Industrial Engineering Reserch Paper Summery

A Multi-Period Inventory Model to Incorporate with  Inventory Age, Accounting Principle, and Product Structure: A Case Study in a Make-to-Stock  Semiconductor Integrated Device Manufacturer

Jei-Zheng Wu 
Department of Business Administration
Soochow University
Taipei, Taiwan
jzwu@scu.edu.tw 

Chen-Fu Chien, Yung-Shang Huang, Hung-Ya Huang
Department of Industrial Engineering and Engineering Management
National Tsing Hua University
Taipei, Taiwan

Since December 31 in 2008, Taiwanese government declared Statement of Financial Accounting Standards (SFAS) No.10 as accounting principle for allowance for reduction of inventory to market. It has become even more difficult for semiconductor integrated device manufacturers, using make-to- stock manufacturing strategy and possessing inventory accounting for about 14% total costs, to maintain robust records in financial statements. It is owing to long manufacturing cycle times for semiconductor industry. For  example, average cycle times of flash memory take about three  months including the wafer fabrication 50 days, circuit probing 5days, chip assembly 7 days and final test 22 days. In addition, 3 months to 1 year of safety stocks are established depending on product market and customer satisfaction level. In this case, all of products will become the amortization item to be written down most inventory loss of work in process (WIP) that results in high recorded inventory cost. Added to this, the variation of financial performance may further cause overreaction in stock market when a large amount of allowance for reduction of inventory to market reported are at one period time. This overreaction normally is not reversible. That is, there will be no compensation even after more inventories is actually sold.

Existing studies have paid attention to echelon inventory policy aiming at optimizing stock levels and order quantity of each echelon supply chain for reducing stock and inventory holding costs . Regarding specific application area, inventory costs of electronic product can be divided into component devaluation costs, price protection costs, product return costs, obsolescence costs and inventory holding costs. Different inventory costs have different proportions with respect to different product categories. Subsequently, one should utilize different supply chain management policies for inventory-driven cost reduction. Since decline of product demand will lead to loss of orders and thus high stock levels, it is necessary to examine the product life cycle of electronic components and accordingly adjust stock levels to reduce inventory overdue cost. Recent inventory management research has taken deterioration and obsolescence into account and developed optimal inventory policies correspondingly. To further improve evaluation of inventory management, effective information systems were suggested including implementation of Just-in-Time (JIT), material requirement planning (MRP), mass customization, and manufacturing flexibility and modularity .

A streamlined procedure is utilized to solve the preemptive goal programming model that simultaneously finds optimal solutions based on relatively high weights of prioritized goals.  The first-priority goal takes the highest weight and the second one takes a smaller one but significant small enough to reveal differences and so on. The proposed model considers sophisticated causal relations by incorporating a number of operation constraints.

This study proposes a holistic inventory cost assessment model to incorporate with inventory ages, accounting principles, product structures, and allowance for inventory  valuation losses. Empirical study with real data shows that the proposed model can help analyze the most beneficial accrue

provision rate setting for robust estimation of inventory cost and to reduce the impact of carrying value fluctuation of  inventory. This study also introduces transition rates and flow  ratios in the proposed model for decision-makers to proactively  control WIP flow and to maintain robust inventory value. On  the other hand, when transition rates and flow ratios are summarized from historical data, the standard inventory costs including allowance for inventory valuation losses can be  estimated based on standard operating costs. In addition, utilizing the proposed model can help explain the gap between  standard and actual inventory costs. Finally, the proposed model can become predictive after further empirical data are  examined and validated. Further research should be done regarding improvement of computation efficiency since it takes more than 30 minutes to assess one product group and there  could be more than one hundred product groups at a time in  practice.

Submitted By
Rupesh Kumar Chauhan
Roll no 79
PGDIE 42