Journals / Turkish Journal of Maritime and Marine Sciences / 2026 / Volume: 12 Issue: 3

Forecasting the Istanbul Freight Index: A Comparative Analysis of SARIMA, XGBoost, and Facebook Prophet Across Multiple Time Horizons

İstanbul Navlun Endeksinin Tahminlenmesi: Çoklu Zaman Aralıklarında SARIMA, XGBoost ve Facebook Prophet Modellerinin Karşılaştırmalı Analizi

Cemile Solak Fışkın(ORDU UNIVERSITY, FATSA FACULTY OF MARINE SCIENCES)0000-0003-3358-0673
Pages
—
Publication Date
2026-09-01
🏛️Bibliographic metadata and full text access for this article are provided via DergiPark.
DergiPark Record ↗
✨

AI Summary & Key Takeaways

AI Insight
⚡ TL;DR(Too Long; Didn't Read)

The results demonstrate that XGBoost achieves superior short-term accuracy with a MAPE of 5.47% at the 1-month horizon, maintaining near-reasonable performance at longer horizons.

📌 3 Key Takeaways

🎯
Objective & Hypothesis:Accurate freight rate forecasting is of critical importance for maritime stakeholders in making informed decisions amid volatile market conditions.
📊
Key Findings & Data:This study evaluates the performance of three forecasting models, Seasonal Autoregressive Integrated Moving Average (SARIMA), Facebook Prophet, and Extreme Gradient Boosting (XGBoost), in predicting the Istanbul Freight Index (ISTFIX), a key benchmark for small coaster and mini-bulk carrier markets across Black Sea and Mediterranean routes.
💡
Conclusion & Impact:The results demonstrate that XGBoost achieves superior short-term accuracy with a MAPE of 5.47% at the 1-month horizon, maintaining near-reasonable performance at longer horizons.
🔬Design:Modeling & Simulation
📍Scope:2008 Verileri
⚙️Variables:Time series forecasting · Facebook Prophet · XGBoost

Abstract

Accurate freight rate forecasting is of critical importance for maritime stakeholders in making informed decisions amid volatile market conditions. This study evaluates the performance of three forecasting models, Seasonal Autoregressive Integrated Moving Average (SARIMA), Facebook Prophet, and Extreme Gradient Boosting (XGBoost), in predicting the Istanbul Freight Index (ISTFIX), a key benchmark for small coaster and mini-bulk carrier markets across Black Sea and Mediterranean routes. Weekly ISTFIX data covering the period from January 7, 2008 to March 17, 2025 are employed in the analysis. A rolling-window split strategy is applied to test forecast accuracy across four time horizons: 1-month, 3-month, 6-month, and 1-year. Model performance is evaluated using MAPE, MAE, and RMSE. The results demonstrate that XGBoost achieves superior short-term accuracy with a MAPE of 5.47% at the 1-month horizon, maintaining near-reasonable performance at longer horizons. SARIMA performs consistently across all horizons with MAPE values ranging from 7.16% to 26.63%. Prophet, by contrast, fails to capture the memory-dependent and volatile nature of ISTFIX, recording a MAPE of 50.09% even at the 1-month horizon. The findings demonstrate that while traditional trend-based models struggle with the extreme volatility of freight markets, tree-based machine learning algorithms excel at capturing complex, non-linear price movements.

Özet

Doğru navlun oranı tahmini, denizcilik paydaşlarının oynak piyasa koşullarında bilinçli kararlar alabilmesi açısından kritik öneme sahiptir. Bu çalışma, İstanbul Navlun Endeksi'nin (ISTFIX) tahminlenmesinde üç farklı modelin, Mevsimsel Otoregresif Bütünleşik Hareketli Ortalama (SARIMA), Facebook Prophet ve XGBoost, performansını karşılaştırmalı olarak değerlendirmektedir. ISTFIX, Karadeniz ve Akdeniz güzergâhlarındaki küçük tonajlı koster ve mini dökme yük gemisi pazarlarının temel göstergesi niteliğindedir. Analizde, 7 Ocak 2008 ile 17 Mart 2025 tarihleri arasını kapsayan haftalık ISTFIX verileri kullanılmaktadır. Model performansı; 1 ay, 3 ay, 6 ay ve 1 yıllık olmak üzere dört farklı tahmin ufkunda kayan pencere yöntemiyle test edilmiş; MAPE, MAE ve RMSE ölçütleriyle değerlendirilmiştir. Bulgular, XGBoost'un 1 aylık aralıkta %5.47 MAPE değeriyle en yüksek kısa vadeli doğruluğa ulaştığını ortaya koymaktadır. SARIMA, tüm tahmin aralıklarında %7,16 ile %26,63 arasında tutarlı bir performans sergilemiştir. Prophet ise hafıza bağımlılığı ve yüksek oynaklık gösteren ISTFIX yapısını yakalayamamış; 1 aylık aralıkta bile %50,09 MAPE değeriyle yetersiz kalmıştır. Elde edilen bulgular, navlun fiyatlarındaki yüksek dalgalanmaların Prophet gibi trend odaklı modellerle tahmin edilmesinin zorluğunu göstermektedir. Buna karşılık, XGBoost gibi ağaç tabanlı (tree-based) makine öğrenmesi yöntemlerinin, piyasanın karmaşık ve geçmişe bağımlı yapısını yakalamada çok daha başarılı olduğu sonucuna varılmıştır.

