Hedging Interest Rate Risk: Evidence from U.S. Nonfinancial Firms (2026)
Abstract: Nonfinancial firms hold roughly $7.7 trillion in dollar interest rate derivatives, yet why some firms hedge while others do not remains an open question. I use a large language model to extract derivatives disclosures from SEC filings, constructing a new dataset on interest rate hedging for 8,792 U.S. public nonfinancial firms from 2014 to 2024. Hedging is uncommon and highly persistent. Hedgers are larger, more leveraged, and have higher operating income than non-hedgers. Firms with more liquid assets hedge less. I explain these facts with a heterogeneous firm model in which debt raises default risk and therefore the value of hedging, but hedging is costly, so only firms with sufficient operating income find it worthwhile. This complementarity between leverage and operating income governs which firms hedge. Counterfactual experiments show that hedging costs are the main barrier to participation. In both the model and data, hedging insulates the investment of highly levered firms from monetary policy shocks, suggesting that the real effects of monetary policy fall disproportionately on unhedged firms.
Decomposing Crop Loss: The Role of Subsidies and Market Price Uncertainty (2025)
Abstract: Farmers choose to leave around 42% of harvestable crops on the field in the United States. This paper studies two of the economic drivers of this crop loss, price uncertainty and subsidies, and underscores their large environmental impacts. I develop a dynamic stochastic structural framework of crop loss in agricultural production, incorporating subsidy distortions and farmer price expectations. Estimated with wheat data, the baseline model indicates that the presence of subsides targeting downside price risk cause crop loss to be higher relative to a no-subsidy world. However, as the signals over prices received by farmers become noisier, the subsidy distortion decreases in importance relative to price uncertainty. The model provides two possible channels of crop loss that I test in an empirical exercise using wheat and wheat futures data. I find that given a high price expectation at planting, a one standard deviation increase in the price expectation at harvest decreases crop loss by around 8-10%. An additional environmental analysis shows that mean greenhouse gas (GHG) emissions due to crop loss increased after the introduction of the PLC/ARC programs by about 1.48% and that eliminating price uncertainty would reduce GHG from crop loss by about 3.78%.
Hedging Currency Risk Exposure and Investment Decisions by U.S. Non-financial Firms (2026) | with Julia Chahine
Interest Rate Risk Hedges by Large Non-financial Firms Effect Monetary Policy Transmission (2026)
Labor Market Demand and Monetary Information Shocks (2024) | with Amy Handlan and Ross Batzer