Poisson INGARCH models: Linear Bayesian estimation and causality testing

Zhongxiu Ma; Zhanshou Chen; Peiyan Qi; Zhongxiu Ma; Zhanshou Chen; Peiyan Qi

doi:10.3934/math.2026913

AIMS Mathematics

2026, Volume 11, Issue 7: 22598-22622. doi: 10.3934/math.2026913

Previous Article Next Article

Research article Special Issues

Poisson INGARCH models: Linear Bayesian estimation and causality testing

1.
School of Mathematics and Statistics, Qinghai Normal University, Xining, 810008, China
2.
The State Key Laboratory of Tibetan Intelligence, Xining, 810008, China
3.
Department of Applied Mathematics, Taiyuan University of Science and Technology, Taiyuan, 030024, China

Received: 29 April 2026 Revised: 05 July 2026 Accepted: 13 July 2026 Published: 27 July 2026
MSC : 62M10, 62F15, 62E20

We develop a linear Bayesian estimator (LBE) for the Poisson integer-valued generalized autoregressive conditional heteroscedastic (INGARCH) model that fuses prior information with sample data, yielding a closed-form solution and thereby bypassing posterior integration. Under the mean squared error matrix (MSEM) criterion, we rigorously establish the theoretical superiority of the proposed estimator over conditional maximum likelihood estimation (CMLE). Exploiting the conditional mean structure of the Poisson INGARCH model, we construct a regression framework for testing the causal effects of exogenous covariates. We derive a least-squares test statistic and prove that it asymptotically follows a chi-square distribution under the null hypothesis of no Granger causality. Simulation studies with prior robustness checks demonstrate that the proposed estimator remains predominantly data-driven and significantly outperforms CMLE in small-to-moderate samples. Applying the method to Chicago crime data, we find that it provides crucial numerical stability when the likelihood geometry is unfavorable. Importantly, we underscore that beyond stabilizing computation, routine distributional diagnostics (e.g., negative binomial (NB) checks) are indispensable for addressing potential overdispersion and ensuring reliable causal conclusions.
- count time-series,
- Poisson INGARCH model,
- linear Bayes estimation,
- Granger causality
Citation: Zhongxiu Ma, Zhanshou Chen, Peiyan Qi. Poisson INGARCH models: Linear Bayesian estimation and causality testing[J]. AIMS Mathematics, 2026, 11(7): 22598-22622. doi: 10.3934/math.2026913

Related Papers:

Abstract

We develop a linear Bayesian estimator (LBE) for the Poisson integer-valued generalized autoregressive conditional heteroscedastic (INGARCH) model that fuses prior information with sample data, yielding a closed-form solution and thereby bypassing posterior integration. Under the mean squared error matrix (MSEM) criterion, we rigorously establish the theoretical superiority of the proposed estimator over conditional maximum likelihood estimation (CMLE). Exploiting the conditional mean structure of the Poisson INGARCH model, we construct a regression framework for testing the causal effects of exogenous covariates. We derive a least-squares test statistic and prove that it asymptotically follows a chi-square distribution under the null hypothesis of no Granger causality. Simulation studies with prior robustness checks demonstrate that the proposed estimator remains predominantly data-driven and significantly outperforms CMLE in small-to-moderate samples. Applying the method to Chicago crime data, we find that it provides crucial numerical stability when the likelihood geometry is unfavorable. Importantly, we underscore that beyond stabilizing computation, routine distributional diagnostics (e.g., negative binomial (NB) checks) are indispensable for addressing potential overdispersion and ensuring reliable causal conclusions.

