From left to right, the journey of old news that would cost $1 billion. A Sun Sentinel story from 2002 resurfaced in an automatically generated area of the Web site, soon landed on Google News before getting posted to Bloomberg terminals.What made a six-year-old article about a bankruptcy filing by United Airlines reappear on Wall Street traders’ screens on Monday as if it were fresh news, prompting a sell-off that erased $1 billion in the company’s market value in a matter of minutes? The path the article followed from forgotten archive entry to present-day stock-killer has begun to emerge, and it raises some interesting questions about how news rockets around the Web.
Both human error and far-from-foolproof technology seem to have played a role in the episode, which involved a 2002 Chicago Tribune report; the web site of the Sun Sentinel, a Florida newspaper owned by the same company; the Bloomberg News financial wire service; and Google, all apparently unwittingly.
The Tribune Company laid out the results of its preliminary investigation into the incident in a news release late Tuesday afternoon, beginning with a line that painted its newspapers as victims of circumstances beyond their control:
The December 10, 2002, Chicago Tribune article on United Airlines’ bankruptcy filing, along with thousands of others stories in the searchable database of the Sun Sentinel website, was accessible for years.
The trouble seems to have started on Sunday, when a link to the old article turned up on a page on the Sun Sentinel site that automatically lists the most popular business stories of the moment. Why did such old news suddenly garner enough attention to land it there? Tribune hasn’t offered an explanation so far, except to say that the story was not promoted elsewhere on the site. So one open question is precisely how many clicks it took to earn a spot on that ranking — dozens? hundreds? thousands?
Any effort to find out more would lead down the many roads that readers may follow before converging anyplace on the Web. Sometimes, there’s an obvious answer tied to the news. Sometimes, it’s seemingly random phenomena. Just one look at Google Trends, the search giant’s version of a most-popular chart, would yield both kinds of items.
Had the old article merely spent a bit of time on that Sun-Sentinel list before returning to the archive, no harm would have been done to United. But a second piece of software apparently stopped by to analyze the work of the first one, and it would call the article to the attention of a much larger audience.
Normally, Web sites welcome the arrival of Google’s “bots,” which “crawl” around the Internet, following links and cataloguing the content they find for users to search. Indeed, many sites have been optimized in recent years specifically to make it easier for Google’s bots to detect new material quickly and bring it to Google’s monstrous audience. Most-popular-articles rankings are one place where Google News bots find the headlines they aggregate for readers. Up it went.
And there lies a rather big catch, which constitutes the biggest disagreement between Google and Tribune. Google believes that the fact that the bankruptcy article dated from 2002 was not obvious to either robot or human:
It has been widely reported that many readers were unable to determine the original date of publication of this article, and our crawling was similarly unable to recognize that the article was old.
For its part, Tribune says that no one who actually read the story could have mistaken it for fresh news:
The December 10, 2002, story contains information that would clearly lead a reader to the conclusion that it was related to events in 2002. In addition, the comments posted along with the story are dated 2002. It appears that no one who passed this story along actually bothered to read the story itself.
Neither assertion contradicts the other, but Tribune seems surprised by a reading style that is commonplace in the age of Google. As Nicholas Carr wrote in an Atlantic essay entitled “Is Google Making Us Stupid?” last month, “Once I was a scuba diver in the sea of words. Now I zip along the surface like a guy on a Jet Ski.”
As for preventing another similar incident, leaving sharper date cues for Google would seem to be far easier than changing the habits of millions of readers. In this case, though, it only took one reader to open the barn door for United stock’s shockingly swift decline — the one working at Income Securities Advisers, who saw the item on Google News and sent out a summary of it on Bloomberg News.
Stripped from its original context, the report gave traders who were already nervous about airline stocks just enough information — “UAL Files for Bankruptcy” over the weekend — to start them pounding the sell button, and not enough time to think twice about salvaging whatever was left of their stake.
Of course, software had a hand in the sell-off as well. “The damage was exacerbated by the growing use on Wall Street of automated programs that trigger stock trades without any human interaction,” The Wall Street Journal reported.
