SEC Chairman Christopher Cox said analysis of the experiment by economists and academics will help the SEC decide what to do once the short-selling pilot program ends next summer.

The SEC ordered the experiment as a part of its Regulation SHO, an overhaul of short-sale rules in 2004. Short sellers sell borrowed shares in hopes of replacing them later a lower price. The practice is legal, but rules adopted in the 1930s by the SEC and U.S. markets restricted short sales to avoid rapid-fire price declines.

Critics say restricting short sales does more harm than good and recommend abolishing restrictions permanently. Yet lifting restrictions on short sales for stocks in the test has produced mixed results, according to researchers who have studied U.S. stock trading since the experiment began in May 2005.

A draft study by SEC economists concludes that order routing, short-selling mechanics and intraday market volatility has been affected by the experiment, with volatility increasing for smaller stocks and declining for larger stocks. Market quality and liquidity don't appear to have been harmed, the SEC economists added.

Overall, short selling seems to be easier and has risen significantly for stocks in the experiment, academics agree. Several found no evidence that these stocks have more downside volatility now than before.

"We didn't find any significant change in volatility," said Gordon Alexander, a finance professor at the University of Minnesota's Carlson School of Management and co-author of a study on the short-selling experiment.

A separate study by professors at Ohio State University's Fisher College of Business concurred. That study, titled "It's SHO Time!," showed the short-sale experiment is having a bigger effect on trading on the New York Stock Exchange than on the Nasdaq Stock Market.

Short-term volatility rose for NYSE-listed shares trading without short-sale restrictions, but has fallen on Nasdaq since the experiment began, including for shares that aren't in the test, the Ohio State researchers said.

Researchers chalk up the different effects to market differences, noting the NYSE's "uptick rule" forbids short sales unless the price of a stock is rising. Nasdaq uses a less restrictive bid test. Another factor: Electronic trading systems such as Archipelago and INET, which account for a big chunk of trading in Nasdaq shares, don't enforce short-selling restrictions, which researchers said may partly explain why lifting short-sale restrictions on some stocks appears to have had less effect on Nasdaq. SEC economists found that short-selling volume increased on Nasdaq since the experiment began, especially for Nasdaq 100 stocks, suggesting market participants had been routing orders away from Nasdaq to avoid restrictions.

Short selling is a staple of U.S. stock trading, with or without restrictions. The Ohio State professors found short sales for NYSE stocks rose to 25.7 percent of trading once short-sale limits were removed, up from 24.8 percent, and rose to 39.2 percent for Nasdaq stocks, up from 36.5 percent. Spreads, the difference between the price quoted to buy and sell a stock, rose on the NYSE after the experiment began, but there was little change in effective spreads, the study also found.

Another study by J. Julie Wu of Texas A&M University's Mays Business School, found significantly more shorting in small-cap stocks trading on the NYSE since the experiment, with large-cap stocks largely unaffected. Given that, Wu questioned "whether one-size-fits-all is the best approach in setting short-selling restrictions."

-- JUDITH BURNS (Dow Jones Newswires)