Korea Capital Market Institute

The purpose of this study is to conduct a rigorous empirical analysis to find evidence supporting arguments in the industry and media that program trading disrupts or causes unnecessary volatility in the stock market, and to consider institutional measures for mitigating the adverse effects on the market, such as stock price surges/crashes, increased volatility and other disruptions. Program trading first attracted attention when it was identified as a major cause of stock price crashes by the Brady Report of the U.S.’s Presidential Task Force on Market Mechanisms at the time of "Black Monday." Ever since, it has been repeatedly argued that program trading amplifies stock price fluctuations and market volatility. The argument that program trading disrupts or causes unnecessary volatility in the market was also continuously raised in Korea by the media and the industry. On the other hand, there is no consistent opinion in the academia regarding the impact that program trading has on stock market volatility, and the views are generally split into two. The first view is that program trading facilitates information transmission between the futures and cash market, thus increases market efficiency and helps reduce unnecessary volatility. The second view is that program trading actually creates supply-demand imbalance and decreases liquidity, thus increases volatility. Foreign literatures on the empirical analysis of these two arguments also fail to arrive at a concurrent view. After the 1987 "Black Monday," NYSE adopted measures to curb program trading, including "sidecar" and "trading collar." Sidecar, which refers to a temporary halt on all program trade orders, was introduced in October, 1988, but was later eliminated in February, 1999, when questions were raised on its effectiveness. Program trading "collar rule" as set forth in NYSE Rule 80A is a price limit imposed on program trades. When the cash market prices seriously declines (surges), program buying (or selling) order price must be below (above) the previous level. The Korean KRX also instituted the program trading control mechanism (sidecar) in order to minimize the price volatility risk generated by program trading. In the early days of adoption, sidecar was triggered frequently, as stock price volatility had increased in the aftermath of the financial crisis. However, the number of sidecar triggers has significantly decreased after the rule was modified twice to raise the threshold and limit the number of triggers to once a day. While the impact of program trading on market volatility is an important issue, there are only a small number of domestic empirical studies on this subject, and the period for analysis is confined to the late 1990s. In order to conduct a rigorous empirical analysis on the impact that program trading has on the market, long-run and short-run volatility were observed separately, and the impact was examined using the VA-CEGARCH model. For daily analysis, the study used daily returns and program trading data from January, 2000 to December, 2004. For intra-day analysis, the study used daily program trading data for 141 component stocks that were included in the KOSPI 200 throughout the period from January, 2003 to December 2004. In summary, the empirical study revealed that program trading and non-program trading both have no impact on the on long-run volatility but did increase short-run volatility. Such results can be explained by the fact that long-run volatility only reacts to information related to corporate fundamentals. In the case of short-run volatility, program trading was found to have a larger impact compared to non-program trading: the impact was larger by 3 folds in the daily analysis, and 1.1 folds in the intra-day analysis. Secondly, net program selling and buying were not found to have a consistent impact on short-run volatility. Program selling, however, was found to have a larger impact in the daily<