Multifractal Analysis and Machine Learning-Based Predictive Modeling of the Tanker Freight Market: Unraveling Complexity Across Four Breaks

This study investigates the tanker freight market and its complex transformation from 1998 to 2024, uncovering a high degree of multifractality and a complex market structure shaped by temporal correlations and inherent volatility. Using multifractal dynamics analyzed through multifractal detrended fluctuation analysis (MF-DFA) and multifractal detrending moving average (MF-DMA), we explore how key external factors drive market complexity, including economic disturbances (the 2008 Financial Crisis), technological innovations (the 2014 Shale Oil Revolution), supply chain disruptions (the COVID-19 pandemic), and geopolitical uncertainties (the Russia-Ukraine conflict). Building on this, a predictive framework is introduced, leveraging the Baltic Dirty Tanker Index (BDTI) to forecast Brent oil prices. By integrating multifractal analysis with machine learning models, such as XGBoost, LightGBM, and CatBoost, the framework captures complexity transformation across these four major global events. Results demonstrate the potential of combining multifractal analysis with advanced machine learning models to improve forecasting accuracy and provide actionable insights during periods of heightened market volatility.