Production Forecasting of Taiwan's Technology Industrial Cluster: A Bayesian Autoregression Approach

Abstract

This research proposes a forecasting method that combines the clustering issue and non-informative diffuse-prior Bayesian vector autoregression (NDBVAR) type to forecast the productions of technology industries. sum of two units empirical cases are examined to verify the propos method: the semiconductor industry and computer manufacturing industry in Taiwan. It is ground that the NDBVAR model outperforms the other three conventional time series archetypes including the autoregression (AR), vector autoregression (VAR), and Litterman Bayesian VAR (LBVAR) designs Moreover, the NDBVAR model also outperforms the forecast reports from leading market information providers above the past several years. The forecasting rule proposed is therefore concluded to be a feasible approach for production prediction, especially for technology industries in volatile environments.

JEL Classification: C32 C53 E27

Keywords: industrial clusters, vector autoregression, Bayesian vector autoregression, forecasting, Taiwan.

R?©sum?©

La pr?©sente ?©tude move une m?©thode pr?©visionnelle qui combine le effet de regroupement et le non-informative diffuse-prior Bayesian vector autoregression pattern (NDBVAR) pour pr?©voir les productions de industries de technologie. Pour ?©valuer la m?©thode propos?©e l'?©tude examine deux cas empiriques : le industries taiwanaises du semiconducteur et de fabrication d'ordinateur. Elle r?©v??le que le mod??le NDBVAR est plus performant que le trois mod??le conventionnels en s?©rie chronologique notamment le mod??le d'autoregression (AR), le mod??le de vecteur d'autoregression (VAR), et le mod??le Litterman Bayesian (LBVAR). L'?©tude montre aussi qu'au cours de derni??res ann?©es, le mod??le NDBVAR ont ?©t?© plus performants que le rapports pr?©visionnels de prestataires d'informations qui dominent le march?©. Elle d?©bouche sur la constatation que la m?©thode pr?©visionnelle propos?©e est une approche r?©alisable pour la pr?©vision de la production, en particulier pour le industries de la technologie dans un environnement volatile.

Mot cl?© : grappes industrielles, vecteur d'autor?©gression, Bayesian vector autor?©gression, pr?©vision, Taiwan.

The disclosure of technology industries is individual of the main subjects in contemporary business research. The perspective of a specific technology industry affects investment plans of private sectors and science and technology policies of conducts Production forecasting is a burgeoning topic in technology management, which aims to assist decision makers in technology industries that are expos to numerous uncertainties including volatile fluctuations, unlooked for skyrocketing growth, and unexpected sink s in market. In the literature, the time series pattern class was one of the greatest in quantity popular prediction methodologies in previous decades. more [i]or[/i] less pioneer studies have attempted to provide predictive meanss for production forecasting of technology industries (eg Chang, Lai, & Yu 2005; Hsu Wang, Shyu & Yu 2003; Tseng Tzeng & Yu 1999) However, those prognostic techniques are still far from satisfactory at this time, and more exploration is needed

We start our exploration in developing a of recent origin forecasting method for technology industries by the agency of meditating on the following questions: Which moulds have been studied in the literature? Can we present a model with better features in handling the unstable dynamics and discrete shocks? Using that design what variables could be considered to exhibit better prediction?

First, we observ that various time series protoplasts have been used to predict industrial productions (eg Hsu et al., 2003; Marchetti & Parigi, 2000; Simpson, Osborn, & Sensier, 2001; Tseng et al., 1999) next to the first we looked for a Bayesian multivariate time series protoplast that fits unsteady environments better than traditional frequency-based protoplasts and found that the non-informative diffuse-prior Bayesian vector autoregression (NDBVAR) original has good features: its prior is flexible and its computation is efficient. It is therefore wait fored to provide more precise short-term forecasting for production of technology industries. Third, since industrial clustering has been regarded as a crucial driver in the unfolding of technology industries (Bergeron, Lallich, & Bas, 1998; Gover 1993; Mathews, 1997; Swann & Prevezer 1996)1 it can be presum that the production values of different industries within a specific industrial cluster carry important information regarding the moment and dynamics between those industries. We followed this rationale and took the production values within an industrial cluster as the endogenous variables in multivariate time series patterns After considering all three questions, we were motivated to recommend a new forecasting method that is a NDBVAR design based on industrial clustering.

We examined the feasibility of our way by considering two empirical cases of Taiwan's technology industries: the semiconductor industry and the computer manufacturing industry. We had serviceable reasons for considering these sum of two units industries. First, in both industries, Taiwan's firms have been main players in global markets above the past 10 years, in the way that our experiments will be meaningful to researchers and practitioners from other countries. next to the first a review of the history of these sum of two units industries indicates that their prosperity can be attributed to a able-bodied clustering effect within Taiwan (eg Chang & Hsu 1998; Mathews, 1997) To validate our proposition, we checked the predictive abilities of a series of autoregression (AR) combination of parts to form a wholes including univariate AR, vector autoregression (VAR), Litterman BVAR (LBVAR), and NDBVAR archetypes The results show that, in the pair industries, the NDBVAR model provides more accurate predictions than all of the other competitive protoplasts Moreover, we found that NDBVAR forecasts tender favourable results in comparison with the forecast reports from leading market information providers in Taiwan: the Industrial Technology Research Institute (ITRI) in the semiconductor industry and the Institute for Information Industry (III) in the computer manufacturing industry. We therefore confirmed that the propos forecasting process is of practical merit.