Taiwanese IC packing and testing industry has contributed more than 44% and 60%
respectively of global market shares since 2006. Data Envelopment Analysis (DEA) is widely used to measure the efficiency of peer decision making units (DMUs). This study employs the two-stage DEA approach, proposed by Chen et al. (2010), to discompose the production processes of the IC packaging and testing industry into two-stage processes, where all the outputs from the first stage are intermediate measures that make up the inputs to the second stage. It can not only guarantee the projection points on the frontier, but also provide the optimal intermediate measures, which can offer useful managerial implications.
The empirical results, based on data obtained from Taiwan Economic Journal over the period 2006-2009, show that the technical efficiency of KH model and CCZ model presents the similar result in the performance evaluation of Taiwanese IC packing and testing industry.
However, only CCZ model can indicate that the adjusted intermediate measures are different between the firms associated with different niche products and target markets. It means that the different types of companies must adopt distinct strategies to achieve their efficiency.
For instance, ASE should increase the current scale of revenue aggressively and reduce the current patents deliberately and SPIL should increase both revenue and patents.
Furthermore, most of small companies should focus on their niche for products differentiation. Other findings include: (1) Even if the firms invest in the foreign market, own more capital or diversify product lines, it is not easy to increase the observed revenue up to the optimal values in a short time. How to increase the current scale of revenue is the main task for most Taiwanese IC packing and testing companies. (2) The obvious discrepancy of the average adjusted intermediate measures percentage with different types happened in 2008 due to global economic recession and the weak demand. The firms without FDI need to increase revenue quicker than the firm with FDI during the economic downturn. The same action needs to take by OTC and non-turnkey company. (3) The non-operating income compared to revenue is less impact on net income so it is only a considerable strategy for the small companies during the economic downturn. (4) Most IC packing and testing firms should increase more or less patents to enhance their technical level or defense tort when they face the international competition and invest the foreign market directly.
It is common that most companies always operate inefficiently and operate efficiently sometimes, i.e., inefficiency is a common problem for IC packing and testing industry even if the industry has been a benchmark of supply chain management in the world.
The shortage of revenue and patents is a common phenomenon in this industry. So this study provides some managerial implications. First, advanced technique is the core value of high-technology industry so patents development is still necessary. However, overmuch development might be waste and conservative strategy might result in the technical capability loss gradually. Second, the risk of markets or customers concentration will impact revenue seriously once any economic recession happens. IC industry is very sensitive to changes in the economy so spreading markets or customers is necessary. If IC packing and testing firms can receive orders from different types of company, like IDM, IC design firms and Fab-lite firms, or different countries and regions, they can reduce the risk of volatility market and benefit greatly in the future. Once Taiwanese IC packing and testing firms can improve their efficiency effectively, they can continue to keep the predominance in the highly competitive industry.
Reference
Banker, R. D., Charnes, A., and Cooper, W.W., 1984, “Some Models for Estimating Technical and Scale Inefficiencies in Data Envelopment Analysis”, Management Science, 30(9), 1078-1092.
Bosworth, D., and Rogers, M., 2001, “Market value, R&D and intellectual property: An Empirical Analysis of Large Australian Firms”, Economic Record 77, 323-337.
Charnes, A., Cooper, W.W., and Rhodes, N., 1978, “Measuring the Efficiency of Decision Making Units”, European Journal of Operational Research, 2(6), 429-444.
Chen, Y., Cook, W.D., and Zhu, J., 2010, “Deriving the DEA Frontier for Two-stage Processes”, European Journal of Operational Research, 202(1), 138-142.
Chen, Y., Liang, L., and Zhu, J., 2009, “Equivalence in Two-stage DEA Approaches”, European Journal of Operational Research, 193, 600-604.
Chen, Y., and Zhu, J., 2004, “Measuring Information Technology's Indirect Impact on Firm Performance”, Information Technology and Management, 5(1-2), 9-22.
