Analysis of the Fluctuation of Chinese Crude Oil Futures- Based on GARCH-type Model

Sun, Hongguo; Li, Wenhui

doi:10.2991/msam-18.2018.25

Cited by 1 publication

(1 citation statement)

References 5 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…At present, more and more scholars study China's crude oil futures, China's crude oil market environment and oil policy. Sun et al (2018) used GARCH and TARCH models to study the fluctuation characteristics of China's crude oil futures returns rate based on high-frequency data, and they found that the changes of China's crude oil futures returns rate in the current as well as the lag period were mainly influenced by itself, and the influence coefficient of one period lag was larger and the influence time was longer. Ji and Zhang (2019) analyzed the initial characteristics of China's crude oil futures market, laying a good foundation for subsequent studies.…”

Section: Literature Reviewmentioning

confidence: 99%

Research on Risk Features and Prediction of China’s Crude Oil Futures Market Based on Machine Learning

Zhang

Liu

2022

Front. Energy Res.

View full text Add to dashboard Cite

Facing the rapidly changing domestic and foreign futures markets, how to accurately and immediately predict the price trend of crude oil futures in order to avoid the risks caused by price fluctuations is very important for all participants in the crude oil futures market. Based on the 5-min high-frequency trading data of China’s crude oil futures market in recent 3 years, this paper uses the EMD-MFDFA model combined with multifractal detrended fluctuation analysis (MF-DFA) and empirical mode decomposition unsupervised K-means clustering and Gaussian mixture model (GMM) to identify the risk status of each trading day. Further, Support vector machine (SVM), extreme gradient lifting (XGBoost) and their improved algorithms are used to predict the risk state of China’s crude oil futures market. The empirical results are as follows: first, There are obvious multifractal features in the return rate series of China’s crude oil futures market and its single trading day; Second, compared with the traditional SVM model, the improved Twin Support Vector Machine (TWSVM) based on solving the sample imbalance issue has better prediction ability for China’s crude oil futures risk.; Third, The XGBoost has a great impact on the prediction of China’s crude oil risk, and the Focal-XGBoost with focal loss function performs the best in predicting the risk of China’s crude oil futures market.

show abstract

Section: Literature Reviewmentioning

confidence: 99%