An empirical investigation of stock returns reaction around earnings announcements for quoted companies at NSE

Abstract

This study investigates stock returns and trading activity reactions around annual earnings announcements for listed companies at the Nairobi Stock Exchange (NSE) to verify" whether these announcements possess informational value. The event study is conducted on a portfolio of thirty (30) securities which made earnings announcements over the period of five calendar years from January 2002 to December 2006. This study contributes to the -:,dy of empirical research focused on the anomalies on the NSE. This study empirically examined if security prices and trading activity proxied by trading volume change in response t<. earnings announcements. For this purpose, the study investigated \ net.. er the magnitude of the price and volume changes (without respect to sign) is larger on the announcement day than during the non-event period. An analytical study was conducted in which quantitative data was collected and analyzed, that is, across the sampled companies an:' then through t::11": (cf the event study). The sn dy relied on secondary data available at NSE database on daily closi.ig share prices and trading volumes. The sample consists thirty (30) active listed companies in the Main Investment Market Segment (MIMS) that made annual ear. ings announcz-nents and with trading records on the NSE. Data was analyzed using event study methodol )g: based on the market model of Sharpe (1964) to eliminate market-wide elements of price and volume changes. Descriptive statistics used are thus, the mean, ordinary least square (OLS) regression analysis and t-statistic to analyze data collected on daily close share prices and trading volumes of sampled companies at NSE. The empirical results is that the average abnormal returns and average abnormal volume on announcements days are significantly larger than zero as compared with the non-event period. Therefore, the null hypothesis that states "stock returns and trading activity for quoted companies,' share' NSE do not react to earnings announcements" is rejected and alternative hypothesis that s "stock returns and trading activity of quoted companies' shares at NSE react to carrings announcements' is accepted". The findings are consistent with the prediction that earnings announcements possess informational value.