| Titel | Estimating Loss Functions of Experts |
|---|---|
| Gepubliceerd in | Report / Econometric Institute, Erasmus University Rotterdam, p.1-28. ISSN 1566-7294. |
| Auteur | Franses, Ph.H.B.F. (Philip Hans); Legerstee, R. (Rianne); Paap, R. (Richard) |
| Datum | 2011-12-15 |
| Trefwoord(en) | econometric models, asymmetry, expert forecasts, model forecasts, loss functions |
| Taal | Engels |
| Type | working paper |
| Uitgever | Econometric Institute |
| Samenvatting | We propose a new and simple methodology to estimate the loss function associated with experts' forecasts. Under the assumption of conditional normality of the data and the forecast distribution, the asymmetry parameter of the lin-lin and linex loss function can easily be estimated using a linear regression. This regression also provides an estimate for potential systematic bias in the forecasts of the expert. The residuals of the regression are the input for a test for the validity of the normality assumption. We apply our approach to a large data set of SKU-level sales forecasts made by experts and we compare the outcomes with those for statistical model-based forecasts of the same sales data. We find substantial evidence for asymmetry in the loss functions of the experts, with underprediction penalized more than overprediction. |
| Publicatie | http://hdl.handle.net/1765/31226 |
| Persistent Identifier | urn:NBN:nl:ui:15-1765/31226 |
| Metadata | XML |
| Repository | Erasmus Universiteit Rotterdam |
Omhoog
Ga terug naar de inhoud
Ga terug naar de site navigatie