Panel Cointegration Tests with Deterministic Trends and Structural Breaks
Author
Summary, in English
This paper proposes Lagrange multiplier (LM) based tests for the null hypothesis of no cointegration in panel data. The tests are general enough to allow for heteroskedastic and serially correlated errors, individual specific time trends, and a single structural break in both the intercept and slope of each regression, which may be located different dates for different individuals. The limiting distributions of the test statistics are derived, and are found to be standard normal and free of nuisance parameters under the null. In particular, the distributions are found to be invariant not only with respect to trend and structural break, but also with respect to the presence of stochastic regressors. A small Monte Carlo study is also conducted to investigate the small-sample properties of the tests. The results reveal that the tests have small size distortions and good power even in very small samples.
Department/s
Publishing year
2005
Language
English
Publication/Series
Working Papers, Department of Economics, Lund University
Issue
42
Full text
- Available as PDF - 230 kB
- Download statistics
Links
Document type
Working paper
Publisher
Department of Economics, Lund University
Topic
- Economics
Keywords
- Panel Cointegration
- Residual-Based Cointegration
- Structural Break
- Deterministic Trend
- LM Principle
Status
Published