Working Paper: NBER ID: w31544
Authors: Oliver de Groot; Ceyhun Bora Durdu; Enrique G. Mendoza
Abstract: Global and local methods used to study open-economy incomplete-markets models yield different cyclical moments, impulse responses, spectral densities and precautionary savings. Endowment and RBC model solutions obtained with first-order, higher-order, and risky-steady-state local methods are compared with fixed-point-iteration global solutions. Analytic and numerical results show that the differences are due to the near-unit-root nature of net foreign assets under incomplete markets and inaccuracies of local methods in computing their autocorrelation. In a Sudden Stops model, quasi-linear methods that handle occasionally binding constraints understate the size of credit constraint multipliers, financial premia and macroeconomic responses.
Keywords: No keywords provided
JEL Codes: F41; F44; G01; G15
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| local approximation methods (C60) | accuracy of NFA dynamics (C69) |
| accuracy of NFA dynamics (C69) | consumption volatility (E20) |
| accuracy of NFA dynamics (C69) | net exports (NX) (F29) |
| local methods (C60) | autocorrelations of NFA (C45) |
| local methods (C60) | autocorrelations of net exports (NX) (F14) |
| raising the persistence of income (D15) | autocorrelation of NFA (C45) |
| raising the persistence of income (D15) | autocorrelation of net exports (NX) (F14) |
| local methods (C60) | precautionary savings (D14) |
| local methods (C60) | macroeconomic responses (E65) |
| stationarity-inducing assumptions in local methods (C51) | accuracy of solutions (C62) |