Modelo de Black-Litterman para la optimización de portafolios con views obtenidos por modelación de volatilidad

Abstract

The Black-Litterman model incorporates the market equilibrium returns and investors views to generate a new prediction of the return of the portfolio -- This model is applied for the optimization of stock portfolios in Colombia -- The main difference compared to the existing literature in Colombia is the use of GARCH processes for forecasting the returns that are used as views in the optimizer -- Portfolios are modeled weekly with a horizon of 20 trading days for the second semester of 2017 and the real returns of those portfolios adjusted by Black-Litterman versus the reference portfolios are compared -- It is found that 58.82% of portfolios outperform COLCAP with the suggested methodology. In addition, comparisons are made with respect to the measure of value aggregation (α), with Black-Litterman presenting a better performance