Find probability density function D of random variable X s.t.
E[X]=0, Var[X]=(sigma)^2 and X<=c for all X, which minimize E[e^tX]
In fact, the answer would be
Prob[X=c]=(sigma)^2/[(sigma)^2+c^2]
Prob[X=-(sigma)^2/c]=c^2/[(sigma)^2+c^2]