Demand forecasting and financial estimation considering the interactive dynamics of semiconductor supply-chain companies

Chih-Hsuan Wang*, Jen Yu Chen

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

3 Scopus citations

Abstract

Demand forecasting and financial estimation are two critical issues in supply chain management. Traditional forecasting techniques are either based on historical data (time-series) or causal predictors (regression). Although numerous schemes have been proposed, most cannot accommodate the time-lag causalities between the predictors and the outcome and the interactive dynamics of the supply-chain members. This research presents a novel framework to highlight the following issues: (1) Time-series models are constructed to accommodate product volatility and conduct demand forecasting. (2) Vector autoregression is used to capture the interactive dynamics of the supply-chain members to conduct financial estimation. (3) Regression methods are applied to conduct sensitivity analyses that can measure the impact on the sales revenue of a firm by increasing or decreasing a specific predictor. Experimental results demand forecast for consumer products can successfully predict the sales revenues of chip-design firms. For chip manufacturers and packaging and testing (P&T) firms, interactive dynamics can be competition (one suffers from the growth of the other) or cooperation (a win-win scenario). If one is strong and the other is weak (asymmetric relationship), the dynamics is cooperative. If two firms perform almost equivalently, the dynamics is competitive.

Original languageEnglish
Article number106104
JournalComputers and Industrial Engineering
Volume138
DOIs
StatePublished - Dec 2019

Keywords

  • Demand forecast
  • Financial estimation
  • Semiconductor
  • Supply chain

Fingerprint Dive into the research topics of 'Demand forecasting and financial estimation considering the interactive dynamics of semiconductor supply-chain companies'. Together they form a unique fingerprint.

Cite this