Cointegration in large VARs

TL;DR

This paper develops a new eigenvalue-based test for cointegration in large VARs where both the number of variables and time periods grow proportionally, improving over existing methods by reducing over-rejection.

Contribution

It introduces a novel asymptotic framework and test for cointegration in large VARs, connecting it to multivariate analysis of variance (MANOVA).

Findings

The new test does not suffer from over-rejection unlike the original Johansen test.

Theoretical asymptotics are supported by Monte Carlo simulations.

Empirical application demonstrates the test's practical usefulness.

Abstract

The paper analyses cointegration in vector autoregressive processes (VARs) for the cases when both the number of coordinates, $N$, and the number of time periods, $T$, are large and of the same order. We propose a way to examine a VAR of order $1$ for the presence of cointegration based on a modification of the Johansen likelihood ratio test. The advantage of our procedure over the original Johansen test and its finite sample corrections is that our test does not suffer from over-rejection. This is achieved through novel asymptotic theorems for eigenvalues of matrices in the test statistic in the regime of proportionally growing $N$ and $T$. Our theoretical findings are supported by Monte Carlo simulations and an empirical illustration. Moreover, we find a surprising connection with multivariate analysis of variance (MANOVA) and explain why it emerges.