rates appreciate until approximately the forth to twelfth month and then
depreciates in our benchmark identification. The delay is slightly shorter for
Germany than for the UK, and considerably longer in the US-Japan case,
with most of the distribution centered around months 8 to 24. Our results
suggest that there is evidence of a delayed overshooting, in contrast to Kim
and Roubini (2000) and Faust and Rogers (2003), but that the delay is
considerably shorter than the three-year horizon found by Eichenbaum and
Evans (1995). The delay is a matter of months, not a matter of years.
It could be that the posterior distribution for the peak months gives
misleading results, if e.g. impulse responses showing mild appreciations peak
early and those with strong appreciations peak late. That this is not the
case can be seen from the posterior joint distributions of the peak and its
altitude both for the real and the nominal exchange rate in figure 6. The
joint distribution is shown both for the first five years as well as for the first
24 months. Also, the distribution of the beyond-impact change of the real
exchange rate q(k)-q(0) is shown. Apparently, the posterior is rather sharply
peaked for the US-Germany and the US-UK case. It is rather remarkable
that there is a sharp peak in the posterior distribution for the change of the
exchange rate compared to the on-impact response, i.e. for q(k) - q(0) for
the US-Germany and the US-UK pair, with greater diffusion in the US-Japan
case. After the on-impact change, the exchange rate drops by a further one
percent for US-Germany within 7 months and for the US-UK within 2 to 5
months. There is also considerable mass on this event in the US-Japan case,
but there is additional mass on early and mild as well as late and somewhat
stronger additional appreciations.
Figure 7 compares the impulse response functions for the three identifi-
cations we investigated, namely the two sign restriction identifications in the
EE and the BIG specification as well as the original recursive Eichenbaum-
Evans identification in the EE specification. There is only a minor difference
17
More intriguing information
1. Towards a Strategy for Improving Agricultural Inputs Markets in Africa2. CONSUMER ACCEPTANCE OF GENETICALLY MODIFIED FOODS
3. PEER-REVIEWED FINAL EDITED VERSION OF ARTICLE PRIOR TO PUBLICATION
4. On the estimation of hospital cost: the approach
5. Washington Irving and the Knickerbocker Group
6. The name is absent
7. ANTI-COMPETITIVE FINANCIAL CONTRACTING: THE DESIGN OF FINANCIAL CLAIMS.
8. Imputing Dairy Producers' Quota Discount Rate Using the Individual Export Milk Program in Quebec
9. The Impact of Optimal Tariffs and Taxes on Agglomeration
10. The Interest Rate-Exchange Rate Link in the Mexican Float