Research

Projects / Work in Progress

Business-Cycle Accounting in Production Networks, with Paulo Lins and Eugenio Flores.

We extend the business-cycle accounting (BCA) methodology to an economy with input-output production networks and propose a Bayesian sampler to estimate the wedges.

The Global Impacts of US Trade Policy Shocks, with Fábio Gomes and Fernando Barros Jr.

We develop a new large, order-invariant panel VAR with stochastic volatility in mean (SVM), in which one country can affect the rest of the world and shocks can be contemporaneously correlated.

The Global Transmission of U.S. Policy Uncertainty: Which Policy Matters, and for Whom?

I develop an order-invariant Global VAR with SVM and a dominant country—the United States—to assess the effects of three U.S. economic policy uncertainty shocks on AEs and EMDEs.

The heterogeneous impact of inflation uncertainty, with Leonardo Ferreira

We use an order-invariant panel VAR with SVM to measure the effects of inflation uncertainty across AEs and EMDEs, finding substantial effects in both groups.

Fiscal shocks and macroeconomic uncertainty: evidence from advanced and emerging countries
Debt-financed fiscal stimuli in advanced and emerging economies
Financial Conditions and Uncertainty: Lessons from a large dataset.

I use a large quarterly dataset covering AEs and EMDEs to examine whether uncertainty is endogenous or exogenous.

Working Papers

Development-Bank Disbursement Shocks and Regional Economic Dynamics: Evidence from Brazil (Submitted)

[Text], [Matlab Codes (soon)]
AbstractDevelopment banks are designed to support long-term investment, yet their lending may also be followed by changes in economic activity at much shorter horizons. We examine this possibility using monthly data for fourteen Brazilian states from 2011 to 2025 and a Bayesian panel VAR for disbursements by the Brazilian Development Bank (BNDES). Disbursement shocks are followed by an immediate increase in capital-goods imports and, within several months, by positive responses in regional activity, services, and formal employment. These broader responses are economically small: at a one-year horizon the identified shock explains less than one per cent of the forecast-error variance of each regional outcome, and they become much weaker when month effects absorb variation common across states. By contrast, the magnitude of the capital-goods-import response is stable across alternative specifications and identification schemes. The evidence therefore indicates that short-run responses to development-bank disbursement extend beyond investment-related expenditure, despite the institution's long-term mandate, but that these broader responses are modest and less robust than the observable equipment-related response.

Development-Bank Credit and Sectoral Investment: Shift-Share Evidence from Brazil (Submitted)

[Text], [Online Appendix], [Data], [R codes]

We estimate the response of sectoral investment to credit from the Brazilian Development Bank (BNDES). Combining administrative lending records with four structural business surveys, we construct a panel of 408 sectors over 2018--2023. We instrument credit with historical sectoral shares of direct BNDES lending interacted with aggregate direct disbursements excluding each sector's own borrowing. Controlling for pre-period sector size, the estimated investment elasticity is approximately 0.75, compared with 1.08 without this control. Alternative exposure windows and a per-worker specification yield similar estimates, and the Anderson--Rubin confidence interval excludes zero. The causal interpretation requires historical exposure to be conditionally unrelated to unobserved investment determinants. Instrument relevance is concentrated in a small number of sectors.

Efficient Equation-by-Equation Coefficient Sampling in Bayesian VARs with Time-Varying Covariances (Submitted)

[Text], [Online Appendix], [Matlab Codes (soon)], [Julia Codes (soon)]

AbstractWe develop an equation-by-equation coefficient sampler for Bayesian VARs with time-varying covariances by representing the Gaussian coefficient block as a systematic-scan Gibbs update. The representation applies to any conditionally known positive-definite precision sequence and covers triangular and unrestricted-impact stochastic volatility and stochastic-volatility-in-mean models with correlated innovations. In benchmarks, coefficient sampling is 4.9-7.5 times faster than the corrected triangular algorithm, while the cost per effective coefficient draw is over two orders of magnitude lower than with the Carter-Kohn algorithm.

