EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

The Loss Function of USDA Forecasters: Evidence from WASDE Animal Product Price Forecasts

Chad Fiechter, Siddhartha S. Bora, Todd H. Kuethe

Journal of Agricultural and Applied Economics · 2026-05 · 58 (2) · 243-260

DOI: 10.1017/aae.2025.10031 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Agricultural and Applied Economics, online erschienen am 23.12.2025. Verfasst von Chad Fiechter und 2 weiteren Autor:innen. Die Arbeit nennt 38 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The loss function is a mathematical representation of the costs experienced by a forecaster when observed outcomes differ from what was predicted. Prior studies suggest that USDA forecasts are not optimal based on an assumed mean-zero quadratic loss function. This study proposes an alternative view of forecast evaluation, which assumes all USDA forecasts are produced to minimize the forecasters’ costs, and searches for the dimensions of the loss function under which optimality holds. We illustrate the degree to which USDA loss functions vary across a series of WASDE price forecasts. A better understanding of USDA forecasters’ costs will benefit forecasters and forecast users.

Zitieren

APA

Fiechter, C., Bora, S. S., & Kuethe, T. H. (2026). The Loss Function of USDA Forecasters: Evidence from WASDE Animal Product Price Forecasts. Journal of Agricultural and Applied Economics, 58(2), 243-260. https://doi.org/10.1017/aae.2025.10031

Vancouver

Fiechter C, Bora SS, Kuethe TH. The Loss Function of USDA Forecasters: Evidence from WASDE Animal Product Price Forecasts. Journal of Agricultural and Applied Economics. 2026;58(2):243-260. doi:10.1017/aae.2025.10031

BibTeX
@article{Fiechter2026,
  author = {Fiechter, Chad and Bora, Siddhartha S. and Kuethe, Todd H.},
  title = {The Loss Function of USDA Forecasters: Evidence from WASDE Animal Product Price Forecasts},
  journal = {Journal of Agricultural and Applied Economics},
  year = {2026},
  volume = {58},
  number = {2},
  pages = {243--260},
  doi = {10.1017/aae.2025.10031}
}
Kostenlos starten →