比例交易成本下的模拟动态资产分配：多阶段随机规划

发布时间：2018-06-13 00:20

本文选题：动态资产分配 + 随机规划　；参考：《厦门大学》2013年硕士论文

【摘要】：本文旨在研究模拟环境下的动态资产分配系统,例如,一个风险厌恶的投资者在选择收益互相依赖的不同资产时面临交易成本。这个动态资产分配系统可以帮助该投资者在短期内实现预期效用最大化。该系统建立在一个用多阶段随机规划来处理市场不确定性的实践模型基础上。这个模型稍微偏离了大多数有关资产分配论文的数学复杂性,而是利用一个能够很好描述市场的简单模型来表现交易成本。为了证明它的有效性,我用GAMS设立一个小的应用来提升它。我通过改变交易成本、风险意识、时间长度和目标在模拟环境中进行了4次独立的实验操作。所有的实验结果表明交易成本和实际环境中是一样的。我将我的研究结果和其他有不同假设环境的学术研究结果进行了比较,他们很好的与主流理论与学术直觉契合。我的模型很好地反应了现实并具有实际意义。我的研究结果进行了稳健检验,证实了其有限性并且可以成为一种可靠的工具。
[Abstract]:The purpose of this paper is to study dynamic asset allocation systems in simulated environments. For example, a risk-averse investor faces transaction costs when choosing different assets with interdependent returns. This dynamic asset allocation system can help the investor maximize expected utility in the short term. The system is based on a practical model of multistage stochastic programming to deal with market uncertainty. This model deviates slightly from the mathematical complexity of most papers on asset allocation and uses a simple model that can describe the market well to represent transaction costs. To prove its effectiveness, I use GAMS to set up a small application to enhance it. I conducted four independent experiments in a simulated environment by changing transaction costs, risk awareness, length of time, and goals. All the experimental results show that the transaction cost is the same as in the real environment. I have compared my findings with those of other hypothetical environments, and they fit well with mainstream theories and academic intuition. My model well reflects the reality and has practical significance. My research results are robust test, which prove its finiteness and can be a reliable tool.
【学位授予单位】：厦门大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：F830.9;F224

【相似文献】