Exacerbated vulnerability of coupled socio-economic risk in complex networks

The study of risk contagion in economic networks has most often focused on the financial liquidities of institutions and assets. In practice the agents in a network affect each other through social contagion, i.e., through herd behavior and the tendency to follow leaders. We study the coupled risk between social and economic contagion and find it significantly more severe than when economic risk is considered alone. Using the empirical network from the China venture capital market we find that the system exhibits an extreme risk of abrupt phase transition and large-scale damage, which is in clear contrast to the smooth phase transition traditionally observed in economic contagion alone. We also find that network structure impacts market resilience and that the randomization of the social network of the market participants can reduce system fragility when there is herd behavior. Our work indicates that under coupled contagion mechanisms network resilience can exhibit a fundamentally different behavior, i.e., an abrupt transition. It also reveals the extreme risk when a system has coupled socio-economic risks, and this could be of interest to both policy makers and market practitioners. Copyright c ⃝ EPLA, 2016 Introduction. – Can the failure of a single element threaten the stability of an economic system? This is the key question when authorities deal with predicting and reducing risk propagation in highly complex economic systems [1]. All complex systems are at risk of cascading failure in which one failed part can cause the failure of linked fractions such that failure propagates throughout the entire system [2–4]. Systemic risks associated with cascading failures have been present in recent and earlier economic crises, e.g., the world financial crisis, the Asianeconomic crisis, and dot-com bubbles [5,6]. Because network science can use both the behavior of economic system participants and the relationships among them to model contagion mechanisms [7–9], it has attracted enormous attention among researchers studying risk propagation in such economic systems as banking networks, buyer and seller credit systems, international trade, capital markets and stock markets [10–14]. In network science vertices represent agents in economic systems and links represent connections among them. The networks can be generated by using real data [15,16] or by developing theoretical networks in which properties are assumed, e.g., random networks or scale-free networks [17]. Two types of network failure mechanism are used in this research. The first is the generic context failure approach based on a network social contagion epidemiological model in which one node fails or survives depending on the state of its neighbors [18–20]. The other is the liquidity approach in which each node is assigned a simplified balance sheet state, and a default in asset results in cascading failures through liquidations. When its asset value drops below its liability value it either fails or survives with an assumed tolerance probability [17,21,22]. Most studies of systemic risk in economic networks focus on economic/liquidity contagion. Yet real-world systems have both economic and social contagion risks [23]