“If something bad is going to happen, it usually happens in September.” The pronouncement may sound like it comes from a Charles Dickens novel, but its source is far more modern. It’s from the financial column “Abreast of the Market” in the Wall Street Journal, a daily article covering the fluctuations of the stock market.
Full of catchy phrases and lively quotes, the column’s language may be memorable. But according to Dr. Paul Tetlock, it may also be predictive.
Dr. Paul Tetlock
Tetlock, assistant professor of finance at the McCombs School of Business, has found that certain media sources, including “Abreast of the Market,” are harbingers of what’s to come in the Dow Jones Industrial Average and Standard & Poor’s 500, even when they don’t talk about hard numbers. In other words, as goes the column, so goes the market.
“In a perfectly rational world, stock prices would only incorporate economic information about firms’ values,” Tetlock says. “But it’s also possible that they respond to psychological information about how traders are feeling.”
To understand those feelings, you don’t need to be on the trading floor probing the psyches of those buying and selling. You need only read the financial press. “Abreast of the Market,” a long-time feature in the Wall Street Journal, is a particularly rich source of psychological information. Appearing daily in the paper’s Money & Investment section—often front and center on the first page—the column provides an overview of the highs and lows of yesterday’s trading.
“It’s kind of a post-mortem of the prior day’s activity,” Tetlock explains. “The column talks a lot about traders’ feelings and expectations for the future: the psychological aspects of the market. Although it does reference firms’ business operations and profitability, that’s not really the main focus of the column. It’s psychologically rich, not economically rich, in content.”
This was important to Tetlock, whose focus on the market is less about rows and rows of numbers and more about the investor sentiment that drives those numbers. It may be because his parents are psychologists, he says, but he’s particularly interested in the people behind the supply and demand curves economists study. If conventional wisdom says a dark mood among investors leads to a drop in stock prices, can that be measured?
“Abreast of the Market,” a daily article in the Wall Street Journal, covers the fluctuations of the stock market.
“Abreast of the Market” gave Tetlock a perfect vehicle for finding out. A typical column includes quotes from management company presidents and investment strategists. It may report that a stock “failed to impress” or that “plenty of dangers still lurk” or it may ponder the moods of money managers as they come home from their summer holidays.
Tetlock analyzed every “Abreast of the Market” column from 1984 to 1999, using a computer to read more than 3,700 editions of the column. He relied on a content analysis program based on the Harvard psychosocial dictionary in which psychologists categorized most of the commonly used words in the English language into one of 77 categories. The two catch-all categories are positive and negative words. The program allowed him to count the positive and negative words in the column.
After a particularly rough day in the market, the column is heavy with negative vocabulary. This may reflect the day that has passed, but Tetlock found it also predicts the days to come. Using a statistical procedure that controls for the influence of other economic variables, Tetlock found direct correlations between language in the column and the movement of the market.
“When there were a lot of negative words in the column, the stock market tended to fall on the day of the column and the next day as well,” Tetlock says. “The decline in the market, however, was subsequently reversed over the next four trading days, so within a week’s time the stock market returned to where it had been.”
With all other things held equal, Tetlock found that a typical increase in the column’s negativity foreshadowed a fall of 0.081 of one percentage point, or 8.1 basis points, in the Dow the next trading day. While that may sound like a subtle change, it’s actually quite significant. The average daily return of the index during the same period was only 5.4 basis points. The 8.1 basis point difference amounts to billions of dollars in firm valuations.
If negative words in a newspaper column can predict a downfall in the market, Tetlock’s findings beg the question: Are investors sitting down to read “Abreast of the Market” with their morning coffee and trading accordingly over the course of the day?
Tetlock admits it’s hard to say. It’s possible that people are reading the “Abreast of the Market” column and responding to its information and mood, a response that is visible in the market by the close of day. But it’s also possible that the column is capturing investor sentiment right before it is reflected in stock prices. Simply put, its reporting is slightly ahead of the curve.
“Unfortunately, I’m not able to know whether it’s the column causing the movement or whether the column is just accurately reflecting investors’ moods,” Tetlock says. “My suspicion is that the column is merely capturing the general mood of the market and how traders are feeling, but I can’t prove that one way or another.”
What he can prove is that investor sentiment is a powerful force. Economic data alone are not driving the price of stocks--the moods of those making the calls on the trading room floor clearly have an impact. And those moods, ultimately, may be measurable.
Tetlock’s findings then beg a second question: Has he uncovered a slick new way for investors to game the market?
Tetlock laughs at the suggestion. Each time he gives a seminar about his findings he receives calls from people asking him that same question. He urges caution.
“My research does show that there is some amount of predictability in stock market returns based on negative words,” he says. “So knowing which words are negative and knowing how many negative words are out there is helpful in guiding you when you are buying and selling. But I wouldn’t use it as the sole basis for a trading strategy.”
Tetlock has found it’s not just the “Abreast of the Market” column that can be predictive of changes in the market. He’s studied the impact of firm-specific articles in the financial press, looking at how reporting in the Wall Street Journal and the Dow Jones News Service influences firms’ stock prices. And he’s now investigating whether the words the Federal Reserve uses in speeches and testimony have an appreciable impact on the market.
“I just read a headline that said, ‘Market goes up on Fisher’s comments,’” Tetlock says. “Richard Fisher is the president of the Dallas Fed, so that’s exactly what I’m trying to study. Why did the stock market go up after Fisher’s remarks? Did they convey economic information? Or did they convey information about what the Fed is going to do? Perhaps studying the language used by the Fed will help disentangle its journalistic reporting role from its interest rate policymaking role.”
Tetlock points out that his work underscores the kind of influence the financial press can have on the country’s economy. The Wall Street Journal, for example, has a circulation of more than two million readers.
“It would be difficult for economic information to become quickly reflected in stock prices unless someone communicates that information to traders,” Tetlock says. “Journalists clearly play that role. But the media also provides an invaluable source of data on the sociology and psychology of the marketplace. Studying the language used in the financial news can improve our understanding of whether economics or psychology is moving markets.”