Information Channel of Monetary Policy in Uncertain Times   (Job Market Paper)   (2024)
   [Draft]

The quantitative importance of the so-called Fed "information effect" has been extensively discussed by recent studies with no clear consensus. This paper examines how the relevance of the information channel varies across different levels of uncertainty. Analyzing revisions in private-sector expectations following monetary policy announcements and the macroeconomic effects of central bank information shocks, I find that the information effect is significant in periods of high uncertainty but negligible when uncertainty is low. A New Keynesian model, featuring a central bank information advantage and Bayesian learning of agents, rationalizes these empirical findings. The results are robust after accounting for the Fed "response to news" channel, which could potentially challenge the relevance of the information channel.

Presented at: American University (2025; Washington D.C.), University of Mississippi (2025; Oxford, MS), SEA (2024; Washington D.C.), KLI (2025), BOK (2025), KIEP (2025), MEG (2024; University of Kentucky), MMM (2024; Purdue University), SNDE (2024; University of Padova), SNDE-IIF Workshop for Junior Researchers (2024; Virtual), UT-TAMU-IH Macro Job Candidate Workshop (2024; Dallas Fed), Rice University (2024), Texas A&M University (2023, 2024) 

The Relevance of Temporal Aggregation for the Propagation of Macroeconomic Shocks   (2024)
    (with Tatevik Sekhposyan)

   [Draft]   [Data Coming Soon: Monetary & Oil Price Shock Series by Aggregation Methods]

High-frequency identification, which derives exogenous variation in macroeconomic variables from financial market surprises during narrow time windows around policy announcements, has become an appealing approach for studying causal effects in macroeconomics. However, while these market surprises are measured at high frequencies, often daily, macroeconomic variables like output and inflation are typically measured at lower frequencies, either monthly or quarterly. Causal identification, therefore, often requires aggregating high-frequency surprises into lower-frequency measures. This paper demonstrates that, in the context of monetary policy and oil supply news shocks, the choice of aggregation method impacts the documented causal effects in a non-trivial manner. Specifically, using a data-driven aggregation approach, we find no evidence of an output puzzle (information channel) in a standard monthly monetary vector autoregression (VAR). Similarly, this aggregation method substantially alters the qualitative and quantitative results in a quarterly monetary VAR and changes the relevance of oil-supply news shocks for inventory dynamics when studying the oil-market dynamics.

Presented at: NBER-NSF Time Series Conference (2025; Rutgers University; Scheduled); Econometric Society World Congress (2025; Korea); North Carolina State University (2025; Raleigh NC), University of Pittsburgh (2025; Pittsburgh, PA), University of Guelph (2025; Canada), City University of New York (2025; New York, NY), SNDE (2025; San Antonio, Texas), George Washington University (2025; Washington D.C.), 12th Workshop on Empirical Monetary Economics (2024; France), Federal Reserve Board of Governors (2024; Washington D.C.), Princeton University (2024; Princeton, NJ), Rutgers University (2024; New Brunswick, NJ), 4th Dolomiti Macro Meetings (2024; Italy)

The Slope of the Yield Curve and Uncertainty News Shock   (2024)    (Submitted)
  [Draft]   [Data: Uncertainty News Shock Series]

Understanding the variation of the yield curve slope is important in economics and finance. A central question in the literature is: what drives the variation of the slope? The state-of-the-art explanation for yield curve slope changes relies on the anticipated changes in total factor productivity (TFP), often referred to as TFP news shocks (Kurmann & Otrok 2013 AER). However, recent findings challenge this view, showing that the relationship becomes weak, particularly when considering the updated TFP measure. This paper proposes an alternative explanation by demonstrating that a negative yield curve slope shock and a positive uncertainty news shock are essentially identical, providing a new perspective on understanding the variation of the slope of the term structure of interest rates. Additionally, this paper explores the unique features of uncertainty news shock and its relevance to the business cycle.

Presented at: MEG (2023; Federal Reserve Bank of Cleveland), MMM (2023; Texas Tech University), SNDE (2023; University of Central Florida), Texas A&M University (2023; College Station, TX)

Effects of a Marginal Tax Rate Shock: the Role of Nominal Wage Rigidity   (2025)    (Submitted)
  [Draft]   [Data: AMTR & IV Series by State]

The stickiness of nominal variables, particularly nominal wage rigidity, plays a crucial role in resource allocation, leading to labor market tightness or unemployment episodes. This paper shows that the macroeconomic effects of federal marginal income tax rate policy changes vary significantly with the degree of nominal wage rigidity. Using state-level variation in wage rigidity and narratively identified state-specific tax shocks, I find that macroeconomic variables respond nearly twice as strongly in states with higher wage rigidity and remain more persistent over time than those in states with more flexible wages. The findings highlight that nominal wage rigidity is a key determinant of the regional transmission of income tax policy.

Presented at: SNDE (2025; San Antonio, Texas), MMM (2022; Dallas, TX), Texas A&M University (2022; College Station, TX)

Demographics and Effects of Government R&D Spending   (2025)  
  [Draft]  

This paper examines the heterogeneity of the effects of government R&D spending across age groups. Using narratively identified exogenous changes in R&D appropriations and age-disaggregated unemployment rates, I find substantial heterogeneity. Following positive R&D shocks, unemployment rates fall most sharply for individuals aged 20–34, decline moderately for those aged 35–54, and show only weak or delayed responses for the 55+ group. These results indicate that government R&D shocks do not affect the labor market uniformly, but instead generate disproportionately larger gains for younger individuals. The evidence highlights the importance of demographic structure in shaping the impact of government R&D policy.