Türkiye’nin GSYİH öngörüsü: Doğrusal (SARIMA) ve doğrusal olmayan (Holt-Winters) zaman serisi yöntemleriyle bir uygulama

Abstract

Bu çalışma, Türkiye'nin 2005Q1-2025Q2 dönemi GSYİH verilerini öngörmede, doğrusal bir model olan SARIMA ile doğrusal olmayan bir model olan Holt-Winters (Additive) arasındaki performans farkını ampirik olarak incelemiştir. RMSE, MAE, MAPE hata metriklerine göre, Holt-Winters modelinin, SARIMA'ya kıyasla daha düşük hata oranları sunarak daha üstün bir öngörü performansı sergilediği tespit edilmiştir. Bu bulgu, gelişmekte olan ekonomilerdeki yüksek volatilite ve karmaşık dinamiklerin modellenmesinde doğrusal olmayan yöntemlerin potansiyelini vurgulamaktadır. Politika yapıcılar için, kısa vadede olası bir yavaşlamaya karşı proaktif politikalar geliştirilmesi ve uzun vadeli büyümenin sürdürülebilirliği için yapısal reformlara odaklanılması gerektiği sonucuna varılmıştır. Gelecek araştırmalar için, aykırı değerlere daha dirençli "robust" Holt-Winters modellerinin ve çok değişkenli VAR, VECM veya hibrit ARIMA-LSTM yaklaşımların kullanımı önerilmiştir.

Keywords

• GSYİH Öngörüsü
• SARIMA
• Holt-Winters
• Zaman Serisi Analizi

Supporting Institution

yok

Ethical Statement

Etik onay gerektiren birli ve içerik bulunmamaktadır. Çalışmada YOK büyük dil modellerinin kullanımına yönelik yayımlanan etik klavıza uyularak dil modellerinden istifade edilmiştir.

Thanks

Katkı sağlayan tüm paydaşlara şükranlarımı sunarım

Forecasting Turkiye’s GDP: an application of linear (SARIMA) and non-linear (Holt-Winters) time series methods

Abstract

This study empirically examines the performance difference between SARIMA, a linear model, and Holt-Winters (Additive), a nonlinear model, in forecasting Turkey's GDP data for the period 2005Q1-2025Q2. Based on error metrics RMSE, MAE, MAPE, the Holt-Winters model was found to exhibit superior forecasting performance by presenting lower error rates compared to SARIMA. This finding highlights the potential of nonlinear methods in modeling the high volatility and complex dynamics of developing economies. For policymakers, it is concluded that proactive policies should be developed against a potential slowdown in the short term, and structural reforms should be focused on to ensure the sustainability of long-term growth. For future research, the use of "robust" Holt-Winters models, which are more resistant to outliers, and multivariate VAR, VECM or hybrid ARIMA-LSTM approaches are suggested.

Keywords

• GDP Forecasting
• SARIMA
• Holt-Winters
• Time Series Analysis

References

1. Ansari Saleh, A., Pawan Kumar, S., Nguyen Van, T., Nguyen Viet, T., ve Vo Minh, H. (2022). Prediction of BRIC stock price using ARIMA, SutteARIMA, and Holt-Winters. Computers, Materials & Continua, 70(1). Doi: 10.32604/cmc.2022.017068
2. Armstrong, J. S., ve Collopy, F. (1992). Error measures for generalizing about forecasting methods: Empirical comparisons. International Journal of Forecasting, 8(1), 69-80. Doi: 10.1016/0169-2070(92)90008-W
3. Atif, D. (2025). Enhancing long-term GDP forecasting with advanced hybrid models: A comparative study of ARIMA-LSTM and ARIMA-TCN with dense regression. Computational Economics, 65(6), 3447-3473. Doi: 10.1007/s10614-024-10683-5
4. Box, G. E. P., ve Jenkins, G. M. (1976). Time series analysis: Forecasting and control (Rev. ed.). San Francisco: Holden-Day.
5. Box, G. E. P., Jenkins, G. M., Reinsel, G. C., ve Ljung, G. M. (2015). Time series analysis: Forecasting and control (5th ed.). John Wiley & Sons.
6. Costa, K. V. S., da Silva, F. L. C., ve da Silva Cordeiro Coelho, J. (2022). Forecasting quarterly Brazilian GDP: Univariate models approach. arXiv. https://arxiv.org/pdf/2010.13259
7. Dickey, D. A., ve Fuller, W. A. (1979). Distribution of the estimators for autoregressive time series with a unit root. Journal of the American Statistical Association, 74(366), 427-431. Doi: 10.1080/01621459.1979.10482531
8. Dritsaki, M., ve Dritsaki, C. (2021). Forecasting of GDP per capita in five Balkan countries. Modern Economy, 12(12), 1972-1998. Doi: 10.4236/me.2021.1212104

