In the data we find the following for several countries. When the domestic rate of interest exceeds the foreign interest rate, then on average the currency appreciates thereafter. This is true when the interest rates considered are short term, such as one year or less. This finding is a problem for the UIP, and is referred to as the ‘UIP puzzle’. When the term of the interest rates is 5 years or more (and the period of the exchange rate appreciation is suitably adjusted), then the data are more closely aligned with what the UIP predicts. (These findings are established in the following paper.)

In class, we discussed what happens when there is uncertainty about the persistence of the shock (the shock we have focused on is a monetary policy shock) that drives the exchange rate. In particular, if the normal shock hitting international financial markets is a temporary one, then when the underlying shock is permanent, it is possible for a rise in the interest rate to be associated with a sustained appreciation in the exchange rate thereafter. This possibility is introduced into a formal mathematical version of the model in this course, and its implications are explored. It is shown that the idea does not help resolve the UIP puzzle. The note in which this is done is attached and is not required for this course. It uses tools from statistics (time series analysis) and tools for solving expectational difference equations. If you are not familiar with these tools, you should not look at this note. The basic intuition for the result was discussed in class.