Open Access Open Access  Restricted Access Subscription Access

Autoregressive Integrated Moving Average Model for Gold Price Forecasting : Evidence from the Indian Market


Affiliations
1 Assistant Professor, Haryana School of Business, Guru Jambheshwar University of Science & Technology, Hisar - 125 001, Haryana, India
2 Research Scholar, Haryana School of Business, Guru Jambheshwar University of Science & Technology, Hisar - 125 001, Haryana, India

   Subscribe/Renew Journal


The present study was conducted to forecast the gold prices in India by employing ARIMA (1, 1, 2) model on time series data for short term. The stationarity of time series data was tested by using the ADF unit ischolar_main test. To overcome the problem of autocorrelation, Breusch - Godfrey serial correlation was conducted. The study forecasted gold prices within sample and post sample forecast. Actual values of gold prices and the forecasted values of gold prices moved in the same direction very closely. The post sample forecasted values of gold prices revealed an increasing trend. The predicted six months values of gold prices probably indicated reasonable returns for investors who held gold in their financial portfolios. Hence, the ARIMA (1, 1, 2) model was found to be the best fit to forecast short term gold prices on time series data.

Keywords

Gold Prices, Time Series, Autoregressive Integrated Moving Average (ARIMA), Investment, Stationary, Forecasting

F470, G170, G130, G150

Paper Submission Date : March 9, 2017 ; Paper sent back for Revision : July 12, 2017 ; Paper Acceptance Date : September 15, 2017.

User
Subscription Login to verify subscription
Notifications
Font Size

Abstract Views: 265

PDF Views: 0




  • Autoregressive Integrated Moving Average Model for Gold Price Forecasting : Evidence from the Indian Market

Abstract Views: 265  |  PDF Views: 0

Authors

Khujan Singh
Assistant Professor, Haryana School of Business, Guru Jambheshwar University of Science & Technology, Hisar - 125 001, Haryana, India
Anil Kumar
Research Scholar, Haryana School of Business, Guru Jambheshwar University of Science & Technology, Hisar - 125 001, Haryana, India

Abstract


The present study was conducted to forecast the gold prices in India by employing ARIMA (1, 1, 2) model on time series data for short term. The stationarity of time series data was tested by using the ADF unit ischolar_main test. To overcome the problem of autocorrelation, Breusch - Godfrey serial correlation was conducted. The study forecasted gold prices within sample and post sample forecast. Actual values of gold prices and the forecasted values of gold prices moved in the same direction very closely. The post sample forecasted values of gold prices revealed an increasing trend. The predicted six months values of gold prices probably indicated reasonable returns for investors who held gold in their financial portfolios. Hence, the ARIMA (1, 1, 2) model was found to be the best fit to forecast short term gold prices on time series data.

Keywords


Gold Prices, Time Series, Autoregressive Integrated Moving Average (ARIMA), Investment, Stationary, Forecasting

F470, G170, G130, G150

Paper Submission Date : March 9, 2017 ; Paper sent back for Revision : July 12, 2017 ; Paper Acceptance Date : September 15, 2017.




DOI: https://doi.org/10.17010/ijrcm%2F2017%2Fv4%2Fi3%2F118920