Supporting purchasing decisions with time series analysis – the case of a plastics company


The role of business forecasts is one of the most important factors of competitiveness in today’s world. There are many tools for estimating future data for planning, among which time series analysis plays a key role. However, multivariate time-series regression analyses have several hidden pitfalls (stationarity, cointegration), making everyday use very difficult, and the desired frequency of data is only sometimes available. In our study, we model the evolution of purchase prices of a Swiss-owned plastics company operating in Hungary. To address the difficulties of stationarity and cointegration in our time series regressions, we construct several models with the FM-OLS estimator after the necessary pre-testing and try to support their long-run usability with bivariate vector error correction models. Our results include a large number of models that are consistent with our expectations. Still, in some cases, further investigation of the economic environment and the inclusion of other methods may be necessary.

X. ÉVF. 2022. Küönszám 19-25

DOI: 10.24387/CI.2022.SI.4


Banque de France (2022): Main monthly average euro foreign exchange rates, (Letöltés dátuma: 2022.09.04.)

EIA (2022): Petroleum & Other Liquids, (Letöltés dátuma: 2022.08.20.)

Karabassov, R. – Fenyves, V. – Túróczi, I. – Tóth, R. (2022): The controlling approach as an integral part of financial culture, Economics & Working Capital 2398-9491 2022 (1-2 issues) pp. 42-47.

Kunststoff Information (2022): Polymerpreise, (Letöltés dátuma: 2022.08.18.)

Statista (2022): Average monthly price of Urals crude oil from January 2018 to September 2022, (Letöltés dátuma: 2022.09.10.)

Wang, Q. – Wu, N. (2012): Long-run covariance and its applications in cointegration regression, Stata Journal, Stata Press, Vol. 12 No. 3, pp. 515-542, ISSN: 1536-867X, Online ISSN: 1536-8734