Research Article A Study of the Strategic Alliance for EMS...

Research ArticleA Study of the Strategic Alliance for EMS IndustryThe Application of a Hybrid DEA and GM (1 1) Approach

Chia Nan Wang1 Nhu Ty Nguyen12 Thanh Tuyen Tran2 and Bui Bich Huong1

1 Department of Industrial Engineering and Management National Kaohsiung University of Applied Sciences415 Jiangong Sanmin District Kaohsiung 807 Taiwan

2 International Relations Office Lac Hong University No 10 Huynh Van Nghe Street Bien Hoa Dong Nai 71000 Vietnam

Correspondence should be addressed to Nhu Ty Nguyen nhutynguyengmailcom

Received 13 April 2014 Revised 15 September 2014 Accepted 25 September 2014

Academic Editor Jung-Fa Tsai

Copyright copy 2015 Chia Nan Wang et alThis is an open access article distributed under the Creative CommonsAttribution Licensewhich permits unrestricted use distribution and reproduction in any medium provided the original work is properly cited

Choosing a partner is a critical factor for success in international strategic alliances although criteria for partner selection varybetween developed and transitional marketsThis study aims to develop effective methods to assist enterprise to measure the firmsrsquooperation efficiency find out the candidate priority under several different inputs and outputs and forecast the values of thosevariables in the future The methodologies are constructed by the concepts of Data Envelopment Analysis (DEA) and grey model(GM) Realistic data in four consecutive years (2009ndash2012) a total of 20 companies of the Electronic Manufacturing Service (EMS)industry that went public are completely collected This paper tries to help target companymdashDMU1mdashto find the right alliancepartners By our proposed approach the results show the priority in the recent years The research study is hopefully of interest tomanagers who are in manufacturing industry in general and EMS enterprises in particular

1 Introduction

The Electronic Manufacturing Service (EMS) also known asContract Manufacturing has become a hot ticket industrynowadays as the peoplersquos demand on using more advancedtechnological electronic devices is significantly increasingThis segment has witnessed the strong development through-out the last ten years Therefore EMS is remaining animportant role in the countryrsquos long term prosperity [1]Specifically ldquoIt creates skilled jobs and generates revenues fornational treasuries in the form of exports and investment Italso has a strong beneficiary role in terms of its contributionto the physical infrastructure of an economy and spill-over effects to other areas such as science construction andlogisticsrdquo according to Global Manufacturing report (2010)

Today the mentioned market is dominated by manyEMSmanufacturers but this study only selects 20 enterpriseswhich play major roles in the industry and can representthe whole industry in stock market among them Hon HaiPrecision Industry Co Ltd is the first rank Founded inTaiwan by its current CEO Terry Gou in 1974 up to now

Hon Hai has more than 1200000 employees The prod-ucts vary in type including connectors cables enclosureswiredwireless communication products optical productsand power supply modules Hon Hai Precision Industry CoLtd offers these products for use in the information tech-nology communications automotive equipment precisionmolding automobile and consumer electronics industriesIts key customers consist almost entirely of the well-knownhigh-tech companies such as Apple Cisco Dell Nokia andSony

According to annual report of MMI 2012 for the firsttime this study found Top 50 sales reached a new highof $2239 billion evidencing an encouraging growth withinthe industry Top 50 sales grew by 48 last year despiteend market weakness However without the contribution ofindustry giant Hon Hai Precision Industry sales would havefallen by 50 Hon Hai represented a 59 majority of Top50 revenue in 2012 Figure 1 demonstrates that the companyis doing a great job which in turn helps strengthen the EMSsector

Hindawi Publishing Corporatione Scientific World JournalVolume 2015 Article ID 948793 14 pageshttpdxdoiorg1011552015948793

2 The Scientific World Journal

0

500000

1000000

1500000

Dec 2011998400 Mar 2012

998400 Jun 2012998400 Sep 2012

998400 Dec 2012

RevenueNet income (MTWD)

Figure 1 Revenue and net income of Hon Hai Precision IndustryCo Ltd in period of 2011-2012 (Source Bloomberg News)

The footprint of HonHai Precision Industry is all over theworld in which China is the operation center the rest of theplants are located in India Malaysia Japan Brazil MexicoHungary Slovakia and Czech Republic and Vietnam Forthe future development planning Foxconn Group expectsto expand its production scale not only by looking forinvestment opportunity in more other territories but also byupgrading and building new factories in countries where ithas already run its business

