Posted on 01/03/2011 9:25:58 AM PST by Arec Barrwin
Last spring, Dow Jones launched a new service called Lexicon, which sends real-time financial news to professional investors. This in itself is not surprising. The company behind The Wall Street Journal and Dow Jones Newswires made its name by publishing the kind of news that moves the stock market. But many of the professional investors subscribing to Lexicon aren’t human—they’re algorithms, the lines of code that govern an increasing amount of global trading activity—and they don’t read news the way humans do. They don’t need their information delivered in the form of a story or even in sentences. They just want data—the hard, actionable information that those words represent.
Lexicon packages the news in a way that its robo-clients can understand. It scans every Dow Jones story in real time, looking for textual clues that might indicate how investors should feel about a stock. It then sends that information in machine-readable form to its algorithmic subscribers, which can parse it further, using the resulting data to inform their own investing decisions. Lexicon has helped automate the process of reading the news, drawing insight from it, and using that information to buy or sell a stock. The machines aren’t there just to crunch numbers anymore; they’re now making the decisions.
(Excerpt) Read more at realclearpolitics.com ...
Reminds me of a conversation between Kyle Reese and Sarah Connor in an underground parking garage.
AI (today) isn't a sentient computer which is indistinguishable from a human. AI isn't a computer which is smarter than a human. AI is not a vast network of intelligence, slowly taking control of everything for some nefarious purpose, known only to the machines.
No. AI is a lot of low-level decision-making programs in fields like Wall St trading, or Supply Chain Management. The algorithms make the decisions, and we don't necessarily understand what is going on anymore. Remember the Wall St Flash Crash? That was basically a bunch of machines all making decisions to sell stuff very quickly. Even today, we don't really know why.
There is a great deal about our society which has passed beyond our understanding. This isn't good.
They have for some time, volume-wise. Traders use computers and software to make buy/sell decisions based on local fluctuations. This story suggests that they're adding in a longer-term prediction aspect, based on sentiment or future prospects. It makes sense as a way to moderate the influence of short-term price fluctuations in the presence of a long-term trend.
I would imagine that the algorithms are based in part on generating correlations between particular phrases and subsequent price changes.
And I will guess that there will soon be a thriving market for those who can manipulate prices by writing press releases with just the right combinations of catch phrases.
Bingo! You can barter with it, you can’t reason with it, etc etc. LOL! Sort of like seniors and baby boomers with Obamacare.
If they are trading carbon credits, wouldn’t that be “Algore-ithms”?
And it absolutely will not stop until you are broke.
Someone was working on a blog sniffer system for hedge funds to try to compile all the blogs to see where stocks and ETFs might go based on the total number of blog posts and weighted by the influence of the blog.
They failed but it was a somewhat interesting idea.
SkyNet is here.
[Bottom line - computers control the capital markets now. ]
I actually started work on such a system in 2002 for a wealthy individual who was too stupid to see the light. However, from my perspective the problem is that once every trader is a computer, the system becomes unstable and there is no longer any real information to suck out of the system.
Eventually individuals realize competing against computers is a suckers game, so they leave the market. Then you are left with algoritms fighting algorithms which is inherently chaotic. In effect, you are also matching computer programmer against programmer and the ties to reality disappear.
I’m glad I wrote that, I was just about to go back to looking at my code, but I have better things to do.
“And it absolutely will not stop until you are broke.”
LOL:
Oh that is The Ben Bernank and his printing presses along with The JP Morgue and The Goldman Sack.
Exactly. Now it is trying to get your 1,000 orders per second in faster than the other guy. Who has the fastest connection and the fastest C++ code. They are also front running the little guy with his eTrade account.
Holy Crap!!
The implications are indeed disquieting. No eeevil Master Control Program like in the original TRON. No Skynet like in the Terminator series. Just a bunch of buy/sell algorithims doing what they’re designed to do. What would cross the line into AI would be if they were communicating with each other, learning, and re-designing themselves.
So the mysterious Flash Crash wasn’t George Soros or the Chinese or some basement genius, it was just the buy/sell algorithms coming to the same conclusion at the same time.
But it was a matter of interpreting statistical number data, not analyzing written language.
Quite a few folks suggested as much at the time.
As is the dog track to the lower class, so is the stock market to the middle class.
Computer trading is operating on a different place in the stock market.
Happy gambling, suckers.
Here’s a little more technical discussion of the problems with algorithmic trading. All such systems, no matter the math used, are really looking at varying orders of the derivatives (the slopes, and slopes of slopes) of different data. When there are only a few computers in te system, this is all well and fine because they don’t influence te entire system. But when a certain critical point is reached, trading on the mathematical derivatives becomes a positive feedback loop.
Now you can try and game this problem with some kind of neural nets, but you always end up evolving towards the same positive feedback situation.
In thermodynamic terms, you can’t get something for nothing. Once everyone has as much information as the system holds, then there is no more advantage to be had and the extra information you think you are getting is actually disinformation. This is also why you don’t generally see more than cubic splines in curve fitting, after that point the equations work only at the data points but divurge elsewhere .
“positive feedback”
words all engineers love sooo much.
I hope they remember to do a backup every now and then.
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.