ABSTRACT
We have developed an expert system, called PORSEL (PORtfolio SELection system), which uses a small set of rules to select stocks. Each year, PORSEL selects twenty stocks from the listed stocks to form an investment portfolio for that year.
This paper improves the PORSEL by incorporating a better method of constructing portfolios. The new PORSEL now consists of two components: the Fuzzy Stock Selector and the Portfolio Constructor. The purpose of the Fuzzy Stock Selector is to evaluate the listed stocks and then assign a composite score for each stock. This component is essentially the old version of PORSEL [ZH96]. The second component, Portfolio Constructor, is a new component designed to generate the optimal portfolios for the selected stocks. It is based on the Markowitz portfolio selection model.
The results of simulation show that our new version of PORSEL outperforms the market every year during the testing period. For 14 years, the average excess return of our system over the S&P 500 Index is 121.38 percent.
