نفوذ و عدم کارایی در بازار سرمایه داخلی

I model inefficient resource allocations in M-form organizations due to influence activities by division managers that skew capital budgets in their favor. Corporate headquarters receives two types of signals about investment opportunities: private signals that can be distorted by managers, and public signals that are undistorted but noisy. Headquarters faces a tradeoff between the cost of attaining an accurate private signal and the value of the information the signal provides. In contrast to existing models of “socialism” in internal capital markets, I show that investment sensitivity to Tobin's Q is higher than first-best in firms where division managers hold equity (a result consistent with evidence presented in Scharfstein, 1998). When managers face high private costs from distorting information (equity holdings), headquarters may commit to investment contracts that place “too little” weight on private signals and “too much” weight on public signals (i.e. Q). This result has implications for managers in the design of capital budgeting processes and incentive compensation systems.

Evidence suggests that diversified conglomerates have active internal capital markets, using cash flow generated by one division to invest in another. While extensive research in financial economics analyzes both the benefits and costs of capital allocation within the firm's hierarchy, a reasonable goal for future research is to understand better what factors explain the variation in efficiency of internal capital markets across firms.1 In pursuit of this goal, I model a particular type of costly behavior by division managers (influence activities that distort private information about investment opportunities) that leads to novel predictions about inefficiencies in internal capital markets.2 I find that asymmetric information can lead to investment sensitivity to public signals (i.e. Tobin's Q) that is higher than first-best. Instead of a measure of efficient capital allocation, increasing sensitivity to Q may be the firm's attempt to mitigate division manager incentives to distort informative private signals (i.e. managerial recommendations). This is in contrast to much of the internal capital markets literature that implicitly assumes that efficiency increases with divisional investment sensitivity to industry Q. Information and knowledge production within the firm's hierarchy is a subject of recent research in organizational economics (e.g. Aghion and Tirole, 1997, Garicano, 2000, Dessein, 2002, Stein, 2002 and Marino and Matsusaka, 2005). In this paper, I consider how the M-form organization creates perverse incentives for division managers to distort private signals that are transmitted to headquarters about competing investment opportunities across divisions within the firm. I identify circumstances under which the problem is most pronounced and subsequently show how investment contracts implicit in capital budgeting processes can mitigate this rent-seeking behavior. While this model makes several novel predictions concerning the cross-sectional variation in the efficiency of internal capital markets across firms, the most important result is the following. When “core” division managers hold equity, firms increase the sensitivity of investment to public signals (i.e. Q) for small divisions above first-best. When managers hold equity and the private costs of investment distortion are large, headquarters attempts to mitigate influence activities by placing “too little” weight on valuable, but distortable, private signals and “too much” weight on noisy, public signals. Headquarters faces a tradeoff between the cost of an accurate private signal and the value of the information the signal provides. The key result of the paper is consistent with evidence presented in Scharfstein (1998). More broadly, the model predicts that inefficiencies are more pronounced in firms with more influential division managers, less firm-level incentive pay for division managers (i.e. smaller equity holdings), and lower quality in public signals about investment opportunities. Bower's classic clinical study (1986) documents how better-informed division managers misrepresent project opportunities to decision makers at headquarters. A more recent example of the potential effect of influence activities in the internal capital market is IBM's inability to capitalize on its early success in the development of its personal computer business. Mills and Friesen (1996; 128–129) argue that: “it was mainframe-myopia that so severely damaged IBM in the 1990s” and that “division executives began to put the welfare of their own organizations above that of the corporation as a whole…in the resistance of the mainframe division to the introduction of new technology.” Based on accounts of IBM's history, it seems that skepticism by the mainframe division about investment opportunities for the IBM PC division was partially to blame for the inconsistent success in personal computers.3 One of corporate headquarters’ primary responsibilities is efficient resource allocation. However, accurate information about relative investment opportunities is typically unavailable, especially in new products and developing businesses. As a part of the capital budgeting process, headquarters relies on several sources of information from division managers: private information such as managerial assessments or recommendations (that may be distorted) and public information such as industry Q (that cannot be distorted, but is noisy). Examples of the private signal include assessments about