Time series forecasting methods for the Baltic dry index

Abstract

This paper forecasts the daily Baltic Dry Index (BDI) using time series and machine learning methods. Significant business cycles and freight rate volatility present in the ocean‐going shipping industry make the ability to forecast freight rates and cycles extremely important for business decisions. Data‐driven model selection based on data characteristics is performed through ARIMA, fractional ARIMA (FARIMA), and ARIMA and FARIMA models with GARCH and EGARCH errors. The corresponding machine learning techniques utilized are feed‐forward fully connected artificial neural networks (ANNs), support vector regression (SVR), and multivariate adaptive regression splines (MARS). Among others, FARIMA and MARS models are used for the first time in forecasting the BDI. Diebold–Mariano tests reveal that time series and machine learning approaches have comparable performance. However, combinations of forecasts of the selected models lead to better forecasting accuracy overall. Bai and Perron tests are utilized to check the robustness of the results over different cycles through the detection of breakpoints in the series.

The influence of COVID-19 epidemic on BDI volatility: An evidence from GARCH-MIDAS model

“…The trend of BDI index indeed reflects the market demands for commodities such as metals, minerals, grains and building materials. Since these “BDI materials” are the basic raw materials for manufacturing final products, the index is regarded by many as a leading economic indicator of industrial production and economic activities, laying the foundation for economic decision-making ( Zeng et al, 2015 ; Makridakis et al, 2020 ; Katris and Kavussanos, 2021 ; Chen et al, 2022a , 2022b ).…”

An evolutionary cost‐sensitive support vector machine for carbon price trend forecasting

“…More and more literature on time series forecasting has emerged in recent years. The research approaches used can be roughly divided into two categories: econometric models (Canakoglu et al, 2018;Katris & Kavussanos, 2021;Ren, Zhu, et al, 2022;Sanin et al, 2015) and artificial intelligence models (Chevallier et al, 2020;Han et al, 2019;Sun & Huang, 2020;Zhu et al, 2017Zhu et al, , 2022.…”

“…Although econometric models represented by the autoregressive integrated moving average (ARIMA) and generalized autoregressive conditional heteroscedastic (GARCH) have some advantages in time series forecasting (Katris & Kavussanos, 2021; Sanin et al, 2015), those models require assumptions of stationarity and linearity, which cannot capture non‐stationary and nonlinear characteristics of the carbon price. Compared with the traditional econometric model, artificial intelligence models represented by an artificial neural network (ANN) and support vector machine (SVM) can reveal the complicated patterns hidden in carbon price through adaptive learning of complex non‐linear mapping relationships (Chevallier et al, 2020).…”

Static Hedging of Freight Risk under Model Uncertainty

“…Forecasting of the freight rate prices is a well studied problem indicading several directions in successfully predicting and modelling different phenomena displayed by the freight rates when the stationarity assumption holds (see e.g. (Batchelor, Alizadeh, & Visvikis, 2007;Chen, Meersman, & Voorde, 2012;Katris & Kavussanos, 2021;Munim & Schramm, 2021)). However, there is not just one model that is universally accepted for modeling the random behaviour of the freight rates in general, since different characteristics may be displayed depending the market condition.…”