An Adaptive Contagion Mapping Methodology

In this paper we augment the CoMap methodology with an advanced set of regulatory distress and default constraints and banks’ behavioural rules which allows us to model and capture the implications for financial stability of dynamic banks’ responses. The result is an adaptive contagion mapping methodology (A-CoMap) - a network-based contagion model where banks dynamically adapt their short-term exposures to news about counterparty risk. The model is calibrated and tested on a unique dataset composed of granular bank and exposure level information on 2,800 consolidated banking groups worldwide in which less-significant institutions (network periphery) play a relevant role. First, we show how banks’ behavioural responses may either increase or decrease the degree of stress to some specific banks. Second, thanks to the multi-complex regulatory environment made of solvency/liquidity and distress/default constraints we shed light upon the effectiveness and calibration of minima and capital buffer requirements, and their trade-off. Third we provide new evidence of the too-many-too fail problem and we compare it with the too-big-too-fail issue. In the end to foster research in the field, which is often limited by the availability of confidential exposure-level data, we provide detailed statistics about the actual network of EA banks’ exposures and the set of calibrated parameters