comparison among high dimensional covariance.pdf

Revista Colombiana de EstadísticaISSN: 0120-1751revcoles_******@ Nacional de ColombiaColombiaGómez, Karoll; Gallón, parison among High Dimensional Covariance Matrix Estimation MethodsRevista Colombiana de Estadística, vol. 34, núm. 3, diciembre, 2011, pp. 567-588Universidad Nacional de ColombiaBogotá, ColombiaAvailable in: ?id=89922501010 How to cite Complete issue More information about this article Journal's homepage in Scientific Information work of Scientific Journals from Latin America, the Caribbean, Spain and PortugalNon-profit academic project, developed under the open access initiativeRevista Colombiana de EstadísticaDiciembre 2011, volumen 34, no. 3, pp. 567 a parison among High Dimensional CovarianceMatrix Estimation paración entre métodos de estimación de matrices de covarianzade alta dimensionalidadKaroll Gómez1,3,a, Santiago Gallón2,3,b1Departamento de Economía, Facultad de Ciencias Humanas y Económicas,Universidad Nacional de Colombia, Medellín, Colombia2Departamento de Estadística y Matemáticas - Departamento de Economía,Facultad de Ciencias Económicas, Universidad de Antioquia, Medellín, Colombia3Grupo de Econometría Aplicada, Facultad de Ciencias Económicas, Universidadde Antioquia, Medellín, urate measures of the volatility matrix and its inverse play a centralrole in risk and portfolio management problems. Due to the accumulationof errors in the estimation of expected returns and covariance matrix, thesolution to these problems is very sensitive, particularlywhen the number ofassets (p) exceeds the sample size (T). Recent research has focused on de-veloping di?erent methods to estimate high dimensional covariance matrixesunder small sample size. The aim of this paper is to examine -pare the minimumvariance optimal portfolio constructed using ?ve di?erentestimation methods for the covariance matrix: the sample covariance, Risk-Metrics, factor model, shrinkage and mixed frequency factor model. Usingthe Monte C