The purpose of this research is to provide an assess-ment model based on Grey theory GM (1 1) and DataEnvelopment Analysis (DEA) helping the target companymdashFoxconn or Hon Haimdashto make a well-considered decisionin finding the right partners At the same time the studyalso provides the prediction about firmsrsquo business in thefuture which is relevant for them when setting strategies forproduction capacity planning and for investment decisionmaking and whether they should expand their business ininternational market or not The results of the case study canbe the reference of the strategic alliances partner selection forworldwide EMS providers

2 Literature Review

21 Strategic Alliances Mockler [2] considers the agreementsbetween companies (partners) to reach objectives of commoninterest Strategic alliances are among the various optionswhich companies can use to achieve their goals they arebased on cooperation between companies This point ofview was shared by Parkhe [3] ldquoStrategic alliance means thepermanent cooperative agreement between the companiesincluding the inflow and linking of resources and its cooper-ative purpose is finishing their company missions in strategicalliancerdquo These definitions emphasize the importance of thecommon business goals of the companies when carrying outalliance strategy

Chan et al [4] see strategic alliance as a cooperativeagreement between various organizations Its purpose aims atreaching the competitive advantage for enterprises that forma partnership Strategic alliance brings together otherwise

independent firms to share resources in product designproduction marketing or distribution In simple words astrategic alliance is sometimes just referred to as ldquopartner-shiprdquo that offers businesses a chance to join forces for amutually beneficial opportunity and sustained competitiveadvantage [5] These definitions attach importance to com-petitive advantages among enterprises in strategic alliance

Wang and Liu [6] researched the evaluation and selectionof partner in logistics strategic alliance The study establishesthe index system of partners in logistics strategic allianceaccording to SCOR index model to evaluate the partnersobjectively Based on AHP method the researchers importthe TOPSIS method to standardize the evaluation result sothat more objective and effective evaluation result could beachieved and the appropriate partner is indicated finally

Oh [7] focused on global strategic alliances in thetelecommunications industry the author points out theglobal marketplace is demanding an increasingly sophisti-cated seamless worldwide communications network alongwith one inexpensive contract for every service GlobalStrategic Alliances is a way of meeting these needs in thecontext of limited available resources

Dittrich et al [8] aimed to show that alliance networkscan play an important role in facilitating large-scale strate-gic change projects They focus on the particular case ofIBM whose radical redirection from an exploitation strat-egy towards an exploration strategy was realized by majorchanges in its network strategy The findings of this papersuggest that the traditional view of large firms as being slowto adapt may not be valid because alliance networks can beused to overcome inertia

22 Grey SystemTheory andDEA Forecast is the explanationof something that has not yet been previously observedor is unknown In recent years many methods have beenproposed for forecasting especially forecasting in businesssuch as fuzzy theory neural networks and grey predictionGrey system theory developed originally in early 1980s by Ju-Long [9] is an interdisciplinary scientific area This theoryhas become a very popular method to deal with uncertaintyproblems under discrete data and incomplete informationHaving superiority to conventional statistical models greymodels require only a limited amount of data to estimate thebehavior of unknown systems [10]

Model Data Envelopment Analysis (DEA) was firstdescribed in a 1978 paper by Charnes Copper and Rhodes(CCR) [11] In that work the authors described a ldquodataorientedrdquo approach for evaluating the performance of a setof peer entities called Decision Making Units (DMUs) whichconvert multiple inputs into multiple outputs The DMU canbe banks managers shipping companies or what we willevaluate in this paper areas within EMS industry Recentyears had a great variety of application of DEA in both publicand private sectors of different countries

Chandraprakaikul and Suebpongsakorn [12] used dataenvelopment analysis to explore the operation performancesof 55 Thai logistics companies from 2007 to 2010 The resultspoint out the reasons for the inefficient Decision Making

