Internet Investment Advice: Investing with a Rock of Salt

Abstract

This study investigates the value contained in stock recommendations posted on two Internet newsgroups. I estimate the quality of the advice using conventional abnormal return methodology around the posting date. I find that stocks recommended do not earn any significant cumulative abnormal return over the next 5, 10 and 20 trading days. I also report that the posters tend to select stocks with extreme recent performance. Sixty percent of the stocks recommended significantly out-performed their benchmark. However, only the stocks in the lower quintiles of prior performance generate significant abnormal returns whereas top prior performers encounter significant under-performance subsequent to the posting. Posters exhibit over-confidence in believing that any new product introduction or possibility of a business alliance will always generate positive returns to companies.