It is widely recognized that social network has a profound influence on the risk of informal finance; however, it is still left unanswered what kind of social network structure controls risk and what kind of social network structure amplifies risk. Moreover, most informal financial activities are at a grass-root level so the hiddenness of informal institution working on them protects them from official monitoring and control, which raises the bar for academic research on its risk mechanisms. This also causes current regulatory policies on informal finance to be ineffective to some extent. Beyond a general discussion on social network and based on social network analysis methods, this paper answers these questions and proposes feasible suggestions to policy-makers regarding the risk control of informal finance. Employing the rotating savings and credit associations (or RoSCAs for short) as one example of informal financial activities, this paper categorizes social network structures into two kinds: the network within one RoSCA and the network among several RoSCAs connected by intermediate members. The former network forms the foundation for monitoring and punishment mechanisms in informal finance, whereas the latter network forms the foundation for synergistic monitoring and synergistic punishment mechanisms in informal finance correspondingly. The informal financial default models in the paper internalize these four mechanisms of the informal finance to be part of default cost for the participants of informal financial activities. Therefore, the cost of default to informal financial contract is much beyond the concept of economic cost. Based on the theoretical models, this paper utilizes the data from the 1999 financial crisis in Wenzhou, Zhejiang to empirically test its hypotheses. For practical purposes, the number of survival periods and the hazard ratios while surviving are set as the dependent variables, respectively, to denote the risk of informal finance. By using the Newey-West robust regression method and the survival analysis (robust Cox Proportional hazard model) method, this paper empirically tests the role played by the social network in informal financial risk and obtains ideal results supporting the hypotheses proposed by theoretical models. The main conclusions include: (1) The tighter a social network is (including no strange member and lower looseness of the network), the more effectively the network can control the financial risk; and (2) the social network among several RoSCAs can either control or amplify risk depending on its specific structure: On one hand, it shapes the foundation of synergistic monitoring and synergistic punishment mechanisms so that it has effect on controlling risk; on the other hand, however, it increases the risk of the RoSCAs that are part of the network by dispersing economic resources into each of them and the risk of the whole market is increased by constituting channels for risk contagion. Combining these effects, the social network is able to control risk only if the intermediate member plays as the head of at least one RoSCA of which he/she participates in. The tightness of social network within a RoSCA plays a role in increasing the monitoring level and default cost faced by its members so as to decrease the motivation of default. On the other hand, the extensiveness of social network among several RoSCAs connected by intermediate members plays a role in increasing the synergistic monitoring level and the default cost faced by all members of those RoSCAs so as to reduce the risk; but meanwhile, it dilutes the intermediate membersresource distributed in each RoSCA so as to increase the risk. This paper theoretically and empirically explores the relationship between the social network structure and informal financial risk, which fills in the theoretical gap in the informal financial research. It dialectically analyzes the function of social network in informal financial activities, based on which the result is constructive for risk control policymaking in informal financial sectors.