| Financial scholars and asset managers have focused extensively on the problem of how to allocate assets in a way that ensures maximum risk control and returns.Since Markowitz proposed the single-period mean-variance model to solve the portfolio problem,portfolio research has grown rapidly.Nowadays,there are many scholars who use intelligent algorithms to solve portfolio problems and conduct simulation experiments.In this paper,we study the multi-period portfolio model,using intelligent algorithms to solve and analyse it empirically,and carry out research on portfolio decision making and risk control.The main content of the research is as follows:1.In order to address the shortcomings of the traditional mean-variance model in practical applications,we introduce transaction costs and basis constraints,develop a probabilistic mean-lower semi-variance multi-period portfolio optimization model in a fuzzy environment,and design an improved differential evolutionary algorithm to solve the model.The numerical results show that the algorithm is effective and the model is feasible.2.In order to address the problem that the assumptions of the traditional mean-variance model on returns cannot be fully adapted to the actual securities market,we introduce return weights,develop a probabilistic mean-lower semi-variance-entropy multi-period portfolio optimization model in a fuzzy environment,and design an improved genetic algorithm to solve the model.The simulation results show that the algorithm is feasible and the model is reasonable,and can better reflect the market conditions.3.The semi-absolute deviation is used to measure the risk in the investment process,in order to address the reality that investors are more concerned about loss of return.The model is solved by a hybrid genetic-differential evolutionary algorithm.The numerical results show that the algorithm is feasible and the model is reasonable,which can help investors to avoid risks and obtain excess returns. |
PDF Full Text Request