Working papers and ongoing work:

5. Brito, Diego S., Marcelo C. Medeiros and Ruy M. Ribeiro (2018). Forecasting Large Realized Covariance Matrices: The Benefits of Factor Models and Shrinkage. (PDF version)
We propose a model to forecast very large realized covariance matrices of returns, applying it to the constituents of the S\&P 500 on a daily basis. To deal with the curse of dimensionality, we decompose the return covariance matrix using standard firm-level factors (e.g. size, value, profitability) and use sectoral restrictions in the residual covariance matrix. This restricted model is then estimated using Vector Heterogeneous Autoregressive (VHAR) models estimated with the Least Absolute Shrinkage and Selection Operator (LASSO). Our methodology improves forecasting precision relative to standard benchmarks and leads to better estimates of the minimum variance portfolios.   
Keywords: Realized covariance, factor models, shrinkage, Lasso, forecasting, portfolio allocation, big data.

4. Medeiros, Marcelo C., Gabriel F. Vasconcelos, Alvaro Veiga and Eduardo Zilberman (2018). Forecasting Inflation in a Data-Rich Environment: The Benefits of Machine Learning Methods. (PDF version)
Inflation forecasting is an important but difficult task. In this paper we explore the advances in Machine Learning (ML) methods and the availability of new and rich datasets to forecast US inflation over a long period of out-of-sample observations. Despite the skepticism of the previous literature, we show that ML models with a large number of covariates are systematically more accurate than the benchmarks for several forecasting horizons both in the 1990s and the 2000s. The ML method that deserves more attention is the Random Forest, which dominated all other models in several cases. The good performance of the Random Forest is due not only to its specific method of variable selection but also the potential nonlinearities between past key macroeconomic variables and inflation. The results are robust to inflation measures, different samples, levels of macroeconomic uncertainty, and periods of recession or expansion.
Keywords: Big data, inflation forecasting, shrinkage, factor models, LASSO, random forests, machine learning.

3. Carvalho, Carlos V., Ricardo P. Masini and Marcelo C. Medeiros (2016). The Perils of Counterfactual Analysis with Integrated Processes. (PDF version)
Recently, there has been a growing interest in developing econometric tools to conduct counterfactual analysis with aggregate data when a "treated" unit suffers an intervention, such as a policy change, and there is no obvious control group. Usually, the proposed methods are based on the construction of an artificial counterfactual from a pool of "untreated" peers, organized in a panel data structure. In this paper, we investigate the consequences of applying such methodologies when the data are formed by integrated process of order 1. We find that without a cointegration relation (spurious case) the intervention estimator diverges resulting in the rejection of the hypothesis of no intervention effect regardless of its existence. Whereas, for the case when at least one cointegration relation exists, we have a $\sqrt{T}$-consistent estimator for the intervention effect albeit with a non-standard distribution. However, even in this case, the test of no intervention effect is extremely oversized if nonstationarity is ignored. When a drift is present in the data generating processes, the estimator for both cases (cointegrated and spurious) either diverges or is not well defined asymptotically. As a final recommendation we suggest to work in first-differences to avoid spurious results.
Keywords: counterfactual analysis, comparative studies, panel data, ArCo, synthetic control, policy evaluation, intervention, cointegration, factor models, spurious regression, nonstationarity.

2. Lichand, Guilherme, Marcos Lopes and Marcelo C. Medeiros (2016). Is Corruption Good for your Health? (PDF Version)
While corruption crackdowns have been shown to effectively reduce missing government expenditures, their effects on public service delivery have not been credibly documented. This matters because, if corruption generates incentives for bureaucrats to deliver those services, then deterring it might actually hurt downstream outcomes. This paper exploits variation from an anti-corruption program in Brazil, designed by the federal government to enforce guidelines on earmarked transfers to municipalities, to study this question. Combining random audits with a differences-in-differences strategy, we find that the anti-corruption program greatly reduced occurrences of over-invoicing and off-the-record payments, and of procurement manipulation within health transfers. However, health indicators, such as hospital beds and immunization coverage, became worse as a result. Evidence from audited amounts suggests that lower corruption came at a high cost: after the program, public spending fell by so much that corruption per dollar spent actually increased. These findings are consistent with those responsible for procurement dramatically reducing purchases after the program, either because they no longer can capture rents, or because they are afraid of being punished for procurement mistakes.

1. Bonomo, Marco, Arnildo Correa and Marcelo C. Medeiros (2016). Estimating Strategic Complementarity in a State-Dependent Pricing Model. (PDF Version)
The macroeconomic effects of shocks in models of nominal rigidity depend crucially on the degree of strategic complementarity among price setters. However, the empirical evidence on its magnitude is indirect - because its based on complete models simulations - and ambiguous. Sources of stragetic complementarities have been divided between "micro" or "macro" type. While sources of the micro type have been found to be at odds with microeconomic evidence because they imply implausibly large variability of microeconomic shocks, sources of the macro type do not face such a hurdle. In this paper we estimate directly the degree of strategic complementarity based on individual price data underlying the CPI-FGV from Brazil during the 1996-2006 period, benefiting from large amount of macroeconomic variation in Brazilian sample. Our identification strategy is to infer the degree of strategic complementarity from the relation between the frictionless optimal price and its macroeconomic determinants. We assume that firms follow an asymmetric Ss pricing rule, which allows us to relate the probability of adjustment to the change in the frictionless optimal price since the last adjustment date. This approach allows us to directly estimate the degree of strategic complementarity from the occurrence of price adjustments. The results - which are based on individual price changes and not on macro effects - indicate a substantial degree of strategic complementarity, contributing to reconcile micro and macro based evidence.