Portfolio Investment Model Based On CVaR And Restrictions Of Average Deviation Volatility

Portfolio is one research focus of the current financial theory, seeking assetsinvestment premise distribution scheme under the expected determine rate of return,or to maximize revenue in the case of identified risks. Preliminary research portfoliofocused on portfolio income measure, Then the transition to portfolio risk measure.At first describes the portfolio theory, analysis of the advantages anddisadvantages of the nature existing portfolio model M-V, MAD, VaR andCvaR,Points out the demand for a multi-stage portfolio model in the investmentprocess.In connection with investment portfolio for a single period, proposed ainvestment portfolio model based on CVaR model. Don’t make normal distributionhypothesis on profit of portfolio, using CVaR model to measure portfolio risk,MADmodel as a limit function to realize the measurement limit of volatility, make use ofconvex utility function to express transaction costs of risk assets as a constraintcondition. Experimental results show that the model satisfy actual investmentrequirements, conform to the actual investment law, compare with M-V model andoriginal CVaR model, it has advantage of volatility and minimum risk value.For multi-stage investment portfolio, a portfolio model based on the model isproposed. Using of discrete dynamic programming to make the investment indifferent stages of the process integration as a whole, make normal distributionhypothesis on profit of portfolio, put average deviation model into the objectivefunction with Lagrange multiplier method to realize the measurement limitofvolatility. Experimental results show that this model do better in allocation offunds and the completion of a multi-stage investment opportunities grasp portfolioinvestment, it is a more practical model.