摘要:The study attempts to investigate the relation between news articles and stock returns, focusing on immediate impact of Pak and Gulf Economist articles. It endeavors to figure out whether such quantification of news articles has a relation with Karachi Stock Exchange (KSE)-100 index. Sample comprised of 511 words picked manually for analysis from 760 weekly issues of Pak and Gulf Economist. Correlation and regression analysis was applied to word indices and KSE excess returns for a period of 16 years from 1999 to 2014. Main finding was that words were not only correlated with KSE- excess returns rather they also had a causal relation with index. Augmenting benchmark models with word indices enhanced the forecasting power of those models.
关键词:Behavioral finance; Karachi Stock Exchange (KSE)-100 index; text mining; R-word index; word count index; efficient markets