Estimation and Inference under Weak Identification and Persistence: An Application to Forecast-Based Monetary Policy Reaction Function

Date

2014-08-05

Journal Title

Journal ISSN

Volume Title

Publisher

Abstract

The reaction coefficients of expected inflations and output gaps in the forecast-based monetary policy reaction function may be merely weakly identified when the smoothing coefficient is close to unity, i.e., the nominal interest rates are highly persistent. Using asymptotic theories for near unit root processes and novel drifting sequence approaches, we modify the method of Andrews and Cheng (2012, Econometrica) on inference under weak identification to accommodate the persistence issue. Large sample properties with a desired smooth transition with respect to the true values of parameters are developed for the nonlinear least squares (NLS) estimator and its corresponding t and Wald statistics of a general class of models.

Despite the not-consistent-estimability when the smoothing coefficient is close to unity, the conservative confidence sets of weakly-identified parameters of interest can be obtained by inverting the t or the Wald tests. We show that the null-imposed least-favorable confidence sets will have correct asymptotic sizes while the projection- based and Bonferroni-based methods may lead to asymptotic over-coverage. An identification-category-selection procedure is proposed to select between the standard confidence set and the conservative one under weak identification. Our empirical application suggests that for the model in which the expected inflations and output gaps have a forecast horizon zero, the NLS estimates for the reaction coefficients in U.S.'s forecast-based monetary policy reaction function for 1987:3{2007:4 are not accurate sufficiently to rule out the possibility of indeterminacy. However, for the model with forecast horizon one, the possibility of indeterminacy may be ruled out.

Description

Citation