Central clearing and pre-emptive liquidity hoarding in financial networks

The impact of central clearing on the structure of financial networks: In this paper I investigate how the introduction of central clearing affects the network structure of over-the-counter (OTC) derivative market participants. Drawing on both modeled and real-life network structures, I study how central clearing affects network centrality | a measure of the systemic importance of market participants in the OTC derivatives network. I find four major results that have policy implications for central clearing. First, I  find that increasing the level of central clearing does not reduce overall network centrality. Second, I present evidence that the impact of central clearing on network centrality is sensitive to how central clearing is implemented and the degree of network interconnectedness. Third, I show that central clearing affects the homogeneity of centrality of individual market participants in the OTC derivatives network. These results hold not only for the U.S. interbank network but also for modeled network structures. Finally, drawing on evidence from network models, I show that the level of interconnectedness within a network impacts network centrality, and that this impact varies with the level of central clearing.

Pre-emptive liquidity hoarding and contagion in the cross-border interbank network: In this paper I examine how pre-emptive liquidity hoarding exacerbates cross-border contagion. I investigate two channels of hoarding | credit default (the \credit" channel) and funding withdrawal (the \funding" channel) and study their relative and combined impact on the extent of contagion in the cross-border interbank network. There are three main results. First, the degree of pre-emptiveness affects how much one hoarding channel dominates the other, but which hoarding channel dominates depends on the level of contagion severity. At low levels of severity, the funding channel increasingly dominates the credit channel as the degree of pre-emptiveness increases. However, at high levels of severity, the opposite holds | the credit channel dominates the funding channel as the degree of pre-emptiveness increases. Second, the credit and funding channels have a mutually reinforcing effect | contagion through a combination of the credit and funding channels has a worse impact than contagion through either one of the channels alone. Third, simulations on the international  financial network between 2006 and 2012 show that global cross-border contagion risk peaked in 2007 at the height of the global  financial crisis, decreased from 2008 to 2011, but showed an uptick in 2012. The simulations also reveal that Eurozone countries were generally more susceptible to contagion risk compared to the rest of the world.