Short-term forecasting of budget tax revenue: methods used and their relevance for Russia

The article identifies relevant methods of short-term forecasting of regional and local budget revenues to improve the accuracy of forecasts. The methodological basis of the study is the analytical review of scientific publications on the methods of forecasting budget revenues, identification of their distinctive features, limitations of their application, as well as justification of the choice of the method that provides high accuracy of forecasting tax revenues of subnational budgets. The authors argue that the accuracy of revenue forecasts is the most important component of economic government planning as it enables the development of effective fiscal policy, resource allocation and strategic financial management. Given that achieving an accurate forecast is a complex task, the paper shows that one of the main sources of error in revenue forecasting is the choice of revenue forecasting method, in addition to a variety of political and institutional factors affecting the accuracy of forecasts. Drawing on comparative analysis of existing forecasting methods, the authors select two methods: time series forecasting method (traditional SARMA/SARIMA method) and forecasting method based on discrete wavelet transform. They are tested on the data of aggregated monthly tax revenues of consolidated budgets of the subjects of the Russian Federation for the period from January 2011 to March 2023. It has been revealed that the forecasting method based on wavelet transformations is superior to the traditional SARMA/SARIMA method by all indicators and allows to achieve a higher level of accuracy of forecasts of monthly aggregated tax revenues of consolidated budgets of Russian regions. The findings demonstrate a high forecasting potential of shortterm forecasting methods with a preliminary decomposition of time series based on wavelet transformations. The obtained results make it possible to improve the accuracy of classical forecasting methods and thus contribute to the growth of efficiency and effectiveness of budget planning and forecasting.

1.

2. Petrosyan, G.A., Karapetyan, N.N., Margaryan, A.A., Sokolov, A.N., Yakovleva I.I., & Votinov A.I. (2024). Bayesian Approach to Forecasting Aggregate Taxes of the Republic of Armenia. Financial Journal, 16(3), 51-67. https://doi.org/10.31107/2075-1990-2024-3-5167.

3. Federal Treasury (n.d.). Consolidated budgets of the constituent entities of the Russian Federation and budgets of territorial state extra-budgetary funds. Retrieved September 26, 2024, from https://roskazna.gov.ru/ispolnenie-byudzhetov/konsolidirovannye-byudzhetysubektov/»

4. Atwood, B., Haddad, C., Knaust, H., & Merkel, J. (2012). Using Wavelets-Based Time Series Forecasting to Predict Oil Prices. Wavelets in Undergraduate Education. http://math.mscd.edu/metadot/index.pl?iid=2553

5. Babii, A., Ghysels, E., & Striaukas, J. (2021). Machine learning time series regressions with an application to nowcasting. Journal of Business and Economic Statistics, 40(3), 1094–1106. https://doi.org/10.1080/07350015.2021.1899933

6. Barnard, J. R., & Dent, W. T. (1979). State tax revenues-new methods of forecasting. The Annals of Regional Science, 13(3), 1–14. https://doi.org/10.1007/bf01287742

7. Bok, B., Caratelli, D., Giannone, D., Sbordone, A. M., & Tambalotti, A. (2018). Macroeconomic Nowcasting and Forecasting with Big Data. Annual Review of Economics, 10(1), 615–643. https://doi.org/10.1146/annurev-economics-080217-053214

8. Chan-Lau, J. A. (2017). Lasso regressions and forecasting models in applied stress testing. SSRN Electronic Journal. https://doi.org/10.2139/ssrn.3053191

9. Chi, D. (2022). Research on electricity consumption forecasting model based on wavelet transform and multi-layer LSTM model. Energy Reports, 8, 220–228. https://doi.org/10.1016/j.egyr.2022.01.169

10. Chicco, D., Warrens, M. J., & Jurman, G. (2021). The coeffi cient of determination R-squared is more informative than SMAPE, MAE, MAPE, MSE and RMSE in regression analysis evaluation. PeerJ Computer Science, 7, e623. https://doi.org/10.7717/peerj-cs.623

11. Chung, I. H., Williams, D. W., & Rok, M., DO. (2022). For Better or Worse? Revenue Forecasting with Machine Learning Approaches. Public Performance & Management Review, 45(5), 1133–1154. https://doi.org/10.1080/15309576.2022.2073551

12. Cirincione, C., Gurrieri, G. A., & Van De Sande, B. (1999). Municipal Government Revenue Forecasting: Issues of Method and data. Public Budgeting &Amp Finance, 19(1), 26–46. https://doi.org/10.1046/j.0275-1100.1999.01155.x

13. Firdawss, T., & Karim, M. (2018). Forecasting Moroccan Tax Revenues: An analysis using international Institutions Methodologies and VECM. Journal of Economics and Public Finance, 4(4), 304. https://doi.org/10.22158/jepf.v4n4p304

14. Greoning, E., Zivanomoyo, J., & Tsaurai, K. (2019). Determinants of Company Tax Revenue in Swaziland (1990–2015). Acta Universitatis Danubius: Oeconomica, 15(5), 7–37.

15. Grizzle, G. A., & Klay, W. E. (1994). Forecasting State Sales Tax Revenues: Comparing the Accuracy of Diff erent Methods. State and Local Government Review, 26 (3),142–152.

