Abstract
Nowcasting can play a key role in giving policymakers timelier insight to data published with a significant time lag, such as final GDP figures. Currently, there are a plethora of methodologies and approaches for practitioners to choose from. However, there lacks a comprehensive comparison of these disparate approaches in terms of predictive performance and characteristics. This paper addresses that deficiency by examining the performance of 17 different methodologies in nowcasting US quarterly GDP growth, including all the methods most commonly employed in nowcasting, as well as some of the most popular traditional machine learning approaches. Performance was assessed over a 20-year period, from 2002 to 2022. This span encompassed two crises, the 2008 financial crisis and the COVID crisis, as well as extended tranquil periods. The two best-performing methodologies in the analysis were long short-term memory artificial neural networks (LSTM) and Bayesian vector autoregression (Bayesian VAR). To facilitate further application and testing of each of the examined methodologies, an open-source repository containing boilerplate code that can be applied to different datasets is published alongside the paper, available at: github.com/dhopp1/nowcasting_benchmark
Similar content being viewed by others
References
Ankargren S, Yang Y, Kastner G (2021). Package ’mfbvar’. https://cran.r-project.org/web/packages/mfbvar/mfbvar.pdf
Antolin-Diaz J, Drechsel T, Petrella I (2020) Advances in nowcasting economic activity: secular trends, large shocks and new data. SSRN. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3669854
Babii A, Ghysels E, Striaukas J (2022) Machine learning time series regressions with an application to nowcasting. J Bus Econ Stat 40:1094–1106. https://doi.org/10.1080/07350015.2021.1899933
Bańbura M, Rünstler G (2011) A look into the factor model black box: Publication lags and the role of hard and soft data in forecasting GDP. Int J Forecast 27:333–346. https://doi.org/10.1016/j.ijforecast.2010.01.011
Bańbura M, Giannone D, Reichlin L (2010) Large Bayesian vector auto regressions. J Appl Economet 25:71–92. https://doi.org/10.1002/jae.1137
Barbaglia L, Frattarolo L, Onorante L, Pericoli FM, Ratto M, Pezzoli LT (2022) Testing big data in a big crisis: nowcasting under covid-19. Int J Forecast. https://doi.org/10.1016/j.ijforecast.2022.10.005
Boehmke B (2018) Gradient boosting machines. http://uc-r.github.io/gbm_regression
Bok B, Caratelli D, Giannone D, Sbordone AM, Tambalotti A (2018) Macroeconomic nowcasting and forecasting with big data. Annu Rev Econ 10:615–643. https://doi.org/10.1146/annurev-economics-080217-053214
Boysel S, Vaughan D (2021) Package ’fredr’. https://cran.r-project.org/web/packages/fredr/fredr.pdf
Brownlee J (2018) Deep learning for time series forecasting: predict the future with MLPs, CNNs and LSTMs in python. Machine Learning Mastery. https://books.google.ch/books?id=o5qnDwAAQBAJ
Buono D, Mazzi G, Marcellino M, Kapetanios (2017) Big data types for macroeconomic nowcasting. Eurona. https://ec.europa.eu/eurostat/cros/system/files/euronaissue1-2017-art4.pdf
Cantú F (2018) Estimation of a coincident indicator for international trade and global economic activity. Technical Report 27 UNCTAD. https://unctad.org/system/files/official-document/ser-rp-2018d9_en.pdf
Carriero A, Galvão AB, Kapetanios G (2019) A comprehensive evaluation of macroeconomic forecasting methods. Int J Forecasting 35:1226–1239. https://doi.org/10.1016/j.ijforecast.2019.02.007
Chen T, Guestrin C (2016) XGBoost. In: Proceedings of the 22nd ACM SIGKDD international conference on knowledge discovery and data mining. ACM. https://doi.org/10.1145/2939672.2939785
