Since Harry Markowitz published his revolutionary paper in 1952 which eventually led to the development of a whole new branch of finance and a Nobel Prize four decades later, Modern Portfolio theory (MPT) has played an important part in investment management. However, what Modern Portfolio Theory promised in its genius of the Efficient Frontier was never borne out in its ability to produce practical portfolios. The failure is mainly due to certain unrealistic assumptions within Markowitz’s solution to the asset allocation problem. More specifically, MPT assumes error-free mean variance inputs as well as perfect or near-perfect beta estimation for using the Capital Asset Pricing Model (CAPM) to price securities. But as with many important and evolving theories, testing and evolution helps genius reach toward its full potential.

In 1992, the Black-Litterman formula appeared on the scene. Their new formula for creating portfolio weights was supposed to address some of the deficiencies encountered with the original Markowitz Optimization algorithm. Indeed, its simplicity, ease of application, and use of subjective investor views helped to make the Black-Litterman Optimization algorithm popular with the investment industry, and for the past twenty years, the B-L model and its products have been applied to the investment of trillions of dollars. However, closer examination of what the Black Litterman formula really does may engender skepticism that it is adding any investment value. Especially since better methods are available for creating portfolios which use investor views and take far better advantage of information available from observation of historical behavior of assets in the portfolio, even when that information is thin, uncertain, or error prone.

In 1998, the New Frontier research team developed and patented a new worthy successor to MPT: the Michaud Resampled Optimization algorithm. During the fourteen years since its introduction, enough evidence has been gathered to make a valid comparison of all three of the optimization algorithms (Markowitz, Black-Litterman, and Michaud Optimization models) and a shot has been fired across the bow of investment managers who use Black-Litterman or any of its associated methods.

In the article “Deconstructing Black-Litterman: How to Get the Portfolio You Already knew You Wanted,” recently published in the Journal Of Investment Management, the authors go into detail about the underlying mathematical, statistical, and financial reasons why Black-Litterman does not adequately address the weaknesses that have plagued the original MPT optimization framework as well as which of the three methods provides the best results.

For the reader of this blog or anyone they know using the Black-Litterman Optimization or any of its products, the new article is a must read.