OPTIMIZING ZERO BETA PORTFOLIOS: A COMPARATIVE ANALYSIS OF ROBUST AND NORMAL PORTFOLIO METHODOLOGIES

Abstract

When building a “zero beta portfolio”, neglecting the parameters’ uncertainty may harm the investor. This paper analyzes a way to build a zero beta portfolio that does not consider only the parameter points estimates, but also the beta and the expected return uncertainties. The stocks’ betas and their uncertainties are calculated using the Kalman Filter and the stocks’ expected returns and their uncertainties are calculated from analysts’ price and dividends estimations. The study applied two different methodologies to build a zero beta portfolio: one that maximizes the ratio between the expected return by the uncertainties of the parameters, called long-short robust portfolio; and another that simply maximizes the expected return, neglecting the uncertainties of the parameters, called as long-short normal portfolio. During the period analyzed, 2015-2022, compared to the long-short normal portfolio, the long-short robust portfolio had a higher realized return and a significantly lower standard deviation.