风险控制下的最优变现问题研究

发布时间：2018-08-02 20:16

【摘要】：在执行大型的机构交易过程中，我们的交易策略必须平衡延迟交易带来的等待风险和快速交易带来的变现费用。在本文中，我们研究风险控制下的最优变现问题。全文分三个部分，以最小化风险和交易费用的组合为目标既研究了单资产的动态变现和静态变现问题，也考虑了投资组合的动态情形。在第一部分，通过最小化清出一大笔某资产所得收益的指数衰减函数，我们得到了动态最优变现策略，其中衰减常数度量了风险厌恶程度。明确地讲，就是给定以股票清出数量和反映市场信息的相关序列为自变量的价格冲击函数，通过最小化在时间T内清空所持全部S支股票得到的收益的指数衰减函数来获取与该相关序列相关的最优变现策略，在某些特殊的情形下，我们可以得到解析表达式。紧接着我们在第二部分考虑了投资组合的动态变现，每个资产的价格变动都要受到其他资产价格变动的影响，这种价格之间的交叉作用对变现费用有着很重要的影响。为了更精确地追踪价格冲击变化轨迹，我们将冲击分为长期和短期价格冲击两部分，分别进行建模。最后，我们研究了单资产的静态变现问题。给定长期冲击和短期冲击模型，通过最小化风险和交易费用的组合，，我们得到静态的变现策略。对于简单的线性成本函数模型，我们构建了有效边界，对于既定的风险偏好，有效边界上的期望交易费用最小。在构建模型的过程中，我们同时进行了仿真实验，并和Bertsimas和Almgren的经典模型进行了性能上的对比。与前者相比，我们的模型对于风险和变现成本的控制都更好；与后者相比，我们的模型展现了更好的鲁棒性。
[Abstract]:In executing large institutional transactions, our trading strategy must balance the waiting risk of delayed trading with the liquidation costs of fast trading. In this paper, we study the optimal liquidity problem under risk control. This paper is divided into three parts. Aiming at minimizing the risk and transaction cost, the dynamic liquidity and static liquidity of single asset are studied, and the dynamic situation of portfolio is also considered. In the first part by minimizing the exponential decay function of the income from a large asset we obtain a dynamic optimal liquidity strategy in which the attenuation constant measures the degree of risk aversion. Specifically speaking, the price impact function is given with the stock clearing quantity and the related sequence reflecting the market information as independent variables. By minimizing the exponential decay function of the return of all S stocks in time T, we obtain the optimal liquidity strategy associated with the correlation sequence. In some special cases, we can obtain an analytical expression. Then in the second part, we consider the dynamic realization of the portfolio. The price change of each asset is affected by the change of other asset prices, which has an important impact on the liquidation cost. In order to track the trajectory of price shocks more accurately, we divide them into two parts: long term and short term price shocks, and model them separately. Finally, we study the static liquidity of single assets. Given the long-term and short-term shock models, we obtain a static liquidity strategy by minimizing the combination of risk and transaction costs. For a simple linear cost function model, we construct an efficient boundary. For a given risk preference, the expected transaction cost at the efficient boundary is the minimum. In the process of constructing the model, we have carried out the simulation experiment at the same time, and compared the performance with the classical model of Bertsimas and Almgren. Compared with the former, our model has better control of risk and realization cost, and our model is more robust than the latter.
【学位授予单位】：上海交通大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：F830.59

【共引文献】