ScholarLens Back to search
2016•AIP conference proceedingsRequires access

Modelling and forecasting monthly crude oil price of Pakistan: A comparative study of ARIMA, GARCH and ARIMA Kalman model

Muhammad Aamir, Ani Shabri

Open publisher page 31 citations

Abstract

Crude oil is one of the most important commodity in the world and it is meaningful for every individual. This study aims at developing a more appropriate model for forecasting the monthly crude oil price of Pakistan. In this study, three-time series models are used namely Box-Jenkins ARIMA (Auto-regressive Integrated Moving Average), GARCH (Generalized Auto-regressive Conditional Hetero-scedasticity) and ARIMA Kalman for modelling and forecasting the monthly crude oil price of Pakistan. The capabilities of ARIMA, GARCH and ARIMA-Kalman in modelling and forecasting the monthly crude oil price are evaluated by MAE (Mean Absolute Error) and RMSE (Root Mean Squared Error). It is concluded that the hybrid model of ARIMA Kalman perform well as compared to the Box-Jenkins ARIMA and GARCH models.

About this research paper

What this paper is about

Crude oil is one of the most important commodity in the world and it is meaningful for every individual. This study aims at developing a more appropriate model for forecasting the monthly crude oil price of Pakistan. In this study, three-time series models are used namely Box-Jenkins ARIMA (Auto-regressive Integrated Moving Average), GARCH (Generalized Auto-regressive Conditional Hetero-scedasticity) and ARIMA Kalman for modelling and forecasting the monthly crude oil price of Pakistan. The capabilities of ARIMA, GARCH and ARIMA-Kalman in modelling and forecasting the monthly crude oil price are evaluated by MAE (Mean Absolute Error) and RMSE (Root Mean Squared Error). It is concluded that the hybrid model of ARIMA Kalman perform well as compared to the Box-Jenkins ARIMA and GARCH models.

Why it matters

OpenAlex reports 31 citations for this work. Citation counts describe recorded attention and do not establish research quality.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

Crude oil is one of the most important commodity in the world and it is meaningful for every individual. This study aims at developing a more appropriate model for forecasting the monthly crude oil price of Pakistan. In this study, three-time series models are used namely Box-Jenkins ARIMA (Auto-regressive Integrated Moving Average), GARCH (Generalized Auto-regressive Conditional Hetero-scedasticity) and ARIMA Kalman for modelling and forecasting the monthly crude oil price of Pakistan. The capabilities of ARIMA, GARCH and ARIMA-Kalman in modelling and forecasting the monthly crude oil price are evaluated by MAE (Mean Absolute Error) and RMSE (Root Mean Squared Error). It is concluded that the hybrid model of ARIMA Kalman perform well as compared to the Box-Jenkins ARIMA and GARCH models.

Key concepts: Autoregressive integrated moving average, Autoregressive conditional heteroskedasticity, Box–Jenkins, Econometrics, Mean squared error, Kalman filter, Crude oil, Mean absolute percentage error

Related papers

Related paper 1Related paper 2Related paper 3
Back to paper searchBrowse research topicsOriginal source
Modelling and forecasting monthly crude oil price of Pakistan: A comparative study of ARIMA, GARCH and ARIMA Kalman model — Research Paper | ScholarLens