We decompose a stationary Markov process (X-t) as: X-t = a(0) + Sigma (infi
nity)(j=1) a(j)Z(j,t), where the Z(j)'s process admit ARMA specifications.
These decompositions are deduced from a nonlinear canonical decomposition o
f the joint distribution of (X-t, Xt-1). (C) 2001 Elsevier Science BN. All
rights reserved.