Impact of Exogeneity and Measurement Errors on Vector Error Correction Model

Abstract

In this thesis we consider two different topics of vector error correction model in cointegration analysis. In chapter 3, we consider the impact of the exogeneity on the vector error correction model. We develop a VECM when vector processes contain exogenous processes. Our model generalize the condition of Hunter (1990)s cointegrating exogeneity, weakly exogeneity of Johansen (1992), Harbo et al. (1998), and Pesaran et al. (2000), and the non-causality of Mosconi and Gianini(1992) and Pradel and Rault (2003). We also develop least square estimator and maximum likelihood estimator of vector error correction model with exogenous processes. Asymptotic properties of estimators are derived and finite sample properties of the estimators are examined through a Monte Carlo simulation. In chapter 4, we investigate the impact of the measurement errors to vector error correction model with general cointegration rank. We study the impact of measurement errors to estimators and likelihood ratio test statistic for cointegration. This is the extension of Hassler and Kuzin (2009)s study which derived asymptotic distribution of LR test statistic for the null hypothesis of no cointegration. We also investigate the impact of measurement errors on the test through a Monte Carlo simulation study.