×

Bias correction for time series factor models. (English) Zbl 07192620

Summary: In this paper we work with multivariate time series that follow a Factor Model. In particular, we consider the setting where factors are dominated by highly persistent AutoRegressive (AR) processes and samples that are rather small. Therefore, the factors’ AR models are estimated using small sample bias correction techniques. A Monte Carlo study reveals that bias-correcting the AR coefficients of the factors allows to obtain better results in terms of prediction interval coverage. As expected, the simulation shows that bias-correction is more successful for smaller samples. We present the results assuming the AR order and number of factors are known as well as unknown. We also study the advantages of this technique for a set of Industrial Production Indexes of several European countries.

MSC:

62-XX Statistics
PDFBibTeX XMLCite
Full Text: DOI

References:

[1] Peña D, Box G.Identifying a simplifying structure in time series. J Amer Statist Assoc. 1987;82:836-843. [Taylor & Francis Online], [Web of Science ®], [Google Scholar] · Zbl 0623.62081
[2] Alonso A, García-Martos C, Rodríguez J, et al. Seasonal dynamic factor analysis and bootstrap inference: application to electricity market forecasting. Technometrics. 2011;53:137-151. doi: 10.1198/TECH.2011.09050[Taylor & Francis Online], [Web of Science ®], [Google Scholar]
[3] García-Martos C, Rodríguez J, Sánchez M.Forecasting electricity prices by extracting dynamic common factors: application to the Iberian market. IET Gener Transm Distrib. 2012;6:11-20. doi: 10.1049/iet-gtd.2011.0009[Crossref], [Web of Science ®], [Google Scholar]
[4] Alonso A, Bastos G, García-Martos C.Electricity price forecasting by averaging dynamic factor models. Energies. 2016;9:1-21. doi: 10.3390/en9080600[Crossref], [Web of Science ®], [Google Scholar]
[5] García-Martos C, Rodríguez J, Sánchez M.Modelling and forecasting fossil fuels, CO2 and electricity prices and their volatilities. Appl Energy. 2013;101:363-375. doi: 10.1016/j.apenergy.2012.03.046[Crossref], [Web of Science ®], [Google Scholar]
[6] Poncela P, Senra E, Sierra L.Common dynamics of nonenergy commodity prices and their relation to uncertainty. Appl Econ. 2014;46:3724-3735. doi: 10.1080/00036846.2014.939377[Taylor & Francis Online], [Web of Science ®], [Google Scholar]
[7] Stock J, Watson M.Forecasting using principal components from a large number of predictors. J Amer Statist Assoc. 2002;97:1167-1179. doi: 10.1198/016214502388618960[Taylor & Francis Online], [Web of Science ®], [Google Scholar] · Zbl 1041.62081
[8] Sargent T, Sims C. Business cycle modeling without pretending to have too much a priory economic theory. Working paper 55 Federal Reserve Bank of Minneapolis; 1977. [Google Scholar]
[9] Forni M, Hallin M, Lippi M, et al. Reference cycles: the NBER methodology revisited. NBER; 1999. [Google Scholar]
[10] Lee R, Carter L.Modeling and forecasting US mortality. J Amer Statist Assoc. 1992;87:659-671. [Taylor & Francis Online], [Web of Science ®], [Google Scholar] · Zbl 1351.62186
[11] Alonso A, Peña D, Rodríguez J. A methodology for population projections: an application to Spain. Working Paper 08 - 45 Statistics and Econometrics Series 12; 2008. Available from: http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.584.2593&rep=rep1&type=pdf[Google Scholar]
[12] Jeong S, Bienkiewicz B.Application of autoregressive modeling in proper orthogonal decomposition of building wind pressure. J Wind Eng Ind Aerod. 1997;69:685-695. doi: 10.1016/S0167-6105(97)00197-9[Crossref], [Google Scholar]