As the search for blame returned once again to something rather than someone, it appeared that United and its shareholders will have a hard time getting satisfaction after an astonishingly bad break.
2008
1:00 pm
This comes as no surprise, or *should* come as no surprise, to journalists. Copy editors are being downsized and demoted, long before their more-glamorous reporter colleagues are affected. They’re being replaced by Web-scraping “bots” programmed by 22-year-olds. Fault for such fiascoes will never, ever, be found with those new employees and their bots, who make so much money for journalism precisely by eliminating journalists.
— Posted by Emily
2008
1:00 pm
The article says: “Sometimes, there’s an obvious answer tied to the news. Sometimes, it’s seemingly random phenomena.” Like the NYT Magazine article a few years back that argued the deaths of 14 noted microbiologists (following the Anthrax scare that preceded the Patriot Act of 2001 vote) was random, sometimes the answer is tied to intent to disinform and obfuscate to protect the perpetrators.
— Posted by RAB
2008
1:08 pm
Readers didn’t read the lede?
— Posted by MARK KLEIN, M.D.
2008
1:10 pm
Communist saboteurs take note!
— Posted by Karl M.
2008
1:16 pm
lack of dating seems endemic to the Internet.
These “dateless” postings seem like a new type of “derivative”: they derive from a real, dated posting, but float widely, and are impossible to value.
— Posted by Paul Antal
2008
1:21 pm
Seems like no coincidence that the Sun Sentinel is one of papers that signed on to have Google archive all of its previously paper only stories online just as an old story surfaces again.
— Posted by Don A.
2008
1:27 pm
What’s missing here is how an article from the archives of the Sun-Sentinel could possibly even end up on the Sun-Sentinel website’s list of most viewed business stories.
First, one must pay to view anything other than an abstract from the Sun-Sentinel’s archive. (One must pay to view any Sun-Sentinel article more than 30 days old.) What’s more, the Sun-Sentinel doesn’t even keep its archived articles on its website, but rather on the ProQuest Archiver site.
So even if a horde of people suddenly paid to view the article contained on the PQ Archiver site, it would not have shown up as one of the Sun-Sentinel site’s most business stories — which are to a one contained not just on the Sun-Sentinel site only, but within the “business” directory of the Sun-Sentinel site.
To the extent there is a stink here, it seems to be emanating from Fort Lauderdale . . . .
— Posted by James W.
2008
1:30 pm
Those who sold now have the luxury of buying back all their shares at pennies on the dollar. UAL has bad Karma anyway, maybe someone is trying to tell them something.
— Posted by osibs
2008
1:33 pm
The line about an employee of Income Securities Advisers sending a summary out on Bloomberg News doesn’t make any sense. The author of this report should get a better handle on what Bloomberg News is. Could a random brokerage employee write a New York Times story? No. So what are they claiming is the role played by Bloomberg News? Unclear.Sounds like Google and Tribune Co. are to blame, if anyone. Nizza should do some background research on Bloomberg LP, Bloomberg News, &c before he makes baseless assertions and inuendoes about it.
— Posted by Bilbobaggit
2008
1:35 pm
This is what happens when people don’t read the WHOLE story and get ALL the information.
Probably the same people who go and get all their information from Wikipedia and believe that it’s the best info. out there.
I do believe the internet is bringing us down intellectually. Most people search the internet for specifics about a topic and do not stop to read the whole article or to see if it’s even a legitimate article or website. For example www.sciam.com (Scientific American magazine) vs. www.dr-bob-is-a-scientist.com.
— Posted by Capt. Concernicus
2008
1:38 pm
Look at the “Related” items on this very post - all six months to a year old, with nary a date in sight…
Good point, Matthew. I’ll pass that on to my editors. — Mike
— Posted by Matthew
2008
1:40 pm
A robot from Google cannot CONTEXTUALLY “read” the story to determine that it is referring to 2002 events. And the comment section is not a standard feature and isn’t looked at by the news robot.
