Evidence of Front-Running on Wall St.

The Securities and Exchange Commission has been actively investigating the illegal practice of front-running. That's when brokers execute trades for their own account, or tip-off favored clients like hedge funds, before filling customers' orders.

In January, the Wall Street Journal reported that regulators were looking into whether several Merrill Lynch employees had stepped in front of trade orders placed by Fidelity Investments. Large stock trades by institutional investors and mutual funds typically move prices, so front-running can be profitable.

Now in a new paper, M.I.T.'s Mozaffar Khan and Hai Lu of the University of Toronto show some compelling evidence that significant front-running does exist.

Khan and Lu looked at the level of short sales between 2005 and 2007 surrounding days when a chief executive sold stock. Short-selling here can be profitable for two reasons:

• if the transaction by the C.E.O. is large enough, it could move the share price
• the C.E.O. could be revealing information about the company's outlook.

On days leading up to the stock sale, the level of short-selling did indeed zoom higher. For the 2,030 insider transactions Khan and Lu examined, short-selling volume on the New York Stock Exchange was on average $3.2 billion higher than normal.

C.E.O.'s are required to notify the S.E.C. within two days of making a stock trade. This information is then immediately revealed to markets, so you'd think that, if front-running is going on, short-selling volumes might dip right around when that happens...and you'd be right:

The first chart above shows that short-selling peaks on the day of the actual stock sale while the second chart shows that the peak comes before the rest of the market is notified of the C.E.O.'s transaction.

Khan and Lu also looked at short-selling volume for when a C.E.O. sold a lot of stock as opposed to just a little. You might expect that if the chief executive is only selling a small amount of stock it won't move prices and doesn't imply that his/her company's future prospects are negative....and you'd be right again:

The two charts above show short-sale volume for large and small insider trades.

Khan and Lu obtained the short-sale information from the N.Y.S.E. which didn't have identifying information so they weren't able to pin-point who was making the trades. The researchers also didn't have information on when the short positions were closed, so they couldn't accurately figure how much, if any, profit was made. However, on insider trading days, stock volatility is higher, so theoretically short positions could be closed profitably on those days. Also, the stocks in the sample that performed the worst after the insider sale also happened to see their short-selling volume peak earlier than other stocks -- also implying that front-running through short-selling is profitable.

What options do regulators have to inhibit front-running? They can either try to scare brokers straight through publicized investigations, or they can change the way exchanges report short-trading:

Currently, exchanges are required to disclose the level of short interest once a month, which only provides a snapshot. Disclosing higher frequency such as daily short sales activity will increase transparency and level the playing field by allowing broader market participants to exploit the information content of this data.

This of course will likely lead short-sellers to try and figure out another way to profit from their private information, which they probably will.