Information acquisition and welfare in network games (Q2195723)

This paper studies the problem of information acquisition and use in economic environments in the presence of various incentives for coordination. It also addresses the inefficiencies that arise. To this end, endogenous information from a common state is incorporated into a network game. Players first invest in the accuracy of a privately observed signal that informs them of the state. In addition, they also use their information to infer the realization of their neighbors' signals. They then simultaneously choose actions. For the analysis, the network scenario of \textit{C. Ballester} et al. [Econometrica 74, No. 5, 1403--1417 (2006; Zbl 1138.91590)] is adopted. In this scenario, an information-response game is derived. The responsiveness to equilibrium signals defines the so-called ``informational centrality'' of each player, given by its weighted Katz-Bonacich centrality [\textit{L. Katz}, Psychometrika 18, 39--43 (1953; Zbl 0053.27606); \textit{P. Bonacich}, ``Power and centrality: a family of measures'', Am. J. Sociol. 92, 1170--1182 (1987)] defined in a fitted network. Information centralities are used to characterize the private and public marginal values of information. The application of analysts acquiring and responding to information about an asset's return is considered. The conclusion reached is that, in general, analysts who seek to coordinate their positions in the asset with those of many colleagues acquire the most information, but ``under'' acquire information. Analysts who compete fiercely among companies acquire less information, but ``over'' acquire information. Thus, there is inefficient conformity in analysts' expertise. Second, a novelty in network environments is the potential to respond to negative signals. When occupying highly adverse network locations, analysts face an incentive to invest in information in order to move in the opposite direction of signal realization and coordinate their asset positions away from those of competing analysts. The result is that these analysts create value for both themselves and their competitors, introducing a positive externality. Finally, having an investment in more information publicly observed encourages other analysts with complementary roles to acquire and use more information as well, while discouraging analysts with competing roles.