References

[1]	D. Andrade, Robust variational Gaussian process regression for count data with the trimmed marginal likelihood, Stat. Comput., 36 (2026), 139. https://doi.org/10.1007/s11222-026-10895-9 doi: 10.1007/s11222-026-10895-9
[2]	E. Bacry, I. Mastromatteo, J. Muzy, Hawkes processes in finance, Market Microstruc. Liquidity, 1 (2015), 1550005. https://doi.org/10.1142/S2382626615500057 doi: 10.1142/S2382626615500057
[3]	Y. Bai, L. Wang, Bayes estimator for the inequality-constrained regression model, Commun. Stat.-Simul. C., 54 (2025), 3401–3414. https://doi.org/10.1080/03610918.2024.2349178 doi: 10.1080/03610918.2024.2349178
[4]	Y. Cui, F. Zhu, A new bivariate integer-valued GARCH model allowing for negative cross-correlation, Test, 27 (2018), 428–452. https://doi.org/10.1007/s11749-017-0552-4 doi: 10.1007/s11749-017-0552-4
[5]	T. Chen, Linear Bayes estimator of the extreme value distribution based on Type II censored samples, Beijing: Beijing Jiaotong University, 2021. https://doi.org/10.1080/03610918.2021.1963454
[6]	E. Cohn, Weather and crime, Brit. J. Criminol., 30 (1990), 51–64. https://doi.org/10.1093/oxfordjournals.bjc.a047980 doi: 10.1093/oxfordjournals.bjc.a047980
[7]	W. Chan, A correlated bivariate Poisson jump model for foreign exchange, Empir. Econ., 28 (2003), 669–685. https://doi.org/10.1007/s00181-003-0153-9 doi: 10.1007/s00181-003-0153-9
[8]	J. Corcoran, R. Zahnow, Weather and crime: A systematic review of the empirical literature, Crime Sci., 11 (2022), 16. https://doi.org/10.1186/s40163-022-00179-8 doi: 10.1186/s40163-022-00179-8
[9]	C. Chen, S. Lee, Bayesian causality test for integer-valued time series models with applications to climate and crime data, J. Roy. Stat. Soc. C-Appl., 66 (2017), 797–814. https://doi.org/10.1111/rssc.12200 doi: 10.1111/rssc.12200
[10]	J. Choi, Y. Ni, Model-based causal discovery for zero-inflated count data, J. Mach. Learn. Res., 24 (2023), 1–32.
[11]	Y. Cheng, L. Li, T. Xiao, Z. Li, J. Suo, K. He, et al., Cuts+: High-dimensional causal discovery from irregular time-series, In: Proceedings of the AAAI Conference on Artificial Intelligence, AAAI, 38 (2024), 11525–11533. https://doi.org/10.1609/aaai.v38i10.29034
[12]	R. Davis, K. Fokianos, S. Holan, H. Joe, J. Livsey, R. Lund, et al., Count time series: A methodological review, J. Am. Stat. Assoc., 116 (2021), 1533–1547. https://doi.org/10.1080/01621459.2021.1904957 doi: 10.1080/01621459.2021.1904957
[13]	R. Davis, H. Liu, Theory and inference for a class of nonlinear models with application to time series of counts, Stat. Sinica, 26 (2016), 1673–1707. https://www.jstor.org/stable/44114353
[14]	J. Davidson, Stochastic limit theory: An introduction for econometricians, Oxford: OUP Oxford, 1994.
[15]	B. Efron, C. Morris, Stein's estimation rule and its competitors–-An empirical Bayes approach, J. Am. Stat. Assoc., 68 (1973), 117–130. https://doi.org/10.1080/01621459.1973.10481350 doi: 10.1080/01621459.1973.10481350
[16]	R. Engle, J. Russell, Autoregressive conditional duration: A new model for irregularly spaced transaction data, Econometrica, 66 (1998), 1127–1162.
[17]	R. Ferland, A. Latour, D. Oraichi, Integer-valued GARCH process, J. Time Ser. Anal., 27 (2006), 923–942. https://doi.org/10.1111/j.1467-9892.2006.00496.x doi: 10.1111/j.1467-9892.2006.00496.x
[18]	S. Field, The effect of temperature on crime, Brit. J. Criminol., 32 (1992), 340–351. https://doi.org/10.1093/oxfordjournals.bjc.a048222 doi: 10.1093/oxfordjournals.bjc.a048222
[19]	K. Fokianos, A. Rahbek, D. Tjøstheim, Poisson autoregression, J. Amer. Stat. Assoc., 104 (2009), 1430–1439. https://doi.org/10.1198/jasa.2009.tm08270 doi: 10.1198/jasa.2009.tm08270
[20]	M. Goldstein, The linear Bayes regression estimator under weak prior assumptions, Biometrika, 67 (1980), 621–628. https://doi.org/10.1093/biomet/67.3.621 doi: 10.1093/biomet/67.3.621
[21]	M. Goldstein, General variance modifications for linear Bayes estimators, J. Am. Stat. Assoc., 78 (1983), 616–618. https://doi.org/10.1080/01621459.1983.10478019 doi: 10.1080/01621459.1983.10478019
[22]	H. Gao, Z. Chen, F. Li, Linear Bayesian estimation of misrecorded Poisson distribution, Entropy, 26 (2024), 62. https://doi.org/10.3390/e26010062 doi: 10.3390/e26010062
