Stock price behavior in the NSE- a test of the predictability and seasonality of stock prices movements

Abstract

This study incorporates a_ predictive regression approach and a centered moving average analysis to test the predictability and seasonality respectively of the stocks on the Nairobi Securities Exchange (NSE) 20 Share Index. The findings reveal that most of the stock prices do not exhibit seasonality with the exception of a few such as Kenya Power, Centum, Bamburi which exhibit repetitive cycles. The t-test of significance portrays the shortcomings of the past stock prices in predicting future stock prices with Sasini showing the closest element of predictability (t-test value of 0.000290608) which is still very low.