How do firms allocate capital? Do units with better investment opportunities receive larger capital allocations and invest more? Are units run by more powerful or better connected managers favored with higher investment budgets? To empirically address these questions, we use a unique and proprietary five-year business-unit panel database on planned and actual capital allocations inside a world-wide conglomerate with 5 divisions and 22 business units.
The efficient internal capital markets view holds that capital allocation inside a conglomerate follows a process by which the firm pools internally generated cash flows and subsequently allocates funds to their best use across units (e.g., Weston, 1970, Williamson, 1975, Matsunaka and Nanda, 2002, Maksimovic and Phillips, 2002). Through winner-picking methods, internal capital markets may add value as the firm gives larger allocations to those units with the highest investment opportunities (e.g., Stein, 1997). In these models, capital allocation is mainly determined by the investment prospects (e.g., marginal Tobin’s Q) of a unit.
An alternative view suggests that investment opportunities may not be the only driving force behind capital allocations and that units run by more powerful or better connected managers may succeed in attracting larger allocations beyond their opportunities. These ideas have been formalized in inefficient internal capital market models such as Meyer, Milgrom, and Roberts (1992), Scharfstein and Stein (2000), Rajan, Servaes, and Zingales (2000), and Wulf (2008) among others. These models assume that unit CEOs have a preference for larger capital allocations (e.g., for rent-seeking or empire building reasons) and thus try to get more funds allocated to their units through the use of so-called influence activities. Influence activities are costly and inefficient due to the amount of resources spent on affecting allocations and the potentially suboptimal final investment decisions. All these models focus on the activities and the power of managers operating below the CEO to explain the allocation of capital inside firms.
In Meyer, Milgrom, and Roberts (1992), unit managers are modeled as rent-seeking agents trying to influence the CEO to get larger capital allocations (e.g., by overstating their own units’ prospects or by exercising bargaining power and lobbying activities). In the same spirit, Scharfstein and Stein (2000) show that division managers can raise their bargaining power to extract larger benefits from the CEO by pursuing rent-seeking activities. As a result of the agency conflict between managers and shareholders, these benefits take the form of preferential capital allocations rather than higher compensation. Unit managers may derive utility from larger capital allocations, as in the classic empire-building view of the agency literature. Rajan, Servaes, and Zingales (2000) show that internal distortions in the allocation of capital can arise even if CEOs are acting on behalf of shareholders. The basic prediction of all of these models is that capital allocation is influenced by the power of unit managers vis-à-vis the CEO rather than based on investment prospects alone.
Although there has been substantial theoretical progress on internal capital markets, empirical evidence on the topic has been hard to come by. Similar to the situation of the internal labor markets literature, described by Baker and Holmstrom (1995) as in a stage of “too many theories, too few facts,” empirical research on internal capital markets may benefit from “…studies of personnel records, supplemented by interviews and institutional facts” (Baker and Holmstrom, 1995). Our paper follows this line of thinking and the work of Baker, Gibbs, and Holmstrom (1994a, 1994b) who test internal labor market theories based on detailed internal data from a single firm.
Download
PDF Ebook Inside a Conglomerate: Internal Capital Markets and Managerial Power
