Simulating small-sample properties of the maximum likelihood cointegration method: estimation and testing

Abstract

This paper analyzes - using Monte Carlo simulation - small-sample properties of the maximum likelihood cointegration method for estimation and inference in cointegrated systems. The simulations of a bivariate system concentrate on the following; the estimator of the cointegrating vector; the trace test for determining cointegrating rank, and the likelihood ratio and Wald tests for linear restrictions on the cointegrating vector: Furthermore, we introduce autoregressive conditional heteroscedasticity, as well as multivariate non-normality in the form of excess skewness and kurtosis, in the error process. All in all, the results suggest that the maximum likelihood method displays desirable features as long as the samples are of reasonable sizes.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 18 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.