Portfolio Construction: From Markowitz to 1/N

Diversification is exact mathematics built on inputs nobody knows precisely. This path derives mean-variance optimization and the limits of diversification, then measures why optimized portfolios so often lose to simply splitting money equally, and finishes with the risk-based rules practitioners use instead: minimum variance, risk parity, shrinkage, constraints and rebalancing. Every worked number is reproducible in numpy, and every claim is traced to its source.