Why the "CointegrationConstant" property doesn't work in the vecm function?

I try to use a constant only in the cointegrating relations, which according to the documentation ( https://www.mathworks.com/help/econ/vecm.html ) should work.

When I try to simulate time-series with the VECM model it has no effect.

I have studied the code in more detail and I haven't find any line in relation with CointegrationConstant property except the validation.

% Parameters
Adjustment = [-0.4; 0];
Cointegration = [1; 0;];
AR = 0.5;
ShortRun = {[AR 0; 0 AR]};
Constant = [0; 0];
Trend = [0; 0];
Covariance = [1 0; 0 1];
Mdl = vecm('Adjustment',Adjustment,'Cointegration',Cointegration,...
    'Constant',Constant,'ShortRun',ShortRun,'Trend',Trend,...
    'Covariance',Covariance);

% Simulation
numpaths = 1;
numobs = 50;

% Without a constant in the cointegrating relation
rng(123)
[Y1,E1] = simulate(Mdl,numobs,'NumPaths',numpaths);


% With a constant in the cointegrating relation
Mdl.CointegrationConstant = [5];
rng(123)
[Y2,E2] = simulate(Mdl,numobs,'NumPaths',numpaths);

% If CointegrationConstant works then the two time-series shouldn't
% be equal(?).
isequal = Y1 == Y2;

Answer 1

[nick]

Hi Mark,

I understand from your query that you want to know why setting the "Cointegration Constant" property in the "vecm" function still results in an identical time series in your code snippet.

According to the example "Specify All Parameter Values of VEC Model" in the documentation for the "vecm" function, the "Cointegration Constant" and "Linear Trend" are not required for simulating observations or forecasting, given that the overall constant and trend parameters are known. This is why you are seeing the same time series being simulated in your code snippet.

You may refer to the following documentation to learn more about “vecm” function : https://www.mathworks.com/help/econ/vecm.html

Hope this helps,

Regards,

Neelanshu