This paper uses higher-order derivatives of the VIX and MOVE indices, more especially, their velocity, acceleration, and jerk, as indicators of sentiment momentum to examine the predictive ability of dynamic financial fear sentiment on crude oil returns. To understand the system-dependent nature of oil market dynamics, we examine the impact of sentiment derivatives on crude oil returns under stable and turbulent conditions using the Markov Switching Regression (MSR) framework. Our empirical study, which uses daily data from 2002 to 2025, shows that sentiment velocity (VIXV, MOVEV) and acceleration (VIXA, MOVEA) have significant predictive value for oil returns under stable market conditions. However, their predictive effectiveness decreases under high volatility conditions, supporting the information saturation hypothesis. Comparing derivatives with fixed sentiment levels reveals a better continuous model fit and early warning signals for increased volatility. The results show the asymmetric and nonlinear flow of fear from financial markets to commodity markets, highlighting the need to incorporate mood momentum into forecasting models. Robustness checks, incorporating out-of-sample prediction tests, macro-financial control variables (DXY, GPR), and Fourier-based seasonal components, confirm the stability of these results. The sentiment momentum remains the dominant driver of regime-dependent return dynamics, with seasonal and cyclical patterns contributing little explanatory power. Compared with stationary sentiment levels, derivative-based measures consistently deliver superior model fit and earlier signals of transitions into high-volatility regimes. The evidence highlights the asymmetric and nonlinear transmission of fear from financial to commodity markets, emphasizing the need to incorporate mood momentum into forecasting frameworks. Policymakers can use sentiment acceleration as a leading indicator to change communication or intervention timing in macroeconomics sensitive to energy consumption; risk managers can include sentiment velocity into real-time monitoring systems to optimise hedging strategies.
Tissaoui et al. (Wed,) studied this question.