Abstract
The Self-Exciting Threshold Autoregressive model (SETAR) is non-linear and considers threshold values to model time series affected by regimes. It is extended through the Multivariate SETAR (MSETAR) model, where the threshold variable can also be a multivariate process. The stationary marginal density (smd) of an MSETAR process of order one corresponds to a Unified Skew-Normal density. In this paper, the smd of an MSETAR of order one process was considered to compute explicit expressions of differential entropy and Kullback–Leibler and Jeffrey’s divergences between two MSETAR(1) processes. In addition, two asymptotic tests based on divergences were built for statistical significance testing of the disparity between MSETAR(1) processes and the threshold coefficient matrix. Information measures considered involved high-dimensional integrals that likewise depended on multivariate cumulative density normal function. To solve these integrals, Genz’s algorithm was considered based on Cholesky decomposition and Monte Carlo approximation. Some numerical experiments and applications to fish condition factor and Chilean economic perception time series illustrate performance.
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs00362-023-01457-6/MediaObjects/362_2023_1457_Fig1_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs00362-023-01457-6/MediaObjects/362_2023_1457_Fig2_HTML.png)
Similar content being viewed by others
Data Availability
Data used in Sect. 4.1 of this paper will be made available upon reasonable request from the corresponding author. Data used in Sect. 4.2 are publicly available from the github repository at https://github.com/percepcioneseconomicas/indices/find/main.
References
Abid SH, Hameed N (2014) The entropy of ARMA\((p, q)\) process. Am J Math Stat 4:12–20
Addo PM (2014) Multivariate self-exciting threshold autoregressive models with exogenous input. http://arxiv.org/abs/1407.7738
Andel J, Netuka I, Zvara K (1984) On threshold autoregressive processes. Kybernetika 20:89–106
Arellano-Valle RB, Azzalini A (2006) On the unification of families of skew-normal distributions. Scand J Stat 33:561–574
Arellano-Valle RB, Contreras-Reyes JE, Genton MG (2013) Shannon entropy and mutual information for multivariate skew-elliptical distributions. Scand J Stat 40:42–62
Arellano-Valle RB, Contreras-Reyes JE, Stehlík M (2017) Generalized skew-normal negentropy and its application to fish condition factor time series. Entropy 19:528
Arnold M, Günther R (2001) Adaptive parameter estimation in multivariate self-exciting threshold autoregressive models. Commun Stat Sim Comput 30:257–275
Aziz MA (2011) Study of unified multivariate skew normal distribution with applications in finance and actuarial science. Ph.D. dissertation thesis, Bowling Green State University, Ohio
Azzalini A, Dalla-Valle A (1996) The multivariate skew-normal distribution. Biometrika 83:715–726
Baragona R, Cucina D (2013) Multivariate self-exciting threshold autoregressive modeling by genetic algorithms. Jahrb Natl Okon Stat 233:3–21
Brockwell PJ, Davis RA (1991) Time series: theory and methods, 2nd edn. Springer-Verlag, New York
Cai TT, Zhang CH, Zhou HH (2010) Optimal rates of convergence for covariance matrix estimation. Ann Stat 38:2118–2144
Chan K, Tong H (1986) A note on certain integral equations associated with non-linear time series analysis. Prob Theor Relat Fields 73:153–158
Chávez D, Contreras-Reyes JE, Idrovo-Aguirre BJ (2023) A threshold GARCH model for Chilean economic uncertainty. J Risk Financ Manag 16:20
Contreras-Reyes JE (2014) Asymptotic form of the Kullback-Leibler divergence for multivariate asymmetric heavy-tailed distributions. Phys A 395:200–208
Contreras-Reyes JE (2015) Rényi entropy and complexity measure for skew-Gaussian distributions and related families. Phys A 433:84–91
Contreras-Reyes JE (2022) Rényi entropy and divergence for VARFIMA processes based on characteristic and impulse response functions. Chaos Solit Fract 160:112268
Contreras-Reyes JE (2023) Information quantity evaluation of nonlinear time series processes and applications. Physica D 445:133620
Cover TM, Thomas JA (2006) Elements of information theory. Wiley, New York
Das S (2021) Stationary and cyclostationary processes for time series and spatio-temporal data. Ph.D. dissertation thesis. King Abdullah University of Science and Technology, Thuwal
Das S, Genton MG (2020) On the stationary marginal distributions of subclasses of multivariate Setar processes of order one. J Time Ser Anal 41:406–420
Francq C, Raïssi H (2007) Multivariate portmanteau test for autoregressive models with uncorrelated but nonindependent errors. J Time Ser Anal 28:454–470
