Limiting distributions of least squares estimates of unstable autoregressive processes (Q1116576)

An autoregressive process is said to be unstable if its characteristic polynomial has all its roots on or outside the unit circle. The paper presents the asymptotic distribution of the least squares estimator of the autoregressive parameters under general assumptions about the positions of the roots on the unit circle and the existence of the \((2+\delta) th\) moment of the innovation distribution. A feature of the paper is the derivation of some convergence results for random variables to stochastic integrals.