Asset allocation: comparison between the Markowitz approach and the Black-Litterman method

In 1952 Harry Markowitz published his best-known article, named Portfolio Selection, in the Journal of Finance. For the first time, the concept of mean-variance optimization was introduced, and this served as the foundation of modern portfolio theory and, later, for the Capital Asset Pricing Model (CAPM). Developed as a solution to practical portfolio optimization problems faced at Goldman Sachs, the Black-Litterman method was firstly published in the Journal of Fixed Income in 1991. This approach overcomes the main limitations of the Markowitz model which tends to create concentrated and unstable portfolios that rely excessively on past performance, without comprising investors’ views. This thesis aims to compare Markowitz’s portfolio allocation method with the one of Black and Litterman, from both a theoretical (chapter 3) and empirical (chapter 5) standpoint. With the aim to highlight the impact of the exogenous shock of Covid-19, this work examines a portfolio of 30 stocks diversified by geography, currency, and industry, over two different time horizons: 2015-2019 and 2020-2021. To capture the effects of the pandemic, certain industries were before-hand selected – airline and pharmaceutical companies as well as firms that exploit intangible economy, such as tech groups. Under the hypothesis of normal distribution of the logarithmic returns, the mean-variance optimization problem for the two analysed methods has been solved. The resulting empirical analysis demonstrated that the dynamic Black-Litterman capital allocation leads to more balanced, diversified, and stable portfolios which comprise investors’ views.