Chernis T, Sekkel R (2017) A dynamic factor model for nowcasting Canadian GDP growth. Empir Econ 53:217–234. https://doi.org/10.1007/s00181-017-1254-1
Chung J, Gulcehre C, Cho K, Bengio Y (2014) Empirical evaluation of gated recurrent neural networks on sequence modeling. CoRR, arXiv:1412.3555
Cimadomo J, Giannone D, Lenza M, Sokol A, Monti F (2020) Nowcasting with large Bayesian vector autoregressions. Working Paper Series 2453 European Central Bank. https://ideas.repec.org/p/ecb/ecbwps/20202453.html
Clements MP, Galvão AB (2008) Macroeconomic forecasting with mixed-frequency data: forecasting output growth in the United States. J Bus Econ Stat 26:546–554
De Mol C, Giannone D, Reichlin L (2008) Forecasting using a large number of predictors: Is Bayesian shrinkage a valid alternative to principal components? J Econom 146:318–328
Dematos G, Boyd MS, Kermanshahi B, Kohzadi N, Kaastra I (1996) Feedforward versus recurrent neural networks for forecasting monthly Japanese yen exchange rates. Financ Eng Japan Markets 3:59–75. https://doi.org/10.1007/BF00868008
Dynan K, Sheiner L, Fiscal BIH Co, Policy M (2018) GDP as a measure of economic well-being. Hutchins Center working paper. https://books.google.co.uk/books?id=rOe3vQEACAAJ
Eickmeier S, Ng T (2011) Forecasting national activity using lots of international predictors: an application to New Zealand. Int J Forecasting 27:496–511. https://doi.org/10.1016/j.ijforecast.2009
Falat L, Pancikova L (2015) Quantitative modelling in economics with advanced artificial neural networks. Procedia Econ Finance 34:194–201. https://doi.org/10.1016/S2212-5671(15)01619-6
Federal Reserve Bank of San Francisco (2005) Why is there such a time lapse getting the latest report on Gross Domestic Product (GDP)? In May 2005, we are just getting the preliminary GDP report for the quarter that ended in March 2005
Ghysels E (2016) Macroeconomics and the reality of mixed frequency data. Econom Anal Mixed Freq Data Sampling 193:294–314. https://doi.org/10.1016/j.jeconom.2016.04.008
Ghysels E, Santa-Clara P, Valkanov R (2004) The MIDAS touch: mixed data sampling regression models. In: CIRANO working papers CIRANO. https://EconPapers.repec.org/RePEc:cir:cirwor:2004s-20
Giannone D, Reichlin L, Simonelli S (2009) Nowcasting euro area economic activity in real time: the role of confidence indicators. Natl Inst Econ Rev 210:90–97. https://doi.org/10.1177/0027950109354413
Giannone D, Reichlin L, Small D (2005) Nowcasting GDP and inflation: the real time informational content of macroeconomic data releases. Technical Report Centre for Economic Policy Research. https://cepr.org/active/publications/discussion_papers/dp.php?dpno=5178
Gluonts (2022) GluonTS - probabilistic time series modeling in python. https://ts.gluon.ai/stable/
Grosse R (2017) Lecture 15: exploding and vanishing gradients. http://www.cs.toronto.edu/~rgrosse/courses/csc321_2017/readings/L15%20Exploding%20and%20Vanishing%20Gradients.pdf
Guichard S, Rusticelli E (2011) A dynamic factor model for world trade growth. In: OECD Economics Department Working Papers, https://doi.org/10.1787/5kg9zbvvwqq2-en. https://www.oecd-ilibrary.org/content/paper/5kg9zbvvwqq2-en
Hallin M, Liška R (2011) Dynamic factors in the presence of blocks. J Econom 163:29–41. https://doi.org/10.1016/j.jeconom.2010.11.004
Hopp D (2021) nowcastLSTM. https://github.com/dhopp1/nowcastLSTM
Hopp D (2022) nowcasting_benchmark. https://github.com/dhopp1/nowcasting_benchmark
Hopp D (2022) Economic nowcasting with long short-term memory artificial neural networks (LSTM). J Off Stat 38:847–873. https://doi.org/10.2478/jos-2022-0037
Hopp D (2022) Performance of LSTM neural networks in nowcasting global trade during the COVID-19 crisis. Stat J IAOS 38:1–14. https://doi.org/10.3233/SJI-210911
