There is widespread agreement that investment decisions are the most important decisions made by firms. In fact, most large firms pay considerable attention to designing capital budgeting systems, and do so in a manner suggesting that firms recognize incentive problems in the internal allocation of capital across organizational units. For example, firms use hurdle rates that are higher than their cost of capital, possibly as a crude measure of addressing managerial attempts to overstate cash flow projections (Poterba & Summers, 1995). In addition, there is a popular trend toward ‘charging’ divisions for the cost of capital by using economic value added (EVA) as a performance measure in determining executive pay (Ittner & Larcker, 1998). Finally, capital rationing is a common practice (Taggart, 1987) and theories suggest that it is used partially to address decentralized information and incentive problems (Harris & Raviv, 1996; Holmstrom & Ricart i Costa, 1986). The purpose of the paper is to explore the use of various instruments by multi-divisional firms as a means to control information and incentive problems in the allocation of capital across divisions.
The theoretical work on the provision of incentives inside firms recognizes that firms frequently use a ‘system’ of incentive instruments and a variety of performance measures (Holmstrom & Milgrom, 1991; Baker, Gibbons & Murphy, 1994). However, the empirical work evaluating this phenomenon is limited (Cockburn, Henderson & Stern, 1999; Ichniowski, Shaw & Prennushi, 1997; Ittner, Larcker & Rajan, 1997). In this paper, I contribute to this literature by evaluating the relationship between compensation contracts for division managers and capital budgeting in multi-divisional firms as alternative instruments to address information and incentive problems in internal capital markets. Recent work in the financial literature explores ‘rentseeking’ behavior by division managers and its effect on interdivisional capital allocation (Scharfstein & Stein, 1996; Rajan, Servaes & Zingales, 1998). Other papers investigate the design of divisional manager compensation contracts (Bushman, Indjejekian & Smith, 1995; Jensen & Meckling, 1999). However, the relationship between the two is a topic that is largely unexplored.
Division managers in multi-divisional firms with preferences for large capital budgets have the incentive to engage in costly influence activities and distort subjective information about investment opportunities in other divisions and thereby skew capital allocations in favor of their division (Wulf, 1999). To mitigate this behavior, firms can either reduce the investment gain or increase the compensation cost to managers of engaging in influence activities. Firms incorporate ‘investment incentives’ into the capital budgeting process by specifying the optimal weights placed on two types of information in allocating capital across divisions: imperfect, objective information (i.e. accounting measures) and distortable, subjective information (i.e. managerial recommendations). To reduce the investment gain from influence activities, firms place less weight on managerial recommendations and more weight on accounting measures relative to first best. Alternatively, to increase the compensation cost from influence activities, firms design ‘compensation incentives’ for division managers that place more weight on firm performance and less on division performance in determining incentive pay. Since influence activities distort investment and reduce firm performance, increasing the weight on firm performance increases the manager’s cost by decreasing compensation.
The principal result in this paper is that multi divisional firms appear to use these two instruments, compensation incentives linked to firm performance and investment incentives through the capital budgeting process, as substitutes to offset incentives to distort information. I combine a panel dataset of segment (or line of business) financial data and firm organizational characteristics from Compustat’s Segment database over the period 1988-1993 with cross sectional data from a proprietary survey about compensation contracts conducted in 1993 by a leading compensation consulting firm. The empirical methodology is based on the estimation of a segment investment equation from a standard investment model. Specifically, I measure the investment incentive by the sensitivity of segment investment to segment profitability (i.e. the weight placed on the imperfect, objective information in allocation of capital) while controlling for investment opportunities. In order to explicitly evaluate the substitutes hypothesis, the specification incorporates the structure of incentive compensation contracts (i.e. the weight placed on firm performance in determining incentive pay) while controlling for additional firm and industry characteristics.
Download
Internal Capital Markets and Firm-Level Compensation Incentives for Division Managers
