Deliverables - D04 Centrality and Groups in Social Networks

To address the issue of this deliverable, we have focussed on two exempla of social networks: the network of directors of large corporations' boards (Board Network) and the network of shareholders in a stock market (Shareholder Network). In both cases we deal with real data collected by the project. Results for the first work have been published in one article. Results for the second work are about to be submitted for publication. In both cases articles are downloadable from COSIN website for prompt dissemination. For the Board Network, in the framework of decision making dynamics, we investigated the role of special subgroups of directors in driving global decisions. Members of boards of directors of large corporations are connected in networks through their multiple appointments on different boards. Members of a board who also serve together on an outside board, form a special subgraph, that we call interlock graph of the board.. We present a first model of decision making dynamics inspired to herd behavior and a second one more realistic in the context of boards. We study how the size and the topology of the interlock graph affect the probability that the board approves a strategy proposed by the Chief Executive Officer. We propose a measure of the impact of the interlock on the decision making, which could help identifying boards susceptible to be influenced by a minority. We apply our models and methodology to the data of the boards of the largest US corporations in 1999. This work is reported in Decision making dynamics in corporate boards by S.Battiston, E.Bonabeau, G.Weisbuch to be published on Physica A. For the shareholders network we have studied the topological properties of the graph in relationship to the ability of firms to control other firms. A financial system forms an intricate network of ownership relationships between owners and controlled companies. On a large scale, such a structure can occasionally experience dramatic cooperative phenomena which follow some empirically well-established scaling laws. It is therefore very important to understand if any underlying structure exists in order to predict when such organized phenomena take place. Despite multiple efforts in describing criteria for the robustness of a financial systems and the best portfolio strategy, poor attention have been devoted to the topology of such a network. This deliverable is formed by the following publications