The Relationship between the Purchasing Managers’ Index (PMI) and Economic Growth: The Case for Poland

Radoslaw Sobko, Maria Klonowska-Matynia
European Research Studies Journal, Volume XXIV, Special Issue 1, 198-219, 2021
DOI: 10.35808/ersj/2038


Purpose: The article aims to analyze and evaluate the relationship between the Purchasing Managers’ Index PMI (economic sentiment indicator) and GDP dynamics in the Polish economy. The subject of detailed research was the possibility of forecasting Poland’s economic situation using a model built based on the PMI sentiment indicator. Approach/Methodology/Design: The study used data on GDP dynamics, EUR/PLN and USD/PLN exchange rates, as well as two indicators of economic sentiment prepared by independent institutions for Poland: the PMI and ESI indicator. The analysis was based on quarterly data for the period from the third quarter of 1998 to the second quarter of 2019 (84 observations). PMI data came from the website, ESI data from the European Commission database, GDP dynamics data from the World Bank database, and exchange rate information was taken from the website. The analysis contained in the article was performed using the ARDL and ECM models. Findings: The analysis showed that the model based on PMI indicator and the model based on ESI indicator is too inaccurate to be considered a tool for forecasting the economic situation in Poland. It also turned out that extension of the model with other explanatory variables increased its accuracy of fitting to real data. Practical Implications: Even though the estimated models were significantly unreliable, it turned out that in periods of greater economic instability, the PMI model showed better forecasting properties. This indicates the possibility of using the PMI model, e.g., in times of recession or economic crisis. Originality/Value: The article broadened the research perspective for forecasting the Polish economy. The results set the directions for further development of research in this aspect. It turned out that probably the optimal solution would be to create different models for different phases of the business cycle, or a different rate of economic growth.

Cite Article (APA Style)