Approximate Method of Estimation of Exponential Trend Parameters for Forecasting Process Purposes

Bednarz-Okrzyńska Kamila ()
Additional contact information
Bednarz-Okrzyńska Kamila: Ph.D. University of Szczecin Faculty of Management and Economics of Services Department of Quantitative Methods Cukrowa 8, 71-004 Szczecin, Poland

Folia Oeconomica Stetinensia, 2018, vol. 18, issue 1, 169-181

Abstract: The paper discusses the issue of estimation of exponential trend parameters in terms of its application in the forecast process. Due to the character of a random element, three models were considered: additive, multiplicative, and mixed. For estimating trend parameters, a log transformation method, least squares method, and approximate methods were applied. As a result of computer simulations, high sensitivity of the log transformation method with regard to the assumed random element model was noticed. This method yields the smallest value of ex post error for the multiplicative model but is burdened with a large error for the additive model, where the estimated parameter B takes large values (B > 0.24). In the paper, a new approximate method of estimation of exponential trend parameters is proposed. The method is compared with approximate formulas presented in the paper by Purczyński (2008).

Keywords: exponential trend; parameter estimation; least squares; approximate methods (search for similar items in EconPapers)
Date: 2018
References: View complete reference list from CitEc
Citations:

Downloads: (external link)
https://doi.org/10.2478/foli-2018-0013 (text/html)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:vrs:foeste:v:18:y:2018:i:1:p:169-181:n:13

DOI: 10.2478/foli-2018-0013

Access Statistics for this article

Folia Oeconomica Stetinensia is currently edited by Waldemar Tarczyński

More articles in Folia Oeconomica Stetinensia from Sciendo
Bibliographic data for series maintained by Peter Golla ().