Portfolio Optimization using Graph Theory

This project explores portfolio optimization using graph theory, contrasting it with traditional Markowitz Optimization. It uses correlation matrices to construct financial networks, where nodes represent assets and edges indicate correlations. Key techniques include:

  1. Degeneracy Ordering: Identifies core assets that are more central and stable within the network.
  2. Clique Centrality: Detects highly interconnected clusters for diversified portfolios.
  3. Correlation Thresholding: Filters out weak correlations, retaining significant relationships to enhance portfolio stability.

The project demonstrates how graph-theoretic metrics can optimize risk-return trade-offs, maximize the Sharpe Ratio, and construct concentrated or diversified portfolios more efficiently than classical methods.