🔖

Cite This Article

Fışkın, C. S., Akgül, G. (2026). Forecasting the Istanbul Freight Index: A Comparative Analysis of SARIMA, XGBoost, and Facebook Prophet Across Multiple Time Horizons. Turkish Journal of Maritime and Marine Sciences, Volume: 12 Issue: 3. https://doi.org/10.52998/trjmms.1911179

📄 Full Text PDF Viewer

📄

PDF görüntüleyici hazırlanıyor...

References (52)

  1. Alizadeh, A.H., Ådland, R.O., Koekebakker, S. (2007). Predictive power and unbiasedness of implied forward charter rates. Journal of Forecasting, 26(6), 385-403. doi: 10.1002/for.1029
  2. Babcock, M.W., Lu, X. (2002). Forecasting inland waterway grain traffic. Transportation Research Part E: Logistics and Transportation Review, 38(1), 65-74. doi: 10.1016/S1366-5545(01)00017-5
  3. Batchelor, R., Alizadeh, A., Visvikis, I. (2007). Forecasting spot and forward prices in the international freight market. International Journal of Forecasting, 23(1), 101–114.
  4. Bao, J., Pan, L., Xie, Y. (2016). A new BDI forecasting model based on support vector machine. In 2016 IEEE Information Technology, Networking, Electronic and Automation Control Conference (pp. 65-69). IEEE.
  5. Bulut, E., Duru, O., Yoshida, S. (2012). A fuzzy integrated logical forecasting (FILF) model of time charter rates in dry bulk shipping: A vector autoregressive design of fuzzy time series with fuzzy c-means clustering. Maritime Economics & Logistics, 14(3), 300-318.
  6. Chen, S., Meersman, H., Voorde, E. van de. (2012). Forecasting spot rates at main routes in the dry bulk market. Maritime Economics & Logistics, 14(4), 498-537. doi: 10.1057/mel.2012.18
  7. Chen, T., Guestrin, C. (2016). XGBoost: A Scalable Tree Boosting System. Cornell University.
  8. Claessens, E.M., Kamel, M., Shanker, A.K. (1984). Traffic and trade projections on the eastbound Suez shipping route. Maritime Policy & Management, 11(1), 45-55. doi: 10.1080/03088838400000033
  9. Cullinane, K. (1992). A short-term adaptive forecasting model for BIFFEX speculation: a Box-Jenkins approach. Maritime Policy & Management, 19(2), 91-114. doi: 10.1080/03088839200000018
  10. Cullinane, K.P.B., Mason, K.J., Cape, M. (1999). A Comparison of Models for Forecasting the Baltic Freight Index: Box-Jenkins Revisited. International Journal of Maritime Economics, 1(2), 15-39. doi: 10.1057/ijme.1999.10
  11. de Gooijer, J.G., Klein, A. (1989). Forecasting the Antwerp maritime steel traffic flow: A case study. Journal of Forecasting, 8(4), 381-398. doi: 10.1002/for.3980080404
  12. de Langen, P.W. (2003). Forecasting Container Throughput: A Method and Implications for Port Planning. Journal of International Logistics and Trade, 1(1), 29-39. doi: 10.24006/jilt.2003.1.1.29
  13. Dinwoodie, J., Landamore, M., Rigot-Muller, P. (2014). Dry bulk shipping flows to 2050: Delphi perceptions of early career specialists. Technological Forecasting and Social Change, 88, 64-75. doi: 10.1016/j.techfore.2014.06.010
  14. Duru, O., Bulut, E., Yoshida, S. (2012). A fuzzy extended DELPHI method for adjustment of statistical time series prediction: An empirical study on dry bulk freight market case. Expert Systems with Applications, 39(1), 840-848. doi: j.eswa.2011.07.082
  15. Duru, O., Yoshida, S. (2009). Judgmental Forecasting in the Dry Bulk Shipping Business: Statistical vs. Judgmental Approach. The Asian Journal of Shipping and Logistics, 25(2), 189-217. doi: 10.1016/S2092-5212(09)80002-3
  16. Fan, Y.H., Xing, Y.W., Yang, H.L. (2014). Prediction of Baltic Capesize Freight Index Based on GARCH Model. Applied Mechanics and Materials, 488-489, 1494-1497. doi: 10.4028/www.scientific.net/AMM.488-489.1494