So the fault that such an old story reappeared as new on Google News lies in that the Sun site linked such an old story prominently on its site, rather than its archive section.
That being said, the ISA researcher that posted it to Bloomberg should have read it contextually prior to posting it to Bloomberg.
— Posted by Joseph
2008
1:47 pm
Seems like a good time to pick up some cheap United Airlines stock.
— Posted by Nancy
2008
1:49 pm
So I presume someone is looking for large short positions or large put holdings in UAL. It would be an interesting way to pull off an inside trade.
— Posted by Lance Labun
2008
1:49 pm
This fiasco points out one of the biggest flaws of the current world wide web - very few web pages are dated. Any Google search (except maybe for Sarah Palin) and the web pages listed are mixed in with pages from as long as ten years ago. The recent hype about Apple’s product annoucements are a good example. A google search last week turned up just as many speculative articles from 2003-2007 predicting what Jobs would announce as it did from 2008. It was often imposible to tell which were new and which were from 5 years ago (which I think tells us a lot about Apple-hype, but that’s another comment entirely).
— Posted by C. Reaves
2008
2:00 pm
What if it was arranged by someone with access to the Sun web site in order to manipulate the market for profit?
— Posted by jay
2008
2:02 pm
I hope someone is checking short sellers of UAL!
— Posted by Tom Severns
2008
2:05 pm
So after some unidentified person hit the old news repeatedly, someone was able to buy UAL stock extremely cheaply.
People better get smarter because, whether or not it was a deliberate scam this time, it would be extremely easy to repeat with a new target.
— Posted by JD
2008
2:07 pm
Wow - whatever firm shorted United before having its monkeys click on the story enough to bring it to the attention of Google news, I salute you.
In any case, fact-checking is *so* 1992.
— Posted by Cletus
2008
2:30 pm
Something vs. Someone. Thanks for filling in how Google and crawlers spread the contagion, but it seems the original reporter deserves more attention. How ignorant is he or she not to pick up the cues that made clear the story was history? I have a low opinion of financial journalism already but this represents a new level of cluelessness.
— Posted by Steve H
2008
2:30 pm
And people wonder why markets rise and fall for, at times, seemingly no good reason.
— Posted by Lou
2008
2:30 pm
Interesting story! Is the SEC or anybody else looking into unusual volumes of call options, futures etc. in the days/weeks before? It must have been worth $$$$$ to somebody for this to happen.
Mishaps such as this can be less innocent or accidential than we would like to believe, given that many millions of dollars can be made (and lost) if they happen.
Peter
— Posted by peter
2008
2:36 pm
What if someone who had access to Sun Sentinel’s system put the story up in order to manipulate the market for profit?
— Posted by Jay
2008
2:38 pm
The sloppy employee at Income Securities Advisors, who posted the article on Bloomberg, should share in the blame. Should he or she have at least read the article over before disseminating it to the public?
— Posted by eskimo
2008
2:57 pm
Could there have been a third party bot that targeted the specific article. Thereafter repeatedly kept on “hitting” it to raise it number of clicks and consequently its visibility.
Has the SEC bother to look into large short positions of UAL?
e-Com Cyber-War anybody?
— Posted by Chaz Ubell
2008
3:05 pm
Time to buy United, I would guess.
— Posted by CalGal
2008
3:06 pm
Ever since reading of this ‘error’ I’ve had one question that no one has answered, “Do we know who was buying United stock when it fell?” and “Were any of those people associated with any of the ‘errors’?”
— Posted by sam
2008
3:31 pm
It used to be that if one selected stocks, perhaps with the assistance of an investment professional, of high-quality companies, bought them and kept them for the long-term, one could expect to prosper.
Now, with the proliferation of information, much of it false or misleading, on the internet and a culture of day trading and “street” earnings expectations, the stock market is far more risky. A person’s net worth, or that of a pension fund, can be wiped out in moments if the market “turns bearish” on a company.
For the small investor, it is very discouraging.
