{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,6]],"date-time":"2026-10-06T21:51:33Z","timestamp":1791323493460,"version":"4.3.1"},"reference-count":36,"publisher":"Institute of Mathematical Statistics","issue":"4","content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Ann. Statist."],"published-print":{"date-parts":[[2026,8,1]]},"abstract":"We consider a k-dimensional multiple-regime vector threshold autoregressive model, in which the regime-switching mechanism is governed by a bivariate threshold variable. Specifically, the regimes are induced by a partition of the threshold space by an unknown number of threshold lines. Within each regime, the process follows a specific vector autoregressive (VAR) model. We formulate the model selection and parameter estimation into a minimization problem based on the Minimum Description Length (MDL) principle and estimate the number of threshold lines, parametric forms of threshold lines and VAR model parameters in each regime simultaneously. Theoretically, we show that the MDL estimators of threshold lines are n-consistent and characterize their limiting distribution. This requires novel proving techniques of introducing a new functional space G for the local MDL difference functions and establishing weak convergence results therein. Finally, we conduct some empirical studies with simulated datasets and perform real data analyses on U.S. interest rates and U.S. GNP Data.<\/jats:p>","DOI":"10.1214\/26-aos2638","type":"journal-article","created":{"date-parts":[[2026,8,18]],"date-time":"2026-08-18T08:17:36Z","timestamp":1787041056000},"page":"2005-2030","source":"Crossref","is-referenced-by-count":0,"title":["Generalized multivariate threshold autoregressive models with linearly partitioned threshold space"],"prefix":"10.1214","volume":"54","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-4200-3811","authenticated-orcid":false,"given":"Gan","family":"Yuan","sequence":"first","affiliation":[{"name":"Department of Biostatistics, City University of Hong Kong"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Chun Yip","family":"Yau","sequence":"additional","affiliation":[{"name":"Department of Statistics and Data Science, The Chinese University of Hong Kong"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"108","reference":[{"key":"34","doi-asserted-by":"publisher","unstructured":"YU, P.<\/small> (2014). The bootstrap in threshold regression. Econometric Theory<\/i> 30<\/b> 676\u2013714.","DOI":"10.1017\/S0266466614000012"},{"key":"5","doi-asserted-by":"publisher","unstructured":"Chan, K. S.<\/small> (1993). Consistency and limiting distribution of the least squares estimator of a threshold autoregressive model. Ann. Statist.<\/i> 21<\/b> 520\u2013533.","DOI":"10.1214\/aos\/1176349040"},{"key":"25","doi-asserted-by":"publisher","unstructured":"Seijo, E.<\/small> and Sen, B.<\/small> (2011). Change-point in stochastic design regression and the bootstrap. Ann. Statist.<\/i> 39<\/b> 1580\u20131607.","DOI":"10.1214\/11-AOS874"},{"key":"3","doi-asserted-by":"publisher","unstructured":"Billingsley, P.<\/small> (1999). Convergence of Probability Measures<\/i>, 2nd ed. Wiley Series in Probability and Statistics<\/i>: Probability and Statistics<\/i>. Wiley, New York.","DOI":"10.1002\/9780470316962"},{"key":"1","unstructured":"ADDO, P. M.<\/small> (2014). Multivariate self-exciting threshold autoregressive models with exogenous input. Available at arXiv:1407.7738<\/a>."},{"key":"2","doi-asserted-by":"publisher","unstructured":"BAI, Q.<\/small> and LING, S.<\/small> (2024). On ergodicity of threshold ARMA<\/mi>(<\/mo>m<\/mi>,<\/mo>p<\/mi>,<\/mo>q<\/mi>)<\/mo><\/math> models. Jpn. J. Stat. Data Sci.<\/i> 7<\/b> 667\u2013675.","DOI":"10.1007\/s42081-024-00248-z"},{"key":"4","doi-asserted-by":"publisher","unstructured":"CHAN, K.-S.<\/small> and GORACCI, G.<\/small> (2019). On the ergodicity of first-order threshold autoregressive moving-average processes. J. Time Series Anal.<\/i> 40<\/b> 256\u2013264.","DOI":"10.1111\/jtsa.12440"},{"key":"6","doi-asserted-by":"publisher","unstructured":"CHAN, K. S.<\/small> and TSAY, R. S.