Statistical Methods in Finance 2016

Abstract

The Black-Litterman Model for Portfolio Optimization

by Amita Sharma

The tendency of having major swings with respect to the small changes in the uncertain expected return in portfolio optimization is one of the big challenge in finance. Meanvariance model which is the first portfolio optimization model in the category of mean-risk models has been criticized for its fragile nature with respect to the changes in expected return. Unlike the simulation or historical technique to estimate expected returns which introduces noise into the efficient frontier of portfolio optimization, the Black-Litterman framework takes an entirely different route to estimate the expected return. The Black- Litterman model (BL), developed by Black and Litterman (1990) at Goldman Saches, uses an CAPM's equilibrium analysis to estimate the uncertain expected return and then blend these equilibrium estimates according to the investor's private information or view (if any) using Bayesian theory. Experimental analysis has shown that the portfolios constructed from BL approach are more balanced and better diversified than those constructed from the classical mean-variance approach. The U.S. investment bank Goldman Sachs regularly publishes recommendations for investor allocations based on the BL model. In this poster presentation, we explain the basics of the BL model and its extensions in direction of relaxing the assumptions of normal-return and the normal-views.

---

---