Generalized Auto-regressive Conditional Heteroscedastic Modeling of the Volatility of Stock Returns in the Nairobi Securities Exchange

Abstract

The most important characteristic of a stock or bond is its return or profit. This return is volatile and tomorrow’s price is uncertain and must be described by a probability distribution. The purpose of this study was to develop a model of stock returns in the Nairobi Securities Exchange (NSE) using the Generalized Autoregressive Conditional Heteroscedastic (GARCH) model. Closing prices of Safaricom and Kenya Commercial Bank (KCB) were obtained from the NSE for the period January 2011 to October 2014 which formed 1000 observations excluding weekends and holidays. Test for normality and stationarity was done using the Shapiro–Wilk test and Augmented Dickey Fuller (ADF) respectively. All the return series exhibited, leptokurtosis, volatility clustering and negative skewness. The estimation results reveal that GARCH (1, 1) best fits both return series over the period of study.

Keywords: Heteroscedasticity. Stock Returns, Volatility