Correlation and network analysis of global financial indices

Area of Science:

• Quantitative Finance
• Network Science
• Econometrics

Background:

• Understanding financial market dynamics is crucial, especially during periods of instability.
• Previous studies have explored market correlations, but comprehensive analysis using both Random Matrix Theory (RMT) and network methods during a crisis is less common.

Purpose of the Study:

• To investigate the correlation and network properties of 20 financial indices.
• To compare these properties before and during the 2008 financial crisis.
• To identify structural changes in the organization of financial indices during a crisis.

Main Methods:

• Application of Random Matrix Theory (RMT) to analyze eigenvector components.
• Network analysis employing the Fruchterman-Reingold layout and minimum spanning tree.
• Hierarchical clustering using the average linkage algorithm and cophenetic correlation coefficients.

Main Results:

• RMT: Eigenvector components show clusters that switch directions during the crisis.
• Network analysis: Regional index clusters (Americas, Europe, Asia-Pacific) merge or separate differently during the crisis.
• Minimum spanning tree structure shifts from starlike to chainlike; hierarchical clustering indicates increased organization during crisis.

Conclusions:

• Significant structural reorganization occurs in financial index networks during crises.
• RMT and network methods provide complementary insights into these dynamic changes.
• The findings highlight increased hierarchical organization and altered interdependencies among global financial markets under stress.