Hai, H.L., 2004, “Evaluating Efficiency of the Semiconductor Manufacturers in Taiwan”, Management Research, 71-89. (in Chinese)
Hung, S.W., and Lu, W.M., 2008, “The Comparative Productivity Efficiency of Taiwan's Integrated Circuits Packing/Testing Firms”, INFOR Journal, 46(3), 189-198.
Jiang, J.C. and Chen, C.A., 2008, “The Supply Chain Patterns and Performance Indicators of IC industry in Taiwan”, Management of Innovation and Technology, ICMIT 2008. 4th IEEE International Conference on
Kao, C., and Hwang, S.N., 2008, “Efficiency Decomposition in Two-stage Data Envelopment Analysis: An Application to Non-life Insurance Companies in Taiwan”, European Journal of Operational Research, 185(1), 418-429.
Kozmetsky G., and Yue P., 1998, “Comparative Performance of Global Semiconductor Companies”, OMAGA-The International Journal of Management Science, 26(2), 153-175.
Liu, F.H. and Wang, P.H, 2008, “DEA Malmquist Productivity Measure: Taiwanese Semiconductor Companies”, International Journal of Production Economics, 112, 367-379.
Pan, W.H., Feng, Y.Y., Huang, Y.C., and Chen, Y.K., 2008, “Performance Review of Taiwanese IC Design Industry: DEA-based Malmquist Productivity Measure”, WSEAS Transactions on Computers, 7(8), 1291-1299.
Seiford, L.M., and Zhu, J., 1999, “Profitability and Marketability of the Top 55 U.S.
Commercial Banks”, Management Science, 45(9), 1270-1288.
Thore, S., Phillips, F., Ruefli, T.W. and Yue, P., 1996, “DEA and the Management of The Product Cycle：The U.S. Computer Industry, Computer and Operation Research”, 23(4), 341-356.
Wang, H.P., and Hong, Z.Y., 2007, “The Innovative Capabilities and Performances of Taiwan’s Semiconductor Packaging and Testing Industry”, National Chiao Tung University, MBA thesis, pp.IV.
Wu, J.Z., and Chien, C.F., 2008, “Modeling Strategic Semiconductor Assembly Outsourcing Decisions Based on Empirical Settings”, OR Spectrum, 30, 401-430.
Yang, G.Q., and Liu, S.Z., 2004, “Patent Strategies and Patent Portfolio Design in High-Tech Industry: An Empirical Study on Semiconductor Package Manufacturers”, National Chiao Tung University, Master's thesis (in Chinese)
Zhu, J., 2000, “Multi-factor Performance Measure Model With an Application to Fortune 500 Companies”, European Journal of Operational Research, 123(1), 105-124.
Zhu, J., 2003, “Quantitative Models for Performance Evaluation and Benchmarking: Data Envelopment Analysis With Spreadsheets”, Kluwer Academic Publishers, Boston.
Others
ASE web site http://www.aseglobal.com/
Annual Statistics of Taiwanese IC Industry, 2008, Taiwan Semiconductor Industry Association (TSIA) http://www.tsia.org.tw/service/news_more.asp?zvpWmcZ=market
observation post system, taiwan stock exchange http://emops.twse.com.tw/emops_all.htm Semiconductor Industry Yearbook, 2009, Industrial Economics & Knowledge Center (IEK),
2-12, 6-40 (in Chinese)
Taiwan Patent Search http://twpat.tipo.gov.tw/tipotwoc/tipotwekm
Table 1 Taiwanese IC packing and testing industry global ranking and market share (Unit: US$ million)
Type Year Global
Ranking Value Market Share
2006 1 6,490 47.8%
2007 1 6,951 44.4%
IC packing
2008 1 7,040 47.3%
2006 1 2,840 60.3%
2007 1 3,119 63.0%
IC testing
2008 1 3,060 65.2%
Table 2 Definition of input and output variables
Variables Definition Inputs of the first stage
Administrative expenditures
consist of selling, general, administrative expenditures and indirect labor cost (NT$ billion)
