This page serves as a comprehensive repository of all my current and past academic research projects. For convenience, abstracts of these papers are available below. In my academic endeavours, I primarily concentrate on Bayesian model averaging techniques, non-parametric estimation, and vector autoregressive models.
Job Market Paper
Justine Lambert Prize 2024 (Department of Logic and Philosophy, University of California, Irvine)
Best Econometric Paper Award 2023 (Department of Economics, University of California, Irvine)
Abstract: The paper considers the efficient estimation of opinion pools in the Bayesian paradigm and extends their application to cases where the number of competing models exceeds the number of observations. An appropriate Bayesian formulation and estimation algorithm is proposed which allows the weights to shrink towards any possible combination. This flexibility makes the Bayesian opinion pool relevant for applications related to model averaging and model selection and improves stability compared to the ones estimated using scoring rules in a small sample setting. Results from a simulation study reveal that the proposed Bayesian opinion pool methodology improves prediction accuracy. An application involving the Survey of Professional Forecasters demonstrates that the Bayesian opinion pool’s inflation forecast competes well with the equal-weight aggregated inflation forecast published by the Federal Bank of Philadelphia. The application showcases the usefulness of the Bayesian solution in situations where optimization-based opinion pools fail.
Extending the Versatility of Non-Parametric VAR Models: Bayesian Specification, Estimation and Dynamic Factor.
Working Paper
Abstract: The paper introduces a non-parametric vector autoregressive model with dynamic factor (NPVAR-DF) through a hierarchical Bayesian approach. A framework is presented that aims at effectively capturing dynamic relationships among variables and enabling the incorporation of extensive information sets. The paper considers the specification, identification and estimation of the NPVAR-DF model, allowing the model to be efficiently fit by MCMC algorithms. Issues related to model comparison and extensions to settings with autocorrelated errors and qualitative variables are also considered. The NPVAR-DF model successfully identifies non-linear associations between macroeconomic variables by employing post-war US data. The dynamic factor captures the business cycle component, which aligns with officially declared recession periods.
with Ivan Jeliazkov
Work in Progress
Abstract: This article revisits the specification and estimation of factor-augmented vector autoregressive models from a hierarchical Bayesian perspective. The formulation emphasizes a dependence structure that allows for joint estimation but, in contrast, is typically ignored in certain traditional plug-in approaches. Efficient state space algorithms are presented for the joint estimation of the dynamic latent factors and model parameters, improving the performance of estimation algorithms as well as eliminating potentially controversial modelling choices or data pre-processing requirements.
with Naveen Srinivasan
Master's Thesis
Abstract: This paper empirically tests whether the inclusion of monetary aggregates in inflation forecasting models helps their forecasting ability or not. We have estimated the P-star model with Divisia M2, Divisia M3, simple sum M2, and simple sum M3 along with the Phillips curve and ARIMA specifications to forecast inflation for India from April 1994 to December 2016. We find that the inflation forecasting ability of both Divisia monetary aggregate and the simple sum monetary aggregates are similar. Though Divisia fits better than the simple sum from 1993-2013, the information contained in Divisia does not explain the behaviour of inflation post-2013.