Nowcasting Finnish real economic activity: a machine learning approach

Abstract

We develop a nowcasting framework, based on microlevel data, to provide faster estimates of the Finnish monthly real economic activity indicator, the Trend Indicator of Output (TIO), and of quarterly GDP. We use firm-level turnovers, which are available shortly after the end of the reference month, and real-time traffic volumes data, to form our set of predictors. We rely on combinations of nowcasts obtained from a range of statistical models and machine learning techniques which are able to handle high-dimensional information sets. The results of our pseudo-real-time analysis indicate that a simple nowcast combination based on these models provides faster estimates of TIO and GDP, without increasing substantially the revision error.

This is a preview of subscription content, access via your institution.

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7
Fig. 8

Notes

  1. 1.

    A description of this indicator is available at http://www.stat.fi/til/ktkk/index_en.html.

  2. 2.

    In our exercise, we compute both nowcasts (predictions of a variable while the reference period is still ongoing) and backcasts (estimates referring to a period which already ended). We refer to our predictions as nowcasts, to be in line with the literature (see Banbura et al. 2011).

  3. 3.

    More details on the techniques we use and on the estimation procedure are provided in online appendix.

  4. 4.

    Alternative estimators of latent factors are presented in Forni et al. (2000) and, more recently, Doz et al. (2011). Bai and Ng (2002) developed a series of information criteria that provide an estimate of the number of static factors r.

  5. 5.

    The data are available at https://aineistot.vayla.fi/lam/rawdata.

  6. 6.

    Statistics Finland adjusts monthly TIO figures so that they are consistent with quarterly GDP growth estimates, once the latter become available. The same adjustment is done to quarterly GDP when yearly GDP figures are released. The practical implication of this procedure is the presence of large revisions of historical growth rates at the monthly and quarterly frequency.

  7. 7.

    In our exercise, this set of models includes 21 specifications, such the factor augmented automated ARIMA, regression splines, tree-based regressions, ridge regressions, support vector machine, k-nearest neighbors, and boosting.

References

  1. Aastveit KA, Trovik T (2014) Estimating the output gap in real time: a factor model approach. Q Rev Econ Finance 54(2):180–193

    Article  Google Scholar 

  2. Altissimo F, Cristadoro R, Forni M, Lippi M, Veronese G (2010) New Eurocoin: tracking economic growth in real time. Rev Econ Stat 92(4):1024–1034

    Article  Google Scholar 

  3. Aruoba SB, Diebold FX, Scotti C (2009) Real-time measurement of business conditions. J Bus Econ Stat 27(4):417–427

    Article  Google Scholar 

  4. Bai J, Ng S (2002) Determining the number of factors in approximate factor models. Econometrica 70(1):191–221

    Article  Google Scholar 

  5. Bai J, Ng S (2009) Boosting diffusion indices. J Appl Econom 24(4):607–629. https://doi.org/10.1002/jae.1063

    Article  Google Scholar 

  6. Baldacci E, Buono D, Kapetanios G, Krische S, Marcellino M, Mazzi GL, Papailias F (2016) Big data and macroeconomic nowcasting: from data access to modelling. Technical report

  7. Banbura M, Giannone D, Reichlin L (2011) Nowcasting. In: Hendry DF, Clements M (eds) Oxford handbook of economic forecasting. Oxford University Press, Oxford

    Google Scholar 

  8. 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(2):318–328

    Article  Google Scholar 

  9. Doz C, Giannone D, Reichlin L (2011) A two-step estimator for large approximate dynamic factor models based on Kalman filtering. J Econom 164(1):188–205

    Article  Google Scholar 

  10. Evans MDD (2005) Where are we now? Real-time estimates of the macroeconomy. Int J Cent Bank 1(2):127–175

    Google Scholar 

  11. Fernandez-Rodriguez F, Sosvilla-Rivero S, Andrada-Felix J (1999) Exchange-rate forecasts with simultaneous nearest-neighbour methods: evidence from the EMS. Int J Forecast 15(4):383–392

