An evolutionary-based approach for optimising diverse group stock portfolio with active and inactive stocks

When a stock portfolio is suggested to inventors, they may need a mechanism to replace stocks when their future prospects are pessimistic. However, existing approaches only consider all assets to find a diverse group stock portfolio (DGSP), which may suffer massive losses as a result. In this paper, an intelligent optimisation algorithm is proposed to obtain a more profitable DGSP with active and inactive stocks. In the coding scheme, not only grouping, stocks, and weighting but also active stock parts are used to represent a DGSP. Two evaluation functions are developed according to five factors, including group balance, modified portfolio satisfaction, price balance, unit balance, and extended diversity factor. These functions are used to assess the fitness of a chromosome. Finally, empirical studies were conducted on two financial datasets to show the merits of the proposed algorithm.