Contributions from large crowds have often shown judging, estimating, or decision making abilities that are of an equal or better quality than those made by domain experts. This effect is called the “wisdom of crowds.” One area where improvement in decision making may be possible is in making stock recommendations.
The authors propose a decision support system (DSS) that enables investors to include crowd-based recommendations in their stock portfolio management investment decisions. The DSS collects crowd votes each day and submits them to the modeling process. There are two phases of the daily modeling process. Crowd votes along with share prices and share metadata are input to a modeling facility and metric computations to yield a rating computation as the output of phase 1. The ratings are input to phase 2, which is the investment phase from which shares of stocks are selected using an investment strategy to build a portfolio.
The authors built a prototype and ran two test cases over a period of two years. The simplistic first case was to invest all available cash into the best-ranked stock each day. The second test case involved creating a portfolio of the crowd’s top ten selections. The authors evaluated both test cases against a market benchmark index over that time. Both test cases outperformed the market index and thus were positive indications of the viability of the authors’ concepts.
Readers who are interested in how DSS can help with investment strategies should find this paper useful.