Working Paper: CEPR ID: DP9381
Authors: Fabio Canova; Filippo Ferroni; Christian Matthes
Abstract: We propose two methods to choose the variables to be used in the estimation of the structural parameters of a singular DSGE model. The first selects the vector of observables that optimizes parameter identification; the second the vector that minimizes the informational discrepancy between the singular and non-singular model. An application to a standard model is discussed and the estimation properties of different setups compared. Practical suggestions for applied researchers are provided.
Keywords: DSGE models; identification; representation; density ratio
JEL Codes: C10; E27; E32
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| certain combinations of observable variables (C39) | optimize parameter identification (C51) |
| choice of observables (C52) | identification of structural parameters (C51) |
| minimizing the informational discrepancy between singular and nonsingular models (C20) | accurate estimation (C13) |
| combination of output, consumption, and investment as observables (E20) | identify intertemporal and intratemporal links (D15) |
| using both inflation and interest rates jointly in estimation (C51) | worsens identification (Y50) |
| careful selection of variables (C39) | avoid significant information loss (L15) |
| the methods effectively rank variable combinations (C52) | improved estimation outcomes (C51) |