INCENTIVE PROPERTIES OF RESIDUAL INCOME WHEN THERE IS AN OPTION TO WAIT
Gunther Friedl
Performance measures based on residual income are increasingly popular. The academic literature shows that residual income has important incentive properties when management bases investment decisions on the net present value (NPV) rule. My analysis focuses on the case in which investment decisions can be postponed, when management must extend the simple NPV rule by considering an option value. My analysis shows that some important incentive properties of residual income still hold when there is an option to wait, but only when the residual income measure is correctly adjusted. I also provide an incentive-based explanation of why the capital charge rate within firms is often significantly higher than the firm's cost of capital.
pp. 3-21
DISCUSSION OF "INCENTIVE PROPERTIES OF RESIDUAL INCOME WHEN THERE IS AN OPTION TO WAIT"
Robert F. Göx
In recent years, the incentive properties of residual income-based performance evaluation systems have become one of the most popular topics in the field of analytical accounting research. The typical model in this area considers a manager who is in charge of making investment decisions on behalf of the firm owners. The owners delegate the investment decision because the manager is privately informed about the project returns. An incentive conflict arises because the manager is assumed to be impatient, i.e., he discounts future profits at a higher rate than would the owners, or he has the option to leave the firm before the investment project ends. Following the papers of Rogerson (1997) and Reichelstein (1997), the incentive conflict can be solved if the manager is compensated on the basis of residual income and the firm calculates the depreciation charges according to the so called relative benefit depreciation rule1.
pp. 22-28
THE IMPACT OF TRUTH-TELLING AND LIMITED WEALTH IN A PRINCIPAL-AGENT MODEL
Eberhard Feess/Sonja Ossig/Markus Walzl
We derive the optimal contract in a model in which the agent's effort is unobservable and his findings are private information. Our focus is on the impact of the agent's wealth constraint. We show that three regions need to be distinguished: if the agent's wealth is above a critical threshold, then the principal implements a first-best effort without rent. In an intermediate wealth region, the principal implements a suboptimally low effort, but does not pay a rent. In the low wealth region, the effort is independent of the agent's wealth and the principal pays a rent. Surprisingly, the principal's profit does not necessarily increase with the percentage of good projects, and the equilibrium effort does not necessarily decrease with the percentage of good projects.
pp. 29-45
DISCUSSION OF "THE IMPACT OF TRUTH-TELLING AND LIMITED WEALTH IN A PRINCIPAL-AGENT MODEL"
Ralf Ewert
The paper by Feess/Ossig/Walzl (FOW) deals with potential interdependencies between information acquisition and a subsequent investment decision that is subject to the agent's wealth constraints. Depending on the amount of the agent's wealth, the authors derive three settings with different structures for the optimal contract, and several relationships that define how the principal's utility and the agent's informational effort depend on the percentage of good projects. Since the paper is written in a very technical style, I focus on presenting a description of the scenario and a more intuitive explanation for the three settings, which in my opinion is the main insight from the paper. I also discuss several limitations of the model that indicate the need for additional research.
pp. 46-54
SCHEDULING INFORMATIONAL ACTIVITIES AND THE PREFERENCE FOR EARLY INFORMATION IN DECISION AND CONTROL PROBLEMS
Anne Chwolka
In this paper, I compare the values of different information systems for a principal in a two-stage production process, in which the principal himself is responsible for the production in the second stage. The available systems are all costless and differ only as to the moment in which information is released. This situation induces a potential tradeoff, because predecision information allows both the principal and agent to ensure a given probability of success with smaller amounts of effort. However, it can also make it costlier for the principal to commit to a prespecified effort level. Thus, delaying information release might be of value. I discuss the implications for the design of information systems and the possibilities for improvement, which also include the mode of delegation and the communication structure.
pp. 55-79
