I did find one truly strong correlation in the data -- with an r of 0.70, no less. That was the link between the total number of words in a shareholder letter and the closing price of the S&P 500 at the end of the year covered by the letter. There's also a pretty strong correlation between the word count and the S&P 500 at the end of the next year. Even looking at the S&P four years later the r is still an impressive 0.62. Which surely tells us something about the limits of exercises such as this.
 

First « 1 2 » Next