EVALUATING THE EFFICIENCY OF PORTFOLIO DIVERSIFICATION STRATEGIES USING STATISTICAL CORRELATION METRICS

This study evaluates the efficiency of portfolio diversification strategies from 2020 to 2024 using statistical correlation metrics. The research objective is to assess the performance of traditional correlation measures, explore advanced statistical models like GARCH and DCC, and examine the impact of market disruptions such as the COVID-19 pandemic on diversification efficiency. A quantitative methodology was employed, utilizing financial data from Bloomberg and Yahoo Finance, analyzed through Pearson correlation, GARCH, and DCC models. The results indicate that while traditional correlation metrics provided a general measure of asset co-movements, their static nature failed to capture dynamic market shifts. The study found that average correlation coefficients ranged from 0.72 to 0.83, with spikes up to 0.85 during crisis periods, reducing diversification benefits. The DCC-GARCH model demonstrated superior adaptability, showing statistically significant improvements (p < 0.001) over static models. Stress test analyses confirmed that diversified portfolios experienced peak losses of -9.1% in 2020, compared to more stable drawdowns in later years. ANOVA results (F = 6.32, p < 0.01) validated significant variations in diversification efficiency across different market phases. The overall correlation coefficient across portfolio types averaged 0.79, suggesting moderate interdependence but insufficient insulation from market volatility. Based on these findings, the study recommends adopting dynamic correlation models, incorporating alternative assets like real estate and cryptocurrencies, leveraging ESG factors, and implementing algorithmic trading for rebalancing to enhance portfolio resilience.