Risk contagion caused by interactions between credit and guarantee networks

Abstract

In this paper, we construct a model of firm-bank multi-networks, which is characterized by the simultaneous integration of the firm-bank credit network and the firm-firm guarantee network. Based on the constructed model, we research risk contagion via computational experiments and simulation methods. The findings suggest that (1) interactions between banks and firms at the micro level lead to the emergence of power-law characteristics in the real word. (2) The parameters of network structures has impacts on risk contagion. Specifically, the parameter relating to the probability of switching to other agents has double effects on risk contagion in the firm-bank system, leading to a downward trend of banks’ bad debt for lower values of this parameter and an upward trend for higher ones; while for the parameter denoting the fraction of guarantors that can be chosen by potential borrowers, bad debt is a non-monotonic function of this parameter, presenting a trend of an increase after a decrease. Finally, for the guarantee amplification parameter, bad debt presents an obvious upward trend with an increase of this parameter.