Not logged in.
Quick Search - Contribution
Contribution Details
Type | Conference Presentation |
Scope | Discipline-based scholarship |
Title | Solving Principal Agent Problems by Polynomial Programming |
Organization Unit | |
Authors |
|
Presentation Type | paper |
Item Subtype | Original Work |
Refereed | No |
Status | Published in final form |
Language |
|
Event Title | OR 2011 |
Event Type | conference |
Event Location | Zurich |
Event Start Date | August 29 - 2011 |
Event End Date | September 2 - 2011 |
Abstract Text | We present a new way to solve principal agent problems by polynomial programming techniques. We study the case where the agent's actions are unobservable by the principal but the outcomes are. We assume that the agent's actions lie in an interval and the space of outcomes is a finite set. Furthermore the agent's expected utility is a rational function in his actions. The resulting problem is a bilevel optimization problem with the principal's problem as the upper and the agent's problem as the lower level. The key idea is to find an exact reformulation of the agent's problem as a semidefinite optimization problem. Since this is a convex optimization problem, we then have necessary and sufficient global optimality conditions for the agent's problem. The reformulation can be done by using classical results from real algebraic geometry linking positive polynomials and semidefinite matrices. We obtain a nonlinear program. If all functions are rational functions, we then can solve it to global optimality. |
Export | BibTeX |