— Posted by famdoc
2008
3:31 pm
This article seems to have completely missed the true culprit here. Don’t blame the newspaper’s archive or an internet search engine. This is clearly a human mistake. Why no mention of the careless employee’s name? His/her negligence has wiped out a lot of stock value. Personally, I would terminate someone whose careless work caused such great financial liability. I see a very-winnable legal action in UAL’s future.
— Posted by end of personal responsibility
2008
3:31 pm
Fortunately for UAL (and others similarly harmed), Google is a deep pocket, along with Tribune and Bloomberg. Considering that these services depend on content, you would think that they would treat content with more respect. Someone should sent them a copy of the copyright laws, under which the date of first publication is meaningful.
— Posted by lkd711
2008
3:37 pm
If this had been an actual bankruptcy the knee-jerk short sellers would claim they were doing a valuable service to society and thus deserved their enormous profits…
So it stands to reason, if they suffered enormous losses because they sold someone else’s shares for no good reason, then they have fully “earned” it.
— Posted by Rwolf01
2008
3:39 pm
Don’t understand — clicking on the story headline should have opened a link to the story archived, or at least a summary of the story archived, on the Tribune web site, which should have shown the date. So what you are saying, without really saying it, is that the problem was caused by people just “reading heads”?
The NYT main page headline for this blog entry is “The Lede: Story Behind the United Stock Plunge.” You aren’t really saying what and why this happened.
— Posted by farhome
2008
3:41 pm
The various implications outlines in this story are truly momentous and frightening.
— Posted by hapinoregon
2008
4:27 pm
Bigger, botter mistakes, faster.
— Posted by cpetrich
2008
4:28 pm
This problem would not have occurred, had there been a good standard of metadata in place for newspaper articles and media reports. That is, a way to store information such as the byline, publication date, etc. Metadata can be added in a large variety of ways. On the most basic and obvious level, HTML — the “language of the Web” — contains a specification of META tags, exactly for this purpose (http://www.w3schools.com/tags/tag_meta.asp). Many other possibilities exist as well.
This same problem has been successfully dealt with in other venues. A good example can be found in ID3 tags in digital-music files (http://www.id3.org/). So there is no technological or practical barrier to implementing a standard, and nothing preventing Google, or any other “crawler,” from taking it into account, if it existed.
In this case, had there been a definite, explicit “publication date” attached to the article, Google would have known this and not presented it as “current news.” This means the problem was completely preventable.
What’s more, the Library of Congress noted this very problem last year (http://www.loc.gov/ndnp/MetadataStandard.html), so it’s not as though no one knew the problem existed. Of course they did. It’s just that no one cared to do anything about it.
Maybe now they will … but I doubt it. This is a problem of volition, not capability. If no one in journalism has managed yet to summon the will do something about this by now, I have no reason to expect they ever will.
— Posted by DJH
2008
4:37 pm
Step 1: Buy tons of put options on United.
Step 2: Have a bot access the old story on the Sun’s website over and over thousands of times until it appears like ‘new news’, which drives the stock down.
Step 3: Profit!
It’s so easy, the underwear gnomes could have done it!
Capitalism might kind of stink…
— Posted by Mike
2008
4:55 pm
So what you are saying, without really saying it, is that the problem was caused by people just “reading heads”?
The sloppy employee at Income Securities Advisors, who posted the article on Bloomberg, should share in the blame.
Yes and no, on both counts. Certainly the sloppy employee is to blame (this is not in any direct way a function of Internet reading habits). But keep in mind that wire services (like Bloomberg’s) function as a matter of quantity, not quality. The remaining copy editors/bot-builders at wire services spit out literally hundreds of stories from dozens of contributors every day, and any dip in that amount is noted and criticized.
I’m not alleviating any responsibility from the low-on-the-totem-pole employee, but this sort of sloppiness exists in context.
— Posted by Emily
2008
5:12 pm
“…You two are going to be in SUCH TROUBLE when Joel finds out!”
