How to loop the number of lags in a VAR?

I am fitting a simple var model. What I would like to do is basically choosing the optimal number of lags. I can do it in two ways: Likelihood test or AIC criteria. In either cases, I need to compute as many VAR as the number of lags p. I would like to avoid doing it one by one. I'd rather write a loop but I am failing at it.

This is an example:

% this is one VAR

dataset = rand(100,4)

model = varm(4, 2);
model.SeriesNames = {'I', 'U', 'O', 'W'};

model = estimate(model, dataset);
results = summarize(model)

% I tried this and variations of this but it doesn't work

for j = 2:10
    model = varm(4, j);
    model = estimate(model, DATASETC.SERIES(:,2:5));
results = summarize(model);
end

% I would like to have all the models stored in 'model' and 'results' 

Is there anyone who can help me with this?