    Article  Google Scholar 

  12. Fornaro P (2016) Predicting Finnish economic activity using firm-level data. Int J Forecast 32(1):10–19

    Article  Google Scholar 

  13. Forni M, Hallin M, Lippi M, Reichlin L (2000) The generalized dynamic-factor model: identification and estimation. Rev Econ Stat 82(4):540–554

    Article  Google Scholar 

  14. Giannone D, Reichlin L, Small D (2008) Nowcasting: the real-time informational content of macroeconomic data. J Monet Econ 55(4):665–676

    Article  Google Scholar 

  15. Hastie T, Tibshirani R, Friedman J (2009) The elements of statistical learning: data mining, inference and prediction, 2nd edn. Springer, Berlin

    Book  Google Scholar 

  16. Hyndman R, Khandakar Y (2008) Automatic time series forecasting: the forecast package for R. J Stat Softw 27(3):1–22

    Article  Google Scholar 

  17. Hyndman RJ, Athanasopoulos G (2018) Forecasting: principles and practice. OTexts, Melbourne

    Google Scholar 

  18. Josse J, Husson F (2016) missMDA: a package for handling missing values in multivariate data analysis. J Stat Softw 70(1):1–31

    Article  Google Scholar 

  19. Matheson TD, Mitchell J, Silverstone B (2010) Nowcasting and predicting data revisions using panel survey data. J Forecast 29(3):313–330

    Google Scholar 

  20. Modugno M (2013) Now-casting inflation using high frequency data. Int J Forecast 29(4):664–675

    Article  Google Scholar 

  21. Nyman R, Ormerod P (2017) Predicting economic recessions using machine learning algorithms. Papers arXiv:1701.01428

  22. Plakandaras V, Papadimitriou T, Gogas P (2015) Forecasting daily and monthly exchange rates with machine learning techniques. J Forecast 34(7):560–573

    Article  Google Scholar 

  23. Stock JH, Watson MW (2002a) Forecasting using principal components from a large number of predictors. J Am Stat Assoc 97:1167–1179

    Article  Google Scholar 

  24. Stock JH, Watson MW (2002b) Macroeconomic forecasting using diffusion indexes. J Bus Econ Stat 20(2):147–62

    Article  Google Scholar 

  25. Stock JH, Watson MW (2004) Combination forecasts of output growth in a seven-country data set. J Forecast 23(6):405–430

    Article  Google Scholar 

  26. Tibshirani R (1996) Regression shrinkage and selection via the lasso. J R Stat Soc Ser B (Methodol) 58(1):267–288

    Google Scholar 

  27. Wohlrabe K, Buchen T (2014) Assessing the macroeconomic forecasting performance of boosting: evidence for the United States, the Euro area and Germany. J Forecast 33(4):231–242

    Article  Google Scholar 

  28. Zou H, Hastie T (2005) Regularization and variable selection via the elastic net. J R Stat Soc Ser B Stat Methodol 67(2):301–320

    Article  Google Scholar 

Download references

Acknowledgements

We gratefully acknowledge the editors Robert M. Kunst and Martin Wagner, as well as two anonymous referees, for providing helpful comments. Moreover, the authors wish to thank the participants at the 1st Vienna Workshop on Economic Forecasting 2018, the BigNOMICS Workshop on Big Data and Economic Forecasting 2019 and the European Conference on Quality in Official Statistics 2018. We are also grateful for the valuable help provided by Faiz Alsuhail, Satu Elho, Samu Hakala, Ville Vertanen, Reetta Moilanen, Markku Kotilainen, Mika Maliranta, Markku Lehmus and Antti Ripatti.

Author information

Affiliations

Authors

Corresponding author

Correspondence to Paolo Fornaro.

Ethics declarations

Conflict of interest

The authors declare that they have no conflict of interest.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

This research is part of the ESSnet Big Data project of Eurostat.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary material 1 (pdf 221 KB)

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Fornaro, P., Luomaranta, H. Nowcasting Finnish real economic activity: a machine learning approach. Empir Econ 58, 55–71 (2020). https://doi.org/10.1007/s00181-019-01809-y

Download citation

Keywords

  • Flash estimates
  • Machine learning
  • Microlevel data
  • Nowcasting

JEL Classification

  • C33
  • C55
  • E37