資料包絡分析法多種模式下之效率值分析-以我國上市、櫃電子零組件廠商為例

Many Different Patterns of The Productivity of Data Envelopment Analysis - Listed and OTC Electronic Component Industries in Taiwan As An Example

巫聖婷(Sheng-Ting Wu)

指導教授 ： 林利萱

元智大學/管理學院/會計學系/碩士(2006年)

摘要

本研究以台灣53家上市電子零組件廠商與70家上櫃廠商為評估對象，探討公司於2002年至2005年間之經營效率。本研究應用資料包絡分析法( Data Envelopment Analysis, DEA)下，對比模式(Bilateral Model)、整體技術效率CCR模式(Charnes-Cooper-Rhodes Model )、純粹技術效率BCC(Banker- Charnes-Cooper Model)模式、非控制變數模式NCN(non-controllable Model)以及窗型分析(Window Analysis)來計算相對效率值。投入變數包括：營業成本、營業外費用合計數、研究發展費用、機械及儀器設備成本、TCRI信用評等及員工人數共六項；產出項目為廠商營業收入與營業外收入兩項。
經實証結果顯示：
1.對比模式(Bilateral Model)下發現，上市電子零組件廠商即便是處於經濟景氣低迷的環境下，其經營績效仍然優於上櫃電子零組件廠商；在景氣復甦的情況下，上櫃廠商相對較易達成其經營效率之目標。
2. 經過對比模式(Bilateral Model)對信用評等指標加以區分為兩群體後發現，信用評等指數在1-5之間的群體，其效率值優於信用評等指數在6-9之間的群體，顯示信用評等指數為一有鑑別力之變數。
3.利用窗型分析所得之結果，可以與其他模型下之效率值相互驗證，可進ㄧ步發現經營績效的確具有穩定性與一致性，此為窗型分析的貢獻之ㄧ。
4.傳統CCR模式、BCC模式以及規模效率下所求得之效率值，已具有一定程度的績效衡量水準，然而，在非控制模式(NCN)下，透過相互比較發現，無論上市、上櫃廠商之效率值，皆有變動之情況產生，原先在傳統資料包絡分析法模式下效率值為1之廠商，在NCN非控制模式下卻不再為1，可見非控制模式(NCN)，考量了更多管理者無法掌控的外在因素，使其效率值發生異動。

關鍵字

資料包絡分析法；績效評估；電子零組件

並列摘要

Utilizing Bilateral model, Charnes-Cooper-Rhodes (CCR) model, Banker-Charnes-Cooper ( BCC) model, non-controllable (NCN) model, and window analysis of data envelopment analysis, this paper examines the relative efficiency of 53 public listed and 70 OTC electronic component industries in Taiwan between 2002 and 2005. The parameters we used for input variables are operating costs, total non-operating expenses, research and develop expense, cost of equipment, TCRI credit rating standard and employee number, and for output variable are operating revenues and total non-operating revenue.
The empirical results are summarized as follows:
1. Under the Bilateral model, we can discover that even the economy environment is murky, the relative efficiency of the public listed electronic component industries are better than the OTC electronic component industries. In the contrary, when the economic environment change for better, the OTC electronic component industries achieve the management performance goal easily.
2. We compartmentalize TCRI credit rating standard for two groups. Group-one is standard first to fifth, group-two is standard sixth to tenth. Under the Bilateral model, group-one’s relative efficiency is better than group-two. As a result, TCRI credit rating standard is a useful variable.
3. The empirical results indicates that the relative efficiency between window analysis and the other models is very similar. Window analysis is more confirmable and steady compared to other models.
4. Under the non-controllable variable model, considered more factors which the superintendents couldn’t control. In this situation, the relative efficiency changed.

並列關鍵字

Data Envelopment Analysis ；performance appraisal；Electronic Component Industries

參考文獻

27.王正義(2005)，「產業情報-電子零組件關聯圖暨廠商名錄」，開文科技股份有限公司。

18.張麗雪(2004)，「我國國際觀光旅館業之經營效率評估」，元智大學會計研究所碩士論文。

28.吳柏樹(2004)，「台灣地區產物保險業經營效率之研究」，元智大學會計研究所碩士論文。

2.Antreas D. Athanassopoulos, and Dimitris Giokas ( 1998 ),“Technical Efficiency and Economies of Scale in State Owned Enterprises: The Hellenic telecommunications organisation”, European Journal of Operation Research, Vol. 107, 62-75.

3.Banker, R.D.,Charnes,A., and Cooper,W.W. (1984), “Some Models for Estimating Technical and Scale Inefficiencies in Data Envelopment Analysis”, Management Science, Vol.30, No.9, 1078-1092.