模糊不确定环境下的多期投资组合模型及算法研究

Securities investment is usually affected by many factors such as political, economic, statutory and investors’ psychological factors and so on. It leads to the fact that the returns and risks on assets are largely associated with fuzzy uncertainty. Thus, investors need to make portfolio decision under fuzzy environment. Multi-period fuzzy portfolio selection is one of topic problems concerned by international academia in recent years. So far, the studies about the kind of aforementioned problem are just beginning. This project investigates some multi-period fuzzy portfolio selection problems, strives to develop the theory about multi-period portfolio selection and studies their applications. The main research contents of the project are summarized as follows. Firstly, we select suitable decision indexes, consider various real investment constraints and propose multi-period fuzzy portfolio selection model based on possibilisic higher moments. Secondly, we introduce risk aversion methods, analyze the influence of historical prediction deviation on current investment decision, construct a multi-period interval portfolio model and a multi-period fuzzy portfolio selection model with feedback control policies. Thirdly, we use fuzzy decision theory to handle the investment preference information and propose a multi-period performance evaluation model for fuzzy portfolio selection. Finally, we design some efficient algorithms to solve these proposed complex portfolio models and make some empirical analysis. The aim of the project is to propose some practical portfolio selection models, design some efficient algorithms to eliminate the gap between the current theoretical research and practical application, and present some feasible policy suggestions to provide scientific support for long term investment decision.

证券投资通常受政治、经济、法律及投资者心理等因素影响，这使得资产收益和风险在很大程度上具有模糊不确定性，投资者需在模糊不确定环境下进行投资决策。多期模糊投资组合选择是近年来国际学术界关注热点问题。目前，有关该问题研究刚刚开始。本项目研究一些多期模糊投资组合选择实际问题，力求发展多期投资组合选择理论并开展应用研究。研究内容包括：首先，选取合适决策指标，考虑各种现实投资约束，建立基于可能性高阶矩多期模糊投资组合模型。其次，引入风险规避方法，分析历史预测偏差对当前投资决策影响，建立多期区间投资组合模型和具有反馈控制策略多期模糊投资组合模型。再次，利用模糊决策理论处理投资偏好信息，建立基于绩效评估多期模糊投资组合模型。最后，设计求解各种复杂模型的高效算法并进行实证分析。本项目研究预期提出贴近实际的模型，设计高效算法消除理论研究与实际应用的差距，提出切实可行的政策建议为长期投资决策提供科学依据。

证券投资受政治、经济、法律及投资者心理等因素影响，这使得资产收益和风险在很大程度上具有模糊不确定性，于是投资者需在模糊不确定环境下进行投资决策。投资者的投资行为通常是多期的，他们需不断地调整投资组合头寸来实现财富快速增长并有效控制风险。多期模糊投资组合选择是近年来国际学术界关注热点问题。目前，有关该问题研究刚刚开始。本项目研究多期模糊投资组合选择问题，力求发展多期模糊投资组合理论并开展应用研究。研究内容包括：首先，选取合适决策指标，考虑各种现实投资约束，建立基于可能性高阶矩的多期模糊投资组合模型。其次，引入风险规避方法，分析历史预测偏差对当前投资决策影响，建立多期区间投资组合模型和具有反馈控制策略的多期模糊投资组合模型。再次，利用模糊决策理论处理投资者偏好信息，建立基于绩效评估多期模糊投资组合模型。最后，设计求解各种复杂投资组合模型的高效算法并开展模型应用研究，为长期投资提供科学方法和技术支持。