Why do SAS and Matlab give me different answers for my OLS regression?

Hi,
i'm trying to run the same OLS regresion in SAS and Matlab but they gave me different answers. The regression is:

R_t = β_0 + B*F_t + β_L*L_t + e_t

Where R_t is a 10X1 vector , F_t is a 4x1 vector, B is a 10x4 matrix, and β_0 and β_L are 10x1 vector.

For SAS I use the next code:

proc reg data=datain outest= dataout tableout noprint;
by id;
model R = F_1 F_2 F_3 L/ hcc adjrsq;
run;

I did the same thing with Matlab (I think??)
with R (300x10 matrix) contains time-series data of 10 different id, and matrix
F = [ones(300,1) F_1 F_2 F_3 L];
betas = inv(transpose(F)*F)*transpose(F)*R;

Shouldn't they give me the same answers or did i do something wrong? Thanks