At the end of 2008, the average number of transactions per day on the NYSE was roughly 90 times what it had been at the beginning of 1993. During that same period, value-weighted monthly share turnover rose from approximately 5 percent to around 26 percent. The authors examine this steep increase in trading activity and its potential effects on market efficiency and pricing. Decreased trading costs are generally assumed to be a major factor in the observed increase in trading activity. The authors explore how this decrease in trading costs has affected different types of investors and how the increase in activity has played out in specific patterns. They also address the consequences of increased trading activity in terms of increased or decreased market quality and efficiency.

First, the authors look at value-weighted monthly turnover from 1993 through 2008, separating S&P 500 Index stocks from non-S&P 500 stocks listed on the NYSE. Non-S&P 500 stocks are smaller companies and less subject to indexation than S&P 500 stocks. The trend terms are positive and significant for both series, which suggests that company size and indexation are not material factors in the recent increased trading activity. Next, the authors examine average trade size and trading frequency. Average trade size has decreased substantially, whereas the average number of transactions per day has increased at an accelerating rate over the time period. The increased trading activity is, therefore, the result of more frequent, smaller trades.

The authors investigate whether the increase in trading activity has been driven by retail or institutional investors. They evaluate turnover for NYSE stocks, which are divided into five groups based on the level of institutional holdings. Next, the authors examine turnover resulting from large (greater than $10,000) trades versus small ones across the quintiles. The authors find that turnover has increased the most for stocks in which a higher percentage of shares are held by institutions. Small trade turnover also has increased much more for the group with the highest percentage of institutional holdings. The overall results suggest that increased turnover activity in recent years is derived mainly from institutions making more frequent, smaller trades. Changes in the number of shareholders, serial correlation in trade imbalances, and turnover patterns among low- and high-priced stocks reinforce the impression that lower trading costs prompted increased trading by institutions rather than by retail investors.

Next, the authors consider three determinants of trading activity—divergence in analysts’ forecasts, return volatility, and money flow into equity funds—and conclude that changes in any of these factors are not large enough or not in the correct direction to explain the steep increase in turnover.

Another important consideration of increased institutional trading activity is whether institutions have been able to trade more effectively on private information because decreased trading frictions may have increased returns from information-based trading. The authors use variance ratios computed from open-to-close and close-to-open returns to explore this idea. They examine the ratios across two subperiods, 1993–2000 and 2001–2008, and across quintiles based on the percentage of institutional holdings. They find that the variance ratios increased in the second subperiod and that the increase in the variance ratios is strongest for the stocks with the highest levels of institutional holdings.

To further explore any changes that may have taken place with regard to market efficiency, the authors investigate trends in intraday price fluctuations and deviations from random walk benchmarks. The results show that intraday volatility has decreased, particularly for stocks with a higher percentage of institutional holdings. The authors also find that prices conform more closely to random walk benchmarks in recent years, which supports the idea that increased institutional trading has led to increased market efficiency.

In an examination of the contribution of increased hedge fund trading, the authors discuss the growth of algorithmic trading by hedge funds and other institutions and suggest that although such algorithms are proprietary, the results are consistent with the use of algorithms by institutions to make smaller, more frequent trades to optimize order submission and efficiently execute trade orders.

In their conclusions, the authors address a potential implication for public policy that stems from their investigations. The idea of introducing a securities transaction tax has been raised in both the United States and the European Union on the assumption that an action intended to decrease trading activity would decrease price volatility. The authors suggest that the opposite may be the case; increased trading activity in recent years may have actually increased market efficiency and decreased intraday price volatility.