gorras ed hardy Transaction costs do not take into

Transaction costs do not take into account a combination of investment model

Volume: model (1): minaz a 7'Vxmaxm an xs. t-1 above the constraints of the model that all investment funds. 3 model into the two-objective programming model for solving more complex, so we try to put two into a single objective goal programming planning to solve it. Investment in securities, in general, along with high-risk high-yield, low risk and only under conditions of low income can be achieved, so we generally difficult to find a maximum ratio of investment to make gains, while at the same time the risk to a minimum. In addition, because different investors attitude towards risk and return is not the same, in order to reflect this in the model investor preferences between individual behavior, we introduce the preference coefficient a ∈ Eo, 1], to model (1 ) are two objective functions in the weight given to O / and 1 a a, then the objective function can be transformed into: minS (z) one by one (1 a) 7'z + O/X7 'z where: 1 A O / and O / investors, respectively, to treat gains and risk aversion. O / larger that the more risk-averse investors, O / A 1 indicates that investors are extremely risk-averse, O / 10 that investors consider only the income but not with risk. Therefore, the above model to the double-objective programming model is transformed into the following single objective programming model: model (2): minS (z) one by one (1 a) 7'z + O/X7 'zs. t-14 to solve the model we can use Lagrange multiplier method to solve the model: For the O / ∈ Eo,[link widoczny dla zalogowanych], 1] Let f one by one (1 a) FTX + port z7'z a 2 (e ~ x-1) is for Lagrange multipliers. One by one (1 a) +2 aVz a 2e a o (1) A -1 A o (2) from (1), we have z A (3) Z -------------- --- A pair of (3) Multiplying both sides,[link widoczny dla zalogowanych], there are: [(1 a) +2 e] z, finishing a ------ a: 2a-(1 a) +2 e ~ Ve] Therefore generation into (3) derived model (2) the optimal solution [(1_ port) + z A -------------- eleven + one V-le. ] On the type of bias coefficient can be determined: if the investment is 3 Li Shiwei: do not consider the transaction cost by 229 investment model is an extremely risk-averse,[link widoczny dla zalogowanych], we take a ∈ Eo. 9,1]; if investors are risk averse, we can take a ∈ [0.7,0.9]; if investors are risk neutral, we can take a ∈ [0.45,0.55] ; If investors are more concerned about the risk of income and not care about people,[link widoczny dla zalogowanych], we can take a ∈ [0.2,0.4]; and if investors only care about income and almost do not care about risk, then we take oral ∈ [0 ,[link widoczny dla zalogowanych], 0.1]. In the optimal solution obtained, we can determine the rate of return on portfolio securities:. TV. 1 a ten [one by one IZPe ~ V a 1-PI___ # V-le. ] The corresponding risk for the :={+[ V-ll ~ a lin V-le. ]) {+ T a E a V-le,]) 5 Special case analysis for the two extreme cases, that is, a = 0 and a a 1, we described separately below. When a 10 (that corresponds to income investors only consider the case without considering the risks), model (2) into the following models: rains (z) = a / ATx solution at this time was that the risk of not considering case, the maximum benefit can be obtained for the analogy. When a-1 (ie, only considering the risk corresponds to the case of investors), model (2) into the following models: minS (z) a zVxS. teTx1 with the Lagrange multiplier method is very easy to find the solution of this model; V a P. Called a P-1e the corresponding minimum risk portfolio investment as a way, we have established a bi-objective optimization model for portfolio investment, and given by way of empowering the model solution under different conditions, This in reality has a certain value. 【References [1] WILLIAMFSHARPE. Portfoliotheoryandcapitalmar-kets [M]. Beijing: Mechanical Industry Press, 2001. [1] TANG Xiao me. Portfolio investment decision-making method of calculating population]. Engineering and Engineering Management, 1990 (3) :45-48. [1] TANG Xiao me. Prediction Theory and Its Application [M]. Chengdu: University of Electronic Science and Technology Press, 1992. [1] Xiao-group. And application of modern statistical analysis [M]. Beijing: China Renmin University Press, 1999.

相关的主题文章：

[link widoczny dla zalogowanych]

[link widoczny dla zalogowanych]

[link widoczny dla zalogowanych]

The post has been approved 0 times