Understanding Principal Component Analysis (PCA) in the Azerbaijan Economy: Case Studies of Vegetable and Fruit Sectors

Is it possible to apply Principal Component Analysis (PCA) in the Azerbaijan economy? Before blindly answering the question, careful examination of the collected data and the relevance of the data set to the mentioned analysis should be checked. PCA is a widely used multivariate dimension reduction tool that statistical analysis employs. Nevertheless, its application area is limited. Also, PCA analysis in macroeconomic studies is not much preferred and in Azerbaijan's case, there is not one. To understand this methods' relevance two case studies, namely vegetable and fruit sectors have been chosen. The aim was to quantify the sub-sectoral performance via vegetable and fruit sectors. The collected data set and PCA analysis on it were evaluated from multiple angles to outline the roadmap for future investigations. The author argues that in line with the method's theoretical expectations, Azerbaijan's vegetable and fruit sectors partially fulfill the planned goals of the analysis. This means PCA is a highly useful analytical tool and might produce valuable sub-sectoral performance evaluations in the case of Azerbaijan. According to the produced indices, the vegetable sector performed better than the fruit sector from 1999 until 2015. However, between 2015-2020 fruit sector outperformed the vegetable sector. A similar analysis can be organized among the other sub-sectors or between sectors in the case of the Azerbaijan economy. Furthermore, this working paper embodies experimentation results, not conclusions about the methods' validity and justification. Hence, this work compares three rotations of PCA analysis (Varimax, Quartimax, and Equamax) and constructed indices separately for vegetable and fruit sectors.