fertile ground for further study. First, anecdotal studies suggest the impact of media on
asset prices could be enormous, with the publication of single articles causing prices to rise
or fall by factors of three to six. But the large-scale evidence from studies using instru-
ments for exogenous changes in media reporting reveals impacts that are smaller by an
order of magnitude. One can reconcile these facts by arguing either the anecdotes are
unusual or the instruments are weak. Future research should determine the merits of
these explanations.
Second, an especially important and unsettling finding in the literature is the weak
link between information arrival and price movement. Part of the explanation seems
to be that market prices underreact