NEW YORK (Reuters) - The U.S. Securities and Exchange Commission on Monday adopted a single set of rules for “clearly erroneous” trades, eliminating a mixed bag of standards that exchanges used to monitor increasingly electronic trading.
So-called clearly erroneous trades can result from human error or computer malfunction, the regulator said. “Because the markets today are so fast, automated and interconnected, an erroneous trade on one market can very rapidly trigger a wave of similarly erroneous trades on other markets,” it added.
SEC Chairman Mary Schapiro said in the statement that consistent standards “will strengthen the resiliency of our markets by reducing the potential for market confusion, especially during periods of high market volatility.”
Exchanges cancel trades determined to be clearly erroneous, relieving firms of obligations that result from the trades. The exchanges, including Nasdaq OMX’s Nasdaq Stock Market, began revealing the new rules last week.
The rules force exchanges to investigate potentially erroneous trades within 30 minutes, and to resolve the matter within 30 minutes thereafter.
As well, exchanges can only consider canceling a trade if the share price exceeds the last public sale price by more than 10 percent for shares priced under $25, by more than 5 percent for shares priced between $25 and $50, and by more than 3 percent for shares priced at more than $50.
The new standard comes amid heightened concern about fairness in markets that rely increasingly on computer algorithms to function smoothly.
John Malitzis, vice president of market surveillance at NYSE Regulation, the on-site New York Stock Exchange monitor, said last month his agency has seen “a number of instances where algorithms have gone wild.
“This is something that is happening not only in the U.S., but it is happening around the world, algorithms malfunctioning,” Malitzis added.
It is estimated that about 60 percent of U.S. equity trading involves so-called high-frequency trading, where firms use algorithms and lightning-quick computer software to earn returns from small spreads in the markets.
Reporting by Jonathan Spicer; editing by Carol Bishopric
Our Standards: The Thomson Reuters Trust Principles.