Working Papers

The current research is supported by NSF grant SES 1424843

Heterogeneity and Aggregate Fluctuations
Abstract: We specify a vector autoregressive model that stacks macroeconomic aggregates and cross-sectional distributions to provide semi-structural evidence about the interaction of aggregate and distributional dynamics. The specification of our functional VAR is motivated by a linearization of a reduced-form model in which dynamics of aggregate variables and a function of the lagged cross-sectional distribution of individual-level decisions or states, and the units (households or firms) base their decisions on lagged macroeconomic aggregates and lagged cross-sectional distributions. To make the functional VAR analysis tractable, we approximate the log-densities of the cross-sectional distributions as well as the transition kernels in the functional VARs by sieves. We apply our techniques to study the dynamics of technology shocks, per capita GDP, employment, and the earnings distribution.
Joint with Minsu Chang (Penn) and Xiaohong Chen (Yale)

Forecasting With a Panel Tobit Model
Joint with Laura Liu (Board of Governors) and Hyungsik Roger Moon (USC)
(Very) Preliminary Version: March 2018

On the Comparison of Interval Forecasts
Joint with Ross Askanazi (Cornerstone), Frank Diebold (Penn), and Minchul Shin (UIUC)
This Version: January 2018

Forecasting with Dynamic Panel Data Models
Joint with Laura Liu (UPenn) and Hyungsik Roger Moon (USC)
This Version: December 2016
Also available at arXiv Link and as PIER Working Paper 16-022