This project explores how mathematical concepts from graph theory can be applied to modern financial markets to improve portfolio diversification and decision-making. Using a dataset of 18 major stocks across multiple sectors, daily returns were calculated and transformed into correlation matrices to measure relationships between assets. These correlations were then converted into distance metrics, allowing the construction of a weighted financial network.
To simplify the complexity of the market structure, a Minimum Spanning Tree (MST) was extracted, revealing the most essential connections between stocks. This network-based approach enabled the identification of central (highly correlated) and peripheral (low-correlation) assets, forming the basis for constructing two contrasting portfolios.
The portfolios were evaluated using key financial performance indicators, including annual return, volatility, correlation, and the Sharpe ratio. Results demonstrated that portfolios composed of less correlated assets achieved superior risk-adjusted performance, highlighting the importance of diversification beyond traditional methods. Overall, the project shows how graph theory can provide a structured, data-driven framework for understanding financial markets and optimizing investment strategies.