Recent studies of business investment (Fazzari, Hubbard, Petersen, 1988), as well as much earlier work (e.g., Tinbergen, 1939;Klein, 1951), have established that investment spending and cash flow are strongly positively correlated. Professors Goergen and Renneboog set for themselves the ambitious task of uncovering the meaning of this fundamental correlation. One interpretation is that it reflects the finance constraints due to asymmetric information between the firm and providers of external finance. An increase in cash flow relaxes finance constraints, thus allowing the firm to undertake more investment. An equally compelling alternative is that managers have a substantial amount of discretion, and undertake actions, such as investment in projects with relatively poor returns, that favor their interests over those of the owners. In this case, additional cash flow leads to additional investment, but for a reason much different than when the firm faces finance constraints. In terms of the labels used by Kathuria and Mueller (1995), we need to test the asymmetric information hypothesis (AIH) versus the managerial discretion hypothesis (MDH). The differences underlying the fundamental correlation are clearly important for understanding firm behaviour. This paper contains three important innovations. First, it joins a small but growing literature that examines both the AIH and MDH hypothesis together. These problems have analyzed in two extensive but largely independent empirical literatures. For example, the recent comprehensive surveys of finance constraints (Hubbard, 1998) and corporate governance (Shleifer, Vishny, 1997) make little mention of the work cited in the other survey. Of the 101 references in the former study and 237 references in the latter, only seven are common to both. This tiny overlap is surprising since both derive from the same basic problem, the uneven distribution of information between firm and participants in external capital markets.Second, the paper uses several detailed measures of ownership concentration. Unlike data from financial statements, this information can not be obtained from computer-accessible files. Rather, one needs to do the careful and painstaking work of reading through annual reports and collecting the information in a form suitable to address the questions at hand.Third, the paper employs a systems GMM estimator. This econometric technique estimates the equation of * Emory University and CESifo. interest in levels with the instrumental variables differenced and with the equation of interest in differences with the instruments in levels. While the estimator is fairly new (Blundell, Bond, 1998), it seems to overcome some problems with the traditional GMM estimator.Given the impressive data and powerful econometric technique, this paper is well-positioned to generate important insights into the relations between the AIH and MDH hypotheses and the fundamental correlation. An Euler equation is chosen as the estimating equation. Although I am persuaded about the importance...