3楼: Originally posted by kainan001 at 2016-10-05 12:50:09
我之前用过蒙特卡洛方法去除数据的异常值，你这里的不确定性是指异常值吗？

     假设有五种资产，其日收益率(%)分别为
      0.0246 0.0189 0.0273 0.0141 0.0311
      标准差分别为
      0.9509 1.4259, 1.5227, 1.1062, 1.0877
      相关系数矩阵为
                1.0000 0.4403 0.4735 0.4334 0.6855
                0.4403 1.0000 0.7597 0.7809 0.4343
                0.4735 0.7597 1.0000 0.6978 0.4926
                0.4334 0.7809 0.6978 1.0000 0.4289
                0.6855 0.4343 0.4926 0.4289 1.0000
      假设初始价格都为100，模拟天数为504天，模拟线程为2，程序如下
%run.m
ExpReturn = [0.0246 0.0189 0.0273 0.0141 0.0311]/100; %期望收益
Sigmas = [0.9509 1.4259, 1.5227, 1.1062, 1.0877]/100;%标准差
Correlations = [1.0000 0.4403 0.4735 0.4334 0.6855
                0.4403 1.0000 0.7597 0.7809 0.4343
                0.4735 0.7597 1.0000 0.6978 0.4926
                0.4334 0.7809 0.6978 1.0000 0.4289
                0.6855 0.4343 0.4926 0.4289 1.0000
];%相关系数
ExpCov = corr2cov(Sigmas, Correlations);%协方差
StartPrice = 100;%初始价格
NumObs = 504;
NumSim = 2;
RetIntervals = 1;
NumAssets = 5;
%开始模拟
randn('state', 0);
RetExact = portsim(ExpReturn, ExpCov, NumObs, RetIntervals, NumSim);
Weights = ones(NumAssets, 1)/ NumAssets;
PortRetExact = zeros(NumObs, NumSim);
for i = 1:NumSim
    PortRetExact(:, i) = RetExact(:,:,i)*Weights;
end
PortExact = ret2tick(PortRetExact, repmat(StartPrice, 1, NumSim));
plot(PortExact, '-r');