Market Volatility can affect investors' investment risk in a stock market. French and Ross (1986) argue that volatility is closely related to information, which can change market expectations and eventually share prices. Epps and Epps (1976) suggest that an increase in the extent to which traders disagree as to the valuation effect of new information can bring about larger absolute price changes and larger volumes. The information that affects stock market volatility comes primarily from financial and economic announcements. Another source of volatility is the trading by liquidity and asset allocation needs, market-timing decisions, and activities of institutional investors and individual investors to take arbitrage and hedge positions, program trading, etc. This type of information is closely related to the volatility attributable to the trading activity of investors. Hence, it is a very interesting and important issue in the behavioral aspect of stock market to empirically examine the relation between investors' different types of trading activities and market volatility. This paper studies the relation between each type of investors' trading activities and the return volatilities in a stock market. In our test, we use daily returns for the KOSPI computed with the use of log difference of daily closing prices from January 4, 1980 to June 30, 2003, for the full period and three sub-periods. First, we divide investor's trading activities into expected and unexpected components, then, examine the relation between each of trade activities and stock return volatilities. The result shows that the coefficients are statistically insignificant in case of expected trade activities, but, in case of unexpected trade activities, the result shows statistically significant positive coefficients. We also examine the effect of program trading activities on stock return volatilities. We can find, here, an interesting result that program trading activities have an effect on stock return volatilities more than 23 times as strong as unexpected trading activities. Therefore, we can draw the conclusion that a program trading activity has a very strong influence on stock return volatilities which have been increased recently. In further examination, program trading activities are divided into buy and sell components, and we test for relation between each volume and volatilities in the stock market. The result shows that when a unit of trade volume is fixed, buy-program trading activities have an influence on stock return volatilities almost 43 times as strong as unexpected trading activities. Consequently, we can find that a program trading activity has a stronger influence on stock return volatilities than other types of investor's trading activities. It is especially true in case of a buy-program trading activity, where the influence is the strongest of all types of trading activities. We also examine down markets based on previous researches which show that the influence of buy-program trading activity on stock return volatilities declines in down markets. Down markets are defined as stock returns decline by at least 1% from the previous day. This test result shows that both sell and buy-program trading activities increase stock return volatilities in a down market. These findings have obvious policy implications. We need more stringent regulations against unfair trading using program trading and systematic improvement of daily trading in order to halt this kind of trading, and this could also weaken the link. Fundamentally, it is necessary to overcome temporary market impact incurred by program trading in order to improve market liquidity which will, consequently, improve market depth.
