The Economics of Capital Allocation in Firms: Evidence from Internal Capital Markets

Abstract

We analyze a unique survey dataset to examine the (micro)foundations of capital allocation in firms. Firms employ systems of interconnected measures to counteract agency problems, including layers of approval, divisional budgets, reporting requirements, and compensation schemes. When making funding decisions, top management relies heavily on top-level, nonfinancial information. However, substantial parts of the capital budget do not require top management approval as firms trade off the benefits and costs of decentralization. Even firms with active internal capital markets tilt capital allocation toward relatively even distributions. Within-firm agency problems may result in divisions' restricted access to internal capital. AbstractWe analyze a unique survey dataset to examine the (micro)foundations of capital allocation in firms. Firms employ systems of interconnected measures to counteract agency problems, including layers of approval, divisional budgets, reporting requirements, and compensation schemes. When making funding decisions, top management relies heavily on top-level, nonfinancial information. However, substantial parts of the capital budget do not require top management approval as firms trade off the benefits and costs of decentralization. Even firms with active internal capital markets tilt capital allocation toward relatively even distributions. Within-firm agency problems may result in divisions' restricted access to internal capital.

Do Nonfinancial Firms Hold Risky Financial Assets? Evidence from Germany

“…Furthermore, to the best of our knowledge, not much is known yet about how nonfinancial firms organize their financial asset management. Qualitative one-on-one interviews (that the authors conducted with senior financial executives of seven sample firms) suggest that industrial firms' financial asset management processes are similar to those of real corporate investments (including approval thresholds, budget limits and different layers of approval; see Hoang et al, 2020). However, the interview evidence also suggests that asset management teams are relatively small and that firms provide financial managers with considerably discretion about how to spend financial resources within asset classes.…”

Optimal Dynamic Capital Budgeting

“…12 The preference for higher spending may re ‡ect perquisite consumption associated with running larger projects as well as an intrinsic preference for empire-building. In line with this assumption, survey evidence shows that divisional managers prefer running large divisions (Hoang, Gatzer, and Ruckes, 2017). I assume that the division manager consumes monetary transfers immediately rather than saving them for the future, as can be seen from…”

“…The division manager has informational advantage over headquarters in that she privately observes both the arrival of projects and their qualities (i.e., dX t ). This assumption has empirical support: for example, in a recent survey of CFOs by Hoang, Gatzer, and Ruckes (2017), more than 70% of CFOs believe that divisional managers have superior information about their businesses compared to the information of headquarters. Headquarters can learn about projects from (and only from) two sources.…”

“…Such separation of authority is common. In a recent survey of capital allocation practices of 115 …rms, Hoang, Gatzer, and Ruckes (2017) show that nearly all (97%) …rms in their survey have formal investment thresholds that feature central approval for investments above the threshold. They also …nd that divisions get considerable discretion over smaller capital expenditures, estimating that 40% of overall capital expenditures are made without central approval.…”

“…Giroud (2013) uses the introduction of new airline routes as a plausibly exogenous shock to the cost of monitoring and …nds that it leads to an increase in investment and productivity. If his research design was applied to the data on the internal resource allocation processes, such as survey evidence in Ross (1986) and Hoang, Gatzer, and Ruckes (2017), Predictions 1 and 3 would be credibly tested.…”