IDEAS home Printed from https://ideas.repec.org/p/arx/papers/2601.12039.html

Nonlinear Dynamic Factor Analysis With a Transformer Network

Author

Listed:
  • Oliver Snellman

Abstract

The paper develops a Transformer architecture for estimating dynamic factors from multivariate time series data under flexible identification assumptions. Performance on small datasets is improved substantially by using a conventional factor model as prior information via a regularization term in the training objective. The results are interpreted with Attention matrices that quantify the relative importance of variables and their lags for the factor estimate. Time variation in Attention patterns can help detect regime switches and evaluate narratives. Monte Carlo experiments suggest that the Transformer is more accurate than the linear factor model, when the data deviate from linear-Gaussian assumptions. An empirical application uses the Transformer to construct a coincident index of U.S. real economic activity.

Suggested Citation

  • Oliver Snellman, 2026. "Nonlinear Dynamic Factor Analysis With a Transformer Network," Papers 2601.12039, arXiv.org.
  • Handle: RePEc:arx:papers:2601.12039
    as

    Download full text from publisher

    File URL: http://arxiv.org/pdf/2601.12039
    File Function: Latest version
    Download Restriction: no
    ---><---

    References listed on IDEAS

    as
    1. Giannone, Domenico & Reichlin, Lucrezia & Small, David, 2008. "Nowcasting: The real-time informational content of macroeconomic data," Journal of Monetary Economics, Elsevier, vol. 55(4), pages 665-676, May.
    2. Minsuk Shin & Anirban Bhattacharya & Valen E. Johnson, 2020. "Functional Horseshoe Priors for Subspace Shrinkage," Journal of the American Statistical Association, Taylor & Francis Journals, vol. 115(532), pages 1784-1797, December.
    3. Thorsten Drautzburg & Jesus Fernandez-Villaverde & Pablo Guerron-Quintana & Dick Oosthuizen, 2024. "Filtering with Limited Information," PIER Working Paper Archive 24-016, Penn Institute for Economic Research, Department of Economics, University of Pennsylvania.
    4. Hauzenberger, Niko & Huber, Florian & Klieber, Karin & Marcellino, Massimiliano, 2025. "Machine learning the macroeconomic effects of financial shocks," Economics Letters, Elsevier, vol. 250(C).
    5. Diebold, Francis X & Mariano, Roberto S, 2002. "Comparing Predictive Accuracy," Journal of Business & Economic Statistics, American Statistical Association, vol. 20(1), pages 134-144, January.
    6. Thomas J. Sargent & Christopher A. Sims, 1977. "Business cycle modeling without pretending to have too much a priori economic theory," Working Papers 55, Federal Reserve Bank of Minneapolis.
    7. Klieber, Karin, 2024. "Non-linear dimension reduction in factor-augmented vector autoregressions," Journal of Economic Dynamics and Control, Elsevier, vol. 159(C).
    8. Philippe Goulet Coulombe, 2022. "A Neural Phillips Curve and a Deep Output Gap," Papers 2202.04146, arXiv.org, revised Oct 2024.
    9. Tony Chernis & Niko Hauzenberger & Haroon Mumtaz & Michael Pfarrhofer, 2025. "A Bayesian Gaussian Process Dynamic Factor Model," Papers 2509.04928, arXiv.org.
    10. Aruoba, S. BoraÄŸan & Diebold, Francis X. & Scotti, Chiara, 2009. "Real-Time Measurement of Business Conditions," Journal of Business & Economic Statistics, American Statistical Association, vol. 27(4), pages 417-427.
    11. Arthur F. Burns & Wesley C. Mitchell, 1946. "Measuring Business Cycles," NBER Books, National Bureau of Economic Research, Inc, number burn46-1, August.
    12. James H. Stock & Mark W. Watson, 1989. "New Indexes of Coincident and Leading Economic Indicators," NBER Chapters, in: NBER Macroeconomics Annual 1989, Volume 4, pages 351-409, National Bureau of Economic Research, Inc.
    13. Diebold, Francis X & Rudebusch, Glenn D, 1996. "Measuring Business Cycles: A Modern Perspective," The Review of Economics and Statistics, MIT Press, vol. 78(1), pages 67-77, February.
    14. Hamilton, James D, 1989. "A New Approach to the Economic Analysis of Nonstationary Time Series and the Business Cycle," Econometrica, Econometric Society, vol. 57(2), pages 357-384, March.
    15. Ben S. Bernanke & Jean Boivin & Piotr Eliasz, 2005. "Measuring the Effects of Monetary Policy: A Factor-Augmented Vector Autoregressive (FAVAR) Approach," The Quarterly Journal of Economics, President and Fellows of Harvard College, vol. 120(1), pages 387-422.
    16. Florian Huber & Gary Koop, 2023. "Subspace shrinkage in conjugate Bayesian vector autoregressions," Journal of Applied Econometrics, John Wiley & Sons, Ltd., vol. 38(4), pages 556-576, June.
    Full references (including those not matched with items on IDEAS)

