Working Papers

2025, Regulation and Intermediation in Over-the-counter Markets (Job Market Paper)

2025, Liquidity Crises and Endogenous Dealer Capacity in Over-the-counter Markets (Under review)

This paper studies how dealers provide liquidity in over-the-counter (OTC) markets during crises when they can choose their intermediation capacity: their order execution speed and inventories. I develop a search model of an OTC market where trade between homogeneous dealers is centralized, but investors face search frictions proportional to dealer order execution rates. Agents engage in bargaining over unit trade when they meet, where dealers charge intermediation fees. The model features a one-time aggregate unanticipated adverse liquidity shock to investors’ private valuations. I show that with more severe shocks, dealers choose lower order execution rates and wait longer to begin taking on inventories. Relative to a constrained planner, I identify a double-sided hold-up problem that can be resolved in the steady state, but not out of the steady state. A calibration to the U.S. secondary corporate bond market highlights that capacity inefficiencies are more consequential following a liquidity shock.

Work in Progress

Diversity and dissent at the board of directors: The case of the Banco de la República, Joint with Cristian Frasser & Juan Acosta 

This work examines how socioeconomic diversity within the Board of Directors at the Banco de la República, the monetary policy committee of Colombia’s central bank, influences consensus on interest rate decisions. Socioeconomic factors can shape the policy preferences of board members, particularly in how each weighs inflation relative to unemployment when setting interest rates. Since interest rate decisions are determined by a majority vote, a more socioeconomically diverse board may find it harder to reach a consensus. To analyze this phenomenon, we hand-collected data on all interest rate votes and board member biographies from June 2007 to September 2024. We construct diversity indices that capture variation over time across three categories: demographic, educational, and political. We show that demographic diversity in the Board of Directors has increased over time, whereas educational diversity has remained stable, and political diversity has declined. We first estimate a forward-looking Taylor Rule augmented with board composition, finding that an older, more tenured board is associated with voting for higher interest rates, whereas a board with more women or more members holding a PhD in economics is associated with voting for lower interest rates. We then study the determinants of dissent, that is, interest rate votes that were non-unanimous. We find that boards with greater demographic and educational variation are more prone to disagreement. Furthermore, disagreement varies non-linearly in the deviation of inflation from target: diverse boards are more likely to disagree when inflation is moderately above target, but as inflation becomes increasingly high, disagreement declines.

Disagreement in Monetary Policy Voting: The Role of Heterogeneous Preferences, Joint with Yelda Gungor

This paper studies how heterogeneous preferences and private information shape voting outcomes in monetary policy committees (MPCs). We utilize a strategic voting model in which each member’s preferences are represented by an individual Taylor rule that places different weights on inflation and the output gap. In a symmetric Bayesian Nash equilibrium, we show that members’ cutoff strategies depend on their inflation weights. We are in the process of hand-collecting interest rate votes and detailed biographical information for MPC members in the United States, the United Kingdom, Sweden, and Mexico, where individual voting records are disclosed. These data will allow us to estimate how members’ characteristics determine their inflation weights, and what the implications are for MPC decision making.

The Role of Exchange Rate Variations in European Firms’ Investment, joint with Luis Reyes

We distinguish between the cost and demand effect of a currency depreciation on firm investment. The former takes place when the cost of existing debt increases with the value of the foreign currency in which it is denominated, whereas from the latter we expect an increase in the demand for exports. We test both and their overall effect on investment for high-income European non-financial firms and find that the positive demand effect outweighs the negative cost effect. We carry out firm-level and aggregate level dynamic linear panel regressions, distinguishing variations in the exchange rate, appreciations, and depreciations. This analysis is complemented by individual country VARs estimated for France, Germany, and the UK, in which we compare their impulse responses.

Rehypothecation, Information Asymmetry, and Endogenous Fragility

Other Research

2025 Series superstars: How streaming-video-on-demand (SVOD) content popularity informs SVOD provider demand, joint with Anthony Palomba (Revise & Resubmit, International Journal on Media Management)

Intense competition among major SVOD providers in the United States—such as Amazon Prime Video, Hulu, HBO Max, Apple TV+, Netflix, and Disney+—has driven a focus on acquiring and creating high-performing superstar series to attract subscribers. However, the impact of superstar series on subscriber numbers and overall audience demand remains unclear. This paper explores the superstar effect, a concept introduced by Rosen (1981), which suggests that top performers can command significantly higher rewards despite only minor differences in talent. This effect results in notable disparities in market dominance and earnings. Using a novel dataset from Parrot Analytics, which measures daily audience demand through downloads, views, searches, and social media posts in the United States, we apply a Berry, Levinsohn, and Pakes (1995) structural demand estimation. We define "superstar series" as those with an average relative demand ranking in the top 1% for each quarter in an annual year (Elberse, 2008). To assess how many superstar series a provider has compared to other providers, we calculate the percentage of a provider's series that are classified as superstars. Our analysis reveals that while superstar series do impact subscriber numbers, consumers generally prefer a catalog with consistently popular series. Furthermore, consumer demand is relatively inelastic to changes in catalog popularity and even less responsive to the proportion of superstar series. Simulations indicate that enhancing the average popularity of content can lead to higher prices and more new subscribers, having a greater effect than merely adding superstar series.