Published
Optimal Detection of Bilinear Dependence in Short Panels of Regression Data
Detección óptima de dependencia bilineal en regresión con datos de panel cortos
DOI:
https://doi.org/10.15446/rce.v43n2.83044Keywords:
Bilinear process, local asymptotic normality, local asymptotic linearity, panel data, pseudo-Gaussian tests, rank tests (en)Datos de panel, Linealidad asintótica local, Normalidad asintótica local, Proceso bilineal, Prueba pseudo-gaussiana, Pruebas de rango (es)
Downloads
In this paper, we propose parametric and nonparametric locally and
asymptotically optimal tests for regression models with superdiagonal bilinear time series errors in short panel data (large n, small T). We establish a local asymptotic normality property– with respect to intercept μ, regression coefficient β, the scale parameter σ of the error, and the parameter b of panel superdiagonal bilinear model (which is the parameter of interest)– for a given density f1 of the error terms. Rank-based versions of optimal parametric tests are provided. This result, which allows, by Hájek’s representation theorem, the construction of locally asymptotically optimal rank-based tests for the null hypothesis b = 0 (absence of panel superdiagonal bilinear model). These tests –at specified innovation densities f1– are optimal (most stringent), but remain valid under any actual underlying density. From contiguity, we obtain the limiting distribution of our test statistics under the null and local sequences of alternatives. The asymptotic relative efficiencies, with respect to the pseudo-Gaussian parametric tests, are derived. A Monte Carlo study confirms the good performance of the proposed tests.
En este artículo, se proponen pruebas paramétricas y no paramétricas locales y asintóticamente óptimas para modelos de regresión con errores de series temporales bilineales superdiagonales en datos de panel cortos (n grande, T pequeño). Se establece una propiedad de normalidad asintótica local con respecto a la intercepción μ, el coeficiente de regresión β, el parámetro de escala σ del error y el parámetro b del modelo bilineal superdiagonal con datos de panel (que es el parámetro de interés) para
una densidad determinada f1 de los términos de error. Se proporcionan versiones basadas en rangos de pruebas paramétricas óptimas. Este resultado permite, por el teorema de representación de Hájek, la construcción de pruebas locales basadas en rangos asintóticamente óptimas para la hipótesis nula b = 0 (ausencia del modelo bilineal superdiagonal con datos de panel). Estas pruebas, en densidades de innovación especicadas f1, son
óptimas (más estrictas), pero siguen siendo válidas en cualquier densidad subyacente. A partir de la contigüidad, se obtiene la distribución limitante de las estadísticas de prueba, bajo la hipótesis nula y una secuencia de alternativas locales. Se deriva eficiencia relativa asintótica de las pruebas, con respecto a las pruebas paramétricas pseudo-Gaussianas. Un análisis basado en simulaciones de Monte Carlo confirma el buen desempeño de las
pruebas propuestas.
References
Akharif, A. & Hallin, M. (2003), Efficient detection of random coefficients in autoregressive models, Annals of Statistics 31, 675–704.
Allal, J. & El Melhaoui, S. (2006), Tests de rangs adaptatifs pour les modèles de régression linéaire avec erreurs arma, Annales des sciences mathématiques du Québec 30, 29–54.
Azzalini, A. & Capitanio, A. (2003), Distributions generated by perturbation of symmetry with emphasis on a multivariate skew t-distribution, Journal of the Royal Statistical Society, Series B 65, 367–389.
Baltagi, B. & Li, Q. (1995), Testing AR(1) against MA(1) disturbances in an error component model , Journal of Econometrics 68, 133–151.
Benghabrit, Y. & Hallin, M. (1992), Optimal rank-based tests against first- order superdiagonal bilinear dependence , Journal of Statistical Planning and Inference 32(1), 45–61.
Benghabrit, Y. & Hallin, M. (1996), Rank-based tests for autoregressive against bilinear serial dependence , Journal of Nonparametric Statistics 6, 253–272.
Bennala, N., Hallin, M. & Paindaveine, D. (2012), Pseudo-gaussian and rank- based optimal tests for random individual effects in large n small t panels , Journal of Econometrics 170(1), 50–67.
Cassart, D., Hallin, M. & Paindaveine, D. (2011), A class of optimal tests for symmetry based on local Edgeworth approximations , Bernoul li 17, 1063–1094.
Chernoff, H. & Savage, I. R. (1958), Asymptotic normality and efficiency of certain nonparametric tests , Annals of Mathematical Statistics 29, 972–994.
Dutta, H. (1999), Large sample tests for a regression model with autoregressive conditional heteroscedastic errors , Communications in StatisticsTheory and Methods A28, 105–117.
Elmezouar, Z. C., Kadi, A. M. & Gabr, M. M. (2012), Linear regression with bilinear time series errors , Panamerican Mathematical Journal 22(1), 1–13.
Fihri, M., Akharif, A., Mellouk, A. & Hallin, M. (2020), Efficient pseudo- gaussian and rank-based detection of random regression coefficients , Journal of Nonparametric Statistics 32(2), 367–402.
How to Cite
APA
ACM
ACS
ABNT
Chicago
Harvard
IEEE
MLA
Turabian
Vancouver
Download Citation
CrossRef Cited-by
1. Slimane Regui, Abdelhadi Akharif, Amal Mellouk. (2024). Locally optimal tests against periodic linear regression in short panels. Communications in Statistics - Simulation and Computation, , p.1. https://doi.org/10.1080/03610918.2024.2314662.
2. Yassine OU LARBİ, Rachid El HALİMİ, Abdelhadi AKHARİF, Amal MELLOUK. (2021). Optimal tests for random effects in linear mixed models. Hacettepe Journal of Mathematics and Statistics, 50(4), p.1185. https://doi.org/10.15672/hujms.773667.
3. A. Lmakri, A. Akharif, A. Mellouk. (2023). Estimation in short-panel data models with bilinear errors. Mathematical Modeling and Computing, 10(3), p.682. https://doi.org/10.23939/mmc2023.03.682.
Dimensions
PlumX
Article abstract page views
Downloads
License
Copyright (c) 2020 Revista Colombiana de Estadística
This work is licensed under a Creative Commons Attribution 4.0 International License.
- Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal.
- Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this journal.
- Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work (See The Effect of Open Access).