The Scientific World Journal 3

Part 1 Data collection

Step 1 Collect the DMUs

Step 2 Select inputoutput

Step 3 Grey prediction

Step 4 Forecasting accuracy

Literature review

Step 5 Pearson correlation

Based on resource dependence theory

Step 6 Choose the DEA model

Step 7 Analysis before alliance

Step 8 Factor modification

Step 9 Analysis after alliance

Step 10 Summary

Part 2 Prediction

Part 3DEA

Part 4Conclusions

If error is too high reselect inputoutput

factors

Add the targetcompany

If not satisfy the DEA assumption reselect inputoutput factors

Figure 2 Research process

Units and provide improving directions for the inefficientcompanies accordingly

Zhao et al [13] made performance benchmarking byusing Data Envelopment Analysis on Chinese coal miningindustry In the same time they measured the changes ofefficiency of coal mining industry byMalmquist ProductivityIndex

A lot of methods are used for ranking and efficiencyevaluation However most of the used models do provideno projection of Pareto efficiencyThus calculating the superefficiency becomes a significant issue [14] Through thisSuper SBM-I-V and grey model are integrated together todeal with the super efficiency and making strategy alliancepartner

Wang and Lee [15] focused on global strategic alliances inthe hi-tech industries in Taiwan By combining grey modeland DEA the researcher develops an effective method tohelp Taiwanrsquos TFT-LCD industry to evaluate the operationefficiency and find the right candidate for alliances Theresults of the study could assist companiesrsquo managers inmaking decisions in business extension

3 Methodology

31 Research Process Figure 2 shows details about this studystep-by-step and through this we can set up a whole imageabout how to integrate DEA and GM approaches Step 1and step 2 are about the setting stage mentioned earlier Instep 3 and step 4 grey prediction has been based on greymodel GM (1 1) to predict the data values on 2013 and 2014However the forecast always show error Therefore in thisstudy the MAPE is applied to measure the forecasting errorIf the forecasting error is too high the study has to reselectthe inputs and outputs

In this paper the software of DEA-Solver is employed tocalculate super-SBM-I-V model During step 5 the efficiencymeasuring by ranking DMUsrsquo performance is then achievedThe formulation of DEA is to measure the efficiency of eachdecision making unit by constructing a relative efficiencyscore via the transformation of the multiple inputs andoutputs into a ratio of a single virtual output to a singlevirtual input Therefore to test whether the data matchwith the basic assumptions of DEA methodology or not


correlation analysis of variables is calculated to verify forpositive relationship between the selected inputs and outputs(step 6) If the variables get with the negative coefficient theyneed to be removed then we will go back to step 2 of theselection process to redo the variable selection until they cansatisfy our condition In this study we employ the PearsonCorrelation Coefficient Test

The purpose of step 7 is to rank the efficiency of eachdecision making unit by applying the super-SBM-I-V modelin the realistic data in 2012 In particular by this way theresearcher also can find out the target companyrsquos position incomparison with the other 19 EMS competitors By step 8a lot of information after analysis is given out since we havethe results of step 7 However the study does not recommendstrategic alliance in this step until further analysis of step9 In this step the researcher has to stand on the side ofthe candidate companies which are selected for the targetcompanyrsquos alliance to find the possible way of cooperation

32 Collecting DMUs After doing the survey the Electronicmanufacturing services (EMS) market segments the studyfinds out 20 enterprises in the MMI Top 50 list of the worldrsquoslargest EMS providersThen the analysis was only conductedon the 20 companies which are stable in market and canprovide the complete data for 4 consecutive years (2009ndash2012) in Bloomberg Business week news Moreover these 20qualified companies playmajor roles in the EMS industry andcan represent for whole industry in stock market (Table 1)

In this study DMU1 is set as the target company with theheadquarters located in Taipei Taiwan In the globalizationand competition environment strategic alliance could be agreat way for DMU1 to require resources and extend itsbusiness map

33 Nonradial Super Efficiency Model (Super-SBM) In thepresent study a DEA model ldquoslack-based measure of super-efficiencyrdquo (super SBM) was usedThis model was developedon ldquoslacks-based measure of efficiencyrdquo (SBM) introduced byTone [16]

In this model with 119899 DMUs with the input and outputmatrices 119883 = (119909

119894119895) isin 119877119898times119899 and 119884 = (119884

119894119895) isin 119877119904times119899 respectively

120582 is a nonnegative vector in 119877119899 The vectors 119878minus isin 119877119898 and 119878+ isin119877119904 indicate the input excess and output shortfall respectively

Themodel formulation provides a constant return to scaleas follows [17]

min 120588 =1 minus (1119898)sum

119898

119894=1119904minus1198941199091198940

1 + (1119904)sum119904

119894=1119904minus1198941199101198940

(1)

subject to

1199090= 119883120582 + 119904

minus

1199100= 119884120582 minus 119904

+

120582 ge 0 119904minusge 0 119904

+ge 0

(2)

