Forum Moderators: open
Computers that automatically scour the Web for news stories may have been behind the resurfacing this week of a 2002 story that panicked investors and cratered UAL Corp.'s stock, according to reports Wednesday.The Wall Street Journal said the old article about a bankruptcy filing that appeared on Google's news service Monday was traced to the Mountain View-based company's "crawler."
Google said its crawler found a new link on Tribune Co.'s South Florida Sun-Sentinel newspaper to the article, which didn't carry a date but actually appeared in December 2002.
Web surfers may have stumbled onto the old UAL story and from there a growing number of hits might have driven it onto a "most read" list, the Journal and Associated Press reported.
It then became available through Google News to searchers looking for keywords such as United Airlines.
After a headline from the old story appeared through Bloomberg mid-morning, UAL shares began to plummet, and 15 minutes later Nasdaq halted trading as they hit $3, dropping from almost $12.50. The stock closed Tuesday at $10.60, almost 13 percent down from Monday's opening.
Article: Bizjournals.com [losangeles.bizjournals.com]
Stock chart: Yahoo! Finance [finance.yahoo.com]
The Trib says:
Tribune Co. said the Google News search engine caused a spike in traffic by adding a Sept. 6, 2008, date to a 6-year-old story about United's bankruptcy on the Web site of the South Florida Sun Sentinel.
While Google says:
...its search engines played no role in prompting Internet users to suddenly begin clicking on the Sun Sentinel's Web site early Sunday morning Florida time.
But goes onto say:
...the episode began because the story did not have a date on it when the Google News search program came across it early Sunday morning. "We indexed the story in Google News," Stricker said. As part of that process, the Google search program assigns a date to a story."
So does the fault lie with Google inserting a date, or the Trib in not having a date on an archived story... interesting
So does the fault lie with Google inserting a date, or the Trib in not having a date on an archived story... interesting
The fault is with the traders who use unverified information to make important decisions of this kind and maybe with stock exchanges that allow quick automated trading - maybe if there was 24h delay on buy/sell orders then such instances of panic would not be possible.
I'm not familiar with how the system works into GoogleNews, but is a news site "required" by Google to have a date appended to the story? If not, then I'd say Google may need to reappraise the practise of automatically inserting a date onto a news story.
there's also a responsibility that should be allocated for how the incorrect news got out there in the first place.
Well, I'd say, IMHO, Google is not at fault unless they had some kind of agreement like Reuters/Bloomberg have - those guys sell newsfeeds for ridiculous amounts of money for a reason - and even they make mistakes.
The newspapers should have certainly dated article, but I don't think they can be blamed since it's not like they pushed it on their own into Google.
The main problem I think is that the Web has become so popular that some people lost sense of reality - apparently some immigration officials were using Wikipedia to make decisions, that's crazy stuff - just like traders using Google News to trade - I bet there are plenty people trading using 15 min delays stock quotes and sooner or later they will get burnt: same situation really.
