How to Represent Mark-to-Market Possibilities with the General Portfolio Selection Model

• Published in: Journal of portfolio management Vol. 39; no. 4; p. 1
• Main Author: Markowitz, Harry M
• Format: Journal Article
• Language: English
• Published: London Pageant Media 01.06.2013
• Subjects: Algorithms; Capital assets; Investments; Optimization; Portfolio management; Selection; Software packages; Stochastic models; United States > US
• DOI: 10.3905/jpm.2013.39.4.001
• ISSN: 0095-4918

This article presents a different way to use general portfolio selection model (GPSM) to represent leverage aversion. Ignoring certain less-important differences between the two approaches, the principal difference is that the proposal presented here requires a separate model -- to evaluate the probability of a margin call and its likely consequences -- to help determine which portfolios are efficient when they consider the possibility of margin calls. Jacobs and Levy, by contrast, use alternate criteria to serve the same purpose. The author assumes that the leveraged investor, like the unleveraged investor, seeks a mean-variance-efficient portfolio. The only difference between the two investors is that the leveraged investor wants the calculation of mean and variance to take into account the possibility of margin calls between now and the next reoptimization. Any such procedure (analytic, numeric, or Monte Carlo) will do, as far as GPSM compatibility is concerned, because it is used in a separate side calculation. The author refers to this as the stochastic margin-call model, or SMCM.