K-Means Clustering Using Principal Component Analysis (PCA) Indonesia Multi-Finance Industry Performance Before and During Covid-19

Abstract

The cluster analysis within specific industry such as in multi finance indsutries is designed to be a tool for accelerating investment decisions, such as whether to buy, sell, or hold stocks in a way to construct an optimized portfolio. The purpose of the study was to apply cluster analysis on multi-finance stock data listed on the Indonesia Stock Exchange in the years 2019 and 2021, before and during Covid-19, using the PCA (Principal Component Analysis) K-means algorithm. The objective of this study is to classify stocks based on PCAs in order to assist investors in segmenting a multi-finance stocks cluster. The clustering is done on the 16 stocks registered in ISE using two-time windows: 2019 data where Covid-19 has not yet occurred and 2021 data where Covid-19 is still ongoing, and the firm is still in the recovery stage. The cluster analysis results show 12 companies worth investing in because they performed well. There is finding that company that have unfavorable Covid-19 externalities since this cluster has worsening performance and is thus not advised as a stock investment. Meanwhile, the others company has neutral externalities because it remains in the same cluster in 2019 and 2021.

Keywords

• pca, k-means, clustering, multi-finance, covid-19, investment