Why does gctest fail when using discontinuous lags?

I'm running a Granger-causality test on a varm model object. The gctest function returns an indexing error when the lags are discontinuous. For example it works when lags = 1,2,3,4. It fails when lags = 1,2,4. Below code recreates the error.

%Following Matlab gctest example

load Data_USEconModel

m1slrate = price2ret(DataTable.M1SL);

inflation = price2ret(DataTable.CPIAUCSL);

rgdprate = price2ret(DataTable.GDP./DataTable.GDPDEF);

tbl = table(m1slrate,inflation,rgdprate);

tbl = rmmissing(tbl);

T = size(tbl,1); % Total sample size

numseries = 3;

% Partition time base.

maxp = 4; % Maximum number of required presample responses

idxpre = 1:maxp;

idxest = (maxp + 1):T;

% Fit VAR models to data.

Y0 = tbl{idxpre,:}; % Presample

Y = tbl{idxest,:}; % Estimation sample

%Two different models. Second model omits lag number 3

lags1=[1:4];

lags2=[1,2,4];

coef0=nan(numseries,numseries);

numlags1=length(lags1);

numlags2=length(lags2);

for i=1:length(lags1)

coef_1{i}=coef0;

end

for i=1:length(lags2)

coef_2{i}=coef0;

end

Mdl_1 = varm('AR',coef_1,'Lags',lags1);

EstMdl_1 = estimate(Mdl_1,Y,'Y0',Y0);

gctest(EstMdl_1)

Mdl_2 = varm('AR',coef_2,'Lags',lags2);

EstMdl_2 = estimate(Mdl_2,Y,'Y0',Y0);

gctest(EstMdl_2) %gctest fails for the second model