“Aw, c’mon, Gypsy, all we did was click a few links…”
“Yeah, Tom and I were just trying to access this article about United going bankrupt in 2002, and the link wouldn’t work, so we kept clicking…”
“…yeah and then like TWENTY-SEVEN MILLION POPUP WINDOWS for Netflix opened up, and…well…it just wasn’t our fault! Right, Crow?”
— Posted by Peter
2008
6:00 pm
The internet provides information quickly and in great amounts and anyone who can connect to it probably has more information available than any person at any time before it existed. It has the potential to reorganize the way people work together and share what they know in more ways than anyone can imagine.
That aside, it’s also a great way to misinform millions and cause people to do silly things.
I’m sure if any person who sold United Airlines stock had checked the UA website or the Federal Court websites, no trace of any bankruptcy would have be found. Those people just acted without checking out the facts. They acted out of panic, unthinkingly reacting from fear without knowing what was going on.
— Posted by Casual Observer
2008
6:09 pm
If anything, it just proves once again that the market sytem works. People that acted without thinking on bad information lost, those that analyzed the information correctly won. Very Darwinian.
— Posted by cb
2008
7:20 pm
This should come as no surprise to anyone in the United States. The media has far too much power, far too broad a reach, and far too little by way of fairness, integrity, and professionalism. And just think, if this degree of damage can be done to a company by three media outlets, imagine the damage being done to our political system on a daily basis? Free speech? No, that’s yet another American lie that very suddenly seems so unbelievable.
— Posted by David L.
2008
7:45 pm
Simple solution:
Google should completely ignore anything that’s not dated (at least news articles). Then, as pagerank falls, the world will “correct” their pages, adding the date to increase the pagerank.
— Posted by BillT
2008
7:55 pm
This is not unusual. During the summer, a NYT article about ferry service from Manhattan to Sandy Hook, NJ, resurfaced on the internet and people called a telephone number that was only relevant in 1997.
— Posted by jmernick
2008
8:09 pm
These particular “somethings” do not spring-forth magically from the void, there is always a “someone” or several “someones” behind each.
— Posted by rick jones
2008
10:02 pm
Why is everyone (#17 Tom) talking about put options? Those would leave an obvious trail. The beauty of this (if it was deliberate) was that is was untraceable. Anyone clever enought to recognize the way the bots worked was probably also savvy enough to spoof their query requests. We’ll probably never know what IP address all the queries originated from. Why would they then go place a bunch of put options which WOULD be tracable. Better to do your misdeed to generate the bad news, then innocently go out and buy when the price is in the tank. You could be fairly certain the price would recover once the corrections hit the wire. Of course you’d have to have capital to actually make the buy, but the price fell to a penny per share and recovered to over $11.00 per share. Someone making a buy of only $1000 would have returned over a million dollars. And there is the beauty - if you investigate all the buyers they will all look absolutely innocent. Even if someone LOOKS guilty proof will be darn near impossible to obtainBut I fail to see how this is an indictment of capitalism (#36 Mike). Go back and do some basic reading - corruption in Soviet Russia was rampant. Look to the middle ground - capitalism works, but the SEC needs suficient funding to investigate and authority to impose changes (if necessary) to the rules.
— Posted by Mark
2008
12:44 am
Hilarious. Let the markets work! Don’t worry, the Harvard, Yale, Princeton graduates manning the Wall Street helm know what they’re doing. Not. What a joke Wall Street is. Yet, the middle class is told that they must invest to get ahead. Invest, to put more money into the pockets of the elites.
— Posted by John
2008
3:17 am
The absence of dates on Web pages is a huge problem. As just one example, it’s easy to use Google (et. al.) to find reference material on all sorts of subjects, e.g., automobiles and software packages. But finding out what *version* of said automobile or software product the material applies to, an absolutely critical piece of information, is generally much more difficult.
— Posted by Greg
2008
6:53 am
Original dates for web articles are difficult to establish for another reason. The “related articles” that pop up nearby are rarely dated themselves. Readers are faced with establishing dates for ALL the stories in front of them. Younger readers don’t have long-term memory to serve them, so they are left to guess at dates and provenance.
