Optimal control systems in managerial accounting

This thesis analyzes two important control systems in managerial accounting practice, cost allocation and transfer pricing. First, I provide a rationale of why a firm might allocate corporate service costs to the division manager who has no control over them using a principal-agent model. In my model, the division's revenue is affected by the quality of the corporate service as well as the division manager's effort level. Responsibility accounting, an important management control principle, appears to imply that this noncontrollable corporate service cost should not be allocated to the division manager. In contrary, we observe that many firms allocate this noncontrollable corporate service cost to the division manager. Interestingly, it is found that the optimal incentive contract to the division manager is increasing in revenue but decreasing in the corporate service cost. In equilibrium, the revenue measure is used to provide a proper incentive to the division manager whereas the noncontrollable corporate service cost is used to provide a proper incentive to the headquarters when the headquarters is also subject to moral hazard. Second, this thesis investigates properties of the optimal allocation of noncontrollable cost and derives empirically testable hypothesis. I find that under-allocation of noncontrollable corporate service cost is optimal because allocation imposes unnecessary risk on the division manager. I also find that less service cost is allocated as the variance of the service cost increases. This result might provide some insight into explaining why we observe that the degree of allocation substantially varies among different kinds of corporate service costs. Moreover, it is found that less service cost is allocated as marginal productivity of the corporate service increases. Third, this thesis analyzes an important and unresolved topic in management accounting, transfer pricing. Most analytical papers on transfer pricing have limited their studies on the characteristics of an optimal transfer pricing scheme in a given organizational structure and failed to answer the fundamental question; Is a transfer pricing valuable? I provide a different view of transfer pricing and identifies view of transfer pricing in a multiple-agent model. I interpret transfer pricing as a contract between two divisions which are related through production flow. I show the conditions under which the firm is made strictly better-off when a transfer pricing scheme is implemented.