The first principal component of a panel of bilateral exchange rates outperforms the
random walk of individual monthly and weekly bilateral exchange rates since the extra
information from the other bilaterals must assist in prediction for each bilateral, given
the importance of the ‘home’ country effect. The benchmark for comparing structural
exchange rate models ought to be ramped up to involve bilateral panels. Will bilateral
panels be more informative for other large or small countries, for developing and developed
ones, and for commodity-exporting and -importing ones? We leave this question
to future research.