12.6 Conclusions

We use macroeconomic panel data from 16 countries to estimate a model that enables us to find evidence for the UIP. Interestingly enough, we find strong evidence for UIP when we use long-term interest rate differentials. Moreover, the corresponding short-term interest rate coefficient turns out to be zero. We conclude that UIP does not hold in the short-run. Since the static fixed effects model may be inefficient and biased because of significant first order autocorrelation, we specify a dynamic panel data model, which is in turn estimated by GMM. Results from the dynamic model are in line with what has been found before from the static model.