change can lead to an underestimation of the impact of monetary policy or even to a false acceptance of
the hypothesis that monetary policy has no impact on exchange rates. Second, we show that for all the 45
displayed lead-models (15 leads estimated separately for each of the three exchange rates in our sample),
no coefficient estimate associated with the surprise component of monetary policy changes appears
significant at the 95% significance level or higher and only one instance of significance at the 90% level
occurs. This absence of delayed effects strongly suggests that the exchange rates under study absorb
monetary policy surprises within the same day as the news are announced.
Comparing our findings to other studies, our findings appear at odds with two related and well-
known studies (neither of which focuses on expectations) of exchange rate responses to actual monetary
policy innovations. Using three measures of monetary policy and a VAR approach for analyzing monthly
data, Eichenbaum and Evans (1995) find that initial USD appreciation in response to a US monetary
contraction is small in comparison with subsequent USD appreciation and for the GBP/USD and the
JPY/USD exchange rates the initial response is insignificant. Similarly, Lewis (1995) uses a VAR
approach and biweekly data and finds no significant immediate reaction to (again, three measures of)
monetary policy for either the DEM/USD or the JPY/USD exchange rate. As shown in section 3 of our
study, it is necessary to disentangle the surprise component from the actual monetary policy change in
order to avoid arriving at incorrect conclusions regarding exchange rate responses to monetary policy. It
is indeed possible that both Eichenbaum and Evans (1995) and Lewis (1995) underestimate the initial
impact of monetary policy due to their focus on actual monetary policy changes rather than on monetary
policy surprises.
Furthermore, our findings appear at odds with Evans and Lyons (2005). Investigating daily
aggregates from an end-user microstructure data-set and employing VAR estimation techniques, they find
that news such as monetary policy surprises induce changes in end-user trading and that these changes
remain significant for several days. In other words, their findings imply that surprises matter but exchange
rates do not absorb news instantaneously. Evans and Lyons (2005) suggest that a possible explanation for
their finding of delayed effects is that in the case of non-financial corporations the “ultimate decision
17