On differentiating cyber-insurance contracts a topological perspective

Recent literature on cyber-insurance has stressed the importance of discriminating network users on insurance contracts for the following reasons: (i) preventing adverse selection, (ii) partly internalizing the negative externalities of interdependent security, (iii) achieving maximum social welfare, (iv) helping a risk-averse insurer to distribute costs of holding safety capital among its clients, and (v) insurers sustaining a fixed amount of profit per contract. Thus, an important problem is studying ways to appropriately execute the user discrimination process. In this paper we take a network topological perspective and propose a technique (mechanism) to pertinently contract discriminate insured network users. We mathematically show that the Bonacich/Eigenvector centralities of network users is an appropriate parameter for differentiating insurance clients.