appendix, Codes and data, Slides
Abstract: I develop a theory of risk diversification through geography. In a general equilibrium trade model with monopolistic competition and stochastic demand, risk-averse entrepreneurs exploit the spatial correlation of demand across countries to lower the variance of their global sales. The model implies that both the probability of entry and trade flows to a country are increasing in the "Diversification Index", which depends on the multilateral covariance of the country's demand with all other markets. The risk diversification behavior can lead to higher welfare gains from trade than the ones predicted by trade models with risk neutrality. Using a panel of international sales of Portuguese firms, I estimate "risk-augmented" gravity regressions, which show that the Diversification Index significantly affects trade patterns at the extensive and intensive margins. In general equilibrium, the risk diversification channel increases welfare gains from trade by 17% relative to trade models with risk neutrality.
Trade Policy Uncertainty and Stock Returns with Marcelo Bianconi and Marco Sammon
Journal of International Money and Finance, December 2021
Data, Working paper version, Slides
Featured on EconBrowser, Financial Times
Abstract: A recent literature has documented large real effects of trade policy uncertainty (TPU) on trade, employment, and investment, but there is little evidence that investors are compensated for bearing such risk. To quantify the risk premium associated with TPU, we exploit quasi-experimental variation in exposure to TPU arising from Congressional votes to revoke China's preferential tariff treatment between 1990 and 2001. A long-short portfolio designed to isolate exposure to TPU earns a risk-adjusted return of 3.6-6.2% per year. This effect is larger in sectors less protected from globalization, and more reliant on inputs from China. Industries more exposed to trade policy uncertainty also had a larger drop in stock prices when the uncertainty began, and more volatile returns around key policy dates. Our results are not explained by the effects of policy uncertainty on expected cash-flows, investors' forecast errors, and import competition from China.
Estimating the welfare costs of autarky: a sufficient statistics approach
Economics Letters, September 2020
Abstract: I use the Jeffersonian Embargo enacted in 1807 to estimate the welfare costs of autarky. I develop an Armington trade model to compute the welfare losses using two sufficient statistics: the share of expenditures on domestic goods and the elasticity of substitution between domestic and imported goods. I use historical data from 1792 to 1807 to estimate the Armington elasticity, using import tariffs as instrument for relative prices. The empirical findings suggest welfare losses of 2.83-8.14% of real income.
Working papers
Import Competition, Trade Credit, and Financial Frictions in General Equilibrium with Fadi Hassan
R&R at American Economic Journal: Macroeconomics, June 2026
CEP Discussion Paper, CEPR Discussion Paper, Slides
Abstract: We analyze the role of trade credit and financial frictions in the propagation of international trade shocks along the supply chain. First, we show empirically that exposure to import competition from China increased the use of trade credit in the U.S. Then, we use a multi-country input-output trade model with borrowing constraints, trade credit, and endogenous employment to quantify the general equilibrium effects of such increase, characterizing the different channels at work. Borrowing constraints amplify the negative consequences of the China shock on employment, but introducing trade credit reduces these losses by 8%-27%, depending on the tightness of the constraints.
Resilience after Climate Shocks: Evidence from SBA Disaster Loans with Saverio Simonelli, October 2026
Abstract: We show that natural disasters generate persistent declines in U.S. county employment and net job creation. We then examine whether emergency credit can mitigate these effects. Linking the universe of SBA disaster-loan applications to Census Bureau records, we exploit a credit-score screening rule introduced in 2016 and implement a fuzzy regression discontinuity design around the score cutoff. Firms just above the threshold are substantially more likely to receive government-backed loans, and approval increases employment relative to declined applicants, with effects persisting over subsequent years. The adjustment operates primarily through lower job destruction, while reduced firm exit and stronger job creation also contribute. Overall, SBA lending preserved an estimated 9.8% of the jobs that would otherwise have been lost in disaster affected counties.
Work in progress
"Social Media Sentiment and Asset Prices" with Marcelo Bianconi
Abstract: This paper develops a structural account of short selling in U.S. equity markets. We combine evidence from 689 S&P 500 firms and 3.39 million Reddit submissions with a continuous-time model in which social media activity shapes disagreement and endogenously determines short-sale constraints. We show that social media coverage is associated with lower short interest and contemporaneous return discounts. The model also generates testable predictions about non-monotonic short selling, expected returns, and threshold effects, with the evidence supporting these predictions.
"Climate Policy with a Carbon Border Adjustment" with Stefan Ambec and Antonia Pacelli
Abstract: Trade policies such as the carbon border adjustment mechanism (CBAM) can mitigate carbon leakage but may reduce gains from trade. We show that when the CBAM reduces emissions, relaxing domestic climate policy—through a lower carbon price or additional emission allowances—can improve both cost-effectiveness and gains from trade while achieving the same emissions target. Calibrating our model to the EU iron and steel sector, we find that a 30% emissions reduction can be achieved with a 10% lower carbon price when the CBAM is implemented.