[13] García-Ferrer A, González-Prieto E, Peña D. Exploring ICA for time series decomposition. Working Paper 11-16 Statistics and Econometrics Series 11 Departamento de Estadística Universidad Carlos III de Madrid May 2011; 2011. [Google Scholar]
[14] Peña D, Poncela P.Nonstationary dynamic factor analysis. J Statist Plann Inference. 2006;136:1237-1257. doi: 10.1016/j.jspi.2004.08.020[Crossref], [Web of Science ®], [Google Scholar] · Zbl 1088.62077
[15] Clements M, Kim J.Bootstrap prediction intervals for autoregressive time series. Comput Statist Data Anal. 2007;51:3580-3594. doi: 10.1016/j.csda.2006.09.012[Crossref], [Web of Science ®], [Google Scholar] · Zbl 1161.62406
[16] Roy A, Fuller W.Estimation for autoregressive time series with a root near 1. J Bus Econom Statist. 2001;19:482-493. doi: 10.1198/07350010152596736[Taylor & Francis Online], [Web of Science ®], [Google Scholar]
[17] Clements M, Taylor N.Bootstrapping prediction intervals for autoregressive models. Int J Forecast. 2001;17:247-267. doi: 10.1016/S0169-2070(00)00079-0[Crossref], [Web of Science ®], [Google Scholar]
[18] Gregory A, Head A.Common and country-specific fluctuations in productivity, investment, and the current account. J Monet Econ. 1999;44:423-451. doi: 10.1016/S0304-3932(99)00035-5[Crossref], [Web of Science ®], [Google Scholar]
[19] Geweke J, Singleton K.Maximum likelihood ‘confirmatory’ factor analysis of economic time series. Int Econ Rev (Philadelphia). 1981;22:37-54. doi: 10.2307/2526134[Crossref], [Web of Science ®], [Google Scholar] · Zbl 0483.90037
[20] Bai J, Ng S.Large dimensional factor analysis. Found Trends Econom. 2008;3:89-163. doi: 10.1561/0800000002[Crossref], [Google Scholar]
[21] Stock J, Watson M Dynamic factor models. In: Clements MP, Hendry DF, editors. Oxford handbook of economic forecasting. New York: Oxford University Press; 2010. p. 35-60. [Google Scholar]
[22] Poncela P, Ruiz E.Small- versus big-data factor extraction in dynamic factor models: an empirical assesment. In: Hillebrand E, Koopman SJ, editors. Dynamic factor models (Advances in econometrics, Volume 35). West Yorkshire: Emerald Group Publishing Limited; 2016. p. 401-434. [Crossref], [Google Scholar] · Zbl 1461.62007
[23] Mardia K, Kent T, Bibby M. Multivariate analysis. San Diego: Academic Press, Inc.; 1979. Chapter 8 and 9. [Google Scholar] · Zbl 0432.62029
[24] Bai J, Ng S.Determining the number of factors in approximate factor models. Econometrica. 2002;70:191-221. doi: 10.1111/1468-0262.00273[Crossref], [Web of Science ®], [Google Scholar] · Zbl 1103.91399
[25] Hurvich C, Tsai C.Regression and time series model selection in small samples. Biometrika. 1989;76:297-307. doi: 10.1093/biomet/76.2.297[Crossref], [Web of Science ®], [Google Scholar] · Zbl 0669.62085
[26] Schwarz G.Estimating the dimension of a model. Ann Statist. 1978;6:461-464. doi: 10.1214/aos/1176344136[Crossref], [Web of Science ®], [Google Scholar] · Zbl 0379.62005
[27] Ahn S, Horenstein A.Eigenvalue ratio test for the number of factors. Econometrica. 2013;81:1203-1227. doi: 10.3982/ECTA8968[Crossref], [Web of Science ®], [Google Scholar] · Zbl 1274.62403
[28] Peña D, Safadi T.Bayesian analysis of dynamic factor models: an application to air pollution and mortality in São Paulo, Brazil. Environmetrics. 2008;19:582-601. doi: 10.1002/env.899[Crossref], [Web of Science ®], [Google Scholar]
[29] Doz C, Giannone D, Reichlin L.A quasi-maximum likelihood approach for large, approximate dynamic factor models. Rev Econ Stat. 2012;94:1014-1024. doi: 10.1162/REST_a_00225[Crossref], [Web of Science ®], [Google Scholar]
