Authors:
Bunting Boruku Paibi, Isaac Didi Essi, John Barinaadaa Nwikpe, Zorle Dum Deebom

Addresses:
Department of Mathematics, Rivers State University, Port Harcourt, Rivers State, Nigeria. Department of Mathematics, Ignatius University of Education, Port Harcourt, Rivers State, Nigeria.

Abstract:

This study undertakes a comprehensive Monte Carlo comparative simulation to evaluate the finite-sample performance of six leading panel cointegration tests, Kao, Pedroni, Hadri, Hoang, Westerlund, and Johansen, to identify the most reliable and robust test across varying data environments. To test the null hypothesis of no cointegration and the alternative of cointegration, the simulation framework adjusts the number of cross-sectional units (N), time dimensions (T), and correlation values (ρ). Compared to Hadri (0.620) and Johansen (0.693), Kao (0.681) and Pedroni (0.715) have moderately higher power on tiny panels (e.g., N=10, T=10). However, the association worsens the problem of size misconceptions. Without cointegration (N=20, T=30, ρ=0.6), Pedroni and Hadri find rejection rates of 0.182 and 0.221, respectively, exceeding the nominal level. Kao shows lesser distortion (0.055). Panel size greatly increases test power. Kao, Hoang, and Westerlund achieved near-perfect power (>0.95) in large panels (N=100, T=500) with controllable size (<0.07), demonstrating asymptotic efficiency (N=20, T=30, ρ=0.0, cointegration). Hadri oversizes in linked panels, while Johansen can handle huge samples but not small ones. The analysis indicated that size control and high power in various scenarios make Kao and Westerlund the most balanced and reliable panel cointegration testing. Pedroni and Hoang use small samples but overstate the significance of their links. Kao and Westerlund excel at analyzing heterogeneous or cross-sectional data. Hadri and Johansen advise caution based on sample structure. 

Keywords: Cross-Sectional; Panel Cointegration Tests; Finite-Sample Performance; Size Distortion; Test Power; Asymptotic Efficiency; Extremely Large Panels; Smaller Configurations.

Received on: 02/08/2024, Revised on: 21/10/2024, Accepted on: 09/03/2025, Published on: 07/05/2026

DOI: 10.69888/FTSASS.2026.000645

FMDB Transactions on Sustainable Applied Sciences, 2026 Vol. 3 No. 1, Pages: 1-14

• Views : 31
• Downloads : 9

Download PDF