| Summary: | 碩士 === 銘傳大學 === 風險管理與保險學系碩士在職專班 === 98 === TFT-LCD process is a rapid growing and technology-transformed industry, heavy investment on new processes is a necessary condition to promote a more competitive company. Semi-quantitative analysis is usually utilized and supplies finite quantitative data for the safety assessment of TFT-LCD industry in past studies; professional veteran approach easily produces bias between practice and theory. The loss of etch process is especially difficult to proceed risk assessment. This investigation tries to supply a quantitative, evaluating, objective approach to assess the risk of TFT-LCD industry with VaR (Value at Risk) method.
Two years of losses data, from January 1, 2008 to December 31, 2009, are used to measure the hazardous risks of etch process equipment with three VaR models, Variance-Covariance method, Historical simulation method and Monte-Carlo simulation method. Back Testing is employed to examine the suitability of VaR model. Furthermore, Risk Matrix was utilized to evaluate the status of etch process equipment and establish appropriate risk response strategies.
Results demonstrate that 1% VaR and CVaR values are applicable for estimating maximum probable loss (MPL). The loss of etch process equipment is suitable to describe with an extreme value distribution (Gamma function) and present a fat-tailed feature, the most serious event of seven types’ losses for TFT-LCD industry. Variance-Covariance method is not applicable in this research. Different loss factors fit different probability distribution functions (PDFs). Monte-Carlo simulation is applicable to estimate specific PDFs with different types of loss factors. Historical loss data and VaR models were employed to estimate MPL of safety accident and offer an objectively and quantitative method to improve the deficiency of traditional semi-quantitative and professional veteran approaches, and provide an objectively better information for decision makers. Furthermore, this study provide a method to assess the loss impacts of safety accident for enterprise business operations and measure Net Loss of financial reports, and these valuable data could be basic information of risk management and risk cost.
|
|---|
