MARKET MICROSTRUCTURE EFFECTS OF SHARE DISTRIBUTIONS

This study investigates various market microstructure effects of share distributions, i.e. effects on volatility and liquidity. Since the two liquidity measures of volume and spread are related, we start by an investigation of components of the bid-ask spread on the Helsinki Stock Exchange (HSE) during different levels of trading activity. The stock price volatility, average trade size, trading turnover, trading volume and a coefficient reflecting the tick sizes were found to be significant variables in explaining the bid-ask spread. Further, the results indicate that the coefficients exhibit variation depending on the volume period. A comparison to results from dealer markets such as the NASDAQ indicates that trade sizes, volume and turnover have a larger effect on the HSE, whereas volatility generally has a smaller effect. Next we investigate liquidity changes after rights issues, institutional placements, and acquisition oriented issues on the HSE. For rights issues, we document a significant spread decrease, large but not significant volume increase and unchanged turnover. The results differ from results in Kothare (1997) for a sample of NASDAQ stocks where a spread increase and no volume change is reported. For institutional placements a significant spread decrease, (median) volume increase, and unchanged turnover is found. The results for acquisition oriented issues yield mostly insignificant results for all liquidity measures. Finally we investigate the effects of stock dividends and splits on bid-ask spreads, trading volumes, trade sizes, and liquidity. We find no change in daily market adjusted volumes, but find decreased market-adjusted spreads. The result for volumes is consistent with results on U.S. data, while the result for spreads is not. Our results further indicate that while aggregate trading volumes do not increase, liquidity improves for small traders. We also find significant volatility increases using daily and weekly close-, bid-, and ask-returns. The result can not be explained by increased noise trading, since weekly volatility increases by an approximately equal amount as daily volatility, and the test for a change in noise trading suggested by French and Roll (1986) is not significant. We find that the volatility shift is positively related to changes in the bid-ask spread, number of trades, and market variance, and negatively related to pre-split variance. The relationship between the bid-ask spread and the vo latility change seems, however be mainly attributed to higher costs of asymmetric information that high volatility causes, and not to larger bid-ask bounces.