Red-Blooded Risk: The Secret History of Wall Street

eRaider.com

I once tried to exploit the knowledge that it's possible to create efficient small portfolios. I was an early computer communications adopter, from the Defense Advanced Research Projects Agency Network (DARPANET) at school to dial-up bulletin boards, to UseNet, to CompuServe, and, in 1995, to Yahoo! and America Online. In these last four forums, I answered financial, statistical, and gambling questions. It was in 1996 that I became impressed with the power of stock bulletin boards. There was a lot of hype and nonsense, of course, but what nonparticipants couldn't see was that it was easy to filter that stuff out and, when you did, boards for some companies had a lot of useful information and discussion. A friend of mine, Martin Stoller, a popular professor at the Kellogg School of Business at Northwestern University, gradually figured out that for some companies a significant fraction of the float of the stock—the shares held by the general public, including institutions but not insiders—could be reached through the boards. Not only that, but there was a lot of tremendously valuable information—valuable because it was specific—from current and former employees, suppliers, customers, and people knowledgeable about the industry.

Martin and I came up with the idea of forming a portfolio of seven stocks with low mutual correlations, so that the combined portfolio had a lower standard deviation than the market. These would be companies with low institutional ownership, so ...