Does the foreign sector help forecast domestic variables in DSGE models? |
| |
Authors: | Marcin Kolasa Micha? Rubaszek |
| |
Institution: | 1. Narodowy Bank Polski, ul. ?wi?tokrzyska 11/21, 00-919 Warsaw, Poland;2. SGH Warsaw School of Economics, Al. Niepodleg?o?ci 162, 02-554 Warsaw, Poland |
| |
Abstract: | This paper evaluates the forecasting performances of several small open-economy DSGE models relative to a closed-economy benchmark using a long span of data for Australia, Canada and the United Kingdom. We find that opening the model economy usually does not improve the quality of point and density forecasts for key domestic variables, and can even cause it to deteriorate. We show that this result can be attributed largely to an increase in the forecast error due to the more sophisticated structure of the extended setup, which is not compensated for by a better model specification. This claim is based on a Monte Carlo experiment in which an open-economy model fails to beat its closed-economy benchmark consistently even if the former is the true data generating process. |
| |
Keywords: | Forecasting New open economy macroeconomics Bayesian estimation |
本文献已被 ScienceDirect 等数据库收录! |
|