new product development, the adoption of a division's product as a standard, or a pending sale to a large customer. To the extent that division managers prefer larger capital budgets, they have the incentive to engage in costly influence activities in order to increase the capital allocated to their division. Division managers of “core” businesses (large, established divisions) are more powerful than managers of developing businesses (small, newer divisions) because they have greater control over valuable firm resources ( Rajan and Zingales, 2000). Core division managers can distort the transmission of private information to headquarters about the investment prospects of developing businesses. Headquarters cannot observe the core division manager's action, but does observe the realizations of both signals about the developing division's investment prospects—the possibly distorted, private signal and the noisy, public signal. 4 Certainly, one extreme way to address the incentive disparity between headquarters and division managers is to eliminate the internal capital market by either spinning off divisions (Gertner et al., 2002 and Inderst and Laux, 2005) or instituting policies that limit the role of headquarters in allocating capital (Ozbas, 2005). Another less extreme approach is to rotate division managers across different divisions (similar to General Electric) to lessen incentives to lobby for capital (Stein, 2003).5 Alternatively, firms can design management processes or incentives to mitigate information distortion. For example, firms design capital budget processes as mechanisms to elicit revelation of private information (Harris and Raviv, 1996, Harris and Raviv, 1998 and Antle and Eppen, 1985) or raise hurdle rates in their evaluation of potential investment projects as a crude method of addressing agency problems (Poterba and Summers, 1995). Finally, firms link division manager compensation to firm performance in bonus contracts and through equity-based incentives to address potential problems in internal capital markets (Palia, 1999 and Wulf, 2002). In this paper, I combine the latter two approaches (the design of capital budgeting processes and division manager compensation schemes) in a model in which firms address incentive concerns in capital allocation. As a solution to mitigate influence activities, headquarters designs ex ante investment contracts that specify the optimal weights placed on private and public information in the capital allocation decision, with the weights determined by the factors that represent the manager's ability to distort information, the private cost of signal-jamming, and the noise in the public signals. The optimal investment contract results in the standard tradeoff between the cost of controlling the division manager's action and the value of that action. In this context, the specific tradeoff is between the cost of attaining an accurate private signal and the value of the information the signal provides. The value, in this case, is more efficient resource allocation. While the model in the paper does not derive optimal compensation contracts, the core division manager's incentive pay consists of an exogenously determined fraction of firm profits. Since compensation of division managers is linked to firm performance, managers are penalized through lower incentive pay because investment distortion leads to lower firm performance. Clearly, there are many reasons why firms offer incentive pay: for example, to induce managerial effort.6 However, such motivations are absent from the present model because I am mainly interested in modeling the effect of information distortion on investment. The paper's main contribution is a model of information distortion about competing investment opportunities by division managers in M-form firms. This paper's model differs from earlier models of internal capital markets in several ways. First, it focuses on the relative quality of different information sources in allocating capital across divisions. In a related paper, Stein (2002) shows that division manager incentives to generate “soft” information in M-form organizations are reduced. 7 While he focuses on incentives to produce information, in contrast, I focus on incentives to prevent distortion of information. Second, the results identify circumstances when firms ignore valuable information and the investment sensitivity to public information (Q) is “too high” relative to first-best. 8 This is in contrast to the “socialism” models that predict a reduced sensitivity to Q relative to standalone firms. Third, since the model identifies the settings in which the influence problem is most severe, the results have implications for firms in the design of optimal capital budgeting processes, a critical administrative process that governs the allocation of resources within M-form organizations ( Simon, 1997). Finally, the paper develops an economic model of a management phenomenon that links together the literatures on influence activities, signal-jamming, and optimal investment contracts in an application to internal capital markets. Existing papers in financial economics investigate the effect of divisional manager power struggles or rent-seeking on capital allocation and generally predict “socialism” in internal capital markets whereby weaker divisions are subsidized by stronger ones (Rajan et al., 2000 and Scharfstein and Stein, 2000). Rajan et al. develop a model in which the principal optimally transfers capital towards the small division with weak opportunities in order to make this division behave more cooperatively in joint