3. Exchange Rate Responses to Actual Monetary Policy Changes
When focusing on exchange rate responses to actual monetary policy changes, the asset-pricing approach
to exchange rate determination suggests that an exchange rate should react to only the surprise component
of such monetary policy news and it should do so “instantaneously”.8
We formally test whether only the surprise element of a monetary policy change is systematically
associated with exchange rate responses by incorporating the Kuttner-decomposition of actual monetary
policy changes into an expected and an unexpected (surprise) component. Using an event study approach
in the tradition of Cook and Hahn (1989), we regress the change in the (log of the) spot exchange rate on
the expected and the unexpected component of the target rate change:
(1) ∆st =α+β1r~te +β2~rtu +CZt +εt
where ∆st is the first-difference in the log of the daily spot exchange rate and r~te (r~tu ) is the expected
(unexpected) target rate change in percentage points, and C is the coefficient vector associated with the
control variables contained in Zt.
The control variable matrix Zt contains the unexpected component of macroeconomic news on
days when an official macroeconomic announcement coincides with a monetary policy change.
Specifically, we control for the types of macroeconomic news that Bonser-Neal and Tanner (1996) and,
more recently, Galati, Melick and Micu (2005) have found to impact exchange rates. We do so by
incorporating several variables that capture the surprise element of US macroeconomic news and policy
developments regarding CPI (CPI-UNEXP), PPI (PPI-UNEXP), industrial production (IP-UNEXP), the
unemployment rate (UNEM-UNEXP) and the trade balance (TRDE-UNEXP).9
8 See Engel and West (2005) for a recent discussion of the asset-market approach to exchange rate determination.
9 Since central bank foreign exchange intervention conducted by central banks may impact day-to-day changes in
the spot exchange rates as well, we also control for the effects of such actions by either the Fed or the foreign central
bank. Similarly, we control for actual monetary policy changes by foreign central banks. However, while we are
able to capture the surprise element of the control variables contained in Zt we do not have corresponding measures