1: \begin{abstract}

2: 	We consider high-dimensional multivariate linear regression models, where the joint distribution of covariates and response variables is a multivariate normal distribution with a bandable covariance matrix.

3: 	The main goal of this paper is to estimate the regression coefficient matrix, which is a function of the bandable covariance matrix. 

4: 	Although the tapering estimator of covariance has the minimax optimal convergence rate for the class of bandable covariances, 

5: 	we show that it has a sub-optimal convergence rate for the regression coefficient; that is, a minimax estimator for the class of bandable covariances may not be a minimax estimator for its functionals.

6: 	We propose the blockwise tapering estimator of the regression coefficient, which has the minimax optimal convergence rate for the regression coefficient under the bandable covariance assumption.

7: 	We also propose a Bayesian procedure called the blockwise tapering post-processed posterior of the regression coefficient and show that the proposed Bayesian procedure has the minimax optimal convergence rate for the regression coefficient under the bandable covariance assumption.

8: 	We show that the proposed methods outperform the existing methods via numerical studies.

9: 	

10: 	% Please include a maximum of seven keywords

11: %	\keywords{ Bandable covariance, Conditional mean, Covariance estimation, Minimax analysis, Post-processed posterior}

12: \end{abstract}

13: