Institutional Repository of Vilnius University
Laiko eilučių išskaidymo ir prognozavimo metodų palyginimas Lietuvos apdirbamojoje pramonėje /
Abstract
dc:descriptionThis paper examines the impact of temporal disaggregation methods on forecasting accuracy in the context of Lithuanian manufacturing. The aim of the study is to compare proportional, Chow-Lin, Litterman and Fernandez disaggregation methods and evaluate their effectiveness in combination with Holt-Winters and SARIMA forecasting models. The study uses annual revenue and monthly sales index data from the Official Statistics Portal covering the period 2001-2024, encompassing the total manufacturing sector and 21 subsectors. It is found that the Chow-Lin autocorrelation parameter is estimated at or near unity in 21 out of 22 sectors, making Chow-Lin, Litterman and Fernandez-IND methods numerically equivalent for these data. The Litterman method outperforms the proportional method in forecasting accuracy, while Fernandez-RW achieves the lowest forecast errors, though for structural rather than methodological reasons. The choice of forecasting model (Holt-Winters or SARIMA) is less consequential than the choice of disaggregation method.
Degree
thesis:*- Grantor dc:publisher
- Institutional Repository of Vilnius University
- Year dc:date
- 2026
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Končius, Aurimas,
Rights
dc:rights- Statement dc:rights
-
- info:eu-repo/semantics/openAccess
- Language dc:language
- lit
Identifiers
dc:identifier.*- Repository record dc:identifier
- https://repository.vu.lt/VU:ELABAETD308334610&prefLang=en_US