Bayesian Multiperiod Forecasting for Indonesia Inflation Using ARMA Model

摘要

This paper present a Bayesian approach to find the Bayesian forecasting using ARMA model under Jeffrey's prior assumption with quadratic loss function applied to monthly national inflation of Indonesia from January 2000 to December 2014. The forecasting model be obtained based the marginal conditional posterior predictive density. To conclude whether the model is adequate is used the Ljung-Box Q statistic, while to look at the accuracy of the model is used the Root Mean Square error (RMSE).