Stock splits in confliction with the economic irrelevance of shares outstanding : An event study on the Stockholm Stock Exchange

Abstract: A survey is conducted through an event study on the Stockholm Stock Exchange based on 119 historical stocks splits with a split factor of at least two, for the years between 1997 and 2012. This study has tested if there is an increase in return variance and systematic risk followed by a stock split. This is a quantitative study with the deductive approach and the positivistic epistemological standpoint. By matching 8925 squared daily returns for 75 days of pre- and post- split data, the sample of stock splits showed an increased return variance 0.515 of the matched squared daily returns, this number is significant at the 1% level in our binomial z-statistics. If the returns are compared on a 15 week interval instead of 75 days, the change in variance disappears; this confirms Dubofsky (1991) findings. When 52 weeks of pre- and post- split data is used, there is an increased variance in a proportion if 0.55 of the 6186 matched observations, this proportion is far greater than our daily sample and tells us that there is a long term effect on the return variance. The systematic risk measured as beta derived from the CAPM, did not show any increase in any of the three different time periods (75days, 16weeks and 52 weeks); the results confirms Wiggins (1992) findings; beta changes are just illusory.  The results suggest that there is an average increase in returns variance in the short and long term after a stock split, that confirms some existing studies by Ohlson and Penman (1985) and Dubofsky (1991). The increase in returns variance can be viewed as the management’s success of signaling the market, enhancing liquidity and reducing information asymmetry without any additional cost of capital. Our findings also contradict the theory of economic irrelevance of shares outstanding. This study is expanding Ohlson and Penman (1985) and Dubofsky (1991) studies, on a European stock market.