Applying multivariate time series models to technological product sales forecasting

International Journal of Technology Management

Volumn 27, Issue 2-3, 2004, Pages 306-319

  Chiu, Yi Chia ^a   Shyu, Joseph Z ^b  

^a CHUNG YUAN CHRISTIAN UNIVERSITY   (Taiwan)

^b NATIONAL CHIAO TUNG UNIVERSITY   (Taiwan)

Author keywords

Forecasting; Litterman bayesian vector autoregression; Technology marketing; Vector autoregression

Indexed keywords

FORECASTING; REGRESSION ANALYSIS; SALES; STATISTICS;

MULTIVARIATE TIME SERIES MODELS; TECHNOLOGICAL PRODUCT SALES FORECASTING; TECHNOLOGY MARKETING; VECTOR AUTOREGRESSION MODEL;

TIME SERIES ANALYSIS;

EID: 1542315542     PISSN: 02675730     EISSN: None     Source Type: Journal    
DOI: 10.1504/ijtm.2004.003957     Document Type: Article

References (25)

1. Sims, C.A. (1980) 'Macroeconomics and reality', Econometrica, Vol. 48, No. 1, pp. 1-48.
2. Doan, T., Litterman, R. B. and Sims, C. (1984) 'Forecasting and conditional projection using realistic prior distributions', Econometric Reviews, Vol. 3, pp.1-100.
3. Litterman, R. B. (1986) 'Forecasting with bayesian vector autoregressions - five years of experience', Journal of Business and Economic Statistics, Vol. 4, No. 1, pp.25-38.
4. Litterman, R.B. (1985) 'How monetary policy in 1985 affects the outlook', Quarterly Review - Federal Reserve Bank of Minneapolis, Vol. 9, No. 4, pp.2-14.
5. Spencer, D.E. (1993) 'Developing a Bayesian vector autoregression forecasting model', International Journal of Forecasting, Vol. 9, pp.407-421.
6. Holden, K. (1995) 'Vector autoregression modeling and forecasting', Journal of Forecasting, Vol. 14, pp.159-166.
7. Dua, P. and Ray, S.C. (1985) 'A BVAR model for the Connecticut economy', Journal of Forecasting, Vol. 14, pp.167-180.
8. Census X-11 methods, including multiplicative and additive methods, are standard methods used by the US Bureau of Census to make seasonal adjustments.
9. Sarantis, N. and Stewart, C. (1995) 'Structural, VAR and BVAR models of exchange rate determination: a comparison of their forecasting performance', Journal of Forecasting, Vol. 14, pp.201-215.
10. Dua, P. and Smyth, D.J. (1995) 'Forecasting US homes sales using BVAR models and survey data on households' buying attitudes for homes', Journal of Forecasting, Vol. 14, pp.217-227.
11. Kumar, V., Leone, R.P. and Gaskins, J.N. (1995) 'Aggregate and disaggregate sector forecasting using consumer confidence measures', International Journal of Forecasting, Vol. 11, pp.361-377.
12. Curry, D.J., Divakar, S., Mathur, S.K. and Whiteman, C.H. (1995) 'BVAR as a category management tool: an illustration and comparison with alternative techniques', Journal of Forecasting, Vol. 14, pp.181-199.
13. Hsu, P.-H., Wang, C.-H., Shyu, J.Z. and Yu, H-C. (2001) 'A Litterman BVAR approach for production forecasting of technology industries', forthcoming, Technological Forecasting and Social Change.
14. Doan, T. (1992) RATS User's Manual, Estima, Evanston, IL, pp.8-20.
15. Moreover, the Chinese (Lunar) New Year vacation may occur in January or in February, and occasionally in March. Such uncertainty and irregularity prompts us to discard monthy data in our consideration
16. Such preliminary deseasonalisation is also found in Doan et al. [3], Dua and Ray [7], Kumar, Leone and Gaskins [10], Salazar and Weale [16], and Marchetti and Parigi [17].
17. Salazar, E. and Weale, M. (1999) 'Monthly data and short-term forecasting: an assessment of monthly data in VAR model', Journal of Forecasting, Vol. 18, pp.447-462.
18. Marchetti, D.J. and Parigi, G. (2000) 'Energy consumption, survey data and the prediction of industrial production in Italy: a comparison and combination of different models', Journal of Forecasting, Vol. 19, pp.419-440.
19. Hamilton, J.D. (1994) Time Series Analysis, Princeton Press, NJ, p. 362.
20. Granger, C.W.J. (1969) 'Investigating casual relations by econometric models and cross spectral methods', Econometrica, Vol. 37, pp.424-438.
21. Many researchers recommended level VAR and opposed the differentiation of sries that present a unit root or trend factors. Based on Enders [22], taking the difference will throw out the co-movement information within a series. Conversely, some researchers considered the VAR model in differentiation ([8,23]). But there are still some arguments among researchers on whether differentiatin for stationary property is necessary in the preliminary transformation. Some researchers set the BVAR only in level data series on the basis of Sims et al's [24] statement "⋯the Bayesian approach is entirely based on the likelihood function, which has the same Gaussian shape regardless of the presence of non-stationarity. A Bayesian inference takes no special account of non-stationarity" [p. 136] (see also [7]). It seems trivial to consider preliminary transformation before the LBVAR model.
22. Enders, W. (1995) Applied Econometric Time Series, John Wiley & Sons, p. 301.
23. Ravishanker, N and Ray, B.K. (1997) 'Bayesian analysis of vector ARMA models using Gibbs sampling', Journal of Forecasting, Vol. 16, pp.177-194.
24. Sims, C.A., Stock, J.H. and Watson, M.W. (1990) 'Inference in linear time series models with some unit roots', Econometrica, Vol.58, pp.113-144.
25. In the differenced VAR model, the differenced series for specification is from 1990:2 to 1997:4. Then the differentiated series from 1998:1 to 2000:1 is used as a prediction assessment.