We consider a class of stochastic mathematical programs
with equilibrium constraints (SMPECs), in which all decisions are
required to be made here-and-now, before a random event is
observed. We show that many problems can be formulated as this
kind of SMPECs. In general, SMPECs are more difficult to deal with
than MPECs. In order to develop an effective approach, we first
give a number of reformulations of the problem and then, based on
these reformulations, we propose some smoothed penalty methods for
solving SMPECs. A comprehensive convergence theory is also
included.