We identify a class of linearly constrained nonlinear optimization
problems with corner point optimal solutions. These include some
special polynomial fractional optimization problems with an objective function equal to the product of some power functions of positive linear functionals subtracting the sum of some power functions of positive linear functionals, divided by the sum of some power functions of positive linear functionals. The powers are required to be all positive integers, and the aggregate power of the product is required to be no larger than the lowest power in both of the two sums. The result has applications to some optimization problems under uncertainty, particularly in finance.