Curious issue generating truncated normals

I try two ways of generating truncated multivariate normals below, which give different results. I'd be interested if anyone could shed light on why this might be. In the code I generate bivariate normal draws with variance c*c', where c=[1 0;1 2], truncated above at (-2,-2).

In the first case, I use simple accept reject sampling. In the second, I sequentially draw truncated normals (this is explained here, for example: http://www.hss.caltech.edu/~mshum/gradio/ghk_desc.pdf).

The means are consistently different. The first variable has a mean of around -2.41 with accept reject and -2.37 with sequential sampling. This seems peculiar to me - any ideas?

u = [-2 -2];
trials = 10000;
c = [1 0;1 2];

i=0;
  AR = zeros(trials,2);
  while i < trials
    t = c*normrnd(0,1,2,1);
    if t' <= u
       i=i+1;
       AR(i,:) = t';
    end
end
mean(AR) 

GHK = zeros(trials,2);
for i=1:trials
    r1 = norminv(rand*normcdf(u(1)/c(1,1)));
    r2 = norminv(rand*normcdf((u(2)-c(2,1)*r1)/c(2,2)));
    GHK(i,:) = (c*[r1;r2])';
end
mean(GHK)