The variables 119878+ and 119878minusmeasure the distance of inputs119883120582and outptut 119884120582 of a virtual unit from those of the unit eval-uated The numerator and the denominator of the objective

function of model (1) measure the average distance of inputsand outputs respectively from the efficiency threshold

Let an optimal solution for SBM be (119901lowast 120582lowast 119904minuslowast 119904+lowast) ADMU (119909

0 1199100) is SBM-efficient if 119901lowast = 1 This condition is

equivalent to 119878minuslowast = 0 and 119878+lowast = 0 no input excesses and nooutput shortfalls in any optimal solution SBM is nonradialand deals with inputoutput slacks directly The SBM returnsand efficiency measures between 0 and 1

The best performers have the full efficient status denotedby unity The super SBM model is based on the SBM modelTone [16] discriminated these efficient DMUs and rankedthe efficient DMUs by super-SBM model Assuming that theDMU (119909

0 1199100) is SBM-efficient 119901lowast = 1 super-SBM model is

as follows

min 120575 =(1119898)sum

119898

119894=11199091198941199091198940

(1119904) sum119904

119903=11199101199031199101199030

(3)

subject to

119909 ge

119899

sum119895=1119895 =0

120582119895119909119895

119910 le

119899

sum119895=1119895 =0

120582119895119909119895

119910 ge 1199090 119910 le 119910

0 119910119910 ge 119910

0 120582 ge 0

(4)

The input-oriented super SBM model is derived frommodel (3) with the denominator set to 1 The super SBMmodel returns a value of the objective function which isgreater than or equal to oneThe higher the value is the moreefficient the unit is

As in many DEA models it is crucial to consider how todeal with negative outputs in the evaluation of efficiency inSBM models too However negative data should have theirduly role in measuring efficiency hence a new scheme wasintroduced inDEA-Solver pro 41Manuel and the schemewaschanged as follows

Let us suppose 1199101199030le 0 It has defined 119910+

119903and 119910+

minus119903by

119910+

119903= max119895=1119899

119910119903119895| 119910119903119895gt 0

119910+

119903= min119895=1119899

119910119903119895| 119910119903119895gt 0

(5)

If the output 119903 has no positive elements then it is definedas 119910+119903

= 119910+minus119903

= 1 The term is replaced by 119904+1199031199101199030

in theobjective function in the following wayThe value 119910

1199030is never

changed in the constraints(1) 119910+119903= 119910+minus119903

= 1 the term is replaced by

119904+119903

119910+minus119903(119910+

119903minus 119910+minus119903) (119910+

119903minus 1199101199030) (6)

(2)

119904+119903

(119910+minus119903)2119861 (119910+

119903minus 1199101199030) (7)

where 119861 is a large positive number (in DEA-Solver 119861 = 100)


Table 1 List of EMS companies

Number order Code DMUs Companies Headquarter addresses1 DMU1 Hon Hai Precision Industry New Taipei Taiwan2 DMU2 Jabil Circuit Inc St Petersburg FL3 DMU3 Sanmina Corporation San Jose CA

4 DMU4 Shenzhen Kaifa Technology CoLtd Shenzhen China

5 DMU5 Benchmark Electronics Inc Angleton TX6 DMU6 Plexus Corp Neenah WI7 DMU7 Kitron Billingstad Norway8 DMU8 Di-Nikko Engineering Co Ltd Nikko Japan9 DMU9 Fabrinet Pathumthani Thailand10 DMU10 Flextronics Intrsquol Ltd Singapore11 DMU11 Hana Microelectronics Pcl Bangkok Thailand12 DMU12 Global Brand Manufacture Ltd New Taipei Taiwan13 DMU13 SIIX Corporation Osaka Japan14 DMU14 Venture Corporation Limited Singapore

