Abstract
Central to many econometric inferential situations is to estimate non-linear functions of parameters. The mainstream in econometrics estimates these quantities based on a plug-in approach, where parameter estimates are just plugged in the objective expressions without consideration of the main objective of the inferential situation. However, this approach suffers of many shortcomings such as infinite moments and unbounded risks. Therefore, we introduce the Bayesian Minimum Expected Loss (MELO) approach using generalized loss functions to estimate functions of parameters, and calculate their frequentist variability to avoid extra computational burden. Simulations exercises show that our proposal outperforms competing alternatives in situations characterized by small sample sizes and noisy models. In addition, we observe in the applications that our approach gives lower standard errors than frequently used alternatives.
Acerca del expositor
Andrés Ramírez Hassan es PhD en Estadística, Master en economía y finanzas y economista. Profesor, investigador y consultor de la universidad EAFIT desde el año 2005. Adicionalmente trabajó en la universidad Nacional de Colombia, mientras que el sector corporativo se desempeñó en Empresas Públicas de Medellín. Ha publicado más de 30 artículos en revistas indexadas nacionales e internacionales, y ha liderado proyectos de consultoría para algunas de las empresas más destacadas en Colombia.