Asymptotic skewness coefficient of maximum likelihood estimators in beta-prime linear
regression models with variable precision
Asymptotic skewness; Maximum likelihood estimators; Beta-prime distribution;
Regression models;
In this work, we derive an explicit closed-form expression, in matrix form, for the
asymptotic skewness of the maximum likelihood estimators in the class of beta-prime linear
regression models proposed by Bourguignon, Santos-Neto e Castro (2021). The general
expression can be easily implemented in any statistical programming environment that
performs matrix operations. We also consider Mont Carlo simulations and present two empirical
applications that illustrate the usefulness of the main result in practice.