This paper examines the implications of using VARs in levels under the Max Share identification approach when variables exhibit unit or near-unit roots. We derive the asymptotic distributions of the Max Share estimator, demonstrating that it converges to a random matrix, resulting in...

This paper examines the implications of using VARs in levels under the Max Share identification approach when variables exhibit unit or near-unit roots. We derive the asymptotic distributions of the Max Share estimator, demonstrating that it converges to a random matrix, resulting in...