台灣印刷電路板產業之營運及整合效率之研究---資料包絡分析法之應用

Abstract

本研究採用資料包絡分析法（Data Envelopment Analysis，簡稱DEA），針對台灣上市、上櫃主要的24家印刷電路板廠商，就其民國88年至90年主要績效指標，進行連續多期績效分析，提供相對無效率廠商改善對策以提昇效率及優化資源配置，並提出其努力的參考標竿。而資料包絡分析法是一種用以衡量具有共同多投入和產出項的事業單位之相對效率及管理績效。本研究更進一步利用多階式效率評選法及多階式最佳整合效率值變化排序法，進行先同類產品群後異類產品群廠商之多階式最佳整合組合模式，即生產同類產品的各個印刷電路板廠商群先進行水平式的整合，再進行異類產品群廠商之垂直式的整合，其過程乃經由計算多種廠商整合組合之個體及產業之相對效率值的變化，以找出印刷電路板產業最適整合組合。由實證研究結果顯示：(1)民國90年之總效率值為0.796，而純粹技術效率和規模效率值分別為0.834及0.984，表示印刷電路板產業之無效率主要源自於技術之無效率，而且有5家廠商處於規模報酬遞減，其餘19家廠商皆處於固定規模報酬，即為最適規模且最具生產力。(2)連續3年之效率分析，從民國88年至90年之總效率值分別為0.889、0.894及0.796，表示有呈現衰退的現象，而其高效率單位家數分別為10家、10家及7家，而非效率單位家數則分別為14家、14家及17家，而在24家印刷電路板廠商中，效率變動呈現進步現象的有4家，維持不變的有7家，呈現上下波動的有5家，及呈現衰退的有8家，此結果可作為經營管理者的一個警惕。(3)以多階式效率評選法，進行先水平後垂直之最佳整合組合，即生產同類產品的印刷電路板廠商群先進行水平式的整合，再進行異類產品廠商群之垂直式的整合，這樣的整合模式是比較容易成功的，並發現產業由24家整合成10家後，即可達到最佳的產業效率。綜觀本研究之結果，不僅有系統的評估產業績效表現，提供改善對策，更為產業整合建構具備科學根據及分析邏輯之架構。 關鍵字：資料包絡分析法、印刷電路板、績效評估、效率、整合、水平購併、垂直購併

In this paper we analyze the PCB enterprises in Taiwan between 1999 and 2001 using the Data Envelopment Analysis (DEA), we measure the relative efficiency of PCB enterprises in Taiwan according to liabilities, common share, operating expenses, number of employees, revenues, operating income and non-operating income in the period from 1999 until 2001.DEA is a methodology for analyzing the relative efficiency and managerial performance of productive units, having the same multiple inputs and multiple outputs, we provide the suggestions of the efficiency improvement for those inefficiency units as to improve the relative efficiency by re-allocate the resources and the reference set of efficiency units for improvement. This paper used the multiple levels of integration method and conducted the horizontal merger first then the vertical merger later to evaluate the possible mergers and acquisition set and conclude the most suitable combination. We have reached the major conclusions as following: (1) The technical efficiency scores, for which is under the constant- returns- to- scale assumption (CCR model) in 2001was 0.796, while the pure technical efficiency scores (variable- returns- to-scale, BCC model) was 0.834 and the scale efficiency was 0.984, indicated that the source of the technical inefficiency was due to the pure technical inefficiency, 5 out of the 24 PCB decision making units (DMUs) were under the decreasing returns-to-scale while others remained as constant returns-to-scale, for which has better productivity. (2) The average technical efficiency in the period of 1999 ~2000 was decline; they were 0.889, 0.894 and 0.796 respectively. The number of efficiency units in the period was 10, 10 and 7 respectively, while the number of the inefficiency units was 14,14 and 17 respectively. For the level classification of the efficiency, among the 24 PCB decision making units, there was 4 units with the efficiency improved continuously, 7 units remained unchanged, 5 units was bouncing and 8 units was decline. This is a sign to the management team of the PCB enterprises. (3) For the part of merger, we used the method of multiple level integration, we started with horizontal merger first, then the vertical merger, and this is the most popular way that the merge most likely to happen and success, we find that the average technical efficiency was optimized by integrated the 24 PCB DMUs to 10. Key words: Data Envelopment Analysis, Printed Circuit Board, Performance Assessment, Efficiency, Integration, Horizontal Merger, Vertical Merger.