【作者】 王平；

【导师】 张维；

【作者基本信息】 天津大学 ， 管理科学与工程， 2008， 博士

【摘要】 随着国内金融改革和经济全球一体化进程的加快,国内金融业在发展的同时也面临着前所未有的风险与挑战。一些金融机构的决策部门已经对金融优化在实践中的应用提出了具体要求。结合我国实际,把金融优化的理论和方法应用到金融决策的实践中去已经成为金融业界和学术界的共识。正确应对高度不确定性环境所带来的风险,优化资产配置,对我国机构投资者的长期发展有着至关重要的意义。传统的投资理论都是假设投资者是风险厌恶的,即在期望收益率相同的情况下,投资者会选择风险较小的投资组合。然而,在现实情况中却并非如此。中国的证券市场中,在投资者的平均持有期内,很多投资者收益率的期望值并不比储蓄的确定收益率高,甚至很多情况下低于储蓄的期望收益率。股市中的大部分投资者并没有因此退出股市,因此可以认为投资者是风险厌恶的并不完全符合实际情况。资产配置是决定一个投资组合收益与风险的最重要因素。动态资产配置,是进行正确的战略资产配置的前提,对于中长期的投资者具有重要的指导意义。损失厌恶是人们普遍存在的一种心理现象。现有的研究对损失厌恶和下侧风险测量方法基本都是独立进行的。将损失厌恶引入动态资产配置中可以更深刻的认识动态资产配置的本质,还可以为投资者提供更好的资产配置决策建议。下侧风险的测量不能充分反映投资者的风险偏好,通过研究两者的一致性,发展能刻画投资者复杂心理的动态的风险测量方法将对投资者的决策提供支持。本文在考虑风险收益分布的厚尾特性和投资者风险偏好的基础上,检验了与现有的研究不同的目标设定情况下的均值-下偏距(M - LPM)组合优化中的两基金分离定理以及其中货币分离的属性。为了在下侧风险控制下也能反映投资者的收益高于目标时的偏好,本文利用UPM /LPM优化模型检验投资者的组合收益偏好,使用模拟比较了在不同收益分布下不同优化模型的优化组合,并用Matlab编程工具对UPM /LPM模型的优化算法进行了实现,得到了一般随机数情境下的资产配置结果,并提出一个特定的指数检验投资者目标的改变对投资者投资偏好的影响。本文引入正态密度函数多项式的扩展即G-C扩展来简化偏度和峰度的定价过程,检验偏度和峰度对优化投资战略的影响。在使用下侧风险方法时,连续相对风险厌恶或均值方差优化的情况中,偏度和厚尾收益分布将会影响股票的优化配置。本文引入扩展的K-T标准求解考虑了最终财富最坏产出的资产配置模型,利用鞅方法和期权定价的方法构造优化战略,投资者的优化组合相当于持有一个共同基金的买权,并检验了下侧风险控制密度改变时对投资战略的敏感程度。更多还原

【Abstract】 With the development of domestic financial reform and economy globalization, the domestic financial industry also face unprecedented risks and challenges. Some decision-making department of financial institutions have put forward specific requirements about the application of Financial optimization in the practice. In light of China’s reality, it has become the consensus of financial industry and the academic community that apply the financial optimization theory and method to financial decision-making practice. On the long-term development of China’s institutional investors, handling risk brought about by the uncertainty environmental correctly and optimizing asset allocation have crucial significance.In the traditional investment theory, investor is supposed to be risk aversion, that is, investor would choose less risky portfolio on the same expect return. However, it is not right in the practice. In the China’s securities market, many investors’expected return is lower than the saving in their average holding period. But most of the investors do not stop investment. So assuming that investors is risk aversion is not entirely consistent with the actual situation.Asset allocation is the most important factors to decide the yield and risk of a portfolio. Dynamic asset allocation is the premise to make correct strategic asset allocation and has important guiding significance to long-term investors. Loss aversion is a widespread psychological phenomenon. Existing research on loss aversion and downside risk are independent. Introducing loss aversion to dynamic asset allocation can have more profound understanding of the nature of dynamic asset allocation and provide better decision-making recommendations. Downside risk can not fully reflect investor’s risk preference. Researching the consistency and finding dynamic risk measurement methods which can describe the complex psychology will provide support to the decision-making of investors.On the base of fat tail and risk preference of investors, we prove two-fund separation theorem with different targets in the context of mean-LPM portfolio optimization and the properties of its monetary separation. In order to explain risk-seeking and risk-averse behavior above as well as below the target return, this paper use a general Upper Partial Moment/Lower Partial Moment (UPM/LPM) model .In particularly, we apply simulation to compare optimization portfolio of the UPM/LPM model and we find result of asset allocation by Matlab. And by choosing appropriate exponential, the UPM/LPM model is able to reflect investors’various asymmetric preferences.In this paper, polynomials extend of normal density function are introduced to explain skewness and kurtosis.We study the impact of skewness and kurtosis on the optimal investment strategic and find fat tail distributed will affect asset allocation of equity capital which is different with others.We study the model with worst-case portfolio outcome. Using extent K-T ,martingale and standard option pricing results, the optimal strategy is derived. The optimal policy is equivalent to the hedging portfolio of a European option on a dynamic mutual fund. We examine the sensitivity of the investment strategy as the downside risk control intensity changes.更多还原