1: \begin{abstract}

2: In this paper, we present the asymptotic distribution of M-estimators for parameters in unstable AR($p$) processes. The innovations are assumed to be in the domain of attraction of a symmetric stable law with index $0<\alpha\le2$. In particular, in the case of  repeated unit roots or conjugate complex unit roots,  M-estimators have a higher asymptotic rate of convergence compared to the least square estimators and the asymptotic results can be written as It\^{o} stochastic integrals.\\

3: 

4: \vspace{9pt}

5: 

6: \noindent{\it AMS classification}: 62M10, 60G52, 62F40.

7: 

8: \vspace{9pt}

9: 

10: 

11: \vspace{9pt}

12: \noindent\textsc{Keywords:}  Autoregressive model, Unit root, Stable process, Non-stationary, Bootstrapping.

13: \end{abstract}