9. Gelper, S., Fried, R., ve Croux, C. (2010). Robust forecasting with exponential and Holt-Winters smoothing. Journal of Forecasting, 29(3), 285-300. Doi: 10.1002/for.1125
10. Granger, C. W. J. (1996). Can we improve the perceived quality of economic forecasts? Journal of Applied Econometrics, 11(5), 455-473. Doi: 10.1002/(SICI)1099-1255(199609)11:5<455::AID-JAE408>3.0.CO;2-E
11. Granger, C. W. J., ve Joyeux, R. (1980). An introduction to long-memory time series models and fractional differencing. Journal of Time Series Analysis, 1(1), 15-29. Doi: 10.1111/j.1467-9892.1980.tb00297.x
12. Hamiane, S., Ghanou, Y., Khalifi, H., ve Telmem, M. (2024). Comparative analysis of LSTM, ARIMA, and hybrid models for forecasting future GDP. Informatica, 29(3), 853-861. Doi: 10.18280/isi.290306
13. Hamilton, J. D. (2020). Time series analysis. Princeton University Press.
14. Holt, C. C. (1957). Forecasting seasonals and trends by exponentially weighted moving averages. Office of Naval Research, Research Memorandum, 52. Carnegie Institute of Technology. (Yeniden basımı: International Journal of Forecasting, 20(1), 5-10, 2004). Doi: 10.1016/j.ijforecast.2003.09.015
15. Hyndman, R. J., ve Athanasopoulos, G. (2021). Forecasting: Principles and practice (3rd ed.). OTexts. https://otexts.com/fpp3/
16. Hyndman, R. J., ve Khandakar, Y. (2008). Automatic time series forecasting: The forecast package for R. Journal of Statistical Software, 27(3), 1-22. Doi: 10.18637/jss.v027.i03
17. Ingale, K. Y., ve Senan, R. (2023). Predictive analysis of GDP by using ARIMA approach. The Pharma Innovation Journal, 12(5), 214-237. https://www.thepharmajournal.com/archives/2023/vol12issue5/PartD/12-4-214-737.pdf
18. Liu, Y. (2022). Prediction for financial time series. Uppsala University. https://www.diva-portal.org/smash/get/diva2:1634907/FULLTEXT01.pdf
19. Ljung, G. M., ve Box, G. E. P. (1978). On a measure of lack of fit in time series models. Biometrika, 65(2), 297-303. Doi: 10.1093/biomet/65.2.297
20. Makridakis, S., Wheelwright, S. C., ve Hyndman, R. J. (1998). Forecasting: Methods and applications (3rd ed.). John Wiley & Sons. http://www.jstor.org/stable/3008101
21. Nwokike, C. C., ve Okereke, E. W. (2021). Comparison of the performance of the SANN, SARIMA and ARIMA models for forecasting quarterly GDP of Nigeria. Asian Research Journal of Mathematics, 17(3), 1-20. Doi: 10.9734/ARJOM/2021/v17i330280
22. Perron, P. (1989). The great crash, the oil price shock, and the unit root hypothesis. Econometrica, 57(6), 1361-1401. Doi: 10.2307/1913712
23. Perron, P. (2006). Dealing with structural breaks. In T. C. Mills ve K. Patterson (Eds.), Palgrave handbook of econometrics (1, 278-352). Palgrave Macmillan.
24. Phillips, P. C. B., ve Perron, P. (1988). Testing for a unit root in time series regression. Biometrika, 75(2), 335-346. Doi: 10.1093/biomet/75.2.335
25. Siami-Namini, S., Tavakoli, N., ve Siami Namin, A. (2019). A comparative analysis of forecasting financial time series using ARIMA, LSTM, and BiLSTM. arXiv. Doi: 10.48550/arXiv.1911.09512
26. Stafiej, J. (2019). NNS: Nonlinear nonparametric statistics and forecasting in R [R package]. Comprehensive R Archive Network (CRAN). https://cran.r-project.org/package=NNS
27. Sumner, S. (2012). The case for nominal GDP targeting. Mercatus Center, George Mason University. https://www.mercatus.org/publications/monetary-policy/case-nominal-gdp-targeting
28. Vinod, H. D. (2008). Hands-on matrix algebra using R. World Scientific Publishing.
29. Winters, P. R. (1960). Forecasting sales by exponentially weighted moving averages. Management Science, 6(3), 324-342. Doi: 10.1287/mnsc.6.3.324
30. Woodford, M. (2012). Methods of policy accommodation at the interest-rate lower bound. Proceedings of the Jackson Hole Symposium. Federal Reserve Bank of Kansas City, 185-288.
31. Xian, X. (2023). Comparison of SARIMA model, Holt-Winters model and ETS model. BMC Infectious Diseases. Doi: 10.1186/s12879-023-08799-4
32. Zivot, E., ve Andrews, D. W. K. (1992). Further evidence on the great crash, the oil-price shock, and the unit-root hypothesis. Journal of Business & Economic Statistics, 10(3), 251-270. Doi: 10.1080/07350015.1992.10509904