Best Estimate Reporting with Asymmetric Loss

16 Pages Posted: 30 Jan 2015

See all articles by Jostein Lillestol

Jostein Lillestol

Norwegian School of Economics (NHH) - Department of Business and Management Science

Richard Sinding-Larsen

Norwegian University of Science and Technology (NTNU)

Date Written: January 30, 2015

Abstract

This paper considers the problem of point prediction based on a predictive distribution, representing the uncertainty about the outcome. The issue explored is the reporting of a single characteristic, typically the mean, the median or the mode, in the context of a skewed distribution and asymmetric loss. Special attention is given to the two-piece normal distribution and asymmetric piecewise linear and quadratic loss. The practical context for the issue is the yearly reporting of remaining petroleum resources given by the authorities to stakeholders that may ask for just a single number.

Keywords: Estimate reporting, asymmetric loss, point prediction, predictive distribution

JEL Classification: C00, C10, C13

Suggested Citation

Lillestol, Jostein and Sinding-Larsen, Richard, Best Estimate Reporting with Asymmetric Loss (January 30, 2015). NHH Dept. of Business and Management Science Discussion Paper No. 2015/7, Available at SSRN: https://ssrn.com/abstract=2557941 or http://dx.doi.org/10.2139/ssrn.2557941

Jostein Lillestol (Contact Author)

Norwegian School of Economics (NHH) - Department of Business and Management Science ( email )

Helleveien 30
Bergen, NO-5045
Norway

Richard Sinding-Larsen

Norwegian University of Science and Technology (NTNU) ( email )

Høgskoleringen
Trondheim NO-7491, 7491
Norway

Here is the Coronavirus
related research on SSRN

Paper statistics

Downloads
27
Abstract Views
274
PlumX Metrics