| Summary: | 碩士 === 世新大學 === 經濟學系 === 92 === Abstact
Taiwanese semiconductor industry is the most competitive sector in global market and the business model is distinctive because of its vertical integration. The semiconductor industry is the major drive to boom Taiwanese tech development. However, companies’ profit is squeezed owing to speedy Semiconductor industry is the major drive to support Taiwan economics growth. However, Taiwanese semiconductor companies face dual pressure from US and Japan applying cutting edge technology and price war raised by Korea and China. This research will focus on key factors that impact on output efficiency and offers an objective reference for improving production efficiency.
The research choose 18 IC design house、6 foundry and 8 assembly & test companies. We pick out fixed asset per capita、common shares per capita、operation expense per capita、R&D expense per capita and average salary as input variables and sales and ROA as output variables. We use DEA to analyze output efficiency and apply MPI to investigate the interactions between productivity and efficiency spanning several periods. In the end, by regression model, we analyze factors affecting output efficiency, which include percentage of doctor and master quantity in total employees、staff working experience、sales from top 3 customers、 share holding on foreign investors and directors、 percentage of building cost in fixed asset、operation income rate and percentage of bonus accounting for net income. We look forward to an answer to enhance business competitiveness from the research result.
The result shows inefficiency in overall industry、IC assembly & test and IC design mainly comes from lack of economics scales so that companies have to allocate their capacity for best scale. IC makers suffer imbalance between resource input and output so that they don’t utilize technical efficiency well. MPI shows decrease of efficiency variation in overall semiconductor industry. IC assembly & test owns highest DEA in single time horizon but performs worst in MPI. IC design houses and makers improve technical efficiency variation and scale efficiency variation respectively, but they don’t work out inefficiency items. Under regression analysis, we notice that there is positive correlation between operation income rate and efficiency, but obviously negative correlation between percentage of building cost in fixed asset and efficiency. Watching efficiency variation in different periods, increasing shares holding on foreign investors and superiors progresses overall technical efficiency； raising percentage of building cost in fixed asset will improve operation efficiency in short time；higher top 3 customers sales hinder MPI production variation efficiency；high bonus to employee and directors from operation income will lose company adapt operation scale.
|
|---|
