Why can't the MSE of lasso() match the theoretical value?

Hi,

I have below MSE1 and MSE2, which I got 13.27 and 5.87. Why don't they match?

clear
%set seed
rng(123);

X=10*rand([30,5]);
y=10*rand([30,1]);

[B,fitinfo]=lasso(X,y,'lambda',0);

MSE1=norm(y-X*B-ones([30,1])*fitinfo.Intercept)
MSE2=fitinfo.MSE

Thanks!