R&D the expenditure on discovering new knowledge about products, processes, and services, and then applying that knowledge to create new and improved products, processes, and services that fill market needs (NT$ billion)
Labor costs the component of costs of goods sold reflect the cost of obtaining labor that is used to manufacture finished goods (NT$ billion) Depreciation of fixed
assets
reflect the execution cost of algorithms over a sequence of operations in fixed assets and intangible assets that are used to manufacture finished goods (NT$ billion)
Raw materials the cost of obtaining raw material and other facility expenditure in the manufacture of finished goods (NT$ billion)
Non-operating costs incurred in performance of activities not directly related to the main business of a firm, such as costs incurred from currency exchange, charges on obsolescence of inventory, interest charges or other costs of borrowing (NT$ billion)
Intermediate measures
Revenue total amount of money received by the company for products sold or services rendered and deducted all return of goods and credit amount (NT$ billion)
Patents a temporary government-granted monopoly right on something made by an inventor (count)
Non-operating income total amount of money received by the company but is not derived from operations and does not occur on an ongoing basis (NT$
billion) Outputs of the second stage
Market value obtained by multiplying the number of common shares outstanding by the price per common share as of December 31 (NT$ billion) ROE measures a corporation's profitability by revealing how
much profit a company generates with the money shareholders have invested (NT$)
Table 3 Descriptive statistics of input and output variables
Variables Mean Std. Dev. Minimum Maximum Administrative
expenditures 0.82 1.62 0.04 7.06
R&D 0.43 0.89 0.01 3.78
Non-operating costs 0.51 1.09 0.00 4.61
Revenue 17.17 27.35 0.16 101.68
Non-operating income 0.50 1.12 0.00 6.43 Intermediate
measures
Patents 18 48 0 263
Market value 30.95 50.45 0.26 180.10 Output of the
second stage ROE 0.14 0.09 0.00 0.38
Table 4 Correlation coefficients of input-output variables analysis Variables Revenue Non-operating
income Patents
The value in parentheses is P-value
Table 5 Average technical efficiency of KH model and CCZ model with different types
Table 6 Mann-Whitney U test of technical efficiency with different types
Table 7 Descriptive statistics of adjusted intermediate measures
Mean Std. Dev. Minimum Maximum
Adjusted revenue 1.71 3.85 -1.17 18.34
Adjusted non-operating income -0.05 0.63 -4.40 0.87
Adjusted patents -5.91 34.37 -199.00 22.00
Adjusted revenue (%) 8.96 14.71 -13.20 67.10 Adjusted non-operating income (%) 34.94 108.20 -93.00 680.20
# of
observation KH model CCZ model
Yes 40 0.088 0.492
FDI No 16 0.156 0.579
Listed 32 0.070 0.467
Status
OTC 24 0.157 0.583
Turnkey 32 0.074 0.472
Type
Non-turnkey 24 0.152 0.576
Average 0.107 0.517
KH model CCZ model FDI Z (P-value) -2.141 (0.032) -1.053 (0.292) Status Z (P-value) -2.902 (0.004) -1.632 (0.103) Type Z (P-value) -1.904 (0.057) -1.367 (0.172)
Table 8 Average adjusted revenue and non-operating income % with different types
Ob: Observed values Op: Optimal values Ad: Adjusted values
Table 9 Average adjusted patents with different types
Revenue Non-operating income
# of observation
Ob Op Ad % Ob Op Ad %
Yes 40 22.60 24.96 9.42 0.67 0.59 23.94
FDI
No 16 3.60 3.68 7.83 0.07 0.08 62.43
Listed 32 28.17 31.00 10.09 0.83 0.75 27.41 Status
