Multi-step-ahead crude oil price forecasting using a hybrid grey wave model

By Yanhui Chen, Chuan Zhang, Kaijian He, Aibing Zheng. Physica A: Statistical Mechanics and its Applications, 2018

49 citations recorded by the source corpus.

Abstract Crude oil is crucial to the operation and economic well-being of the modern society. Huge changes of crude oil price always cause panics to the global economy. There are many factors influencing crude oil price. Crude oil price prediction is still a difficult research problem widely discussed among researchers. Based on the researches on Heterogeneous Market Hypothesis and the relationship between crude oil price and macroeconomic factors, exchange market, stock market, this paper proposes a hybrid grey wave forecasting model, which combines Random Walk (RW)/ARMA to forecast multi-step-ahead crude oil price. More specifically, we use grey wave forecasting model to model the periodical characteristics of crude oil price and ARMA/RW to simulate the daily random movements. The innovation also comes from using the information of the time series graph to forecast crude oil price, since grey wave forecasting is a graphical prediction method. The empirical results demonstrate that based on the daily data of crude oil price, the hybrid grey wave forecasting model performs well in 15- to 20-step-ahead prediction and it always dominates ARMA and Random Walk in correct direction prediction.

Subfields: Economics and Econometrics, Management Science and Operations Research

View Paper: https://www.semanticscholar.org/paper/7f0684a998cc8d4c83daa82d500b006b300496c3