Non-fixed Assets Index Fund Optimization by Using Evolutionary Algorithms

本文利用进化算法对由非固定资产组成的指数基金进行了组织。非固定资产是指指数基金动态选取的资产。基于现代投资组合理论的投资组合由固定资产组成，其优化问题是确定固定资产的比例加权组合。我们提出的优化方法利用了进化算法，利用了评价函数的理论性质，可以动态地从所有资产中选择资产到指数基金，并确定这些选择资产的权重。在东京证券交易所的数值实验表明，与传统的优化方法相比，本文提出的方法能够更有效地优化指数基金，并且具有更小的计算时间。

In this paper, we organized the index funds consisting of the non-fixed assets by using evolutionary algorithms. Non-fixed assets mean the assets that are dynamically selected to the index funds. The portfolios based on the Modern Portfolio Theory consist of the fixed assets and their optimization problems are to determine the proportion-weighted combinations for the fixed assets. Our proposed optimization methods, that use the evolutionary algorithms employing the theoretical properties of the evaluating functions, can dynamically select assets from all the assets to the index funds and can determine the weights of these selected assets. In the numerical experiments on the Tokyo Stock Exchange, we showed that our proposed methods can optimize the index funds that have more effective performances and smaller computational time than those of the traditional optimization methods do.