Portfolio Construction and Global Asset Allocation : A Practitioner Solution to a Black-Litterman Flaw

This paper presents a variation of the Black-Litterman Model (B&L) for portfolio construction and global asset allocation practices. The methodology proposed retains the Bayesian approach of the original B&L model and, in particular, the derivation of the Posterior Vector of expected market returns incorporating subjective investors' views. The variation from the canonical model we are presenting is in the construction of the Prior Equilibrium Vector of implied market return (B&L Market Equilibrium Starting Point), which we derive without the help of questionable models (e.g. the CAPM), without restrictive and subjective assumptions, without the need to pre-determine an all-encompassing investment universes/global benchmarks, and without forecast. Our version of the vector of market implied expected returns provides an unbiased and more robust equilibrium starting point, thus generating superior portfolios as measured by the Sharpe Ratio when applied to the Black-Litterman construct to determine the Posterior Vector. In addition, our methodology can start from any subjective investment universe, thus completely removing the logistical hurdle of defining a global universe