<\/small> (1998). Limiting properties of the least squares estimator of a continuous threshold autoregressive model. Biometrika<\/i> 85<\/b> 413\u2013426.","DOI":"10.1093\/biomet\/85.2.413"},{"key":"7","doi-asserted-by":"publisher","unstructured":"CHEN, C. W. S.<\/small>, SO, M. K. P.<\/small> and LIU, F.-C.<\/small> (2011). A review of threshold time series models in finance. Stat. Interface<\/i> 4<\/b> 167\u2013181.","DOI":"10.4310\/SII.2011.v4.n2.a12"},{"key":"8","doi-asserted-by":"publisher","unstructured":"CHEN, H.<\/small>, CHONG, T. T.-L.<\/small> and BAI, J.<\/small> (2012). Theory and applications of TAR model with two threshold variables. Econometric Rev.<\/i> 31<\/b> 142\u2013170.","DOI":"10.1080\/07474938.2011.607100"},{"key":"9","doi-asserted-by":"publisher","unstructured":"DAVIS, R. A.<\/small>, LEE, T. C. M.<\/small> and RODRIGUEZ-YAM, G. A.<\/small> (2008). Break detection for a class of nonlinear time series models. J. Time Series Anal.<\/i> 29<\/b> 834\u2013867.","DOI":"10.1111\/j.1467-9892.2008.00585.x"},{"key":"10","doi-asserted-by":"publisher","unstructured":"DAVIS, R. A.<\/small> and YAU, C. Y.<\/small> (2013). Consistency of minimum description length model selection for piecewise stationary time series models. Electron. J. Stat.<\/i> 7<\/b> 381\u2013411.","DOI":"10.1214\/13-EJS769"},{"key":"11","unstructured":"EMILSSON, V.<\/small>, THORLEIFSSON, G.<\/small>, ZHANG, B.<\/small>, LEONARDSON, A. S.<\/small>, ZINK, F.<\/small>, ZHU, J.<\/small>, CARLSON, S.<\/small>, HELGASON, A.<\/small>, WALTERS, G. B.<\/small> et al. (2008). Genetics of gene expression and its effect on disease. Nature<\/i> 452<\/b> 423\u2013428."},{"key":"12","doi-asserted-by":"publisher","unstructured":"GIANNERINI, S.<\/small>, GORACCI, G.<\/small> and RAHBEK, A.<\/small> (2024). The validity of bootstrap testing for threshold autoregression. J. Econometrics<\/i> 239<\/b> Paper No. 105379.","DOI":"10.1016\/j.jeconom.2023.01.004"},{"key":"13","doi-asserted-by":"publisher","unstructured":"GIANNERINI, S.<\/small>, MAASOUMI, E.<\/small> and BEE DAGUM, E.<\/small> (2015). Entropy testing for nonlinear serial dependence in time series. Biometrika<\/i> 102<\/b> 661\u2013675.","DOI":"10.1093\/biomet\/asv007"},{"key":"14","doi-asserted-by":"publisher","unstructured":"GORACCI, G.<\/small> (2021). An empirical study on the parsimony and descriptive power of TARMA models. Stat. Methods Appl.<\/i> 30<\/b> 109\u2013137.","DOI":"10.1007\/s10260-020-00516-8"},{"key":"15","doi-asserted-by":"publisher","unstructured":"GORACCI, G.<\/small>, FERRARI, D.<\/small>, GIANNERINI, S.<\/small> and RAVAZZOLO, F.<\/small> (2025). Robust estimation for threshold autoregressive moving-average models. J. Bus. Econom. Statist.<\/i> 43<\/b> 579\u2013591.","DOI":"10.1080\/07350015.2024.2412011"},{"key":"16","doi-asserted-by":"publisher","unstructured":"HANSEN, B. E.<\/small> (2011). Threshold autoregression in economics [discussion of MR2812802]. Stat. Interface<\/i> 4<\/b> 123\u2013127.","DOI":"10.4310\/SII.2011.v4.n2.a4"},{"key":"17","doi-asserted-by":"publisher","unstructured":"HANSEN, M. H.<\/small> and YU, B.<\/small> (2001). Model selection and the principle of minimum description length. J. Amer. Statist. Assoc.<\/i> 96<\/b> 746\u2013774.","DOI":"10.1198\/016214501753168398"},{"key":"18","doi-asserted-by":"crossref","unstructured":"LEE, T. C. M.<\/small> (2001). An introduction to coding theory and the two-part minimum description length principle. Int. Stat. Rev.<\/i> 69<\/b> 169\u2013183.","DOI":"10.1111\/j.1751-5823.2001.tb00455.x"},{"key":"19","doi-asserted-by":"crossref","unstructured":"LEWIS, P. A. W.<\/small> and STEVENS, J. G.<\/small> (1991). Nonlinear modeling of time series using multivariate adaptive regression splines. J. Amer. Statist. Assoc.<\/i> 86<\/b> 864\u2013877.","DOI":"10.1080\/01621459.1991.10475126"},{"key":"20","doi-asserted-by":"publisher","unstructured":"LI, D.<\/small>, LI, W. K.<\/small> and LING, S.<\/small> (2011). On the least squares estimation of threshold autoregressive and moving-average models. Stat. Interface<\/i> 4<\/b> 183\u2013196.","DOI":"10.4310\/SII.2011.v4.n2.a13"},{"key":"22","doi-asserted-by":"publisher","unstructured":"NEUHAUS, G.<\/small> (1971). On weak convergence of stochastic processes with multidimensional time parameter. Ann. Math. Statist.