Working Paper: NBER ID: w3955
Authors: Zhaohui Chen; Alberto Giovannini
Abstract: This paper develops a simple econometric procedure for estimating expected exchange rate under target zones. We employ the linear projection methodology to make predictions without relying on any prior structural or distributional assumptions, and at the same time demonstrate that such a methodology has to be modified in an important way to account for the presence of the fluctuation band. Our empirical results show that the band effect is nontrivial for narrow target zones such as the Bretton Woods system. We also develop a method to estimate the shapes of the unconditional distributions of exchange rates under target zones. The empirical results show that the unconditional distributions of exchange rates can take several different shapes, which may correspond to possibly widely different monetary and exchange-rate intervention policies. We also show how to use the projection equations and the information about the band to test the credibility of the exchange rate regimes.
Keywords: exchange rates; target zones; monetary policy
JEL Codes: F31; F33
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| fluctuation band (E32) | expected exchange rate (F31) |
| monetary intervention policies (E63) | expected exchange rate (F31) |
| band restrictions (L96) | estimation bias (C51) |
| monetary intervention policies (E63) | distribution of exchange rates (F31) |
| fluctuation band (E32) | distribution of exchange rates (F31) |
| modified projection methodology (C59) | unbiased estimates (C51) |