Template-type: ReDIF-Article 1.0
Author-Name: Ekaterina V. Astafyeva
Author-Workplace-Name: Russian Presidential Academy of National Economy and Public Administration
Author-Name: Maria Yu. Turuntseva
Author-Workplace-Name: Russian Presidential Academy of National Economy and Public Administration; Gaidar Institute for Economic Policy
Title: Analysis of Opportunities to Improve the Quality of Natural Resource Price by Combining Forecasts Resulting from Methods Based on Regression Estimates of Weights
Title: Анализ возможностей улучшения качества прогнозов цен на природные ресурсы методами комбинирования на основе регрессионных оценок весов
Abstract: Numerous empirical works indicate that combining (aggregation) forecasts can improve forecast accuracy compared to individual forecasts. This paper investigates the possibilities of regression methods of forecast aggregation to improve the forecast quality of prices for oil, aluminum, gold, nickel and copper. The calculations are based on the forecast database of the Gaidar Institute for Economic Policy. E.T. Gaidar Institute for Economic Policy, which provides an array of individual (aggregated) forecasts. All calculations are performed in (pseudo) real time. 
Based on the findings obtained in the paper, it can be argued that for resource prices, regardless of the period under consideration, there is a regression method of aggregation that provides qualitative advantages relative to all primary forecasts. At the same time, summarizing the results of qualitative characteristics of regression and simple aggregation methods, it should be noted that the choice of the best method of combining forecasts (and even a group of methods) is ambiguous and depends on the indicator.
Keywords: forecasts combination, forecast aggregation, oil prices, aluminum prices, gold prices, nickel prices, copper prices
Classification-JEL: C53, E31
Journal: Russian Economic Developments
Year: 2023
Issue: 12
Month: December
Pages: 24-33
File-URL: http://www.iep.ru/files/RePEc/gai/recdev/r23100.pdf
File-Format: Application/pdf
Handle: RePEc:gai:recdev:r23100