Hopp D, Cantú F (2020) nowcastDFM. https://github.com/dhopp1-UNCTAD/nowcastDFM publication Title: GitHub repository
IMF (2020) Gross domestic product: an economy’s all. https://www.imf.org/external/pubs/ft/fandd/basics/gdp.htm
Jansen WJ, Jin X, Winter JMd (2016) Forecasting and nowcasting real GDP: comparing statistical models and subjective forecasts. Int J Forecast 32:411–436. https://doi.org/10.1016/j.ijforecast.2015.05.008
Kapoor A, Debroy B (2019) GDP is not a measure of human well-being. https://hbr.org/2019/10/gdp-is-not-a-measure-of-human-well-being
Kohzadi N, Boyd MS, Kermanshahi B, Kaastra I (1996) A comparison of artificial neural network and time series models for forecasting commodity prices. Neurocomputing 10:169–181. https://doi.org/10.1016/0925-2312(95)00020-8
Kronenberg P, Mikosch H, Neuwirth S (2021) The nowcasting lab. Technical Report KOF Swis Economic Institute Zurich. https://www.gpeari.gov.pt/documents/35086/229852/2022-15.11-nov-The+Nowcasting+Lab_Paper.pdf/456c3af1-ce10-1b73-494e-b840b8a0d05d?t=1668621627180
Kuzin VN, Marcellino M, Schumacher C (2009) MIDAS versus mixed-frequency VAR: nowcasting GDP in the euro area. Discussion Paper Series 1: Economic Studies 2009,07 Deutsche Bundesbank. https://ideas.repec.org/p/zbw/bubdp1/7576.html
Kuzin V, Marcellino M, Schumacher C (2011) MIDAS vs. mixed-frequency VAR: nowcasting GDP in the euro area. Int J Forecast 27:529–542. https://doi.org/10.1016/j.ijforecast.2010.02.006
Kvedaras V (2021) Package ’midasr’. https://cran.r-project.org/web/packages/midasr/midasr.pdf
Liew V (2004) Which lag selection criteria should we employ? Econ Bull 3:1–9
Loermann J, Maas B (2019) Nowcasting US GDP with artificial neural networks. MPRA Paper 95459 University Library of Munich, Germany. https://ideas.repec.org/p/pra/mprapa/95459.html
Mahaboob B, Bhumireddy V, Narayana C, Sankar J, Balasiddamuni P (2018) A treatise on ordinary least squares estimation of parameters of linear model. Int J Eng Technol (UAE), 7. https://doi.org/10.14419/ijet.v7i4.10.21216
Marcellino M, Schumacher C (2010) Factor MIDAS for nowcasting and forecasting with ragged-edge data: a model comparison for German GDP*. Oxford Bull Econ Stat 72:518–550. https://doi.org/10.1111/j.1468-0084.2010.00591.x
Mariano RS, Murasawa Y (2003) A new coincident index of business cycles based on monthly and quarterly series. J Appl Economet 18:427–443. https://doi.org/10.1002/jae.695
Mills TC (2019) Chapter 3: ARMA models for stationary time series. In: Mills TC (ed), Applied time series analysis (pp 31–56). Academic Press. https://doi.org/10.1016/B978-0-12-813117-6.00003-X
Morgado AJ, Nunes LC, Salvado S (2007) Nowcasting an economic aggregate with disaggregate dynamic factors: an application to Portuguese GDP. GEE Papers 0002 Gabinete de Estratégia e Estudos, Ministério da Economia. https://ideas.repec.org/p/mde/wpaper/0002.html
Natekin A, Knoll A (2013) Gradient boosting machines, a tutorial. Front Neurorobotics. https://doi.org/10.3389/fnbot.2013.00021
Patel H, Prajapati P (2018) Study and analysis of decision tree based classification algorithms. Int J Comput Sci Eng 6:74–78. https://doi.org/10.26438/ijcse/v6i10.7478
Porshakov A, Ponomarenko A, Sinyakov A (2016) Nowcasting and short-term forecasting of Russian GDP with a dynamic factor model. J New Econ Assoc 30:60–76
Probst P, Bischl B, Boulesteix A-L (2018) Tunability: importance of Hyperparameters of machine learning algorithms. arXiv:1802.09596
Richardson A, Mulder TvF, Vehbi T (2021) Nowcasting GDP using machine-learning algorithms: a real-time assessment. Int J Forecast 37:941–948. https://doi.org/10.1016/j.ijforecast.2020.10.005
Salinas D, Flunkert V, Gasthaus J, Januschowski T (2020) DeepAR: Probabilistic forecasting with autoregressive recurrent networks. Int J Forecast 36:1181–1191. https://doi.org/10.1016/j.ijforecast.2019.07.001
