This thesis conducts three different empirical studies and finds that some of the pre-2007 risk assessment model could underestimate the systemic risk of the banking sector and justifies an overhaul. First, it simulates the contagion impact of the UK interbank market. Subject to a number of assumptions (netting agreement, seniority, etc), it finds that the contagion is much severer if the simulation uses consolidated data than using unconsolidated data. Second, the thesis tests whether the riskiness of banks can be mitigated by peer interbank monitoring. Applying to UK market, the thesis finds little evidence of market discipline. The results are attributed to the lenders’ assumption of “too-big-to-fail” and the shortness of loan maturity. Last, the thesis investigates whether banking sector difficulties are preceded by macroeconomic distress. In contrast to most existing studies, the thesis finds that economy still thrives in the “pre-crisis” in terms of increasing GDP growth and the recession is generally associated with the “post-crisis” period. The inconsistency of results is very likely due to imprecise crisis identification of earlier studies which identify crises too late on the basis of “event studies”.