Liberal Arts Major
...Wall Streets computer scientists and linguists keep trying to find quicker ways to react to the news by creating ever-more complicated algorithms, the mathematical formulas that execute stock trades automatically based on such criteria as headlines and news stories. The idea is to buy or sell at a faster clip than the guy or computer at a rival trading desk.
Sounds like a plan, right? Mine news stories for weighted price-change signals (which, as the article explains later, are the result of a statistical study of past news stories indexed against the price movements that occurred for the relevant stocks on that day) and execute trades immediately as the story is made public, preferably before anyone else gets to do so. It's genius, except when it isn't.
...last week.... a six-year-old story about the [United Airlines]s 2002 bankruptcy filing gained new life on the Internet, triggering a cascade of stock sales. In a matter of about 12 minutes more than $1 billion in stock-market value evaporated.
I think Professor Andrew W. Lo said it best:
"The downside of the technology is occasional glitches like these."
Of course, and why not? If you're on the cutting edge of technological progress, what's a billion dollars between friends?
How do you go about engineering a glitch like this? It's easy! All you have to do is take advantage of the commonly-known fact that nobody ever reads past the headline:
At 1:36 a.m. E.D.T. last Sunday, Sept. 7, Googles search "crawler" picked up a 2002 news article about United filing for bankruptcy from the Web site of The South Florida Sun-Sentinel; for some reason the outdated story had been listed on The Sun-Sentinels list of most popular business stories. (United emerged from bankruptcy protection in 2006.) The next morning, an employee of the investment advisory firm Income Securities Advisors saw the story and posted it to the companys own wire service, which is available over Bloombergs trading terminals. Uniteds stock plummeted soon after. In a statement, the Tribune Company, which owns The Sun-Sentinel, said the bankruptcy story "contains information that would clearly lead a reader to the conclusion that it was related to events in 2002. ... It appears that no one who passed this story along actually bothered to read the story itself."
And if do you manage to Google-bomb your way to NYSE-rending publicity, you don't even have to worry about software patches and intelligent programmers ruining your fun - if even one faulty news-parsing algorithm picks up the story and enters an order of significant magnitude, all the other algorithms will trigger off that. It's the perfect recipe for mayhem!
How could they have resolved this problem? It seems some software developer may have overlooked the fact that news aggregators might rank by popularity rather than by date. The solution is either (a) to check the date-created value or the "Posted on" advisory of the page you're mining and/or (b) do a context-based search of other references to the company or companies in question and compare the price signals, ruling out the article if its signal differs significantly from all the others and/or (c) pick your news services more carefully, perhaps even abandoning the time-honored practice of throwing darts on a billboard.
Of course, (b) would take a lot of milliseconds, and (a) won't inoculate you against recent articles about past events - assuming these even exist - but dear lord, you'd think when you're dealing with millions of dollars you'd have the sense to make sure you account for time in some way rather than put your full faith and credit in some other company's aggregation algorithm. Apparently computational linguists don't have to be competent to work for the big boys.
___
~ Ogordemir ~
"The sciences have hitherto harmed us little; but some day the piecing together of dissociated knowledge will open up such terrifying vistas of reality that we shall either go mad from the revelation or flee from the light into the peace and safety of a new dark age." ~ H.P Lovecraft, The Call of Cthulhu
Legend
___
(9:33:53 PM) Alestra77: you seem to have glossed over the fact that you treat all women like prostitutes
(9:34:10 PM) Alestra77: "k, so, I bought you coffee... when do I get my handjob?"
Goddammit, I hate you so much. ~ Kenri to me (3 times)
~~WINNER OF WORST TOPIC OF THE YEAR '08~~