This paper studies asymptotic likelihood inference on cointegration paramet
ers in systems integrated of order two. We start with so-called triangular
systems and then extend the analysis to vector autoregressions. We show tha
t even when all unit root restrictions have been imposed, the asymptotic ob
served information is not (locally) ancillary, which implies that the log-l
ikelihood ratio is not locally asymptotically mixed normal. The results are
applied to inference on polynomial cointegration. Some similarities and di
fferences with I(1) systems are also discussed.