Fabio Gómez-Rodríguez

Develops an endogenous regime-switching SVAR to model nonlinear exchange rate pass-through dynamics in a small open economy, allowing regime transitions to depend on macroeconomic conditions.

Develops a structural framework to decompose yield curve movements into expectations and term-premium components, identifying fiscal and monetary shocks and quantifying their differential effects across maturities.

Introduces a macro-accessible functional time series framework to analyze how structural shocks affect the entire distribution of inflation expectations, not just the mean, highlighting heterogeneous responses across households.

Decomposes the term structure of inflation expectations into permanent and transitory components using functional principal components and structural identification, linking movements to macroeconomic shocks.

Studies how systematic monetary policy and exchange rate regimes shape exchange rate pass-through to inflation, using a structural VAR approach across Latin American economies.

This article proposes a structural vector autoregression (SVAR) to study the forces driving inflation in Costa Rica, a small open economy. We estimate the model using Bayesian methods and combine several identification strategies. The resulting historical decomposition attributes inflation dynamics to structural drivers, including supply, demand, exchange-rate, and monetary-policy shocks, classified according to their global or domestic origin. The historical decomposition attributes the inflation surge from 2021Q4 to 2023Q1 primarily to global factors. The subsequent disinflation from 2023Q2 to 2024Q1 reflected both a decline in global inflationary pressures and additional downward pressure from domestic factors. Domestic monetary-policy shocks contributed little to realized inflation, averaging about +0.07 percentage points per quarter. Long-run expectations remained considerably more stable than headline inflation, while the model-implied inflation target displayed a gradual downward drift and its associated shock contributed to the initial disinflation. Taken together, these results identify relative resilience, but not perfect alignment, of the perceived nominal anchor and highlight the implicit inflation target as a useful indicator for evaluating Costa Rica’s inflation-targeting regime.

How do customer transactions and central-bank operations affect Costa Rica’s exchange rate? We study this question using daily data from MONEX and a Bayesian structural model that separates stabilization operations from other BCCR market operations. The distinction matters because an unexpected dollar purchase and the bank’s regular response to exchange-rate movements can have different effects. In the reference model, switching off the same-day stabilization response raises the uncertainty in predicting daily exchange-rate changes five trading days ahead by 1.40 percent, with a 90 percent credible interval of 0.93–1.94 percent. The direction of this result holds across the alternatives examined, but its size increases substantially when the model allows volatility to change over time. An analysis of 2022–2023 shows that several forces contributed to the exchange-rate path, often in opposite directions. A complementary monthly model suggests that the transmission of exchange-rate movements to consumer prices depends on the shock behind those movements. Taken together, the results support a stabilizing role for the BCCR response represented in the model, while showing that the size of the estimated effect depends on the assumptions used to describe market volatility.

Develops unit root tests for functional autoregressive processes in Hilbert space, providing theoretical foundations and simulation evidence for persistence analysis in functional macroeconomic data.

Economic uncertainty is measured using a wide range of news-based, financial-market, survey, geopolitical, policy, and topic-specific indices. Because these measures differ substantially in construction and interpretation, it remains unclear whether they capture common underlying forces or distinct forms of uncertainty. We develop a Bayesian stochastic-volatility factor model for a large, harmonized panel of U.S. uncertainty measures. The model decomposes each observed series into exposure to latent uncertainty factors, observed macroeconomic conditions, and persistent series-specific variation. The latent factors and macroeconomic variables evolve jointly in a structural vector autoregression with time-varying shock volatility, allowing both the levels and volatilities of common uncertainty components to change over time. This framework provides a unified method for identifying the principal dimensions of uncertainty, comparing the information contained in different measures, and studying how structural monetary, fiscal, and productivity shocks affect common uncertainty factors. Our preliminary analysis applies the model to a quarterly panel of U.S. uncertainty indicators and evaluates the extent to which their comovement can be summarized by a small number of dynamic latent factors.