A note on sensitivity of value functions of mathematical programs with complementarity constraints (Q1396215)

In the introduction to this note, the authors give an ample review of recent results on mathematical programs with complementarity constraints (MPCC). They consider parametric MPCCs. Their main result is that under a linear independence condition on active gradients (LICQ), standard nonlinear programming (NLP) sensitivity analysis yields sensitivity results for the optimal value function of the MPCC as a natural consequence, despite the combinatorial properties of the MPCC feasible set. The authors also show that unlike NLP, second order directional derivatives of the MPCC optimal value function show combinatorial structure. Finally, an example is included which shows that the global optimal solution set of a parametric MPCC need not have a continuous selection.