FULL PDF
Abstract
Ocean freight rates (hereafter referred to as ocean rates) have seen unprecedented growth (over 150-175% increase) and volatility in recent years due to many factors, including energy prices, global supply chain logistics and transportation challenges. In this paper, we use machine learning methods (regularized regression and support vector machine regression) to predict ocean rates using daily data between January 2015 and May 2022. The models include global supply chain pressure index, Baltic Dry Exchange Index, Brent crude oil prices, time charter rates, total bulker sales, commodity price, and several other global trade indices as features to predict ocean rates. For model selection, evaluation, and improving accuracy, we employed time series cross validation as well as hyperparameter tuning. Predictive accuracy results of ocean rates will help trading firms in their risk management strategies and strategic decisions.
View more >>