Adding the pre-line “Originally reported _____” would help solve this problem.
— Posted by Emil
2008
8:13 am
This is nothing in comparison to the housing mortgage twins nationalised by the Administration,just track how many who knew this coming had made money before the decision sold their holdings.
— Posted by Arun Mehta
2008
1:09 pm
The dating problem is hardly confined to the electronic media. Once upon a time, newspaper articles started with a dateline. I am continually reading undated articles in printed newspapers with a sense of déjà vu because I read the same article weeks or months previously in another newspaper. The only defense is to bring a healthy skepticism to the table before you read or listen to anything. This is as true now as it has ever been.
— Posted by Chris in WA
2008
7:21 pm
Can’t United sue for damages? It’s obvious, even admitted, that the reason the stock plummeted was a DIRECT result of the error.
I guess the issue is, who/what exactly would United sue? Would United sue both Google and the Tribune, or one over the other, or even the trader, or Bloomberg? I don’t know.
But if I were United, I would consider at least suing in the hopes of reaching a settlement, even if it’s a preventative settlement.
— Posted by Dan
2008
2:32 am
This news story represents another example that disproves the “efficient market hypothesis”.
According to Wikipedia: “The efficient-market hypothesis states that it is impossible to consistently outperform the market by using any information that the market already knows, except through luck.”
What happens in reality, as this example shows, is that market players have very short term memories and have to keep re-learning the same old things over and over again.
— Posted by Robert Laden
2008
1:08 am
An even more serious example of what happens when you merely regurgitate old news (extracted from Wikipedia):
“On 2 August 1964, the destroyer USS Maddox (DD-731) engaged three North Vietnamese P-4 torpedo boats, resulting damage to the three boats. Two days later the Maddox (having been joined by the destroyer C. Turner Joy (DD-951) reported a second engagement with North Vietnamese vessels. This second, however, later discovered to be in error.[1] Together, these two incidents prompted the first large-scale involvement of U.S. armed forces in Southeast Asia.”
The people on the Maddox mistook a Vietnamese radio conversation about the original attack and reporte it to Washington as if it were another attack! Result: The Vietnam War.
— Posted by Mike
2008
9:12 pm
It proves one thing, “Market is a stupid animal”.
If fly gets in the ear, the World is coming to an end. If you want to make money, misinform the market, then take the money and run.
— Posted by Mukul Kantharia
2008
9:32 am
Perhaps, though the internet revels in its “free status” the “industry leaders” - google and their ilk, should show that, they are worthy of their position as information disseminators of global prominence, by introducing date lining of articles. This, comming of honesty, may consolidate their global status as presenters of up to date information, and by the same token relieve the general populace, of the tedium, of recycled, hackneyed dross. A sort of self regulation, even if it means re-training their “Bots”.
— Posted by the prof
2008
9:25 am
Response to #55.
Google’s algorithms are good, but they are not that good. For example, if you qualify a search by date and say you only want information that has appeared in the last year, you always get back a large number of search returns that are many years old, but have somehow gotten a more recent date stamp on them.
In fact, you get back some some search returns that have no relevance whatever to your key words. Those pages may have been relevant before they were refreshed.
You have to compensate for the limitations of the retrieval system.
Even a library has books have been misfiled on the wrong shelves.
— Posted by Robert Laden
2008
2:52 am
I constantly have the problem arise of the date of a page, and also the geographical location of a company not being immediately apparent. I recently bought on Amazon UK a BBC TV wildlife documentary DVD from a so called UK company. It never arrived. The tracking showed that it was despatched from Shanghai by a Chinese shipping company and had been delivered! I suspected a scam utilizing the internet’s opacity, and the reluctance of many customer’s to spend hours getting their $30 returned. I did spend hours and reported what I found to Amazon who removed the company from their listing. This may seem unrelated, but I see it as part of the greater situation of the loss of “hands on” human control and accountability.
— Posted by barry beckett