| Financial markets are a complicated and uncertain system. In this complex and uncertain financial environment, the investment institutions and individual investors will be faced with how to realize the effective allocation of capital and they also prospect to maximize the benefits and risk minimization of equilibrium balance. In many cases,investors are often not able to accurately predict the probability distribution of asset returns,the description of the benefits and risks normally only be summed up with some vague language. For this reason, investors will make investment decisions in a vague uncertain the investment environment. Therefore, we will discuss portfolio optimization model at the possibility of a theoretical framework in order to make a good solution to solve this problem.With portfolio theory developing, the scholars found that it is not a good definition of investor risk characterization using variance to measure the investment risk, but the research have shown that the return rate of financial assets tend to have a non-symmetrical peak and thick distribution of the end. The mean and variance of the asset are not well characterize distribution of asset returns when the return on assets is not normally distributed. In order to overcome the mean-variance model deficiencies, this paper will be based on the skewness of the return of assets, the semi absolute deviation distribution as the investment risk function to research the portfolio optimization problem. At the same time, liquidity as the basis and prerequisite for the existence of the stock market is an important attribute of financial markets. There’s always having transaction cost in the transaction process, such as taxes, transaction costs, trading volume limits, bonuses and other friction and transaction costs are often ignored securities will lead to ineffective investment decisions in the stock market. Existing research shows that ignore financial assets transaction costs will lead to ineffective investment decisions. Therefore, we consider the impact of the stock market risk measurement, distribution yield, liquidity and transaction costs and other factors for investment decisions to establish a theoretical framework with mean-semi absolute deviation-skewness multi-objective portfolio optimization model which contains transaction costs and liquidity constraints.Due to the multi-objective portfolio optimization problem is a kind of complicated and intractable problem, it is hard to find a satisfactory solution within a limited time, this problem is also proved to be NP problem with constraints. It is more complex and difficult to use conventional method to solve this problem.Therefore, this article will focus on theportfolio optimization model with complex constraints, we propose an improved constraint artificial bee colony algorithm to solve this constraints multi-objective optimization problem, and by using the real data on the securities market model. Finally comparing the improved multi-objective artificial colony algorithm with other multi-objective algorithm,it is verified that the practicality of the model and the validity of the improved algorithm. |
PDF Full Text Request