Previous work on information and financial markets has focused on a special
set of assumptions: agents have exponential utility, and random variables
are normally distributed. These assumptions are often necessary to obtain c
losed-form solutions. We present an example with alternative assumptions, a
nd demonstrate that some of the conclusions from previous literature fail t
o hold. In particular, we show that in our example, as more agents acquire
information, prices do not necessarily become more informative, and agents
may have greater incentive to acquire information. Learning can therefore b
e a strategic complement, allowing for the possibility of multiple equilibr
ia.