Abstract
The fast-paced and ever changing freight market compels maritime executives to use sound forecasting tools. This paper aims to enhance the forecasting accuracy of the Baltic Dry Index (BDI) by means of developing a multivariate Vector Autoregressive model with exogenous variables (VARX). The proposed model incorporates the Chinese steel production, the dry bulk fleet development and a new composite indicator, the Dry Bulk Economic Climate Index (DBECI). The predictive power of this approach is evaluated against a univariate ARIMA framework, which serves as a benchmark model. The selection of explanatory variables and the model specification are validated using a series of pertinent tests. The results demonstrate that the VARX model outperforms the ARIMA approach, suggesting that the selected independent variables can substantially improve the accuracy of BDI forecasts. The present study is of interest to maritime practitioners, as it provides useful insights into the direction of the freight market and allows them to make informed decisions.