The aim of this letter is to formulate the statistically stationary DIA problem, which requires a careful choice
of the appropriate (i.e. physically meaningful) boundary conditions. It turns out that the resulting equations
are obviously simpler than the corrispoiuling nonstationary ones, in the perspective of a numerical solution.
The fluctuation-dissipation ansatz (ID) and the markovian closures arc also discussed in the stationary DIA.