RT Journal Article
SR Electronic
T1 Geometric Mean Maximization: Expected, Observed,
and Simulated Performance
JF The Journal of Investing
FD Institutional Investor Journals
SP 109
OP 119
DO 10.3905/joi.2013.22.2.106
VO 22
IS 2
A1 Rafael De Santiago
A1 Javier Estrada
YR 2013
UL https://pm-research.com/content/22/2/109.abstract
AB Portfolios can be optimized in a wide variety of ways, depending on the definition of risk and the goal stated. Although the traditional criterion of maximizing a portfolio’s Sharpe ratio remains the standard, many other alternatives exist and are currently used by practitioners. One of those alternatives is to maximize a portfolio’s geometric mean return, which amounts to maximizing the expected growth of the capital invested, or, similarly, the capital expected at the end of a holding period. In this article, we assess the expected, observed, and simulated performance of this criterion, and we compare it to those of the traditional criterion. We find that geometric mean maximization outperforms Sharpe ratio maximization in more than one dimension, ultimately providing investors with higher growth, much higher upside potential, and rather limited downside potential.TOPICS: Portfolio construction, simulations