    Most related items

    These are the items that most often cite the same works as this one and are cited by the same works as this one.
    1. Chauvet, Marcelle & Potter, Simon, 2013. "Forecasting Output," Handbook of Economic Forecasting, in: G. Elliott & C. Granger & A. Timmermann (ed.), Handbook of Economic Forecasting, edition 1, volume 2, chapter 0, pages 141-194, Elsevier.
    2. Dennis J. Fixler & Jeremy J. Nalewaik, 2007. "News, noise, and estimates of the \"true\" unobserved state of the economy," Finance and Economics Discussion Series 2007-34, Board of Governors of the Federal Reserve System (U.S.).
    3. Marcos Dal Bianco & Jaime Martinez-Martín & Maximo Camacho, 2013. "Short-Run Forecasting of Argentine GDP Growth," Working Papers 1314, BBVA Bank, Economic Research Department.
    4. Stock, J.H. & Watson, M.W., 2016. "Dynamic Factor Models, Factor-Augmented Vector Autoregressions, and Structural Vector Autoregressions in Macroeconomics," Handbook of Macroeconomics, in: J. B. Taylor & Harald Uhlig (ed.), Handbook of Macroeconomics, edition 1, volume 2, chapter 0, pages 415-525, Elsevier.
    5. Catherine Doz & Peter Fuleky, 2019. "Dynamic Factor Models," Working Papers halshs-02262202, HAL.
    6. Brandyn Bok & Daniele Caratelli & Domenico Giannone & Argia M. Sbordone & Andrea Tambalotti, 2018. "Macroeconomic Nowcasting and Forecasting with Big Data," Annual Review of Economics, Annual Reviews, vol. 10(1), pages 615-643, August.
    7. Vegard H. Larsen & Leif Anders Thorsrud, 2026. "Using Transformers and Reinforcement Learning as Narrative Filters in Macroeconomics," CESifo Working Paper Series 12454, CESifo.
    8. Cimadomo, Jacopo & Giannone, Domenico & Lenza, Michele & Monti, Francesca & Sokol, Andrej, 2022. "Nowcasting with large Bayesian vector autoregressions," Journal of Econometrics, Elsevier, vol. 231(2), pages 500-519.
    9. Kholodilin, Konstantin A. & Yao, Vincent W., 2005. "Measuring and predicting turning points using a dynamic bi-factor model," International Journal of Forecasting, Elsevier, vol. 21(3), pages 525-537.
    10. Arias, Maria A. & Gascon, Charles S. & Rapach, David E., 2016. "Metro business cycles," Journal of Urban Economics, Elsevier, vol. 94(C), pages 90-108.
    11. Maximo Camacho & Gabriel Perez-Quiros & Pilar Poncela, 2010. "Green shoots in the euro area. A real time measure," Working Papers 1026, Banco de España.
    12. Bae, Juhee, 2024. "Factor-augmented forecasting in big data," International Journal of Forecasting, Elsevier, vol. 40(4), pages 1660-1688.
    13. Aruoba, S. BoraÄŸan & Diebold, Francis X. & Scotti, Chiara, 2009. "Real-Time Measurement of Business Conditions," Journal of Business & Economic Statistics, American Statistical Association, vol. 27(4), pages 417-427.
    14. Catherine Doz & Laurent Ferrara & Pierre-Alain Pionnier, 2026. "Switching Macroeconomic Growth and Volatility: Evidence from a Mean-Variance Markov-Switching Dynamic Factor Model," PSE Working Papers halshs-02443364, HAL.
    15. Francis X. Diebold, 1998. "The Past, Present, and Future of Macroeconomic Forecasting," Journal of Economic Perspectives, American Economic Association, vol. 12(2), pages 175-192, Spring.
    16. Antolín-Díaz, Juan & Drechsel, Thomas & Petrella, Ivan, 2024. "Advances in nowcasting economic activity: The role of heterogeneous dynamics and fat tails," Journal of Econometrics, Elsevier, vol. 238(2).
    17. Harm Bandholz, 2005. "New Composite Leading Indicators for Hungary and Poland," ifo Working Paper Series 3, ifo Institute - Leibniz Institute for Economic Research at the University of Munich.
    18. Carriero, Andrea & Marcellino, Massimiliano, 2007. "A comparison of methods for the construction of composite coincident and leading indexes for the UK," International Journal of Forecasting, Elsevier, vol. 23(2), pages 219-236.
    19. Ángel Cuevas & Enrique Quilis, 2012. "A factor analysis for the Spanish economy," SERIEs: Journal of the Spanish Economic Association, Springer;Spanish Economic Association, vol. 3(3), pages 311-338, September.
    20. Carriero, Andrea & Marcellino, Massimiliano, 2007. "A comparison of methods for the construction of composite coincident and leading indexes for the UK," International Journal of Forecasting, Elsevier, vol. 23(2), pages 219-236.

    More about this item

    NEP fields

    This paper has been announced in the following NEP Reports:

    Statistics

    Access and download statistics

    Corrections

    All material on this site has been provided by the respective publishers and authors. You can help correct errors and omissions. When requesting a correction, please mention this item's handle: RePEc:arx:papers:2601.12039. See general information about how to correct material in RePEc.

    If you have authored this item and are not yet registered with RePEc, we encourage you to do it here. This allows to link your profile to this item. It also allows you to accept potential citations to this item that we are uncertain about.

    If CitEc recognized a bibliographic reference but did not link an item in RePEc to it, you can help with this form .

    If you know of missing items citing this one, you can help us creating those links by adding the relevant references in the same way as above, for each refering item. If you are a registered author of this item, you may also want to check the "citations" tab in your RePEc Author Service profile, as there may be some citations waiting for confirmation.

    For technical questions regarding this item, or to correct its authors, title, abstract, bibliographic or download information, contact: arXiv administrators (email available below). General contact details of provider: http://arxiv.org/ .

    Please note that corrections may take a couple of weeks to filter through the various RePEc services.

    IDEAS is a RePEc service. RePEc uses bibliographic data supplied by the respective publishers.