Words matter: a simple, timeless fact that recently, with the advent of data science, has begun to take on a whole new meaning.
Every three months, publicly traded companies announce their earnings and revenue results for the previous quarter. Executives for these companies share this information with the public through press releases and earnings conference calls. Investors and analysts carefully read each line, and often between the lines, looking for clues as to the future revenue streams of these companies.
This isn’t purely an exercise in reading tea leaves. At the time of these calls, the executives speaking likely know far more information about the health of their companies than is reflected in the contents of their earnings. After all, these earning calls typically occur one month following the end of the preceding quarter. Or put it another way, at the time of these calls executives have a third of a quarter’s worth of new information in their heads – customer interactions, sales, and strategy meetings – not reflected in the previous quarter’s immediate financials. Executives may make subtle changes in their language choice, either consciously or unconsciously, during these calls that reflects that additional information. For example, a CEO who has since received good news and knows good times are ahead for the company may use bolder words that reflect his positive sentiment. By contrast, the CEO of a company now seeing a lot more risk since the close of the previous quarter might use more conservative words that reflect his tempered outlook.