<\/i> 42<\/b> 1285\u20131295.","DOI":"10.1214\/aoms\/1177693241"},{"key":"23","doi-asserted-by":"publisher","unstructured":"QIAN, L.<\/small> (1998). On maximum likelihood estimators for a threshold autoregression. J. Statist. Plann. Inference<\/i> 75<\/b> 21\u201346.","DOI":"10.1016\/S0378-3758(98)00113-X"},{"key":"24","unstructured":"RISSANEN, J.<\/small> (1989). Stochastic Complexity in Statistical Inquiry<\/i>. World Scientific Series in Computer Science<\/i> 15<\/b>. World Scientific Co., Inc., Teaneck, NJ."},{"key":"26","doi-asserted-by":"crossref","unstructured":"TIAO, G. C.<\/small> and TSAY, R. S.<\/small> (1994). Some advances in non-linear and adaptive modelling in time-series. J. Forecast.<\/i> 13<\/b> 109\u2013131.","DOI":"10.1002\/for.3980130206"},{"key":"27","doi-asserted-by":"crossref","unstructured":"TONG, H.<\/small> (1978). On a threshold model. Pattern Recogn. Signal Process.<\/i> 29<\/b> 575\u2013586.","DOI":"10.1007\/978-94-009-9941-1_24"},{"key":"29","doi-asserted-by":"publisher","unstructured":"TONG, H.<\/small> (2015). Threshold models in time series analysis\u2014some reflections. J. Econometrics<\/i> 189<\/b> 485\u2013491.","DOI":"10.1016\/j.jeconom.2015.03.039"},{"key":"30","doi-asserted-by":"crossref","unstructured":"TONG, H.<\/small> and LIM, K. S.<\/small> (1980). Threshold autoregression, limit cycles and cyclical data. J. R. Stat. Soc. Ser. B. Stat. Methodol.<\/i> 42<\/b> 245\u2013292.","DOI":"10.1111\/j.2517-6161.1980.tb01126.x"},{"key":"31","doi-asserted-by":"publisher","unstructured":"TSAY, R. S.<\/small> (1998). Testing and modeling multivariate threshold models. J. Amer. Statist. Assoc.<\/i> 93<\/b> 1188\u20131202.","DOI":"10.2307\/2669861"},{"key":"32","doi-asserted-by":"publisher","unstructured":"TSAY, R. S.<\/small> (2021). Multivariate hysteretic autoregressive models. Statist. Sinica<\/i> 31<\/b> 2257\u20132274.","DOI":"10.1002\/rnc.5392"},{"key":"33","doi-asserted-by":"publisher","unstructured":"YAU, C. Y.<\/small>, TANG, C. M.<\/small> and LEE, T. C. M.<\/small> (2015). Estimation of multiple-regime threshold autoregressive models with structural breaks. J. Amer. Statist. Assoc.<\/i> 110<\/b> 1175\u20131186.","DOI":"10.1080\/01621459.2014.954706"},{"key":"35","doi-asserted-by":"crossref","unstructured":"YUAN, G.<\/small> and YAU, C. Y.<\/small> (2026). Supplement to \u201cGeneralized multivariate threshold autoregressive models with linearly partitioned threshold space.\u201d https:\/\/doi.org\/10.1214\/26-AOS2638SUPP<\/a>","DOI":"10.1214\/26-AOS2638SUPP"},{"key":"36","doi-asserted-by":"publisher","unstructured":"ZHANG, X.<\/small>, LI, D.<\/small> and TONG, H.<\/small> (2024). On the least squares estimation of multiple-threshold-variable autoregressive models. J. Bus. Econom. Statist.<\/i> 42<\/b> 215\u2013228.","DOI":"10.1080\/07350015.2023.2174124"},{"key":"28","doi-asserted-by":"publisher","unstructured":"TONG, H.<\/small> (1983). Threshold Models in Nonlinear Time Series Analysis<\/i>. Lecture Notes in Statistics<\/i> 21<\/b>. Springer, New York.","DOI":"10.1007\/978-1-4684-7888-4"},{"key":"21","doi-asserted-by":"publisher","unstructured":"LI, D.<\/small> and LING, S.<\/small> (2012). On the least squares estimation of multiple-regime threshold autoregressive models. J. Econometrics<\/i> 167<\/b> 240\u2013253.","DOI":"10.1016\/j.jeconom.2011.11.006"}],"container-title":["The Annals of Statistics"],"original-title":[],"deposited":{"date-parts":[[2026,10,6]],"date-time":"2026-10-06T21:12:20Z","timestamp":1791321140000},"score":1,"resource":{"primary":{"URL":"https:\/\/projecteuclid.org\/journals\/annals-of-statistics\/volume-54\/issue-4\/Generalized-multivariate-threshold-autoregressive-models-with-linearly-partitioned-threshold-space\/10.1214\/26-AOS2638.full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,8,1]]},"references-count":36,"journal-issue":{"issue":"4","published-online":{"date-parts":[[2026,8,1]]}},"URL":"https:\/\/doi.org\/10.1214\/26-aos2638","relation":{},"ISSN":["0090-5364","2168-8966"],"issn-type":[{"value":"0090-5364","type":"print"},{"value":"2168-8966","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,8,1]]}}}