Fast-Reading Computers Are About To Drink Your Trading Milkshake

Tyler Durden's picture

Add speed reading to the set of mad skills that rich robotic algos are now much better at than humans. The latest craze among the computerized trading community is using Johnny 5 to read through thousands of press releases in combination with some fuzzy logic and the 80/20 rule, to trade stocks pronto, even as plan vanilla Homo Sapiens are still stuck on footnote 1. The WSJ reports that: "Researchers have been working on an artificial-intelligence computer program designed to mimic the way an analyst uses financial news. In simulated trading, the program beat the S&P 500, and when combined with quantitative stock-picking techniques, it saw a return on trades of more than 20%." Unlike traditional number crunching methods to front run those armed with less than the latest SPARCs, this latest development will instead try to determine what the subtext of a given PR narrative is and trade it accordingly. It is unclear whether these programs are already in place, and whether they account for the almost universally wrong knee jerk reaction post any earnings and press release by public companies. We can't wait for corporate counterespionage divisions to reverse engineer these algos, and determine just what keywords and phrases set off Buy programs, and flood their press releases which announce 100% misses to expectations, yet result in flash dashes and inexplicably push stock prices up a few million percent. In other news, the market is now no longer about forecasting, predicting, and, generally, investing, but merely about being faster to frontrun/react/buy than your immediate competitor.

More from the WSJ:

To make stock predictions, the program does what is known as “text mining” — scanning large volumes of content and analyzing the words in it. Computer-aided quantitative funds already are plentiful, but they analyze numerical data rather than text. The new program is different because it attempts to simulate what has traditionally been a human activity.

“Our approach is more like the analyst approach, simulated by a program,” said Hsinchun Chen, director of the University of Arizona’s Artificial Intelligence Lab, in an interview with Digits. “You have an analyst reading papers, looking for clues that others have not observed.”

The program, which was first reported by MIT’s Technology Review, scans stock prices and financial news and buys or shorts stocks it believes will move more than 1% in the next 20 minutes. The system sells the stocks after 20 minutes.

“When you do long-term predictions, there are many variables,” Dr. Chen said. “But … you can have an advantage if you look at five minutes, 10 minutes.”

And before one says this is the dumbest thing in the world, here is the retort: the program's creator, fully aware that past performance is a guarantee of future results, note that based on backchecks Johnny 5 would already be richer than Jim Simons:

The system’s creators tested it using data from five weeks in the fall of 2005 — more than 9,000 news articles and 10 million stock quotes. During that test, the system, called AZFinText, had an 8.5% return on trades, beating the S&P 500 index as well as six of the top 10 quantitative funds. When the researchers used quantitative strategies to select a portfolio and then used AZFinText to decide which trades to make, they saw a return of more than 20%.

But that test period was selected for its lack of unusual market conditions, and Dr. Chen said the evidence is still early. He expects the system to be in use in the next two to five years.

Dr. Chen and Robert Schumaker, an AZFinText creator who has since become an assistant professor at Iona College in New York, also have been working on systems that analyze finance writers’ sentiment. The researchers also are expanding beyond Yahoo Finance and traditional news outlets to analyze blogs and investor and employee forums.

We hope these binary librarians learn to avoid the pages of Zero Hedge early on.