Strategic Bidding of Private Information for Dynamic LQ Networks under Moral Hazard

Abstract

This paper investigates the whole system behavior caused by the influence of the agents’ strategic behavior while utilizing their individual control and private information for a dynamic linear-quadratic (LQ) network in the presence of a principal-agent relationship. The principal aims at integrating agents’ behavior into the socially optimal one based on private information bid by the agents. To avoid a moral hazard on agents’ controls, the principal must give a reward to the agents. The reward induces the agents to choose their controls achieving the social objective under the true private information case, but the reward cannot prevent the strategic bidding of the agents’ private information. Under this situation, the case is considered that all the agents minimize their net cost composed of their own private cost and the reward from the principal, which is called the strategic bidding problem under moral hazard. Then, the strategic bidding problem is formulated and the optimal design of the problem is analytically derived. Their effectiveness and limitations are also discussed through a simulation.