University of Southern California
Department of Economics
University Park Campus
Los Angeles, CA 90089
Institutional Affiliation: University of Southern California
NBER Working Papers and Publications
|December 2019||Forecasting with a Panel Tobit Model|
with Laura Liu, Frank Schorfheide: w26569
We use a dynamic panel Tobit model with heteroskedasticity to generate point, set, and density forecasts for a large cross-section of short time series of censored observations. Our fully Bayesian approach allows us to flexibly estimate the cross-sectional distribution of heterogeneous coefficients and then implicitly use this distribution as prior to construct Bayes forecasts for the individual time series. We construct set forecasts that explicitly target the average coverage probability for the cross-section. We present a novel application in which we forecast bank-level charge-off rates for credit card and residential real estate loans, comparing various versions of the panel Tobit model.
|September 2018||Forecasting with Dynamic Panel Data Models|
with Laura Liu, Frank Schorfheide: w25102
This paper considers the problem of forecasting a collection of short time series using cross sectional information in panel data. We construct point predictors using Tweedie's formula for the posterior mean of heterogeneous coefficients under a correlated random effects distribution. This formula utilizes cross-sectional information to transform the unit-specific (quasi) maximum likelihood estimator into an approximation of the posterior mean under a prior distribution that equals the population distribution of the random coefficients. We show that the risk of a predictor based on a non-parametric kernel estimate of the Tweedie correction is asymptotically equivalent to the risk of a predictor that treats the correlated-random-effects distribution as known (ratio-optimality). Our empiric...
|June 2011||Inference for VARs Identified with Sign Restrictions|
with Frank Schorfheide, Eleonora Granziera, Mihye Lee: w17140
There is a fast growing literature that partially identifies structural vector autoregressions (SVARs) by imposing sign restrictions on the responses of a subset of the endogenous variables to a particular structural shock (sign-restricted SVARs). To date, the methods that have been used are only justified from a Bayesian perspective. This paper develops methods of constructing error bands for impulse response functions of sign-restricted SVARs that are valid from a frequentist perspective. We also provide a comparison of frequentist and Bayesian error bands in the context of an empirical application - the former can be twice as wide as the latter.
Published: Eleonora Granziera & Hyungsik Roger Moon & Frank Schorfheide, 2018. "Inference for VARs identified with sign restrictions," Quantitative Economics, Econometric Society, vol. 9(3), pages 1087-1121, November. citation courtesy of
|April 2009||Bayesian and Frequentist Inference in Partially Identified Models|
with Frank Schorfheide: w14882
A large sample approximation of the posterior distribution of partially identified structural parameters is derived for models that can be indexed by a finite-dimensional reduced form parameter vector. It is used to analyze the differences between frequentist confidence sets and Bayesian credible sets in partially identified models. A key difference is that frequentist set estimates extend beyond the boundaries of the identified set (conditional on the estimated reduced form parameter), whereas Bayesian credible sets can asymptotically be located in the interior of the identified set. Our asymptotic approximations are illustrated in the context of simple moment inequality models and a numerical illustration for a two-player entry game is provided.
Published: Hyungsik Roger Moon & Frank Schorfheide, 2012. "Bayesian and Frequentist Inference in Partially Identified Models," Econometrica, Econometric Society, vol. 80(2), pages 755-782, 03. citation courtesy of