15 DMU15 Orient SemiconductorElectronics Ltd Kaohsiung Taiwan

16 DMU16 Universal Scientific Industrial(Shanghai) Co Ltd Shanghai China

17 DMU17 VS Industry Berhad Senai Malaysia

18 DMU18 Integrated Micro-ElectronicsInc Laguna Philippines

19 DMU19 Celestica Inc Toronto Canada20 DMU20 PartnerTech AB Vellinge SwedenSource the MMI Top 50 for 2012

In any case the denominator is positive and strictlyless than 119910+

minus119903 Furthermore it is inversely proportion to

the distance 119910+119903minus 1199101199030 This scheme therefore concerns the

magnitude of the nonpositive output positively The scoreobtained is units invariant that is it is independent of theunits of measurement used

34 Grey Forecasting Model Although it is not necessary toemploy all the data from the original series to construct theGM (1 1) the potency of the series must be more than fourIn addition the data must be taken at equal intervals and inconsecutive order without bypassing any data [17] The GM(1 1)model constructing process is described as follows

Denote the variable primitive series119883(0) as formula

119883(0)

= (119883(0)

(1) 119883(0)

(2) 119883(0)

(119899))

119899 ge 4

(8)

where119883(0) is a nonnegative sequence 119899 is the number of dataobserved

Accumulating Generation Operator (AGO) is one ofthe most important characteristics of grey theory with theaim at eliminating the uncertainty of the primitive data

and smoothing the randomnessThe accumulated generatingoperation (AGO) formation of119883(0) is defined as

119883(1)

= (119883(1)

(1) 119883(1)

(2) 119883(1)

(119899))

119899 ge 4

(9)

where

119883(1)

(1) = 119883(0)

(1)

119883(1)

(119896) =

119896

sum119894=1

119883(0)

(119894) 119896 = 1 2 3 119899

(10)

The generated mean sequence 119885(1) of119883(1) is defined as

119885(1)

= (119885(1)

(1) 119885(1)

(2) 119885(1)

(119899)) (11)

where 119885(1)(119896) is the mean value of adjacent data that is

119885(1)

(119896) =1

2(119883(1)

(119896) + 119883(1)

(119896 minus 1))

119896 = 2 3 119899

(12)


From the AGO sequence 119883(1) a GM (1 1) model whichcorresponds to the first order different equation1198831(119896) can beconstructed as follows

1198891198831 (119896)

119889119896+ 119886119883(1)

(119896) = 119887 (13)

where parameters 119886 and 119887 are called the developing coefficientand grey input respectively

In practice parameters 119886 and 119887 are not calculated directlyfrom (13) Hence the solution of above equation can beobtained using the least squares method That is

119883(1)

(119896 + 1) = [119883(0)

(1) minus119887

119886] 119890minus119886119896

+119887

119886 (14)

where119883(1) (119896 + 1) denotes the prediction119883 at time point 119896+1and the coefficients [119886 119887]119879 can be obtained by the OrdinaryLeast Squares (OLS) method

[119886 119887]119879= (119861119879119861)minus1

119861119879119884

119884 =

[[[[[

[

119909(0)

(2)

119909(0) (3)

119909(0)

(119899)

]]]]]

]

119861 =

[[[[[

[

minus119911(1)

(2) 1

minus119911(1) (3) 1

minus119911(1) (119899) 1

]]]]]

]

(15)

where 119884 is called data series 119861 is called data matrix and[119886 119887]119879 is called parameter series

We obtained 119883(1) from (14) Let 119883(0) be the fitted andpredicted series

119883(0)

= 119883(0)

(1) 119883(0)

(2) 119883(0)

(119899)

where 119883(0)

(1) = 119883(0)

(1)

(16)

Applying the inverse accumulated generation operation(IAGO) namely

119883(0)

(119896 + 1) = [119883(0)

(1) minus119887

119886] 119890minus119886119896

(1 minus 119890119886) (17)

The grey model prediction is a local curve fitting extrap-olation scheme At least four data sets are required by thepredictor (14) to obtain a reasonably accurate prediction [18]

35 Establishing the Input and Output Variables In orderto apply DEA model it is particularly vital that inputsand outputs considered for the study be specified Besidesusing appropriate inputs and outputs should be consideredcarefully so that conclusions drawn may not be misledBy investigating some DEA literature reviews mentionedpreviously and the elements of the operation for EMS theresearchers decided to choose three inputs factors directlycontributing to the performance of the industry includingfixed assets operating expenses and cost of goods sold Theresearch selected the revenues operating income retainedearnings as output factors because they are the important

indexes to measure the performance of enterprises both incurrent and future situation (Table 2)