  17. Goulielmos, A.M., Psifia, M.-E. (2009). Forecasting weekly freight rates for one-year time charter 65 000 dwt bulk carrier, 1989-2008, using nonlinear methods. Maritime Policy & Management, 36(5), 411-436. doi: 10.1080/03088830903187150
  18. Guo, H., Kuang, H., Sui, C., Wang, L. (2025). On predicting ocean freight rates: a novel hybrid model of combined error evaluation and reinforcement learning. Maritime Economics & Logistics, 27(2), 350-372.
  19. Han, Q., Yan, B., Ning, G., Yu, B. (2014). Forecasting Dry Bulk Freight Index with Improved SVM. Mathematical Problems in Engineering, 2014(1). doi: 10.1155/2014/460684
  20. Huang, J., Chu, C.W., Hsu, H.L. (2026). A Comparative Study of Univariate Models for Baltic Dry Index Forecasting. Forecasting, 8(1), 11.
  21. Hyndman, R.J., Athanasopoulos, G. (2018). Forecasting: principles and practice (2nd ed.). OTexts.
  22. Hyndman, R.J., Koehler, A.B. (2006). Another look at measures of forecast accuracy. International Journal of Forecasting, 22(4), 679-688. doi: 10.1016/j.ijforecast.2006.03.001
  23. Kavussanos, M.G., Alizadeh-M, A.H. (2002). Seasonality patterns in tanker spot freight rate markets. Economic Modelling, 19(5), 747-782. doi: 10.1016/S0264-9993(01)00078-5
  24. Kavussanos, M.G., Nomikos, N.K. (2003). Price Discovery, Causality and Forecasting in the Freight Futures Market. Review of Derivatives Research, 6(3), 203-230. doi: 10.1023/B:REDR.0000004824.99648.73
  25. Kim, N., Cha, J., Jeon, J. (2025). A comparative evaluation of machine learning approaches for container freight rates prediction. The Asian Journal of Shipping and Logistics, 41(2), 99-109. doi: 10.1016/j.ajsl.2025.05.001
  26. Kjeldsberg, F., Munim, Z.H., Schramm, H.J. (2026). Machine learning in freight rate forecasting: a systematic literature review. Maritime Economics & Logistics, 1-31.
  27. Klein, A. (1996). Forecasting the Antwerp maritime traffic flows using transformations and intervention models. Journal of Forecasting, 15(5), 395-412. doi: 10.1002/(SICI)1099-131X(199609)15:5 3.0.CO;2-7
  28. Leonov, Y., Nikolov, V. (2012). A wavelet and neural network model for the prediction of dry bulk shipping indices. Maritime Economics & Logistics, 14(3), 319-333. doi: 10.1057/mel.2012.10
  29. Lewis, C.D. (1982). “Industrial and Business Forecasting Method”. Butterworth Scientific Publishers: London.
  30. Li, J., Parsons, M.G. (1997). Forecasting tanker freight rate using neural networks. Maritime Policy & Management, 24(1), 9-30. doi: 10.1080/03088839700000053
  31. Li, M.-W., Han, D.-F., Wang, W. (2015). Vessel traffic flow forecasting by RSVR with chaotic cloud simulated annealing genetic algorithm and KPCA. Neurocomputing, 157, 243-255. doi: 10.1016/j.neucom.2015.01.010
  32. Lyk-Jensen, S.V. (2011). Forecasting Freight Flows. Transport Reviews, 31(5), 603-624. doi: 10.1080/01441647.2011.566984
  33. Lyridis, D.V., Zacharioudakis, P., Mitrou, P., Mylonas, A. (2004). Forecasting Tanker Market Using Artificial Neural Networks. Maritime Economics & Logistics, 6(2), 93-108. doi: 10.1057/palgrave.mel.9100097
  34. Mo, H., Sun, H., Liu, J., Wei, S. (2019). Developing window behavior models for residential buildings using XGBoost algorithm. Energy and Buildings, 205, 109564. doi: 10.1016/j.enbuild.2019.109564
  35. Mostafa, M.M. (2004). Forecasting the Suez Canal traffic: a neural network analysis. Maritime Policy & Management, 31(2), 139-156. doi: 10.1080/0308883032000174463
  36. Munim, Z.H., Fiskin, C.S., Nepal, B., Chowdhury, M.M.H. (2023). Forecasting container throughput of major Asian ports using the Prophet and hybrid time series models. The Asian Journal of Shipping and Logistics, 39(2), 67-77. doi: 10.1016/j.ajsl.2023.02.004
