This paper presents a stochastic linear programming formulation of a firm's short term financial planning problem. This framework allows a more realistic representation of the uncertainties fundamental to this problem than previous models. In addition, using Wets's algorithm for linear simple recourse problems, this formulation has approximately the same computational complexity as the mean approximation (i.e., the deterministic program obtained by replacing all random elements by their means). Using this formulation we empirically investigate the effects of differing distributions and penalty costs. We conclude that even with symmetric penalty costs and distributions the mean model is significantly inferior to the stochastic linear programming formulation. Thus we are able to demonstrate that ignoring the stochastic components in linear programming formulations can be very costly without having significant computational savings.
Stored only when both papers are in the corpus — 114 citations exist upstream.
Imported from OpenAlex (CC0) — nothing here is generated.