Abstract
This paper develops a Bayesian analysis for the multiplicative seasonal moving average model by implementing a fast, easy and accurate Gibbs sampling algorithm. The proposed algorithm does not involve any Metropolis-Hastings generation but is generated from normal and inverse gamma distributions. The problem of forecasting multiple future observations is considered. The proposed algorithm is illustrated using a simulated example and airline data. Unlike the classical approach, by using the airline data, the proposed algorithm is easily used to test the significance of the interaction parameter.
