[R] link between arima and arma fit

[STOLIAROFF VINCENT]

Hi dear sirs,
I am wondering why the fit of the time serie x with an arima and the fit of
diff(x) with an arma (same coeff p & d) differ one from another
here are the output of R:
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
> modelarma<-arma(diff(x),c(7,5))
> modelarma
Call:
arma(x = diff(x), order = c(7, 5))
Coefficient(s):
ar1 ar2 ar3 ar4 ar5 ar6 ar7 ma1 ma2 
0.06078 -0.44774 0.41881 0.47624 0.01406 0.06565 -0.06167 -0.01294 0.31313 
ma3 ma4 ma5 intercept 
-0.49027 -0.55461 -0.11520 -0.10692 
> modelarima<-arima(x,c(7,1,5))
> modelarima
Call:
arima(x = x, order = c(7, 1, 5))
Coefficients:
ar1 ar2 ar3 ar4 ar5 ar6 ar7 ma1 ma2 ma3 ma4
0.1244 -0.1149 -0.2283 0.5209 0.4135 -0.0072 0.0263 -0.087 0.0038 0.1868
-0.5477
s.e. NaN NaN 0.0696 0.1554 NaN NaN NaN NaN NaN 0.1079 0.1369
ma5
-0.5179
s.e. NaN
sigma^2 estimated as 149417: log likelihood = -129578.0, aic = 259181.9
Warning message: 
NaNs produced in: sqrt(diag(x$var.coef)) 
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
What method between the two should be considered the best one?
Thanks for your help
Vincent



*************************************************************************
Ce message et toutes les pieces jointes (ci-apres le "message") sont
confidentiels et etablis a l'intention exclusive de ses destinataires.
Toute utilisation ou diffusion non autorisee est interdite. 
Tout message electronique est susceptible d'alteration. 
SG Asset Management et ses filiales declinent toute responsabilite au titre
de ce message s'il a ete altere, deforme ou falsifie.

Découvrez l'offre et les services de SG Asset Management sur le site
www.sgam.fr 

				********

This message and any attachments (the "message") are confide...{{dropped}}