论文部分内容阅读
在给定相关假设的基础上,考虑了在某一交易时间段的初始时刻,“庄家”型封闭式基金投资者初始持有资金额度有限的状况,将“庄家”型基金投资者的基金价格控制决策分为两个阶段:第一阶段为在确定各时刻期望收益率及其收益离差的条件下求得该投资者在对应时刻的最优投资值,第二阶段以各时刻的最优投资值为约束来获得最优基金价格控制序列。对于第一阶段可建立使“庄家”型基金投资者期望收益最大、收益平均绝对离差最小(风险最小)的双目标模型,变换为一个单目标模型后利用库恩-塔克条件进行求解,从而获得各时刻的最优投资值;对于第二阶段则以第一阶段得到的各时刻最优投资值为约束,以“庄家”型基金投资者在该交易时间段内的现金支付为极小值(现金收益的极大值)建立目标函数,而其为一个带约束的非线性规划问题,对此采用了一种改进的遗传算法进行求解,最终获得基金的最优价格控制序列。
Given the relevant assumptions, considering the initial moment of a trading session, “Banker ” type closed-end fund investors initially held a limited amount of funds, the “Makers ” fund investment Of the fund price control decision-making is divided into two stages: the first stage in determining the expected rate of return of each moment and its yield deviation under the conditions of the investor in the corresponding moment optimal investment value, the second phase of each The optimal investment value at the moment is a constraint to obtain the optimal fund price control sequence. For the first stage, we can set up a dual-objective model that allows the investors of the “banker” type to expect the highest return and the smallest average absolute deviation of returns (the least risk), transform them into a single-objective model and then use the Kuhn-Tucker condition For the second stage, the optimal investment value at each moment obtained in the first stage is used as the constraint, and the cash of the “Investor” fund investor during the trading period An objective function is established for the minimum value (the maximum of cash returns), which is a constrained nonlinear programming problem, and an improved genetic algorithm is used to solve the problem. Ultimately, the optimal price control of the fund is obtained sequence.