Sarker IH (2021) Machine learning: algorithms, real-world applications and research directions. SN Comput. Sci. 2:160. https://doi.org/10.1007/s42979-021-00592-x
Sazli M (2006) A brief review of feed-forward neural networks. Commun. Faculty Sci. Univ. Ankara 50:11–17. https://doi.org/10.1501/0003168
Schorfheide F, Song D (2015) Real-time forecasting with a mixed-frequency VAR. J Bus Econ Stat 33:366–380. https://doi.org/10.1080/07350015.2014.954707
Scikit learn (2021a) 1.10. Decision Trees. https://scikit-learn.org/stable/modules/tree.html
Scikit learn (2021b) 3.1. Cross-validation: evaluating estimator performance. https://scikit-learn.org/stable/modules/cross_validation.html
Scikit learn (2021c) scikit-learn machine learning in python. https://scikit-learn.org/stable/index.html
Shi X, Chen Z, Wang H, Yeung D-Y, Wong W-k, Woo W-C (2015) Convolutional LSTM network: a machine learning approach for precipitation nowcasting. arXiv:1506.04214
Sims CA (1980) Macroeconomics and reality. Econometrica 48:1–48. https://doi.org/10.2307/1912017
Singh R, Prajneshu (2008) Artificial neural network methodology for modelling and forecasting maize crop yield. Agric Econ Res Rev 21
Smith T (2021) pmdarima: ARIMA estimators for Python. https://alkaline-ml.com/pmdarima/index.html
Soley-Bori M (2013) Dealing with missing data: Key assumptions and methods for applied analysis. Technical Report No. 4 Boston University Boston. https://www.bu.edu/sph/files/2014/05/Marina-tech-report.pdf
Soybilgen B, Yazgan E (2021) Nowcasting US GDP using tree-based ensemble models and dynamic factors. Comput Econ 57:387–417. https://doi.org/10.1007/s10614-020-10083-5
Stock JH, Watson MW (2001) Vector autoregressions. J Econ Perspect 15:101–115. https://doi.org/10.1257/jep.15.4.101
Stock JH, Watson MW (2002) Forecasting using principal components from a large number of predictors. J Am Stat Assoc 97:1167–1179. https://doi.org/10.1198/1016214502388618960
Stratos K (2020) Feedforward and recurrent neural networks. http://www1.cs.columbia.edu/~stratos/research/neural.pdf
Striaukas J, Babii A, Ghysels E, Kostrov A (2022) midasml: Estimation and prediction methods for high-dimensional mixed frequency time series data. https://cran.r-project.org/web/packages/midasml/index.html
Taylor R (2016) PyFlux. https://pyflux.readthedocs.io/en/latest/index.html
Tibshirani R (1996) Regression shrinkage and selection via the Lasso. J R Stat Soc Ser B (Methodological) 58:267–288
Tiffin A (2016) Seeing in the dark: a machine-learning approach to nowcasting in Lebanon. IMF Work Pap 16:1. https://doi.org/10.5089/9781513568089.001
WMO (2017) Guidelines for nowcasting techniques. Technical Report 1198 WMO. https://library.wmo.int/doc_num.php?explnum_id=3795
XGBoost Developers (2022) XGBoost Documentation. https://xgboost.readthedocs.io/en/stable/index.html
Zou H, Hastie T (2005) Regularization and variable selection via the elastic net. J R Stat Soc Ser B (Stat Methodol) 67:301–320
Acknowledgements
The author would like to thank Anu Peltola for her valuable comments and feedback.
Funding
No funds, grants, or other support were received.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
The author has no relevant financial or non-financial interests to disclose.
Ethical approval
This article does not contain any studies with human participants or animals performed by any of the authors.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
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
Hopp, D. Benchmarking econometric and machine learning methodologies in nowcasting GDP. Empir Econ 66, 2191–2247 (2024). https://doi.org/10.1007/s00181-023-02515-6
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00181-023-02515-6