16. Gumbo, V., & Dhliwayo, L. (2018). VAT Revenue Modelling: The Case of Zimbabwe. Retrieved September 19, 2024, from https://www.researchgate.net/publication/326517998_VAT_revenue_modelling_the_case_of_zimbabwe

17. Karaev, A. K., Gorlova, O. S., Ponkratov, V. V., Sedova, M. L., Shmigol, N. S., & Vasyunina, M. L. (2022). A comparative analysis of the choice of mother wavelet functions aff ecting the accuracy of forecasts of daily balances in the Treasury single account. Economies, 10(9), 213. https://doi.org/10.3390/economies10090213

18. Krol, R. (2010). Forecasting State Tax Revenue: A Bayesian Vector Autoregression Approach. Retrieved August 30, 2024, from https://www.csun.edu/~hcecn001/published/BVAR_Forecast.pdf

19. Lahiri, K. & Yang, C. (2022). Boosting tax revenues with mixed-frequency data in the aftermath of COVID-19: The case of New York. International Journal of Forecasting, 38(2), 545–566. https://doi.org/10.1016/j.ijforecast.2021.10.005

20. Makananisa, M. P. (2015). Forecasting annual tax revenue of the South African taxes using time series Holt-Winters and ARIMA/SARIMA Models. https://uir.unisa.ac.za/bitstream/10500/19903/1/dissertation_makananisa_mp.pdf

21. Molapo, M. A., Olaomi, J. O., & Ama, N. O. (2019). Bayesian Vector Auto-Regression Method as an alternative technique for forecasting South African tax revenue. Southern African Business Review, 23. https://doi.org/10.25159/1998-8125/4416

22. Nandi, B. K., Chaudhury, M., & Hasan, G. Q. (2015). Univariate Time Series Forecasting: A study on monthly tax revenue of Bangladesh. East West Journal of Business and Social Studies, 4, 1–28. https://doi.org/10.70527/ewjbss.v4i.113

23. Noor, N., Sarlan, A., & Aziz, N. (2022). Revenue Prediction for Malaysian Federal Government Using Machine Learning Technique. (р. 143–148). https://doi.org/10.1145/3524304.3524337

24. Ofori, M. S., Fumey, A., & Nketiah-Amponsah, E. (2020). Forecasting Value Added Tax Revenue in Ghana. Journal of Economics and Financial Analysis, 4(2), 63–99. https://doi.org/10.1991/jefa.v4i2.a37

25. Qasim, M., & Khalid, M. (2016). Accuracy of revenue forecast: Analysis of Pakistan’s federal revenue receipts. Forman Journal of Economic Studies, 12, 41–56. https://doi.org/10.32368/fjes.20161203

26. Reed, D. A. (1983). A Simultaneous Equation Tax Revenue Forecasting Model for the State of indiana. Indiana: University. Retrieved August 30, 2024, from https://file.pide.org.pk/pdfpideresearch/wp-23-12-federal-tax-revenue-forecasting-of-pakistan-alternativeapproaches.pdf

27. Shaikh, W. A., Shah, S. F., Pandhiani, S. M., & Solangi, M. A. (2022). Wavelet Decomposition Impacts on Traditional Forecasting Time Series Models. Computer Modeling in Engineering & Sciences, 130(3), 1517–1532. https://doi.org/10.32604/cmes.2022.017822

28. Simonov, J., & Gligorov, Z. (2021). Customs Revenues Prediction Using Ensemble Methods (Statistical Modelling vs Machine Learning). World Customs Journal, 15(2). https://doi.org/10.55596/001c.116452

29. Streimikiene, D., Ahmed, R. R., Vveinhardt, J., Ghauri, S. P., & Zahid, S. (2018). Forecasting tax revenues using time series techniques — a case of Pakistan. Economic Research-Ekonomska Istraživanja, 31(1), 722–754. https://doi.org/10.1080/1331677x.2018.1442236

30. Sаyeda, U. B. (2023). Federal Tax Revenue Forecasting of Pakistan: Alternative Approaches. PIDE-Working Papers 12, Pakistan Institute of Development Economics. Retrieved July 6, 2024, from https://ideas.repec.org/p/pid/wpaper/202312.html

31. Ticona, W., Figueiredo, K., & Vellasco, M. (2017). Hybrid model based on genetic algorithms and neural networks to forecast tax collection: Application using endogenous and exogenous variables (p. 1–4). https://doi.org/10.1109/intercon.2017.8079660

32. Urrutia, J. D., Mingo, F. L. T., & Balmaceda, C. N. M. (2015). Forecasting income tax revenue of the philippines using autoregressive integrated moving average (ARIMA) Modeling: A time series analysis. [Dataset]. Figshare. https://doi.org/10.6084/m9.figshare.1469539

33. Waciko, K. J., & Ismail, B. (2020). Performance of Shrinkage Methods for Forecasting GDP. International Journal of Advanced Science and Technologie, 29(5), 7792–7799.

34. Wolfram Mathematica.Wolfram Language Documentation Center. (n.d.). ArrayPad. Retrieved July 6, 2024, from https://reference.wolfram.com/language/ref/ArrayPad.html?q=ArrayPad

35. Yousefi , S., Weinreich, I., & Reinarz, D. (2005). Wavelet-based prediction of oil prices. Chaos, Solitons and Fractals, 25(2), 265–275. https://doi.org/10.1016/j.chaos.2004.11.015

36. Zhang, X., Kim, D., & Wang, Y. (2016). Jump Variation Estimation with Noisy High Frequency Financial Data via Wavelets. Econometrics, 4(3), 34. https://doi.org/10.3390/econometrics4030034

37. Zhang, Q., Ni, H., & Xu, H. (2023). Nowcasting ChineseGDP in a data-rich environment: Lessons from machine learning algorithms. Economic Modelling, 122, 106204. https://doiorg/10.1016/j.econmod.2023.106204