Credit default swaps drawup networks: too interconnected to be stable?

Area of Science:

• Quantitative Finance
• Financial Econometrics
• Network Analysis in Finance

Background:

• Analysis of Credit Default Swaps (CDS) spreads is crucial for understanding institutional financial health.
• The recent financial crisis highlighted the need for better methods to assess interconnectedness and systemic risk.
• Existing methodologies for analyzing market movements, such as ε-drawdowns, require extension to capture joint behaviors.

Purpose of the Study:

• To extend the ε-drawdown methodology to analyze joint ε-drawups in CDS spreads.
• To estimate conditional probabilities of synchronized, spike-like movements (co-movements) among pairs of CDS spreads.
• To apply network analysis to model financial dependencies and assess systemic risk during a crisis period.

Main Methods:

• Time series analysis of CDS spreads for major US and European institutions.
• Extension of ε-drawdown methodology to joint ε-drawups for estimating conditional probabilities.
• Network analysis using joint drawup probabilities as a proxy for financial dependencies.
• Introduction of novel centrality measures to analyze systemic impact and vulnerability.

Main Results:

• Over 50% of institutional CDS spread pairs exhibit high probabilities of joint drawups, indicating significant co-movement.
• A majority of CDS spread series are trend-reinforced, suggesting that upward movements tend to persist within a series.
• Network analysis reveals evolving systemic impact and vulnerability of financial institutions over time.

Conclusions:

• Joint ε-drawups provide a robust method for quantifying synchronized CDS spread movements.
• The findings underscore the interconnected nature of financial institutions, especially during crisis periods.
• Novel network centrality measures offer valuable insights into dynamic systemic risk assessment.