Answers are grounded in this paper and cite their sources.
The paper proposes a novel way to test a unit root for the process with structural breaks. Because estimating the structural break points is challenging, researchers often approximate the break process using a Fourier function. However, abrupt or short-lived changes in the intercept cannot be accurately captured by a Fourier function, which tends to overfit when data contain perturbations or noise. To improve the performance of unit root testing, we adapt the modified Sargan–Bhargava (MSB) statistic, originally proposed for panel data with structural breaks, and derive its asymptotic distribution under the null hypothesis. Its limit distribution is a Brownian motion function that does not depend on any parameters. Through Monte Carlo simulation, the critical values of the test statistic are given and the test results are investigated without estimating the break points. The results show that the MSB test statistics are better and more effective than the conventional EL and FF test statistics based on Fourier approximation. In the empirical analysis, the new test is applied to relative commodity prices, yielding a more suitable conclusion.
No takes yet. Share an insight, caveat, or question.
Junjuan Hu (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: