An Extension of Markowitz' Modern Portfolio Theory for Long-Term Equity Investors

Abstract

In this work, I introduce a new investment strategy for long-term equity investors. In particular, my proposed portfolio relies on the mean-variance approach as in Markowitz (1952), and it requires the same input estimates and is vividly located on the efficient frontier. The analytic (in-sample) formula of my portfolio is thus derived from a rather simple principle in finance: 'High potential returns require high risk'. Based on this fact, I propose a theory assuming that the importance of a portfolio standard deviation must somehow decrease for expanding investment horizons, and I therefore imply that risk-adjusted metrics such as Sharpe ratio and information ratio cannot adequately be used for judging a portfolio strategy over the long-term. Specifically, I compare my new strategy with other traditional portfolios for a set of 14 US stocks, during an out-of-sample time period from 1/1975 until 6/2019. For all portfolios, I consider the buy & hold-, a naïvely rebalanced- and a strategic procedure, where in the latter case the portfolios are rebalanced annually based on an expanding estimation window. The empirical results prove that my proposed strategy significantly outperforms all other traditional portfolios in the long run, in terms of cumulative performance.

Keywords: Portfolio Theory, Markowitz-Extension, Factor Portfolio, Equity Strategy, Long-Only

JEL Classification: G11, G12, G17