Boundedness of the value function of the worst-case portfolio selection problem with linear constraints
Financial Economics. FE. Высшая школа экономики, 2017. No. WP BRP 59/FE/2017.
We study the boundedness properties of the value function for a general worst-case scenario stochastic dynamic programming problem. For the portfolio selection problem,we present sufficient economically reasonable conditions for the finitness and uniform boundedness of the value function. The results can be used to decide if the problem is ill-posed and to correctly solve the Bellman-Isaacs equation with an appropriate numeric scheme.