Removing the necessity of simplifications in large-scale portfolio selection and implications to Chinese portfolio theory and management

Purpose - The purpose of this paper is to argue that simplifications of portfolio selection may no longer be necessary, based on computational advancements of portfolio theory and powerful computers. Design/methodology/approach - First, the paper reviews the two branches of portfolio optimization and second, presents the results of large-scale portfolio selection, based on exhaustive sampling in China. Some speedy results support removing the simplification. Findings - The paper finds that for some simplification techniques, the results of simplified models and original models are quite alike, while for other techniques, the results are strikingly distinctive. Moreover, the performance of portfolio optimizers varies from being instantly fast to being unbearably slow, so it pays to be picky. Originality/value - This paper reports for large-scale portfolio selection the results of kinds of software and this alone makes the paper unique. Based on the leading software and exhaustive sampling in China, for the first time the difference between the original and simplified models is studied.