The Black-Litterman model is an asset allocation framework that blends global market equilibrium returns with an investor's subjective views to produce well-behaved portfolios. Black and Litterman, in "Global Portfolio Optimization" (1992), used monthly returns for seven countries from January 1975 to August 1991 to test the model. A globally diversified, currency-hedged bonds-and-equities portfolio earned a 5.61% expected excess return versus 4.76% for a domestic-only portfolio at the same 10.7% risk level. The 85 basis-point gain reflects equilibrium expected returns anchored to an 80% universal currency-hedging constant before any tilt toward investor views.
What the Study Found
Without currency hedging, a globally diversified equities-only portfolio earned a 5.48% expected excess return versus 4.72% for a domestic-only portfolio. That gap equals a 76 basis-point gain at a constant 10.7% risk level. The equilibrium optimal portfolio allocated 29.7% to U.S. equities and 16.3% to U.S. bonds, with 80% of currency exposure hedged. The annualized equilibrium risk premium for U.S. equities was 7.32%, compared with an annualized historical mean excess return of 5.2% over the same period. A moderate three-month view shifted expected U.S. returns: bonds up 0.8 percentage points, equities down 2.5 percentage points.
Methodology
The analysis uses monthly excess returns on equities, bonds, and currencies across a seven-country model. The full sample period runs from January 1975 through August 1991. The model treats market-capitalization weights and an 80% universal currency-hedging constant as the equilibrium benchmark, with no explicit statistical controls.
Key Statistics
| Metric | Finding | Context |
|---|---|---|
| Universal hedging constant | 80% | Equilibrium degree of currency hedging used in the model |
| Global vs. domestic (bonds + equities, hedged) | 5.61% vs. 4.76% (85 bp gain) | Constant 10.7% risk, full sample |
| Global vs. domestic (equities only, unhedged) | 5.48% vs. 4.72% (76 bp gain) | Constant 10.7% risk, full sample |
| Equilibrium risk premium (U.S. equities) | 7.32% | Annualized, full sample |
| Black-Litterman expected returns formula | E[R] = [(τΣ)⁻¹ + P′Ω⁻¹P]⁻¹[(τΣ)⁻¹Π + P′Ω⁻¹Q] | Combines investor views with equilibrium to produce posterior expected returns |
| Equilibrium risk premium vector | Π = δΣW | Derives neutral expected returns from the CAPM equilibrium condition |
Why This Matters
Portfolio managers can use equilibrium risk premiums as a neutral starting point instead of unstable historical-average forecasts. Tilting those neutral weights toward specific, confidence-weighted views avoids the extreme long and short positions that plague standard mean-variance optimizers. The framework therefore suits global asset allocators who hold opinions about only a few markets rather than every asset and currency. Treating the degree of confidence in a view as an input also lets managers control how strongly any single opinion shifts the final portfolio.