production with other divisions. The authors find support for inefficient cross-subsidies and that the extent of inefficiency is positively related to the diversity of resources and investment opportunities across divisions. Similarly, this paper's model also predicts inefficiencies, and a key assumption about differential power between the division managers of large, core businesses versus small, developing businesses is a more restricted version of their finding on diversity of resources.9 However, I explicitly model how influence activities distort the information that is available to headquarters which in turn affects investment efficiency. Moreover, Rajan et al. make no predictions about how division manager compensation (or equity ownership) or the uncertainty in divisional investment opportunities affect internal capital market efficiency across firms. The model in this paper allows division managers to be compensated in terms of capital allocations (because they prefer managing larger divisions) in addition to being penalized by the reduced value of equity holdings. In featuring two forms of compensation, this model is similar to Scharfstein and Stein's (2000) two-tiered agency model that predicts that large socialist-type inefficiencies are more likely when the CEO has low-powered incentives. Providing empirical support for their model, Scharfstein (1998) finds that investment of small divisions of multi-business firms is less sensitive to investment opportunities than stand-alone firms, but it is greater when top management owns more equity. While this finding is broadly consistent with my model's main result, the explanation for the result is quite different: when division managers face high private costs from distorting information (i.e. high equity), headquarters commits to investment contracts that place “too little” weight on private signals and “too much” weight on public signals (i.e. Q). As such, higher investment sensitivity to public signals is a second-best solution to an asymmetric information problem in which firms face a tradeoff between the cost of an accurate private signal and the value of the information that the signal provides. Adding to the evidence that division manager incentives are important, both Palia (1999) and Wulf (2002) find evidence that there are greater inefficiencies in internal capital markets when division manager compensation is less closely linked to overall firm performance. While this paper's model makes some predictions similar to Scharfstein and Stein, their model makes no predictions about division manager equity. Moreover, the results in this paper also highlight the importance of relative bargaining power of division managers and noise in public signals about investment opportunities. The remainder of the paper is organized into four parts. Section 2 presents the model of influence and signal distortion. Section 3 solves for the value-maximizing investment rules under influence and derives specific predictions regarding investment sensitivity to signals. Section 4 discusses existing evidence in the literature. Section 5 concludes.

The efficiency of resource allocation decisions within an organization ultimately depends on the quality of the information available to the decision-maker. The intent of this paper is to develop an economic model that considers division manager incentives to influence capital allocation decisions based on different types of information in M-form organizations. Decision-makers place less weight on subjective, distortable information (e.g. managerial recommendations) and more weight on objective, but noisy information (e.g. industry Q) to offset incentives for division managers to distort information about investment opportunities. Investment inefficiency depends on the ability of managers to distort information, the presence of firm-level compensation incentives for division managers, and the quality of public signals of investment opportunities. By arguing that managerial ability to distort private signals is greater in certain types of firms, the model leads to testable implications for division investment sensitivity to public signals across firm characteristics. I highlight several empirical findings of related work that are generally consistent with the model's predictions. One important point that this paper makes is that high investment sensitivity to public information may be a sign of inefficiency in internal capital markets. Firms that face asymmetric information problems attempt to mitigate incentives for managers to distort more informative private signals by increasing investment sensitivity to public information above the first-best level. This is in contrast with much of the internal capital markets literature, which implicitly assumes that divisional investment sensitivity to industry Q increases with investment efficiency. The importance of influence activities in the allocation of capital across divisions may help explain empirical findings about large firms. For example, large multi-business firms have difficulty in creating a desirable entrepreneurial climate and are generally less successful than small firms in developing new products and businesses. One caveat is that this paper has not explained why multi-divisional firms should exist if investment inefficiencies are prevalent. It must be that other advantages outweigh the costs of an internal capital market or that firms learn of these inefficiencies over time and eventually correct them. Clearly, there is more work to be done to understand the decisions firms make regarding organizational structure.