Soutenance de thèse de Norfaryanti Kamaruddin

Ce travail permet de mettre en place des outils qui décrivent la structure financière des grandes corporations de l’agribusiness qui exploitent les forêts et développent les plantations de palmier à huile. Ces outils permettent de discerner des facteurs financiers dont le rôle sur les stratégies et les décisions était insoupçonné jusqu’ alors. C’est un travail tout à fait novateur qui ouvre des champs d’application qui vont bien au delà de l’Asie du sud-est, et aussi dans d’autres secteurs touchés par la financiarisation et la consolidation des grands groupes.

Résumé de la thèse :

Decisions made by the government-linked companies (GLCs) and family- owned businesses (FOBs) are seen to be fundamentally different. This issue has become more pertinent now that ownership and control structures of major oil palm corporations have become far more complex in the context of growing and intense competition. This study analysed whether the control and decision-making is determined by the type of owner, i.e. the government or a family. In the palm oil, forestry and biomass sectors, where GLCs and family-owned companies are present as big businesses, there is a major gap in understanding what drives their decision-making. Agribusiness corporations deal with challenging decisions that have economic, social and environmental implications, regardless of their ownership. The global agriculture sector is under the process of more financialisation, precisely through the extension of major agribusiness corporations. Understanding what really drives theirdecision making, be it their control patterns or other factors is of extreme importance for the sustainability of agricultural sectors worldwide.
The study was designed to analyse how the ownership structure inform decision-making behaviour by the of oil palm corporations owned by the government and family. This is a pioneering study that seeks to quantify and to assess the ownership and control patterns of two very different types of plantation companies. This study comprises two ownership groups of companies, GLCs and family-owned firms. Companies within each group are compared, to assess their similarities and differences, and to analyse their decision-making behaviour.
A network analysis was employed in quantifying and analysing the corporate structures of eight major Malaysian oil palm corporations. The GLCs are Sime Darby, Boustead, IJM Plantations, and Kulim. The FOBs are IOI Corp, KLK, Genting Plantations, and Jaya Tiasa. The data were obtained from various reputable sources. There are 4,331 companies’ shareholding data gathered, covering ten levels of shareholding. 
The shareholdings data generated network topology graphs with its simple and advanced network centrality metrics. The simple network metrics explained the characteristic (pyramid levels, average ownership tier, subsidiaries degree, shareholding degree, and hierarchy index) of the eight corporate networks. The advanced centrality metrics are, structural control (betweenness) and decisions load (stress). These two advanced network metrics were used to assess the decision-making behaviour. T-test was done to respond to whether GLCs and FOBs are similar or different. Linear regression was run using local and global regression models to obtain further insights on the companies’ decision-making behaviour pattern. The model was validated to further understand the decision-making behaviour.
Based on the simple network centrality metrics comparison, the eight companies displayed various designs of ownership structure. Both GLCs and FOBs showed a range of four to eleven levels of pyramiding in their ownership structure. Among GLCs and FOBs, on average a company owned one to four companies. The subsidiaries degree ranges from 20 to 499 for GLCs, and 28 to 116 for FOBs. The shareholders degree ranges from 27 to 44 within GLCs, and 33 to 43 within FOBs. Both GLCs and FOBs have potential to be highly hierarchical, with a range of 75% to 94% hierarchic.
The advanced network metrics comparison was done using t-test. Results showed that the structural control for all companies have similarities and differences. Not all GLCs structural control is significantly different from FOBs. Within GLCs, structural control of Sime Darby and Boustead were different from IJM Plantations and Kulim. Within FOBs, structural control of IOI Corp, and Genting Plantations were different from Jaya Tiasa. However, Kulim and Jaya Tiasa shared similarity in terms of structural control. Sime Darby’s structural control was similar to IOI Corp and KLK. The structural control did not appear to have any association with the ownership identity. There is a high similarity pattern of decisions load amongst the eight companies. Boustead is the only company that showed different decisions load from the other six companies, excluding IJM Plantations. Decisions load of Sime Darby is onlydifferent from IJM Plantations. Within FOBs, the decisions load for all four companies are similar.
In bringing the analysis further, this research run local and global model of linear regression. Local model is the individual/unique model that reflects each company’s decision-making behaviour. Global model is the communal model that reflects the whole plantation sector. Both models showed a linear pattern of increasing decisions load as the structural control increases. Global model presented a better fitted model to understand their decision-making behaviour.
The global model explained that the eight companies were scattered and not inclined to their ownership identity. Three groups had appeared: Boustead was in Group 1; Sime Darby, KLK, IOI Corp, IJM Plantations and Genting Plantations were in Group 2; and Jaya Tiasa and Kulim were in Group 3. The groupings signify their similarity in decision-making behaviour pattern. However, each companies were different in terms of the decisions load given the same subsidiaries’ structural control and vice versa. Even though the companies belonged to the same group, followed the same linear pattern of increasing decisions load when structural control increased, their decisions load are all different given the same structural control. There was an emerging behaviour derived from the intercepts of the decisions load and structural control relationship. This research regards the behaviour as structural flexibility. The structural flexibility of the companies was highly dependent on the number of subsidiaries, hierarchy index and pyramid size.
Based on the results, the analysis showed that each company, whether a GLC or a family-owned enterprise, functioned differently. Their decision-making behaviour depended less on ownership type and more on the topology or design of the structure, such as the number of the companies, hierarchy index, and the pyramid size. The different levels of decisions load and control denote a variety of flexibility patterns.
The analyses proved that the ownership structure of a company influenced their decision-making behaviour. This research concluded that their decision- making behaviour depended less on ownership type and more on the topology or design of the ownership structure. Both of the GLCs and FOBs have similarities and differences in their decision-making behaviour. The results may contradict with some other studies and it opens a new field of research and analysis of corporations, regardless of their ownership type/identity (government-owned or family-owned). This method allows us to evaluate the ownership topology structure quantitatively and qualitatively. Furthermore, these companies can be ranked based on the analysis used in this research.