Forecasting Crude Oil Prices Using Wavelet ARIMA Model Approach

The increase in crude oil prices can give a significant impact on worldwide economic activities. Therefore, an accurate forecasting is crucial so that necessary precaution steps can be planned earlier. This paper proposed a Wavelet-ARIMA approach to improve the forecasting accuracy of the crude oil price series. In this approach, wavelet transform decomposes the original series into a more stable constitutive series while ARIMA model is utilized to forecast each individual series and lastly inversed back to original series. Dataset of crude oil spot prices from West Texas Intermediate and the Brent North Sea are used. Direct application of ARIMA model is used as a benchmark for effectiveness measurement with the proposed approach by comparing Mean Absolute Error and Root Mean Square Error obtained. R software is used to implement ARIMA while MATLAB is used for dataset pre-processing. The result has proven that Wavelet-ARIMA approach demonstrates better prediction performance than ARIMA model.

[1]  Stephan Schlüter,et al.  Using wavelets for time series forecasting: Does it pay off? , 2010 .

[2]  A.J. Conejo,et al.  Day-ahead electricity price forecasting using the wavelet transform and ARIMA models , 2005, IEEE Transactions on Power Systems.

[3]  Ruhaidah Samsudin,et al.  CRUDE OIL PRICE FORECASTING WITH AN IMPROVED MODEL BASED ON WAVELET TRANSFORM AND SUPPORT VECTOR MACHINES , 2013 .

[4]  Rob J Hyndman,et al.  Automatic Time Series Forecasting: The forecast Package for R , 2008 .

[5]  Crude Oil Price Forecasting Based on the ARIMA and BP Neural Network Combinatorial Algorithm , 2012 .

[6]  Patrick M. Crowley,et al.  A Guide to Wavelets for Economists , 2007 .

[7]  Z. Tan,et al.  Day-ahead electricity price forecasting using wavelet transform combined with ARIMA and GARCH models , 2010 .

[8]  L. Campos,et al.  Combination of forecasts for the price of crude oil on the spot market , 2016 .

[9]  Ruhaidah Samsudin,et al.  Crude Oil Price Forecasting Based on Hybridizing Wavelet Multiple Linear Regression Model, Particle Swarm Optimization Techniques, and Principal Component Analysis , 2014, TheScientificWorldJournal.

[10]  Gillian Dooley,et al.  An assessment of time series methods in metal price forecasting , 2005 .

[11]  Ruhaidah Samsudin,et al.  Daily Crude Oil Price Forecasting Using Hybridizing Wavelet and Artificial Neural Network Model , 2014 .

[12]  H. Mohammadi,et al.  International evidence on crude oil price dynamics: Applications of ARIMA-GARCH models , 2010 .

[13]  Lean Yu,et al.  A New Method for Crude Oil Price Forecasting Based on Support Vector Machines , 2006, International Conference on Computational Science.

[14]  Nazuha Muda Yusof,et al.  Malaysia crude oil production estimation: an application of ARIMA model , 2010, 2010 International Conference on Science and Social Research (CSSR 2010).

[15]  Mohammad Kazem Sheikh-El-Eslami,et al.  Price forecasting of day-ahead electricity markets using a hybrid forecast method , 2011 .

[16]  Kwok-wing Chau,et al.  Improving Forecasting Accuracy of Annual Runoff Time Series Using ARIMA Based on EEMD Decomposition , 2015, Water Resources Management.

[17]  G. Box,et al.  On a measure of lack of fit in time series models , 1978 .

[18]  Farshid Keynia,et al.  Day ahead price forecasting of electricity markets by a mixed data model and hybrid forecast method , 2008 .

[19]  David J. Parsons,et al.  An improved wavelet–ARIMA approach for forecasting metal prices , 2014 .