Financial Conditions and Endogenous Uncertainty in Emerging Markets

[Text (soon)], [Matlab Codes - empirical analysis (soon)], [Julia Codes - theoretical model (soon)]
AbstractThis paper investigates the relationship between financial conditions, macroeconomic uncertainty, and business cycles in emerging markets. For a set of 15 emerging economies, using a nonlinear VAR model that allows shocks to affect second moments, we show that macroeconomic uncertainty increases after some macroeconomic shocks under financial stress regime but not in normal times. We then show that a model for a sudden stop-prone small open economy featuring a Fisherian debt-deflation mechanism and Epstein-Zin preferences helps to rationalize the empirical findings. If the economy is in a region sufficiently close to a binding constraint, the risk of a sudden stop induces a deterioration in the forecast accuracy, endogenously increasing macroeconomic volatility. However, when the economy is sufficiently far from a binding constraint, macroeconomic volatility is mainly driven by exogenous shocks.

Selected policy functions from the theoretical model can be found here.
Here are some plots from the empirical model showing the linkage between financial stress and uncertainty in EMs.
And here are the responses of macroeconomic uncertainty to selected shocks.

Fiscal Stance and Macroeconomic Volatility: Lessons From Brazil

[Text ] - [Online Appendix] - [Fortan Codes]
AbstractDoes the fiscal stance influence macroeconomic volatility in emerging markets? We address this question through the lens of a Bayesian Vector Autoregression model in which shocks to the level of the variables might also affect the variables' volatility. We examine the impact of fiscal policy shocks on macroeconomic uncertainty in Brazil, an emerging economy that has experienced several episodes of fiscal instability in recent years. Our findings reveal that an unanticipated deterioration in the government's primary fiscal balance is negatively related to industrial production, investment, and stock market performance, while being positively correlated with sovereign spreads. Notably, these shocks result in increased volatility across all examined variables, suggesting that a worsening fiscal stance leads to heightened macroeconomic uncertainty in Brazil. Our findings highlight the crucial role played by fiscal management in emerging economies.

Published Papers

In English

Flexible Markov-Switching Models With Evolving Regime-Specific Parameters: An Application To Brazilian Business Cycles (with Lívia Melo and Fábio Gomes, 2024 - AE)

[Text] - [WP version] - [Online Appendix] - [Matlab Codes]

A Panel Threshold VAR With Stochastic Volatility-In-Mean Model: An Application To The Effects Of Financial And Uncertainty Shocks In Emerging Economies (2023 - AE).

[Text], [WP Version], [Matlab Codes]

Geopolitical Risk Shocks and the Brazilian Economy (with Fernando Barros Jr. and Fábio Gomes, 2023 - AEL).

[Text], [Online Appendix], [Matlab Codes]

Government Spending Multipliers in Good Times and Bad Times: The Case of Emerging Markets (with Fabio Gomes and Sergio Sakurai, 2022, MD).

[Text], [Matlab Codes], [WP version]

The Role of Credit and Housing Shocks in Emerging Economies (with Fabio Gomes, 2021, AEL)

[Text], [WP version] [Online Appendix], [Matlab Codes]

Financial Conditions and the Business Cycles in Emerging Markets (2021, AEL).

[Text], [Online Appendix], [Matlab Codes]

International Drivers of Policy Uncertainty in Emerging Economies (2020, EB).

[Text], [Online Appendix], [Matlab Codes]

Convergence in income inequality: revisiting the case of Brazilian municipalities (with Fabio Gomes, 2019, EB).

[Text]

In Portuguese

Previsão do Consumo Agregado: O Papel dos Índice de Confiança do consumidor (com Patrícia Feline e Fábio Gomes, 2022, EE).

[Text]

Determinantes Robustos do Crescimento nos Estados Brasileiros: Uma Abordagem Bayesiana (com Fábio Gomes, 2021 - RBE).

[Text]

Desigualdade e Desenvolvimento: Revisitando a hipótese de Kuznets após a redução da desigualdade nos municípios brasileiros (com Fernando Barros Jr. e Fabio Gomes, 2019 - RBERU).

[Text]

Efeitos do gasto do governo sobre o consumo privado: Evidências de países desenvolvidos e em desenvolvimento (com Fabio Gomes e Sergio Sakurai, 2016, EE).

[Text]

In English

Page updated

Google Sites

Report abuse