The study also applied DEA-based testing the correlationbetween input and output factors correlation which willclearly show whether those variables are suitable or not Theresult is indicated clearly in the next section

4 Empirical Results and Analysis

41 Forecasting Results The researchers use GM (1 1)modelto predict the realistic inputoutput factors for the next twoyears 2013 and 2014 In the Table 3 the study takes companyDMU1 as an example to understand how to compute in GM(1 1)model in period 2009ndash2012

This research selects the fixed assets of DMU1 as exampleto explain for calculation procedure other variables arecalculated in the same way The procedure is carried out stepby step as follows

First the researchers use the GM (1 1) model for tryingto forecast the variance of primitive series

1st create the primitive series

119883(0)

= (78714 913060 1192280 1309450) (18)

2nd perform the accumulated generating operation(AGO)

119883(1)

= (78714 17002 289248 420193)

119909(1)

(1) = 119909(0)

(1) = 78714

119909(1)

(2) = 119909(0)

(1) + 119909(0)

(2) = 17002

119909(1)

(3) = 119909(0)

(1) + 119909(0)

(2) + 119909(0)

(3) = 289248

119909(1)

(4) = 119909(0)

(1) + 119909(0)

(2) + 119909(0)

(3) + 119909(0)

(4) = 420193(19)

3rd create the different equations of GM (1 1)To find 119883(1) series and the following mean obtained by

the mean equation is

119911(1)

(2) =1

2(78714 + 17002) = 124367

119911(1)

(3) =1

2(17002 + 289248) = 229634

119911(1)

(4) =1

2(289248 + 420193) = 3547205

(20)

4th solve equationsTo find 119886 and 119887 the primitive series values are substituted

into the Grey differential equation to obtain

913060 + 119886 times 124367 = 119887

1192280 + 119886 times 229634 = 119887

1309550 + 119886 times 3547205 = 119887

(21)

Convert the linear equations into the form of a matrix


Table 2 Inputs and outputs data of all DMUs in 2012

DMUs Inputs (by 1000000 US dollars) Outputs (by 1000000 US dollars)(I) Fixed assets (I) Cost of goods sold (I) Operating expenses (O) Revenues (O) Operating Income (O) Retained earning

DMU1 130945 1191445 74205 1300887 36136 148261DMU2 17792 158434 6871 171519 6214 7669DMU3 5694 56576 2658 60933 170 minus49609DMU4 2255 26074 602 26766 72 3788DMU5 1761 22914 90 24682 868 4937DMU6 2652 20868 1158 23067 1042 5969DMU7 219 1751 101 2773 113 minus207DMU8 43 3416 182 3682 117 237DMU9 979 5028 235 5647 384 1849DMU10 21746 221874 8348 235695 5473 minus53027DMU11 2263 4976 376 5247 319 4026DMU12 2831 11413 654 12124 68 82DMU13 1473 18135 71 19126 47 2409DMU14 1121 14665 323 18747 1028 11052DMU15 2157 3244 231 3548 83 minus1164DMU16 1797 19126 1423 21738 1214 2905DMU17 1572 3252 263 3926 205 47DMU18 881 6092 44 6618 87 931DMU19 337 60688 2636 65072 1748 minus2097DMU20 302 3255 144 3418 19 112Sources Bloomberg News

Table 3 Inputs and outputs factors of DMU1 in period of 2009ndash2012

DMU1Inputs (by 1000000 US dollars) Outputs (by 1000000 US dollars)

(I) Fixed assets (I) Cost ofgoods sold

(I) Operatingexpenses (O) Revenues (O) Operating

Income (O) Retained earning

2009 78714 5906400 343310 6526040 278190 8833502010 913060 9173010 530280 9983690 287040 10488102011 1192280 10616750 615770 11500880 276040 12561002012 1309450 11914450 742050 13008870 361360 1482610Sources Bloomberg news

Let

119861 = [

[

minus124367 1

minus229634 1

minus3547205 1

]

]

120579 = [119886

119887]

119910119873= [

[

91306011922801309550

]

]

(22)

And then use the least squares method to find 119886 and 119887

[119886

119887] = 120579 = (119861

119879119861)minus1

119861119879119910119873= [

minus0169642810

737498] (23)

Use the two coefficients 119886 and 119887 to generate the whiteningequation of the differential equation

119889119909(1)

119889119905minus 0169642810 times 119909

(1)= 737498 (24)