Genz A (1992) Numerical computation of multivariate normal probabilities. J Comput Graph Stat 1:141–149
Genz A (1993) Comparison of methods for the computation of multivariate normal probabilities. Comput Sci Stat 25:400–405
Gupta AK, Aziz MA, Ning W (2013) On some properties of the unified skew normal distribution. J Stat Theor Pract 7:480–495
Kharazmi O, Contreras-Reyes JE, Balakrishnan N (2023) Optimal information, Jensen-RIG function and \(\alpha \)-Onicescu’s correlation coefficient in terms of information generating functions. Phys A 609:128362
Lee S, Karagrigoriou A (2011) A divergence test for autoregressive time series models. Stat Methods 8:442–450
Loges W (2004) The stationary marginal distribution of a threshold AR(1) process. J Time Ser Anal 25:103–125
R Core Team (2022) A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna
Salicrú M, Menéndez ML, Pardo L, Morales D (1994) On the applications of divergence type measures in testing statistical hypothesis. J Multivar Anal 51:372–391
Tong H (1993) Non-linear time series: a dynamical system approach. Clarendon Press, Oxford, Oxford statistical science series
Troncoso N, Contreras-Reyes JE, Idrovo-Aguirre BJ (2023) Measuring economic uncertainty synchrony with cross-sample entropy under common external factors: the case of Chile. Fluct Noise Lett. https://doi.org/10.1142/S0219477523500463
Tsay RS (1998) Testing and modeling multivariate threshold models. J Am Stat Assoc 93:1188–1202
Wong SF, Tong H, Siu TK, Lu Z (2017) A new multivariate nonlinear time series model for portfolio risk measurement: the threshold copula-based TAR approach. J Time Ser Anal 38:243–265
Acknowledgements
Contreras-Reyes’s research was funded by FONDECYT (Chile) grant No. 11190116. Special thanks to Soumya Das for providing the R codes related to maximum likelihood estimation method. The author also thanks the editor and two anonymous referees for their helpful comments and suggestions. All R codes used in this paper are available upon request from the corresponding author.
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendix
Appendix
Proof of Proposition 2.2
The differential entropy (Cover and Thomas 2006) of a random vector \(\textbf{X}\) with pdf (2.1) is defined by
Based on (5.1), the logarithm is
Considering that differential entropy is invariant under translations of \(\textbf{x}\) (Cover and Thomas 2006), through Proposition 3 of Arellano-Valle et al. (2013) we get
where \(\textbf{Z}=\varvec{\Omega }^{-1/2}(\textbf{X}-\varvec{\xi })\sim N(\textbf{0},\textbf{I})\) and \(\textbf{X}_N\sim N(\textbf{0},\varvec{\Omega })\). Thus, \(H(\textbf{X}_N)=\frac{1}{2}\log \{(2\pi e)^{k_1}\textrm{det}(\varvec{\Omega })\}\) and the result is obtained. \(\square \)
Proof of Proposition 2.3
Consider the pdf of \(\textbf{X}_i\) given by (2.3), \(f_{\textbf{X}_i}(\textbf{x}_i|\textbf{0},\textbf{P}_i,-\textbf{B}_i,\textbf{0},\textbf{L}_i^{-1})\), \(i=1,2\). The KL divergence between \(\textbf{X}_1\) and \(\textbf{X}_2\) is
The second term of the last equation is the so-called cross-entropy and depends on the logarithm of the pdf of \(\textbf{X}_2\) due to (2.3),
and
Given that mean vector of \(\textbf{X}_i\) is zero, by part (ii) of Lemma 2.1 the expected value of the quadratic form of the latter equation is \(\mathbb {E}[\textbf{X}_1^{\top }\textbf{P}_2^{-1}\textbf{X}_1]=\textrm{tr}(\textbf{P}_2^{-1}\textbf{P}_1)\) and
\(\square \)
Proof of Proposition 3.1
Considering the proof of Lemma 6 of Cai et al. (2010), it is proved that the KL divergence based on covariance matrix estimates \(\textbf{P}_1\) and \(\textbf{P}_2\) has the following upper bound:
where \(c_1\) is a positive constant, \(a=n^{1/(2\gamma +1)}\) and \(b=a^{-(\gamma +1)}\). \(\gamma \) is a smoothness parameter whose optimal value found by Cai et al. (2010) was \(\gamma =1/2\). Thus,
Considering Corollary 2.4, the J divergence (3.8) can be expressed in terms of KL divergences of two multivariate normal random variables given by (see also Arellano-Valle et al. 2017). Therefore,
where the triangular inequality was used considering that KL divergence is non-negative. Using (3.4) and (5.3) and applying the expected value to both sides of (5.4), we get
where \(c_2\) and \(c_3\) are two positive constants. Thus, the result is obtained. \(\square \)
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Contreras-Reyes, J.E. Information quantity evaluation of multivariate SETAR processes of order one and applications. Stat Papers 65, 1553–1573 (2024). https://doi.org/10.1007/s00362-023-01457-6
Received:
Revised:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00362-023-01457-6