How to constrain this optimisation to be long-only?

6 次查看（过去 30 天）
Luciano Lilloy Fedele
编辑： Luciano Lilloy Fedele 2018-6-11
Hi everyone, I'm using a library that optimise a portfolio based in past information and my views on best stocks. However, I need the final weights to be long-only (i.e. positives). Any idea if this is possible? See below the code:
function Allocation = DecisionBayesParameters(Market,InvestorProfile,Prior)
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% estimate market parameters
Exp_LinRets_Hat=mean(Market.LinRetsSeries)';
Cov_LinRets_Hat=cov(Market.LinRetsSeries);
T=Market.T;
T_1=Prior.T_0+T;
Exp_LinRets_Post=(Prior.LinRets_EV*Prior.T_0+Exp_LinRets_Hat*T)/T_1;
Nu_1=Prior.Nu_0+T;
Cov_LinRets_Post=(Prior.LinRets_Cov*Prior.Nu_0+Cov_LinRets_Hat*T+(Exp_LinRets_Hat-Prior.LinRets_EV)*(Exp_LinRets_Hat-Prior.LinRets_EV)'*Prior.T_0*T/T_1)/Nu_1;
Exp_Prices_Hat=diag(Market.CurrentPrices)*(1+Exp_LinRets_Post);
Cov_Prices_Hat=diag(Market.CurrentPrices)*Cov_LinRets_Post*diag(Market.CurrentPrices);
% compute allocation
S=inv(Cov_Prices_Hat);
A_Hat=Market.CurrentPrices'*S*Market.CurrentPrices;
B_Hat=Market.CurrentPrices'*S*Exp_Prices_Hat;
Gamma = (InvestorProfile.Budget - InvestorProfile.RiskPropensity*B_Hat)/A_Hat;
Allocation = InvestorProfile.RiskPropensity*S*Exp_Prices_Hat + Gamma*S*Market.CurrentPrices;
This is the library: https://uk.mathworks.com/matlabcentral/fileexchange/9061-risk-and-asset-allocation

请先登录，再回答此问题。

回答（0 个）

类别

在 Help Center 和 File Exchange 中查找有关 Portfolio Optimization and Asset Allocation 的更多信息

标签

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by