Comparing grey system and ARIMA models in forecasting baltic exchange dry index

Bi-Huei Tsai*, Chi Chen Wang, Chun Hsien Chang

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

1 Scopus citations


The freight rate is generally affected by the uncertainty in bulk shipping market. Since Baltic Dry Index (BDI) is a combined index of the freight rate, BDI trend precisely reflects bulk shipping market variations. The purpose of this paper is to forecast the BDI, helping the bulk shipping companies to reduce risks. This work applies Grey- Markov model, GM(1,1), ARIMA model, and ARIMA model considering the structural change to predict BDI, while comparing prediction capability among the four models. The empirical results show that over all the Grey-Markov model performs superior to the other models in BDI prediction, suggesting that the Grey-Markov model specializes in making predictions with incomplete and uncertain raw data.

Original languageEnglish
Pages (from-to)693-698
Number of pages6
JournalICIC Express Letters
Issue number3 A
StatePublished - 28 Mar 2013


  • Baltic Dry Index
  • Chow test
  • GM(1,1)
  • Grey-Markov model

Fingerprint Dive into the research topics of 'Comparing grey system and ARIMA models in forecasting baltic exchange dry index'. Together they form a unique fingerprint.

Cite this