jensen's alpha

[Eltun]

Dear All,

I am going to measure the performance of funds based on Jensen's alpha.
Rpt - Rft = alpha + beta*(Rmt-Rft) + Et (Model)

a) Rpt minus Rft - is initially calculated and imported into Eviews as series and named "RpRf"
b) Rmt minus Rft is calculated and imported into Eviews as series and named "RmRf"

Regression in Eviews: RpRf c RmRf.
And the coefficient for RmRf (BEta) is different from the one calculated in Excel.
Coefficient for C(Constant - Alpha) is different from the calculated one in Excel.

Am I just asking, the way that I have introduced above is correct way to do?