Working Paper: CEPR ID: DP9858
Authors: Anindya Banerjee; Massimiliano Marcellino; Igor Masten
Abstract: Starting from the dynamic factor model for non-stationary data we derive the factor-augmented error correction model (FECM) and, by generalizing the Granger representation theorem, its moving-average representation. The latter is used for the identification of structural shocks and their propagation mechanism. Besides discussing contemporaneous restrictions along the lines of Bernanke et al. (2005), we show how to implement classical identification schemes based on long-run restrictions in the case of large panels. The importance of the error-correction mechanism for impulse response analysis is analysed by means of both empirical examples and simulation experiments. Our results show that the bias in estimated impulse responses in a FAVAR model is positively related to the strength of the error-correction mechanism and the cross-section dimension of the panel. We observe empirically in a large panel of US data that these features have a substantial effect on the responses of several variables to the identified real shock.
Keywords: Cointegration; Dynamic Factor Models; Factor-Augmented Error Correction Models; FAVAR; Structural Analysis
JEL Codes: C32; E17
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Omission of the error correction term in FAVAR models (C22) | biased impulse responses (E71) |
| Bias in estimated impulse responses in a FAVAR model (C51) | strength of the error correction mechanism (C20) |
| Bias in estimated impulse responses in a FAVAR model (C51) | cross-section dimension of the panel (C23) |
| Inclusion of ECM terms (E01) | enhances the accuracy of responses to identified real shocks (E13) |
| FECM can be approximated by a FAVAR with a large lag order (C22) | introduces estimation uncertainty (C51) |
| Estimation uncertainty (C13) | distorts the impulse response analysis (C22) |
| Responses of several variables to identified monetary policy shocks (E39) | differ significantly when comparing results from FECM and FAVAR models (C22) |
| FECM framework (E17) | yields impulse responses that align more closely with economic theory and observed macroeconomic behavior (E13) |
| FAVAR models that neglect the error correction mechanism (C22) | produce impulse responses that are less accurate (C45) |