I am a fifth-year Ph.D. candidate in Economics at the University of Pennsylvania.
My research interests include time series econometrics, Bayesian analysis, panel data models, applied macroeconomics, and machine learning. Most of my work focus on econometric forecasting. My recent projects are focused on grouped heterogeneity in dynamic panel data models. In particular, I’m interested in bring prior knowledge on the latent group structure into the model and explore the predictive gain over prevailing panel data estimators. It is an exciting topic that involves panel data model estimation and semi-supervised learning.
Ph.D. in Economics, 2017 - 2023 (Expected)
University of Pennsylvania
M.S. in Statistics, 2016
University of Illinois at Urbana-Champaign
B.S. in Physics, 2014
Renmin University of China
The diminishing extent of Arctic sea ice is a key indicator of climate change as well as an accelerant for future global warming. Since 1978, Arctic sea ice has been measured using satellite-based microwave sensing; however, different measures of Arctic sea ice extent have been made available based on differing algorithmic transformations of the raw satellite data. We propose and estimate a dynamic factor model that combines four of these measures in an optimal way that accounts for their differing volatility and cross-correlations. We then use the Kalman smoother to extract an optimal combined measure of Arctic sea ice extent. It turns out that almost all weight is put on the NSIDC Sea Ice Index, confirming and enhancing confidence in the Sea Ice Index and the NASA Team algorithm on which it is based.
We leverage a data rich environment to construct and study a measure of macroeconomic uncertainty for the Korean economy. We provide several stylized facts about uncertainty in Korea from 1991M10–2016M5. We compare and contrast this measure of uncertainty with two other popular uncertainty proxies, financial and policy uncertainty proxies, as well as the U.S. measure constructed by Jurado et al. (2015). We find that neither financial nor policy uncertainty proxies capture economy-wide uncertainty. Unlike our measure or financial uncertainty, policy uncertainty does not have much effect on real variables in Korea.
We propose methods for constructing regularized mixtures of density forecasts. We explore a variety of objectives and regularization penalties, and we use them in a substantive exploration of Eurozone inflation and real interest rate density forecasts. All individual inflation forecasters (even the ex post best forecaster) are outperformed by our regularized mixtures. The log scores of the Simplex and Best-Average mixtures, for example, are approximately 7% better than that of the ex post best individual forecaster, and 15% better than that of the median forecaster. From the Great Recession onward, the optimal regularization tends to move density forecasts’ probability mass from the centers to the tails, correcting for overconfidence.
In this paper, we estimate and leverage latent constant group structure to generate the point, set, and density forecasts for short dynamic panel data. We implement a nonparametric Bayesian approach to simultaneously identify coefficients and group membership in the random effects which are heterogeneous across groups but fixed within a group. This method allows us to flexibly incorporate subjective prior knowledge on the group structure that potentially improves the predictive accuracy. In Monte Carlo experiments, we demonstrate that our Bayesian grouped random effects (BGRE) estimators produce accurate estimates and score predictive gains over standard panel data estimators. With a data-driven group structure, the BGRE estimators exhibit comparable accuracy of clustering with the Kmeans algorithm and outperform a two-step Bayesian grouped estimator whose group structure relies on Kmeans. In the empirical analysis, we apply our method to forecast the investment rate across a broad range of firms and illustrate that the estimated latent group structure improves forecasts relative to standard panel data estimators.