Business Week: Risky Business, Risky Words
In the 12-June-2006 edition of Business Week there is a neat little blurb about Professor Feng Li. He’s counted up the number of times that words like “risk” and “risky” and “uncertain” show up in 34,180 annual reports from 1993 through 2004. Then he looked at those numbers in comparison with the previous year.
Let me just stop you for a moment. This guy simply counted up words in annual reports. Counting isn’t too hard and annual reports are freely available. However, the idea is still genius. It gets at the heart of investing at so many levels.
Professor Li discovered that a “big jump in words related to risk is usually followed by poor share performance” which makes a ton of sense. He built a model portfolio based on this data. The punch line is that he would have outperformed the S&P 500 index by 6% per year since 1995. Smashing!
This story is really interesting but I strongly feel it has implications beyond investing. I’ve mapped out 2-3 business models which are based on the principles outlined above. One model is tied to blog postings and content valence and the other is tied to RSS feeds and stock prices. If I had time and a programming team, I could do damage.
But I digress. My point is there is a vast pool of publicly available information; free and open data. We’ve moved from a world of collecting information to using information. I mean, really using it. Not searching for it. Not mining it. Not sharing it. I’m talking about really using it in new ways. There are so many combinations just waiting to explode.
What is the future of search? Where are the innovations? Who is making the magic happen? Find out what I think in Found and Lost.
July 10th, 2006 at 11:22 am
I had a longer comment on this which, by the looks of it didn’t make it through. The gist of it was…
This kind of stuff is cool.
Here is a link to a guy at Accenture who works on stuff like this, though he hasn’t posted lately: http://www.accenture.com/Global/Research_and_Insights/Accenture_Blogs/Mike_Bechtel/default.htm
And we should definately work on this sometime, John. I’m freeing up in mid August.
July 10th, 2006 at 6:08 pm
[…] WebWord has a good write up about Professor Feng Li who data mined the annual reports of 34,180 companies with some interesting results. Professor Li discovered that a “big jump in words related to risk is usually followed by poor share performance” which makes a ton of sense. He built a model portfolio based on this data. The punch line is that he would have outperformed the S&P 500 index by 6% per year since 1995. Smashing! […]