Performance Measures of Univariate Time Series Methods for Price Forecasting — An Empirical Study of Gram Price Forecast Bhardwaj S.P., Singh K.N., Kumar Amrender Indian Agricultural Statistics Research Institute (ICAR), Library Avenue, New Delhi-110 012 Online published on 9 October, 2012. Abstract The commodity-price forecasts constitute a vital input of macroeconomic planning. The present study is based on the time series forecast models which are non-structural-mechanical in nature. The model that has been selected for forecasting the prices of gram is autoregressive integrated moving average [ARIMA (0, 1, 1)] model which gives reasonable and acceptable forecasts; it does not perform very well when there exists volatility in the data series. For this reason, generalized autoregressive conditional heteroskedasticity [GARCH (1, 1)] model has been used to forecast gram prices. The model performs better than ARIMA (0, 1, 1) because of its ability to capture the volatility by the conditional variance of being non-constant throughout the time. The study has concluded that GARCH (1,1) is a better model than ARIMA (0, 1, 1) in forecasting spot price of gram. The values for RMSE, MAE and mean absolute percentage error (MAPE) obtained were smaller than those in ARIMA (0,1,1) model. The AIC and SIC values from GARCH model were smaller than that from ARIMA model. It shows that GARCH is a better model than ARIMA for estimating daily prices of gram. Top Keywords Price forecasting, ARIMA model, GARCH model, gram price. Top |