When a Corporate Investment Gets a Go

Capital budgeting frequently involves multiple stages at which firms can decide to continue or abandon ongoing projects. Capital budgeting practices are important for firms because they can limit opportunistic behavior on the part of managers: for instance, these systems can easily uncover, and thereby preclude, a manager's diversion of capital funds from appropriate investments to private uses such as vacations or excess compensation. What capital budgeting practices cannot do as effectively, however, is prevent a manager from investing in unnecessary assets, provided that they fall within the general class of appropriate purchases; this is, for instance, the case of empire-building activities from which the manager derives private benefit. Unlike monetary compensation or vacation packages, a manager can only consume the benefit of these latter assets over time – the manager's consumption of his informational rent or slack is delayed. As a result, if the principal abandons the project before completion, he both loses part of his initial investment and denies the manager the full consumption of his private benefits.

In a recent research article titled Multistage Capital Budgeting with Delayed Consumption of Slack, forthcoming in Management Science (doi: 10.1287/mnsc.1120.1621), Stanley Baiman, adjunct professor at Bocconi University's Department of Accounting, and his coauthors Mirko Heinle (Wharton) and Richard Saouma (UCLA) study the implications of this abandonment option and "delayed slack consumption" on optimal capital allocation schemes. Building upon extant single-period, adverse selection capital budgeting models, the authors consider a setting in which, in the first stage, a manager holding private information proposes an initial research budget to develop a new project that the principal either accepts or rejects. If the principal accepts, the funds are made available to the manager, who engages in the research required to discover the feasibility and net benefit of implementing the project. In the second stage, if the manager's research efforts are successful, the continuation value of the project becomes known to both the manager and the principal; the latter then determines whether to implement or abandon the project. It is assumed that, in the first stage, the manager cannot fully take advantage of her informational rents (the delayed consumption of slack). As a result, if the project is terminated at the second stage, the manager forgoes some of the benefits associated with his informational rent, while the principal in unable to fully recoup her initial investment.

Results show that the optimal capital allocation scheme uses a single hurdle rate in the first stage that results in underinvestment or capital rationing, while in the second stage it applies different hurdle rates depending on the manager's first-stage report and the outcome of his research. Another result is that it is optimal for the principal to commit to continue some projects even if they have a negative continuation value (i.e. second-stage overinvestment), and it is also optimal to commit to forgo other projects that have a positive continuation value (i.e. second-stage underinvestment). The results provide an explanation for previously reported empirical findings. The implications of the above results are particularly relevant for R&D-intensive industries, whose projects require significant capital investments and span multiple phases, as well as industries that rely on multiple-stage regulatory approval, such as the pharmaceutical industry.