Contemporary press frequently attaches significance with specific numerical values in exchange rates. This research empirically tests for the presence of clustering and “psychological barriers” (transgressional effects) on various exchange rates involving the Australian dollar. Overall, we find a widespread clustering effect, which accordingly means there is partial information content in the actual numbers of the exchange rates themselves. Furthermore, there is some, but not strong, evidence that perceived “psychological barriers” do exist. Surprisingly, the location and form of both the clustering and transgressional effects and hence the information content differ across different exchange rates and in most instances, they are not in the expected direction.
|Journal||Journal of International Financial Markets, Institutions and Money|
|Publication status||Published - 2006|