[23]	C. Granger, Investigating causal relations by econometric models and cross-spectral methods, Econometrica, 37 (1969), 424–438.
[24]	J. Hartigan, Linear Bayesian methods, J. Roy. Stat. Soc. B, 31 (1969), 446–454. https://doi.org/10.1111/j.2517-6161.1969.tb00804.x doi: 10.1111/j.2517-6161.1969.tb00804.x
[25]	J. Higgins, C. Tsokos, A study of the effect of the loss function on Bayes estimates of failure intensity, MTBF, and reliability, Appl. Math. Comput., 6 (1980), 145–166. https://doi.org/10.1016/0096-3003(80)90039-9 doi: 10.1016/0096-3003(80)90039-9
[26]	B. Hansen, A. A. Pacheco, M. Russo, R. De Vito, Fast variational inference for Bayesian factor analysis in single and multi-study settings, J. Comput. Graph. Stat., 34 (2025), 96–108. https://doi.org/10.1080/10618600.2024.2356173 doi: 10.1080/10618600.2024.2356173
[27]	J. Kang, S. Lee, Parameter change test for Poisson autoregressive models, Scand. J. Stat., 41 (2014), 1136–1152. https://doi.org/10.1111/sjos.12088 doi: 10.1111/sjos.12088
[28]	M. Liu, F. Zhu, J. Li, C. Sun, A systematic review of INGARCH models for integer-valued time series, Entropy, 25 (2023), 922. https://doi.org/10.3390/e25060922 doi: 10.3390/e25060922
[29]	A. M. Novelo, C. Anderson, Climate change and psychology: Effects of rapid global warming on violence and aggression, Curr. Clim. Change Rep., 5 (2019), 36–46. https://doi.org/10.1007/s40641-019-00121-2 doi: 10.1007/s40641-019-00121-2
[30]	A. M. Novelo, C. Anderson, Climate change and human behavior: Impacts of a rapidly changing climate on human aggression and violence, Cambridge: Cambridge University Press, 2022. https://doi.org/10.1017/9781108953078
[31]	W. Neuhaus, Choice of statistics in linear Bayes estimation, Scand. Actuar. J., 1985 (1985), 1–26. https://doi.org/10.1080/03461238.1985.10413774 doi: 10.1080/03461238.1985.10413774
[32]	K. Papaspyropoulos, D. Kugiumtzis, On the validity of Granger causality for ecological count time series, Econometrics, 12 (2024), 13. https://doi.org/10.3390/econometrics12020013 doi: 10.3390/econometrics12020013
[33]	L. Qian, Q. Zuo, W. Liu, H. Zhu, MSDG: A multiscale dynamic graph neural network for inferring dynamic Granger causality in multivariate time series, Inform. Sci., 681 (2026), 123287. https://doi.org/10.1016/j.ins.2026.123287 doi: 10.1016/j.ins.2026.123287
[34]	E. Stein, Interpolation of linear operators, T. Am. Math. Soc., 83 (1956), 482–492. https://doi.org/10.1090/s0002-9947-1956-0082586-0 doi: 10.1090/s0002-9947-1956-0082586-0
[35]	Z. Su, F. Zhu, S. Liu, Local influence analysis in the softplus INGARCH model, Test, 33 (2024), 951–985. https://doi.org/10.1007/s11749-024-00930-0 doi: 10.1007/s11749-024-00930-0
[36]	G. Trenkler, L. Wei, The Bayes estimator in a misspecified linear regression model, Test, 5 (1996), 113–123. https://doi.org/10.1007/bf02562684 doi: 10.1007/bf02562684
[37]	P. Tsamtsakiri, D. Karlis, On Bayesian model selection for INGARCH models via trans-dimensional Markov chain Monte Carlo methods, Stat. Model., 23 (2023), 81–98. https://doi.org/10.1177/1471082x211034705 doi: 10.1177/1471082x211034705
[38]	C. Weiß, Stationary count time series models, WIREs Comput. Stat., 13 (2021), e1502. https://doi.org/10.1002/wics.1502 doi: 10.1002/wics.1502
[39]	C. Weiß, F. Zhu, A. Hoshiyar, Softplus INGARCH models, Stat. Sinica, 32 (2022), 1099–1120. https://doi.org/10.5705/ss.202020.0353 doi: 10.5705/ss.202020.0353
[40]	H. Xuan, L. Maestrini, F. Chen, C. Grazian, Stochastic variational inference for GARCH models, Stat. Comput., 34 (2024), 45. https://doi.org/10.1007/s11222-023-10356-7 doi: 10.1007/s11222-023-10356-7
[41]	F. Zhu, Modeling overdispersed or underdispersed count data with generalized Poisson integer-valued GARCH models, J. Math. Anal. Appl., 389 (2012a), 58–71. https://doi.org/10.1016/j.jmaa.2011.11.042 doi: 10.1016/j.jmaa.2011.11.042
[42]	F. Zhu, Zero-inflated Poisson and negative binomial integer-valued GARCH models, J. Stat. Plan. Infer., 142 (2012b), 826–839. https://doi.org/10.1016/j.jspi.2011.10.002 doi: 10.1016/j.jspi.2011.10.002

Reader Comments

Your name:*

Email:*
© 2026 the Author(s), licensee AIMS Press. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0)