Modelling the volatility of the tanker freight market based on improved empirical mode decomposition

Abstract

ABSTRACTIn this article, a method based on empirical mode decomposition (EMD) and BEKK–Multivariate GARCH (MVGARCH) is developed to analyse the volatility of the tanker freight market. First, the initial time series of tanker freight rates is decomposed into several independent intrinsic mode functions (IMFs). Next, the IMFs are composed as three components by an improved EMD: the long-term trend time series that represents the benchmark freight rates of the tanker freight market, the high-frequency time series that reflects the short-term supply–demand relation and the low-frequency time series caused by extreme events. Based on the results of EMD, the volatility spillover effects between the freight rates of Aframax, Suezmax and Very Large Crude Carrier (VLCC) markets are tested by the BEKK–MVGARCH model. The results indicate that there are volatility spillover effects between the reconstructed components, although the volatility spillover effects between the original freight series are not significant. The improved EMD method contributes to the retention of the economic characteristics of the original time series, thereby providing a vital approach for tanker freight market analysis. Furthermore, the potential volatility spillover among different sub-markets can be investigated through the integration of EMD and the BEKK–MVGARCH.