The Black-Litterman Model

Fixing the flaws of Mean Variance Optimization using market equilibrium and investor views.

How It Works

1. Start with Equilibrium: Instead of estimating historical returns (which causes MVO to break), Black-Litterman starts with the assumption that the "Market Portfolio" (e.g., global market cap weights) is perfectly balanced and represents equilibrium.
2. Reverse Engineer Returns: It reverse-engineers the expected returns that would justify these market-cap weights. This forms the baseline expected return vector.
3. Add Investor Views: The investor then specifies their own absolute or relative views on specific assets (e.g., "Tech will outperform Energy by 3%").
4. Bayesian Updating: The model mathematically blends the equilibrium expected returns with the investor's views based on the investor's level of confidence in those views.

Why It Matters

By doing this, the Black-Litterman model produces highly stable, diversified portfolio weights that smoothly tilt away from the global market cap purely based on where the investor actually expressed a confident active view. The rest of the portfolio remains balanced.