THE APPLICATION OF PRINCIPAL COMPONENT ANALYSIS AND FACTOR ANALYSIS TO STOCK MARKETS RETURNS

It is evident from literature that few Asian markets have strong linkage with US markets. The movement in one market often affects the other market if they have linkage. Current study explores the linkage among the major Asian and US stock markets using PCA and FA techniques. PCA identifies patterns in series on the variability while FA defines the structures using covariance and correlation. In this study, weekly closing returns of fourteen stock markets namely: KSE-100 (Pakistan), Nikkie225 (Japan), S&P 500(US), Nasdaq Composite and DJI (US), KLSE (Malaysia), BSESN (India), HIS (Hong Kong), JKSE (Indonesia) SSE(China), KS11(Korea), TWII (Tiwan), CSE(Sri Lanka) and TASI (Saudi Arabia) spanning from 1st January , 2001 to 14th January, 2019 are used as case study. Only first nine PCA?s are constructed and from them first six PCA?s are chosen as they contain almost 79.4% of total variability. First two PCA provide important information, such that PC1 consist of a group of US and few Asian stock markets (BSESN, NIKKIE 225, HIS, KS11 and TWII). Whereas, PC2 contains all US and Asian markets (JKSE, KLSE and SSE). Furthermore, no relationship of KSE-100, CSE and SSE with US markets is found. Moreover, Factor analysis with VARIMAX method gives quite different results as compare to PCA. FA1 comprises of a group of US markets and FA2 represents only two Asian stock markets (KS11 and TWII). Finally, FA method is found more appropriate as it utilizes correlation/covariance as well.