A panel data model with interactive effects characterized by multilevel non-parallel factors

ABSTRACT

Current research only considers parallel factors in factor models. In this article, we provide an algorithm based on cross principal component analysis that identifies and estimates a panel data model with interactive effects characterized by multilevel and non-parallel factors. The simulation results show that our estimator is consistent, converges quickly and outperforms other estimators that identify the factor structure incorrectly.

KEYWORDS:

• Panel data
• multilevel factors
• non-parallel factors
• cross PCA

JEL CLASSIFICATION:

• C13
• C33
• C51

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This article is supported by NSFC [Grant No. 71271096].