| Portfolio theory try to solve the following problem: how to allocate investment capital,maximize the return and minimize the risk. Since stock market is changeable and full ofdifferent kinds of uncertainty, portfolio is also full of a variety of uncertainty. Uncertainty isperformance at randomness and fuzziness. The paper will unite both randomness andfuzziness, then we will assume that asset return is fuzzy random variable. In order to reflectthe actual situation, investor’s subjective factors, expert advice and transaction costs are alsoadded to our model.The paper can be divided into the following three parts:(1)Assume the investment return is fuzzy random variable, research the mean-varianceportfolio model under the fuzzy condition. In order to defuzzification of our model, we givethe Lambda ranking method to realize our goal.(2)Common measurement of investment risk are variance and semivariance, Roy alsogave the safety first model, he researched the risk in another way. In our paper, we will givethe new investment risk measurement, which is credibility theory. And then we also putforward the new portfolio model, in which investors’ subjective opinions, expert advice andtransaction cost are added.(3)Generally speaking, our model is difficult to solve in using the traditional method. Inthis case, we offer the Hybrid Intelligent Algorithm which is the combination of Monte CarloAlgorithm and Genetic Algorithm. Monte Carlo Algorithm is used to calculate the Expectedvalue and Genetic Algorithm is used to find out the best portfolio solution. Our real stockexchange data study shows that the Hybrid Intelligent Algorithm has very good convergenceand stability. |
PDF Full Text Request