The role of fluctuating modes of autocorrelation in crude oil prices

Autocorrelation exists in the crude oil price due to price inertia, the cobweb theorem, model errors, etc. Many researchers have studied the fluctuation of the crude oil price, but few have focused on the autocorrelation fluctuation in crude oil prices. Exploring the fluctuating rules of autocorrelation can aid in understanding the fluctuating mechanism of crude oil prices. To study the role of fluctuating modes of autocorrelation in crude oil prices, which have time series characteristics, this study selected international crude oil spot prices as sample data to employ the methods of statistical physics. The fluctuating modes of autocorrelation were defined by the autocorrelation coefficient, symbolization, and a coarse-graining process. We set the modes as nodes and the transformation between modes as edges; the fluctuating mode weight network of autocorrelation was then built. Thus, the study of autocorrelation fluctuation was transformed to a network study. Then, certain aspects, such as the statistical properties, the “small-world” behavior, and the transmission medium in the network, could be analyzed using complex network theory and analytical methods. The periodicity of the fluctuation was calculated using a spectral analysis method. This study not only describes the fluctuation of the time series more precisely than other methods but also provides ideas for methods of studying the fluctuation of univariate autocorrelations. •We define the fluctuating modes of autocorrelation of time series.•We transform the fluctuating modes of time series into complex network.•The node strength of the fluctuating modes follows a power-law distribution.•Some fluctuating modes play the role of a transmission medium.•The fluctuation of autocorrelation in oil price time series has periodic feature.