A1360
Title: Estimation of risk aversion coefficient for tangency portfolio
Authors: Stepan Mazur - Orebro University (Sweden)
Stanislas Muhinyuza - Linnaeus University (Sweden) [presenting]
Abstract: The purpose is to investigate the distributional properties of the estimated risk aversion coefficient for the tangency portfolio (TP), assuming that the asset returns follow a multivariate normal distribution and/or a matrix-variate closed skew-normal distribution. Under the normality assumption, a stochastic representation, density function, characteristic function, and high-dimensional distribution are constructed to fully characterize the estimated risk aversion coefficient distribution for both finite and high-dimensional settings. For a matrix-variate closed skew-normal distribution assumption, a stochastic representation is provided that is later used to derive the first and second moments and the high-dimensional distribution of the estimated risk aversion coefficient. Through a simulation study, it is shown that the derived high-dimensional asymptotic distributions provide a good approximations to the exact ones for the finite sample sizes under both distributional assumptions considered.
