In this project, we investigate portfolio construction methods and replicate the comparison done in ”Optimal Versus Naive Diversification: How Inefficient is the 1/N Portfolio Strategy?” by DeMiguel, Garlappi, and Uppal. We evaluate various strategies, including the equally weighted 1/N portfolio, sample-based mean-variance portfolio, other moment estimators, and shrinkage estimation. The analysis employs a rolling-window estimation approach to assess the out-of-sample performance of these strategies using multiple metrics, including Sharpe ratio. Our analysis spans multiple empirical datasets, including S&P 500 sector portfolios (1981–2002), industry portfolios (1963–2004), and factor-based portfolios (1963–2004).