Automated trading bites some UAL investors
NEW YORK (Reuters) - Computers, unable to see the mold on an outdated UAL Corp (UAUA.O) article, sparked confusion, a mass sell-off, and an emergency trading halt earlier this week that highlights the pitfalls of increasingly automated financial markets.
Shares in the parent of United Airlines plunged about 76 percent after a nearly 6-year-old news story on its 2002 bankruptcy filing appeared online on Monday.
UAL, which is no longer in bankruptcy, scrambled for a retraction and its stock later bounced back -- but for some investors, especially frequent traders, the damage was done.
The use of algorithms -- which allow computers to make decisions in fractions of a second -- appears to be a main culprit in the UAL case.
Experts said the automated programs were applied to both the reading of the outdated news story and the trading of shares based on that information.
"You're taking human judgment out of the news processing. Ninety-nine times out of 100 you get good results, but this is a pretty glaring exception." said Jamie Selway, managing director of institutional broker White Cap Trading.
"Savvier people will benefit, and the art is knowing when someone is trading on bad information," said Selway, who focuses on market structure.
So-called algos are now used to scour the Internet, combing through news for market-moving information.
On Monday, the Chicago Tribune's UAL story first appeared on the Florida Sun-Sentinel Web site and later found its way to the Bloomberg financial news service, which is when the sell-off
began in earnest. In many instances, the story was said to not include a dateline.
Some analysts said the stock drop and related surge in volume was in part due to the lack of confidence investors have in the troubled U.S. airline industry.
But there's little doubt the problem was exacerbated by automated trading -- and probably hit high-frequency traders, and those who missed the 2002 bankruptcy article. harder than those who hang onto stocks for longer periods.
"Word travels fast on the Street, and a lot of trading desks are tied together ... so it's very possible for people to panic," said Matt Samelson, senior analyst at Aite Group. "I'd say it's a combination of algorithms and humans."
ALGOS GROWING
Last year, algorithms handled some 30 percent of all equity trading volume, according to a recent study by Aite Group.
The study projected that algos would grow to handle half of equity trading by 2010, and noted similar growth in derivatives and other asset classes as a hunger for faster trading grows.
"Algos as a strategy has certainly been around for some time, but what's new is that more and more people now get more and more information," said Selway.
"That democratization is probably hugely beneficial to the market, but it does have negative externalities."
Trading halts due to erroneous trading are rare, but last year, NYSE Euronext's (NYX.N) New York Stock Exchange halted the morning dealings in Jefferies Group Inc (JEF.N), AT&T Inc (T.N) and Wyeth WYE.N shares after a technical glitch. In that instance, Nasdaq OMX (NDAQ.O) was forced to cancel some off-market trades.
The example of UAL, which is traded on the Nasdaq Stock Market, once again highlights the difference between Nasdaq's all-electronic platform and the hybrid platform operated by NYSE Euronext.
On Nasdaq, erroneous trades are corrected afterward, while on the NYSE, specialists on the trading floor can halt abnormal stock drops as they happen.
The latter -- human oversight -- is disappearing from the landscape of financial markets, although it may have softened the blow for UAL on Monday.
A Nasdaq spokeswoman said the exchange operator halted trading in UAL stock after it was contacted by the company.
(Reporting by Jonathan Spicer, editing by Gerald E. McCormick)
