I am trying to replicate a study applied to my own data.
This is the case:
I have a portfolio which contains a risky momentum (MOM) asset and a risk free asset. According to Brandt et al (2007) one of the steps that must be done is the simulation of the path of the risky asset.
In the case of Brandt et al(2007) the return of the Value weighted CRSP index is used. Since dividends have to be accounted for, a Var model must be used.
Assuming that I need a VAR model as well, I modelled the following:
Mom(t+1) = Average excess return momentum + epsilon(in this case two random numbers [0.006,-0.0051])
I tried the following code but this does not give me the desired results:
%Specify initial parameters M = 10000; % Number of simulations T = 20; % Time horizon gamma = 5; % Risk aversion r_f = (1.06)^0,25; % Risk-free rate
%% epsilon = zeros(M,T,1) r= zeros(M,T);
for m=1:M epsilon(m,:,:) = mvnrnd([0;0],[0.006,-0.0051],T); end
for m=1:M r(m,1)=0,027 + epsilon(m,i+1,1); end
Next to that, I doubt if I need a VAR model in this case, but what should be the alternative? Next to that, any suggestions how to modify this code? It yields me errors of dimension inequalities...