Active portfolio management using the black-litterman model

Abstract

In this project, we are studying the application of the Black-Litterman model for active portfolio management. We focus on three elements that are important for a real-world application of the model. First, we present a robust construction of the variance-covariance matrix that is used as input for the model. Second, we present a rigorous explanation of the model as it was presented initially by Black and Litterman, which is important to highlight the driving mechanisms inside the model. Third, we describe how to adjust the model to an application in an active setting, where the investor is evaluated relative to a benchmark. Empirical evidence for the active case is limited to date. Working with a range of benchmarks (relevant for various types of investors), we derive a portfolio that includes multiple asset classes and present the results obtained under the application of the model. Finally, a simulation exercise was developed to understand how the model reacts to its inputs.