Search

The majority of Australian firms listed on the Australian Stock Exchange are multi-segment rather than single-segment organisations. Relative to focused firms, multi-segment firms have advantages such as potential intra-firm synergies and flexibility in capital budgeting through the use of their internal capital markets. On the other hand, it is often argued that multi-segment firms are more prone to agency problems such as divisional rent-seeking, which could render their internal capital markets inefficient. On balance, it is an empirical question as to whether or not the benefits of multi-segment structure outweigh its costs. Compared to multi-segment firms in other developed economies, Australian firms have been doing quite well recently. Can one take this as evidence in favour of the multi-segment structure in Australia and, if so, what is special about multi-segment firms in Australia? Or is their relatively superior performance largely due to exogenous factors? This thesis empirically examines these issues by studying corporate diversification in Australia. Multi-segment firms can be subject to more influence activities by influential division managers, who may be able to affect the capital budgeting process. This thesis first analyses how influence activities in the form of signal jamming affect the capital budgeting process in corporate organisations in Australia. Following Wulf (2002), the specific focus is on how corporate headquarters allocates capital budget to a small division based on two types of information. The first type of information is the past performance of the small division, which is a noisy, public signal of its future performance, but which is not subject to manipulation. The second type of information is the private report about the small division’s future prospects made by the manager of a large division, which may be more informative than the public signal but also subject to influence by the large division manager. Investment sensitivity is defined as how capital budget allocated to the small division depends on its past performance. The main findings are as follows. First, investment sensitivity is found to be positive, indicating that headquarters invests more in the small division as its past performance improves. Second, investment sensitivity decreases as influence problems become more severe, where the severity of influence problems is measured by several firm-level proxies. The second finding, which is counter-intuitive, may be due to the fact that, as influence problems become more severe, headquarters may proactively counter influence activities through explicit incentives given to the manager of the large division. In that case, headquarters can rely on a more informative private signal. This is examined by studying how compensation incentives for the large division manager are related to investment sensitivity. A negative relationship is found between short-term incentives and investment sensitivity, which indicates that firms that provide large short-term incentives rely more on managerial recommendations than on noisy accounting measures. Finally, the empirical analyses are repeated using five new measures of diversification that are constructed based on information such as relatedness between segments, number of segments in the firm, and Herfindahl indices constructed from sales and assets. Estimations using these new measures also show similar results. The second theme of this thesis is the Australian evidence on diversification discount/premium. It examines whether the existence of diversification discount/premium is a measurement issue. First, the existing methodologies (Lang and Stulz, 1994; Berger and Ofek, 1995) are followed and mixed evidence is found: diversified firms in Australia trade at a discount in some cases and at a premium in others. Second, new and more informative measures of diversification are constructed as compared to the existing measures, based on which the diversification discount/premium is re-estimated. Mixed results are found when using industry adjusted q and the sales multiplier as dependent variables. When the asset multiplier is used as a dependent variable, however, a significant premium is found throughout. This premium obtained using the asset multiplier as the dependent variable is robust to standard control variables such as firm size, profitability and growth opportunity. Collectively, these results may imply that diversified firms in Australia trade at a premium and the existence of diversification discount may be a measurement issue. Finally, compensation incentives, which have not been explicitly considered in existing studies, are incorporated into the study. Incorporating compensation incentives for CEOs and division managers, long-term versus short-term in particular, shows that the diversification premium is robust after controlling for long-term incentives. However long-term incentives do contribute to the diversification premium. Indeed, effective long-term incentives positively affect this premium. In the sample of Australian firms in this study, such effective long-term incentives are shown to be 30% or more of total remuneration. Short-term incentives, on the other hand, are shown to be at best neutral and in some cases reduce the size of the diversification premium. In particular, the diversification premium switches to a discount in firms paying 90% or more as short-term incentives to division managers. Overall, the results suggest that at least part of diversification discount/premium can be explained by compensation incentives; without explicitly incorporating compensation incentives, the reported diversification discount/premium can be either over- or under-estimated.

This paper performs a rigorous survey of existing measures of diversification and fills a number of missing gaps between the theoretical and empirical literature surveyed by previous authors. In addition, the study outlines the areas where future work is possible and in which direction researchers may advance their thoughts. A detailed discussion on the various methodologies employed by earlier researchers and their findings are presented. This survey indicates that there is further scope for development of new theories and the empirical literature suggests that there is scope for development of new measures of diversification.

This paper analyses how influence activities in the form of signal jamming affect the capital budgeting process of corporate organisations in Australia. Empirical results suggest that investment sensitivity (the relationship between investment in the smallest segment and its past performances) is positive for Australian firms. However, when influence problems within a firm become more severe, mixed evidence is obtained for different measures of influence activity. With an increase in the number of segments, influence activity becomes more severe and headquarters relies more on a public signal. In contrast, with the increase in relatedness across segments, the influence problem increases and headquarters relies more on private information from the manager of the large segment. Evidence suggests that Australian firms provide high short-term incentive payments to managers of large segments to mitigate the influence activity problems, and thus rely more on managerial recommendations for investing in the smallest segment. Refereed/Peer-reviewed