Construcción de un portafolio de inversión con instrumentos de renta variable en mercado de valores estadounidense

Abstract

Portfolio construction is the process of selecting and combining financial assets with the aim of achieving a specific economic goal. To achieve such objective, contextualizing portfolio construction through a quantitative model is a valid approach, however, complexity quickly increases as the number of variables that mimic real investment scenarios are accounted for. In this thesis, portfolio construction is solved though classic one-period portfolio theory, such that, this process can be stated as a static portfolio choice problem. It will be shown that by using the assumption of Log-Normality of returns and that a risk averse individual possesses a CRRA utility function, an optimal portfolio is defined by maximizing an expected utility function of M-V parameters. With this theoretical model at hand, the main objective becomes to construct portfolios using financial data of 49 U.S industry-portfolios taken from the Kenneth French Repository. The optimization process is handled by two algorithms that were built in order to fasten calculations and to allow us to analyze a set of risk preferences and the effects of diversification.