Fluctuations in the crude oil price allied to risk have increased significantly over the last decade frequently varying at di↵erent risk levels. Although existing models partially predict such variations, so far, they have been unable to predict oil prices accurately in this highly volatile market. The development of an e↵ective, predictive model has therefore become a prime objective of research in this field. Our approach, albeit based in part on previous research, develops an original methodology, in that we have created a risk forecasting model with the ability to predict oil price fluctuations caused by changes in both fundamental and transient risk factors. We achieve this by disintegrating the multi-scale risk-structure of the crude oil market using Variational Mode Decomposition. Normal and transient risk factors are then extracted from the crude oil price using Variational Mode Decomposition and modelled separately using the Quantile Regression Neural Network (QRNN) model. Both risk factors are integrated and ensembled to produce the risk estimates. We then apply our proposed risk forecasting model to predicting future downside risk level in
Abstract:For the modeling of complex and nonlinear crude oil price dynamics and movement, wavelet analysis can decompose the time series and produce multiple economically meaningful decomposition structures based on different assumptions of wavelet families and decomposition scale. However, the determination of the optimal model specification will critically affect the forecasting accuracy. In this paper, we propose a new wavelet entropy based approach to identify the optimal model specification and construct the effective wavelet entropy based forecasting models. The wavelet entropy algorithm is introduced to determine the optimal wavelet families and decomposition scale, that will produce the improved forecasting performance. Empirical studies conducted in the crude oil markets show that the proposed algorithm outperforms the benchmark model, in terms of conventional performance evaluation criteria for the model forecasting accuracy.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.