Job Market Paper

Decomposing the Investment Channel of Monetary Policy (with M. Momm) [PDF] 

European Central Bank's Young Economist Prize 2026

Cambridge Finance Best Student Paper Award 2026

Abstract. Monetary policy announcements simultaneously change borrowing costs and firms' expectations about future economic conditions. Conventional estimates of the investment response to monetary policy therefore conflate price effects with belief revisions and fail to identify the structural sensitivity of firm investment to financing conditions. We exploit quasi-experimental variation from the ECB's 2016 Corporate Sector Purchase Programme (CSPP), which created persistent cross-sectional differences in firms' financing costs through bond market segmentation, to isolate the causal effect of borrowing costs on firm investment. We find that investment elasticities estimated from aggregate monetary policy shocks understate the effect of a pure cost-of-debt shock by 38 percent over four quarters. Our results imply that the structural sensitivity of firm investment to borrowing costs is larger than existing estimates indicate. The gap reflects a sizable firm-level information effect. A simple model of firm investment with an information effect rationalizes this attenuation. Our findings carry direct implications for heterogeneous real transmission of monetary policy and the calibration of quantitative macroeconomic models.

Presentations: ECB (Seminar), Chicago FED (Seminar), ASSA 2027* (Washington DC), ECB Forum on Central Banking 2026 (Sintra), EEA 2026 (Dublin), PSE-CEPR Policy Forum 2026, RAPS/RCFS Europe Conference 2026 (Cambridge), Theories and Methods in Macro 2026 (Montreal), Midwest Macro Spring 2026 (Milwaukee), EDGE Conference 2025 (Cambridge). 

Working Papers

The Role of Penalties for Sovereign Default Expectations: Evidence from the Courts [PDF]

Runner-up of the Cambridge Finance Best Student Paper award 2024

Abstract. Why do sovereigns repay foreign creditors when external enforcement is weak? This paper shows that direct economic default penalties - not just exclusion from financial markets - are a key determinant of repayment incentives. I exploit quasi-exogenous variation in the cost of default generated by US court rulings in sovereign default litigation. Identification comes from high-frequency responses in third-country sovereign bonds governed by US law whose issuers are not parties to the litigation. I find that a one percentage point increase in the recovery rate lowers the default probability by 19 basis points. I interpret these results in a dynamic model of international borrowing with endogenous default and endogenous debt recovery rates. Calibrated to match the empirical relationship between recovery and default, the model implies that a one percentage point increase in the cost of default lowers the default probability by 7.9 basis points. I find that higher default penalties can be welfare improving. This offers an explanation for why many emerging markets have been pursuing high-debt-high-default-cost regimes, heightening financial instability under probabilistic tail events.

Presentations: EEA-ESEM 2026 (Dublin), Midwest Macro Spring 2026 Conference (Milwaukee), RIEF Meeting 2026 (Paris), ASSA 2027 (poster; Washington DC), CEPR Paris Symposium 2024 (poster), 96th IAEC (Philadelphia), EUI (seminar).

The Network Origins of Financial Cycles (with G. Carboni) [PDF]

Abstract. We use proprietary European holdings data to show that common ownership is informative for how shocks transmit between securities. Holder-base overlap predicts cross-sectional variation in abnormal excess yield changes, conditional on a rich set of bond- and bond-pair characteristics. We provide evidence consistent with a causal interpretation by exploiting a major fiscal surprise from Germany in 2025. We show that non-German sovereign bonds with greater pre-existing holder-base overlap with German Bunds experience larger abnormal excess-yield spillovers over a tight event window. Motivated by this, we construct pairwise measures of holder-base overlap and characterize the resulting network of the European sovereign bond market. We then calibrate a common-ownership network to the German event study and find that security-interconnectedness amplifies aggregate volatility from security-specific shocks by roughly 2.5 to 4 times relative to a disconnected market.

Work in Progress

Call Me Maybe: Debt Management with Callable Perpetuities (with M. Ellison, E. Faraglia and F. Velde) [Draft Coming Soon]

Abstract. Why did governments historically issue callable long-term debt, and when can callability improve welfare? We study optimal debt management when the government can redeem long-duration bonds at a fixed price. Under rational pricing, callability provides poor fiscal insurance: bonds are called precisely in states with low financing needs, generating state-contingent transfers from households to the government that amplify rather than smooth fiscal shocks. This lowers welfare relative to otherwise similar noncallable debt. The conclusion changes when investors underprice sufficiently out-of-the-money call options. Mispricing raises issue prices, relaxes current financing needs, and can make callable debt welfare-improving despite its adverse state contingency. We establish these mechanisms in a tractable Ramsey model and extend them to an infinite-horizon economy with callable perpetuities calibrated to Britain, 1700 - 1900. The analysis rationalizes the historical use of call provisions as a debt-management instrument and highlights asset-pricing frictions as a determinant of optimal sovereign debt design.

Presentations: 2025 AI and Learning the Macroeconomy Workshop (†), 2025 Financial History Workshop (†).

Pushing on a String: The Role of the Information Effect

Abstract. Transmission of monetary policy to the real economy is weakest in a recession when stimulus is most needed. This paper studies how aggregate state dependency in transmission can be explained by cyclicality in the information effect - belief revisions about economic fundamentals around monetary policy shocks. We disentangle monetary policy shocks from their information component using a structural vector autoregression. We find that investment responds less to monetary policy shocks with an information component during recessions than during booms. In contrast, we find no such difference for monetary policy shocks without an information component. We develop these results in a model of rational inattention, where we show that in bad times, firms are more incentivized to pay attention to information transmitted alongside monetary policy surprises.

(* = scheduled, † = presented by co-author)