makers” are not the people who are in charge of continuous monitoring of markets. Therefore, argues
Evans and Lyons (2005), these corporations will not respond to news until the time of their, say, weekly
“currency strategy meeting” when the ultimate decision makers are present. If this description of an
important institutional aspect of decision making structures is accurate, it would indeed induce response
lags to news. However, delayed responses to news are costly and the argument offered by Evans and
Lyons (2005) does not explain why corporations with less frequent strategy meetings are not over time
driven out of the market by corporations with more frequent strategy meetings.
Our findings appear in line with Bonser-Neal, Roley and Sellon, Jr. (1998). They use an event
study approach and the Fed funds target rate as a measure of monetary policy actions and show that
exchange rates generally respond immediately to changes in US monetary policy. Their work, however,
does not focus on expectations. As such, their findings may have some resemblance to ours when we
analyze the GBP/USD exchange rate. Even without isolating the monetary policy surprise component we
show that monetary policy changes are associated with same-day GBP/USD exchange rate changes.
However, we also show that without disentangling the surprise from the actual change it is possible that
the impact of a monetary policy change is underestimated.
Despite our study being different due to, in particular, our use of separate measures of the surprise
component, the expected component, and the news announcement itself and, furthermore, our use of a
market based measure of expectations, our findings seem consistent with the high-frequency analysis by
Andersen, Bollerslev, Diebold and Vega (2003) and the daily data analysis by Simpson, Ramchander and
Chaudry (2005). Both of these papers focus on the exchange rate responses to only the surprise
component of news and both use survey data for measuring expectations. In the context of a two-stage
weighted least squares time-series analysis and, subsequently, an event study analysis, Andersen,
Bollerslev, Diebold and Vega (2003) show that exchange rates generally respond instantaneously
(characterized by a jump immediately following the announcement and little movement thereafter) to
news such as Fed fund target rate changes. Simpson, Ramchander and Chaudry (2005) use a VECM
18