Newsletter

Noticias

Conferencia: Using Factor Portfolios to Reduce Estimation Risk

Expositor: Francisco Rosales

Departamento Académico de Finanzas y Centro de Investigación de la Universidad del Pacífico

Universidad del Pacífico (UP)

Resumen: Under the expected loss function of Kan and Zhou (2007), we determine the impact on out-of-sample performance when the sample mean and sample covariance matrix are used to construct the optimal mean-variance (MV) portfolio but when it satisfies a set of linear constraints. Then, we propose a novel portfolio rule that is theoretically superior to the three-fund rule of Kan and Zhou (2007). This rule combines the risk-free asset with two sample portfolios with zero expected out-of-sample covariance: a portfolio with minimum variance and full exposition to the factors and a portfolio with zero exposition to the factors and maximum ex-ante performance. The theoretical implication of our results is that an optimal combination of two uncorrelated portfolios generated by adding constraints to the MV has an important impact in reducing estimation risk in portfolio optimization. 

* La conferencia se dará en castellano

Fecha y lugar: IMCA, Calle Los Biólogos 245, La Molina

Viernes 16 de junio de 2017, a las 16:00