eInformation: a clinical study of investor discussion and sentiment

Financial Management (Financial Management Association), Autumn, 2005 by Sanjiv Das, Asis Martinez-Jerez, Peter Tufano

Using traditional measures, the boards do not appear concentrated. In Table II, Panel A, we calculate Herfindahl indices for the 16 stock boards (four stocks times four board vendors). In only four cases is the share of message concentration at about the level that the Department of Justice would consider "mildly concentrated" in product markets (i.e., 1,000), and in each the number of postings is small. However, the distribution of posting activity is highly skewed. Table II, Panel B reports the number of postings by screen name. There is a relatively small and vigorous core of frequent posters, surrounded by a large number of occasional posters and by unobserved "lurkers," who only read the postings.

While we might understand why someone might post a few messages and then lose interest, it is less clear what motivates someone to post over 5000 messages about a single stock in a bit more than half a year. The time and effort expended by this person was considerable. Why?

B. Profile of an Active Poster

In a large empirical study, it is normal to discard anomalous observations. In contrast, in a clinical study, we can and should understand outliers. It is in this spirit that we interviewed Glenn R., the most prolific poster on the Amazon boards. Glenn gave an interesting interpretation of membership in a posting group, which we report at length to give readers a first-hand look at an active poster.

Glenn was in his late 40s when he was posting the messages we studied. He has an undergraduate degree in engineering from a large Midwestern university, and had completed some of the requirements for a business degree. He owned a small chain of jewelry stores, including an on-line jewelry store, and was self-employed. He did most of his postings on nights and weekends, when he was not otherwise busy at work. He estimated that he spent approximately 30 hours a week interacting on the boards.

He was a client of a large brokerage firm and read the professional analyst reports he received. A few years later, when we interviewed him, he was still able to cite analysts by name. He was interested in stocks and in technology, so he gravitated toward tech stocks. He also actively searched the web for news stories about these stocks. He provided four explanations for his activity.

1. Learning

"I wanted to learn." Glenn repeatedly emphasized that he lived in a small town of 15,000 people and that there were no investment clubs in his town. He reports, there were "not many people in town that he could talk to about investing." His activity in the Silicon Investor board was equivalent to membership in an investing club. Glenn was keen on learning from "people who had more experience than (he) had." In particular, he felt the boards were quite good in providing information on market microstructure details and technical analysis, especially the nuances of shorting stocks and the daily fluctuations in the outstanding float of the stock. Glenn approached stocks from the perspective of fundamental analysis, but was intrigued by the approaches of technical analysts that seemed to give them "a better batting average."