ok, I posted it ;p
This is from THE WSJ, not me
Some information is usually better than none, but investors who track a new set of short-selling data would be well-advised to proceed with care.
U.S. stock exchanges on Aug. 3 started to track and publish "short volume" statistics -- a measurement of daily short sales of stock -- as part of a Securities and Exchange Commission initiative to curb abusive short sales. Potentially abusive sales, which include "naked shorting," have been blamed for turmoil in the stock market and even the 2008 demise of Bear Stearns Cos. and Lehman Brothers Holdings Inc.
The problem with the new information is that some experts think the data, when used alone, could be misleading. That is largely because the numbers don't show new short positions or the reasons why the sellers made their moves. In other words, not all short sellers are betting against a stock.
Unlike short-interest data, which are distributed twice a month and show the percentage of shares that have been sold short and not repurchased, short volume simply tracks the number of shares that are sold short over the course of a day.
On Sept. 10, for example, 391,032 shares of Google Inc. were sold short on the Nasdaq Stock Market, out of 880,911 shares of Google traded that day.
Investors often sell stock short -- or sell stock they don't already own -- in hopes the stock price will fall, allowing them to buy it back for a cheaper price. But the short-volume data in Google, and other stocks, don't disclose whether the short sales establish new positions, the red flags that typically signal bearish expectations for a stock.
For example, if a trader sells 100 shares short at 2 p.m. and then buys them all back an hour later, the short volume will still record 100 shares sold short. It won't account for the trader covering his short position.
"One of the concerns in the discussions so far involves the ability [of the data] to send false signals," said Frank Hathaway, chief economist for Nasdaq OMX Group Inc. "Investors would need to learn over time whether the pickup in short selling was normal or whether it was significant."
Market makers, for example, routinely sell stock short as they attempt to meet customer orders. Money managers also sell stock to hedge positions they take in other stocks or to take advantage of arbitrage situations. In neither of those cases are they making a directional bet on the price, so it isn't a bearish indicator.
To be sure, some experts think the data could be useful in some way but are wondering how investors will use the information when making trading decisions. There is some evidence that short-volume statistics can be a useful indicator. A 2008 study found that stocks heavily shorted tended to underperform those lightly shorted by an average of 1.2% over the course of 20 trading days.
"It's interesting that, despite all the weakness in the data, there is a very strong uniformity and very strong predictability," said Charles Jones, a professor of finance and economics at Columbia University and one of the authors of the study.
But Prof. Jones warned that this may not last. Traders using algorithmic and high-frequency strategies are big short sellers, weaving in and out of stocks on a short-term basis with no long-term view on the stock. "If that's the case, then there could be a lot more noise in this data than there used to be," he said.
Brokerage firms, meanwhile, say they prefer to provide customers with as much data as possible, even if it doesn't give clear signals on investment decisions.
"Is this going to be the magic sauce? I don't think so," said Tom Sosnoff, a senior vice president of TD Ameritrade Holding Corp. "But could this be another indicator we look at? Yes. It's 100 times better to have information out there than to regulate it."
AI