[30] García-Ferrer A, González-Prieto E, Peña D.A conditionally heteroskedastic independent factor model with an application to financial stock returns. Int J Forecast. 2012;28:70-93. doi: 10.1016/j.ijforecast.2011.02.010[Crossref], [Web of Science ®], [Google Scholar]
[31] Fiorentini G, Galesi A, Sentana E.A spectral EM algorithm for dynamic factor models. Documentos de Trabajo, Banco de España. 2016;1619:1-62. [Google Scholar]
[32] Alonso A, Bastos G, García-Martos C. Bias correction for dynamic factor models. UC3M Working Paper 17 - 02 Statistics and Econometrics; 2017. Available from: http://e-archivo.uc3m.es/handle/10016/240290:1-44[Google Scholar]
[33] Kilian L.Accounting for lag-order uncertainty in autoregressions: the endogenous lag order bootstrap algorithm. J Time Ser Anal. 1998;19:531-548. doi: 10.1111/1467-9892.00107[Crossref], [Google Scholar] · Zbl 0913.62083
[34] MacKinnon J, Smith J.Approximate bias correction in econometrics. J Econometrics. 1998;85:205-230. doi: 10.1016/S0304-4076(97)00099-7[Crossref], [Web of Science ®], [Google Scholar] · Zbl 0961.62111
[35] Kilian L.Small-sample confidence intervals for impulse response functions. Rev Econ Stat. 1998;80:218-230. doi: 10.1162/003465398557465[Crossref], [Web of Science ®], [Google Scholar]
[36] Pascual L, Romo J, Ruiz E.Bootstrap predictive inference for arima processes. J Time Ser Anal. 2004;25:449-465. doi: 10.1111/j.1467-9892.2004.01713.x[Crossref], [Web of Science ®], [Google Scholar] · Zbl 1062.62199
[37] Pascual L, Romo J, Ruiz E.Effects of parameter estimation on prediction densities: a bootstrap approach. Int J Forecast. 2001;17:83-103. doi: 10.1016/S0169-2070(00)00069-8[Crossref], [Web of Science ®], [Google Scholar]
[38] Fresoli D, Ruiz E, Pascual L.Bootstrap multi-step forecasts of non-Gaussian VAR models. Int J Forecast. 2015;31:834-848. doi: 10.1016/j.ijforecast.2014.04.001[Crossref], [Web of Science ®], [Google Scholar]
[39] Ruiz E, Poncela P.More is not always better: Kalman filtering in dynamic factor models. In: Koopman S J, Shephard N, editors. Unobserved components and time series econometrics. New York: Oxford University Press; 2015; Chapter 11. [Google Scholar]
[40] Alonso A, Peña D, Romo J.Forecasting time series with sieve bootstrap. J Statist Plann Inference. 2002;100:1-11. doi: 10.1016/S0378-3758(01)00092-1[Crossref], [Web of Science ®], [Google Scholar] · Zbl 1007.62077
[41] Marcellino M, Stock J, Watson M.A comparison of direct and iterated multistep AR methods for forecasting macroeconomic time series. J Econometrics 2006;135:499-526. doi: 10.1016/j.jeconom.2005.07.020[Crossref], [Web of Science ®], [Google Scholar] · Zbl 1418.62513
[42] Robinson P.Conditional-sum-of-squares estimation of models for stationary time series with long memory. IMS Lecture Notes-Monograph Series Time Series and Related Topics. 2006;52:130-137. doi: 10.1214/074921706000000996[Crossref], [Google Scholar] · Zbl 1268.62117
[43] Egrioglu E, Aladag C, Kadilar C.The CSS and the two-staged methods for parameter estimation in SARFIMA models. J Probab Statist. 2011;2011:1-11. doi: 10.1155/2011/691058[Crossref], [Google Scholar] · Zbl 1229.62121
[44] Alonso A, Peña D, Romo J.Introducing model uncertainty in time series bootstrap. Statist Sinica. 2004;14:155-174. [Web of Science ®], [Google Scholar] · Zbl 1035.62088
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. In some cases that data have been complemented/enhanced by data from zbMATH Open. This attempts to reflect the references listed in the original paper as accurately as possible without claiming completeness or a perfect matching.