International finance

.

Section 3. Currency Analysis

This part will include most of your statistical analysis. Depending on the currency model assigned to you, you will work only on the model that you are assigned.

The International Fisher Effect (IFE): As discussed in class, this model focuses on the difference between domestic and foreign interest rates. For this purpose, you will need monthly data on the short-term interest rates for the U.S. and the foreign market (try to find the shortest term interest rate possible for each country). Start with a plot of monthly ef,t against the interest rate differential (it – if,t), i.e. the short-term interest rate for the U.S. and the foreign market. You will then estimate

ef,t (it if,t)t

where it (if,t) is the interest rate for the U.S. and the foreign market in month t, respectively. As discussed in class, under the IFE model, the hypothesis to test is: H0: α=0 and β=1. If these two conditions hold, then you will conclude that the IFE holds.

Separately, you will test another hypothesis to see if the interest rate spread is statistically significant, i.e. H0: β=0. Doing so, you will test whether the interest rate spread has any explanatory power over exchange rate movements.

Forecasting with the IFE: You'll forecast St+1 using E(St1)St (1E[ef,t1])

where the expected percent change in the exchange rate is E (e f ,t 1 ) ( E[it 1 ] E[i f ,t 1 ]) . The

forecasts for interest rates for individual countries will be part of your research to find out. Again, forecasts of these variables might be available from the IMF or World Bank, so you will have to research it. In the worst case, you can use a random walk model to get forecasts for inflation rates.

the model, you will test the significance of each x-variable by testing H0: βk=0, k=1,2,3. Doing so, you will test whether the variable xk has any predictive power over exchange rate movements. Forecasting with the predictive model: Having estimated the best fitting predictive model, you will then forecast St+1 using