The effect of stock splits on return volatility of firms listed at the Nairobi Securities Exchange

Abstract

The overall objective of this study was to examine how the risk factors associated with investing in stocks are affected by stock splits. Previous studies on the effect of stock splits at the Nairobi Securities Exchange failed to examine the effect of stock split on return volatility which could have resulted in the rejection of the hypothesis of no abnormal returns subsequent to stock splits. An event study methodology was used for the splitting stocks at the Nairobi Securities Exchange over the period 2004-2012. A census was conducted on the twelve stock splits executed over the period. Volatility was measured using the standard deviation of return and the beta coefficient used as a measure of systematic risk. Stock returns were calculated over a 150 days rolling period with 75 days pre split and 75 days post split. A graphical observation of the daily standard deviation for an equally weighed portfolio of pre and post split returns standard indicated a temporary increase in standard deviation that faded away shortly. When the returns were tested for increase in the beta coefficient it was found that seven out of the twelve splitting stocks exhibited an increase in post split beta. A t-test for the significance of the increase in post split beta for the seven stock splits resulted in a t statistic of 3.922 which was significant at the 5% level. When the t-test was performed on the entire splitting group a t statistic of 1.118 was obtained which was not significant at the 5% level. The study found an increase in post split standard deviation of returns in six of the twelve (representing 50%) of the splitting stocks. An F-test for the significance of increase in post split standard deviation showed that only four of the six stock splits that exhibited an increase in standard deviation were significant at the 5% level representing just a third of the splitting stocks. The study concludes that there occurred a temporary increase in post split return volatility following a stock split at the Nairobi Securities Exchange. The increase in volatility cannot be generalized since not all splitting stocks exhibited the increase at 5% significance level. Also the increase in post split beta coefficient was observed in some but not all splitting stocks. The temporary increase in returns volatility on the ex-split date seems to render support to the optimal trading range hypothesis and signaling hypothesis. Managers can effectively use stock splits to restore stock prices to a trading range and also signal the market on future prospects thus increase the liquidity of their stocks. However they should carefully examine the consequences of such corporate action as it may result in an increase in cost of capital induced by an increase in the beta coefficient as observed in most of the splitting stocks. Investors and securities analysts should carefully examine the information content of stock splits in order to evaluate the risk effect of such stocks in their portfolio. Further study may seek to evaluate the causes of increase in return volatility in some of the splitting stocks since a stock split is an event that do not alter the firm’s fundamentals. This study used inter-day stock prices in measuring return; a further study may use intraday stock prices to evaluate the volatility effect of stock splits.