Univariate and Multivariate ARIMA Versus Vector Autoregression Forecasting

performance The purposes o f t h i s study are two: 1) t o compare t h e f o r e c a s t i n g a b i l i t i e s o f t h e t h r e e methods: u n i v a r i a t e a u t o r e g r e s s i v e i n t e g r a t e d moving average (ARIMA), m u l t i v a r i a t e a u t o r e g r e s s i v e i n t e g r a t e d moving average (MARIMA), and v e c t o r a u t o r e g r e s s i o n ( b o t h unconstrained--VAR--and Bayesian--BVAR) and 2 ) t o study t h e i d e a t h a t one advantage o f v e c t o r autoregressions i s t h a t t h e models can e a s i l y and i n e x p e n s i v e l y be r e e s t i m a t e d a f t e r each a d d i t i o n a l d a t a p o i n t .A l l o f these methods have been shown t o provide f o r e c a s t s t h a t a r e more accurate than many econometric methods, which r e q u i r e more resources t o implement.These methods were a p p l i e d t o seven economic v a r i a b l e s : r e a l GNP, annual i n f l a t i o n r a t e s , unemployment r a t e , t h e money supply ( M I ) , gross p r i v a t e domestic investment, t h e r a t e on f o u r -t o six-month commercial paper, and the change i n business i n v e n t o r i e s .The major r e s u l t s o f t h i s study a r e : 1) on average, t h e method t h a t performs b e s t i n terms of t h e r o o t mean square e r r o r (RMSE) i s t h e m u l t i v a r i a t e ARIMA model; 2) t h e u n i v a r i a t e ARIMA and BVAR methods p e r f o r m approximately t h e same on average; 3 ) r e e s t i m a t i n g t h e VAR model a f t e r each d a t a p o i n t increases t h e accuracy of t h i s method; 4 ) r e e s t i m a t i n g t h e BVAR model a f t e r each d a t a p o i n t becomes a v a i l a b l e decreases t h e accuracy o f t h i s method; and 5) the VAR method u s i n g r e e s t i m a t i o n i s approximately as accurate as t h e BVAR method.