不确定环境下的养老保险基金投资组合模型研究

发布时间：2018-06-13 23:18

本文选题：投资组合选择 + 无风险保障系数　；参考：《华北电力大学》2014年硕士论文

【摘要】：养老保险基金投资的目标是在可允许的风险下获取尽可能高的收益。现代投资组合理论的核心问题是处理不确定性。经典的投资组合理论均将研究中涉及的不确定性视为随机性来处理,然而,随着行为金融学的发展,证券市场中的模糊不确定性也逐渐被人们所关注和认识。模糊理论及可信性理论随之发展起来,并被运用到投资组合模型的研究建立当中。但模糊性理论和可信性理论在描述投资的主观预期收益时都存在着一定的缺陷。清华大学刘宝碇教授在2007年提出了不确定测度,进而建立起了一套不确定性理论。许多学者在其基础上进行研究,并将其应用于了投资分析中。本文利用不确定理论进行了不确定环境下的养老保险基金投资组合模型的研究。首先,本文在不确定理论的框架下,考虑到养老基金投资时其安全性是首要的准则,针对无风险资产收益为保障整体投资组合达到预定收益目标值所起的类似安全防护网的作用,提出了无风险保障系数(Risk-free protection index,RFPI)。无风险保障系数可以度量在一定置信水平下,无风险资产收益为投资组合面临的损失所提供的保障程度,因此可以衡量无风险资产收益对风险资产收益所起的保护作用。进而,本文分别建立了基于无风险保障系数的不确定投资组合模型和双目标无风险保障系数模型,并给出各自的具体计算步骤。随后,对模型进行了熵优化。每个投资组合模型既考虑到了投资组合整体收益的波动情况,又可以在一定水平的无风险资产收益保障的作用之上最大化投资期望收益,这为机构投资者进行投资组合的选择提供了一种新的建模方法。最后,通过算例分析,验证了模型的实用性和可行性。并归纳了养老保险基金无风险保障系数投资组合模型的规律,具体如下：(1)随着无风险保障系数的增大,投资于无风险资产的比例将随之提高。(2)随着无风险保障系数的增大,投资组合的总体期望收益率将逐渐下降。(3)随着无风险保障系数的增大,投资组合的总体不确定在险价值(Value at Risk in Uncertainty,VaRU)将逐渐减小。(4)随着无风险保障系数的增大,方差最初稳定在临界值处,但当风险保障系数继续增大到一定程度以后,方差将逐渐降低。此外,通过与不加入无风险保障系数的模型进行比较,发现无风险保障系数模型在风险厌恶程度较高时可以得出更好的结果,验证了本文提出的模型对于养老保险基金的适用性。
[Abstract]:The purpose of the pension fund investment is to obtain the highest possible return at allowable risk. The core problem of modern portfolio theory is dealing with uncertainty. The classical portfolio theory treats the uncertainty involved in the research as randomness. However, with the development of behavioral finance, the fuzzy uncertainty in the securities market is gradually concerned and recognized. The fuzzy theory and credibility theory have been developed and applied to the establishment of portfolio model. However, both fuzzy theory and credibility theory have some defects in describing the subjective expected return of investment. Professor Liu Bao Anding of Tsinghua University put forward the uncertainty measure in 2007, and then established a set of uncertainty theory. Many scholars have studied it and applied it to investment analysis. In this paper, the investment portfolio model of endowment insurance fund under uncertain environment is studied by using uncertainty theory. First of all, under the framework of uncertainty theory, considering the security of pension fund investment is the first criterion, In view of the role of riskless asset income as a similar safety protection net to ensure the overall portfolio to reach the target value of return, a risk free protection index is proposed. The risk-free security coefficient can measure the degree of protection that risk-free asset returns provide for portfolio losses at a certain confidence level, so we can measure the protective effect of risk-free asset returns on risk asset returns. Furthermore, the uncertain portfolio model based on risk-free security coefficient and the two-objective risk-free guarantee coefficient model are established in this paper, and their calculation steps are given. Then, the model is optimized by entropy. Each portfolio model not only takes into account the volatility of the overall return of the portfolio, but also maximizes the expected return on the basis of a certain level of risk-free asset income protection. This provides a new modeling method for institutional investors to choose their portfolios. Finally, the practicability and feasibility of the model are verified by an example. It also sums up the law of the risk free security coefficient portfolio model of the pension fund, which is as follows: 1) with the increase of the risk-free security coefficient, the proportion of investment in the risk-free assets will increase with the increase of the risk-free security coefficient, and the risk-free security coefficient will increase with the increase of the risk-free security coefficient. The overall expected return of the portfolio will decrease gradually.) with the increase of the risk-free security coefficient, the overall uncertainty of the portfolio will decrease gradually with the increase of the risk-free security coefficient, and the overall uncertainty of the portfolio will decrease gradually with the increase of the risk-free security coefficient. The variance is stable at the critical value at first, but when the risk protection coefficient continues to increase to a certain extent, the variance will gradually decrease. In addition, by comparing with the model without risk-free security coefficient, it is found that the risk-free security coefficient model can get better results when the risk aversion degree is higher. The applicability of the proposed model to pension funds is verified.
【学位授予单位】：华北电力大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：F224;F840.61

【参考文献】