  37. Munim, Z.H., Schramm, H.-J. (2017). Forecasting container shipping freight rates for the Far East - Northern Europe trade lane. Maritime Economics & Logistics, 19(1), 106-125. doi: 10.1057/s41278-016-0051-7
  38. Munim, Z.H., Schramm, H.-J. (2021). Forecasting container freight rates for major trade routes: a comparison of artificial neural networks and conventional models. Maritime Economics & Logistics, 23(2), 310-327. doi: 10.1057/s41278-020-00156-5
  39. Nielsen, P., Jiang, L., Rytter, N.G.M., Chen, G. (2014). An investigation of forecast horizon and observation fit’s influence on an econometric rate forecast model in the liner shipping industry. Maritime Policy & Management, 41(7), 667-682. doi: 10.1080/03088839.2014.960499
  40. Niu, M., Hu, Y., Sun, S., Liu, Y. (2018). A novel hybrid decomposition-ensemble model based on VMD and HGWO for container throughput forecasting. Applied Mathematical Modelling, 57, 163-178. doi: 10.1016/j.apm.2018.01.014
  41. Randers, J., Göluke, U. (2007). Forecasting turning points in shipping freight rates: lessons from 30 years of practical effort. System Dynamics Review, 23(2-3), 253-284. doi: 10.1002/sdr.376
  42. Santos, A.A.P., Junkes, L.N., Pires Jr, F.C.M. (2014). Forecasting period charter rates of VLCC tankers through neural networks: A comparison of alternative approaches. Maritime Economics & Logistics, 16(1), 72-91. doi: 10.1057/mel.2013.20
  43. Solak Fışkın, C., Akgül, E.F. (2022). Understanding the Influencers of Freight Rate Forecasting Accuracy: A Meta-Regression Analysis of the Literature. Journal of ETA Maritime Science, 10(1), 86-95. doi: 10.4274/jems.2022.87894
  44. Tavasszy, L., Minderhoud, M., Perrin, J.-F., Notteboom, T. (2011). A strategic network choice model for global container flows: specification, estimation and application. Journal of Transport Geography, 19(6), 1163-1172. doi: 10.1016/j.jtrangeo.2011.05.005
  45. Taylor, S.J., Letham, B. (2018). Forecasting at Scale. The American Statistician, 72(1), 37-45. doi: 0.1080/00031305.2017.1380080
  46. Tsioumas, V., Papadimitriou, S., Smirlis, Y., Zahran, S.Z. (2017). A Novel Approach to Forecasting the Bulk Freight Market. The Asian Journal of Shipping and Logistics, 33(1), 33-41. doi: 10.1016/j.ajsl.2017.03.005
  47. Veenstra, A.W., Franses, P.H. (1997). A co-integration approach to forecasting freight rates in the dry bulk shipping sector. Transportation Research Part A: Policy and Practice, 31(6), 447-458. doi: 10.1016/S0965-8564(97)00002-5
  48. Veenstra, A.W., Haralambides, H.E. (2001). Multivariate autoregressive models for forecasting seaborne trade flows. Transportation Research Part E: Logistics and Transportation Review, 37(4), 311-319. https://doi.org/10.1016/S1366-5545(00)00020-X
  49. Wang, W., He, N., Chen, M., Jia, P. (2024). Freight rate index forecasting with Prophet model based on multi-dimensional significant events. Expert Systems with Applications, 249, 123451. doi: 10.1016/j.eswa.2024.123451
  50. Winston, C. (1981). A Disaggregate Model of the Demand for Intercity Freight Transportation. Econometrica, 49(4), 981. https://doi.org/10.2307/1912514
  51. Zeng, Q., Qu, C., Ng, A.K.Y., Zhao, X. (2016). A new approach for Baltic Dry Index forecasting based on empirical mode decomposition and neural networks. Maritime Economics & Logistics, 18(2), 192-210. doi: 10.1057/mel.2015.2
  52. Zhang, X., Chen, M.Y., Wang, M.G., Ge, Y.E., Stanley, H.E. (2019). A novel hybrid approach to Baltic Dry Index forecasting based on a combined dynamic fluctuation network and artificial intelligence method. Applied Mathematics and Computation, 361, 499-516. doi: 10.1016/j.amc.2019.05.043