1: \begin{abstract}

2: We consider a class of vector autoregressive models with

3: banded coefficient matrices. The setting represents a type of sparse

4: structure for high-dimensional time series, though the implied autocovariance

5: matrices are not banded. The structure is also practically

6: meaningful when the order of component time series is arranged

7: appropriately. The convergence rates for the estimated banded autoregressive

8: coefficient matrices are established.  We also propose

9: a Bayesian information criterion for determining

10: the width of the bands in

11: the coefficient matrices, which is proved to be consistent.

12: By exploring some approximate banded structure for the auto-covariance

13: functions of banded vector autoregressive processes, consistent estimators for the

14: auto-covariance matrices are constructed.

15: \end{abstract}