OTC 24 2.50 2.71 7.46 0.05 0.05 44.98
Turnkey 32 27.07 29.67 9.87 0.80 0.70 23.22 Type
Non-turnkey 24 3.98 4.49 7.75 0.09 0.11 50.56
# of observation
(with ASE)
Ob Op Ad
# of observation
(without ASE)
Ob Op Ad
Yes 40 24.85 16.45 -8.40 36 8.39 11.16 2.77
FDI
No 16 0.81 1.35 0.54 16 0.81 1.35 0.54
Listed 32 30.91 19.49 -11.42 28 10.61 13.12 2.51 Status
OTC 24 0.75 2.33 1.58 24 0.75 2.33 1.58
Turnkey 32 30.16 18.96 -11.19 28 9.75 12.52 2.77 Type
Non-turnkey 24 1.75 3.03 1.28 24 1.75 3.03 1.28
Table 10 Mann-Whitney U test of adjusted intermediate measures with different types
Revenue Revenue (%)
Non-operating income
Non-operating
income (%) Patents Patents
(without ASE)
Z -2.452 -1.217 -0.856 -0.581 -2.034 -2.870
FDI
P-value 0.014 0.224 0.392 0.561 0.042 0.004
Z -3.002 -1.799 -1.579 -0.638 -0.473 -1.407
Status
P-value 0.003 0.072 0.114 0.523 0.637 0.159
Z -2.562 -1.683 -0.439 -0.439 -0.870 -1.848
Type
P-value 0.010 0.092 0.946 0.660 0.384 0.065
Table 11 Average adjusted revenue and non-operating income % by year
Average adjusted revenue Average adjusted non-operating income
2006 2007 2008 2009 2006 2007 2008 2009
Yes 4.57 3.45 16.07 13.59 -0.71 17.46 39.28 39.75
FDI
No 2.69 -0.79 5.69 23.72 30.39 -14.08 27.04 206.35
Listed 5.19 4.58 16.86 13.73 5.20 15.96 27.57 60.90 Status
OTC 2.49 -0.90 8.10 20.16 12.15 -1.57 46.73 122.61
Turnkey 6.43 3.47 16.97 12.63 -5.10 8.70 20.30 69.00 Type
Non-turnkey 0.84 0.58 7.95 21.63 25.88 8.11 56.44 111.81
Figure 1 IC Industry Diagram (draw by author)
20.6 19.8
15.2 18.2 20.6 23.5
27.4 24.2
17.8 19.8 22.4
24.5
0 10 20 30 40 50 60
2007 2008 2009(E) 2010(F) 2011(F) 2012(F)
Year Output Value
US$Billon SATS IDM
Figure 2 Global output value of IDMs and SATSs
IC Main Manufacturing Flow
Design Fabrication Packing Testing
IC Fabless IC Foundry
Semiconductor Assembly and Test Services (SATS)Integrated Device Manufacturer (IDM) Company Type
Global IC Packing and Testing Capacity
IC Main Manufacturing Flow
Design Fabrication Packing Testing
IC Fabless IC Foundry
Semiconductor Assembly and Test Services (SATS)Integrated Device Manufacturer (IDM) Company Type
Global IC Packing and Testing Capacity
-5%
Figure 4-1 Average adjusted revenue % by FDI Figure 5-1 Average adjusted non-operating income % by FDI
Figure 4-2 Average adjusted revenue % by status Figure 5-2 Average adjusted non-operating income % by status
Figure 4-3 Average adjusted revenue % by type Figure 5-3 Average adjusted non-operating income % by type
Appendix
Appendix 1 Basic data sets of Taiwan IC packing and testing firms
Company
KYEC 1987 7 12,808,540 yes yes Listed testing
Powertech 1997 6 6,308,000 yes yes Listed packing &
Data resource: Summary according to TEJ, Market Observation Post System of Taiwan stock exchange
Appendix 2 Literature Summary of Electronic Industry and Two-stage DEA
Authors Title Inputs Outputs
Thore et al.
Selling, general, and administrative expenditures
Labor force, in thousands of workers
Plant, property, and equipment, gross
Capital expenditures
Expenditures on R&D
gross sales revenues
income before taxes
market capitalization
selling and operational expense
Gross sales revenue
Income before taxes
Market Value
Authors Title Inputs Intermediates Outputs
Seiford and Hai (2004) Evaluating Efficiency
of the Semiconductor
Appendix 3 Accumulative FDI in China (Unit: NT$K)
Year ASE SPIL KYEC
2006 6,251,913 1,620,500 1,483,887 2007 8,869,916 2,271,150 1,476,376 2008 11,809,884 2,947,500 2,149,220