Find the prediction model from equation

119883(1)

(119896 + 1) = (119883(0)

(1) minus119887

119886) 119890minus119886119896

+119887

119886

119909(1)

(119896 + 1) = (78714 minus 737498

minus0169642810) 1198900169642810times119896

+737498

minus0169642810

= 5134498 sdot 1198900169642810times119896

minus 434736

(25)

Substitute different values of 119896 into the equation

119896 = 0 119883(1)

(1) = 78714

119896 = 1 119883(1)

(2) = 1736414

119896 = 2 119883(1)

(3) = 2861192


119896 = 3 119883(1)

(4) = 4193920

119896 = 4 119883(1)

(5) = 5773045

119896 = 5 119883(1)

(6) = 7644121

(26)

Derive the predicted value of the original series accordingto the accumulated generating operation and obtain

119909(0)

(1) = 119909(1)

(1) = 78714 for the year 2009

119909(0)

(2) = 119909(1)

(2) minus 119909(1)

(1) = 949274

forecasted for 2010

119909(0)

(3) = 119909(1)

(3) minus 119909(1)

(2) = 1124778

forecasted for 2011

119909(0)

(4) = 119909(1)

(4) minus 119909(1)

(3) = 1332728

forecasted for 2012

119909(0)

(5) = 119909(1)

(5) minus 119909(1)

(4) = 1579125

forecasted for 2013

119909(0)

(6) = 119909(1)

(6) minus 119909(1)

(5) = 1871076

forecasted for 2014

(27)

In the same with above computation process the studycould get the forecasting result of all DMUs in 2013 and 2014the detail numbers are shown in Table 4

42 Forecasting Accuracy It is undeniable that forecastingalways remains some errors they are essentially about pre-dicting the future in uncompleted information Thus in thispaper theMAPE (Mean Absolute Percent Error) is employedto measure the accuracy of a method for constructing fittedtime series values in statisticsMAPE is often used tomeasureforecasting accuracy In the book of Stevenson [19] it statedout clearly that MAPE is the average absolute percent errorwhich measures of accuracy in a fitted time series value instatistics specifically trending

MAPE =1

119899sum

|Actual minus Forecast|Actual

times 100

119899 is forecasting number of steps(28)

The parameters of MAPE state out the forecasting abilityas follows

MAPE lt 10 ldquoExcellentrdquo10 ltMAPE lt 20 ldquoGoodrdquo20 ltMAPE lt 50 ldquoReasonablerdquoMAPE gt 50 ldquoPoorrdquo

The results of MAPE are displayed as in Table 5

The calculations of MAPE are almost smaller than 10especially the average MAPE of 20 DMUs reaches 5822(below 10 as well) it strongly confirms that the GM (1 1)

model provides a highly accurate prediction

43 Pearson Correlation To apply DEA model we haveto make sure the relationship between input and outputfactors is isotonic whichmeans if the input quantity increasethe output quantity could not decrease under the samecondition [20] In this study firstly the researcher conducteda simple correlation testmdashPearson correlationmdashto measurethe degree of association between two variables [21] Highercorrelation coefficient means closer relation between twovariables while lower correlation coefficient means that theyare less correlated [22]

The interpretation of the correlation coefficient isexplained in more detail as follows The correlationcoefficient is always between minus1 and +1 The closer thecorrelation is to plusmn1 the closer to a perfect linear relationshipIts general meaning was shown in Table 6

In the empirical study the bellowing results in Tables 78 9 and 10 indicate that the correlation well complies withthe prerequisite condition of the DEA model because theircorrelation coefficient shows strong positive associationsTherefore these positive correlations also demonstrate veryclearly the fact that the researcherrsquos choice of input and outputvariables at the beginning is appropriate Obviously none ofvariables removal is necessary

44 Analysis before Alliance This study executes the softwareof Super-SBM-I-V for the realistic data of 2012 to calculate theDMUsrsquo efficiency and get their ranking before alliances Theempirical results are shown in Table 11

The result indicated that the DMU14 has the best effi-ciency (the first ranking with the score = 39649155) 14other companies including the target DMU1 also have goodoperation efficiencymdashin the 15th row This ranking provesconclusively again that it is necessary for the target companyto conduct strategic alliance to improve its performance

45 Analysis after Alliance According to the above calculatedresult before alliance the target company got the score equalto 1 interpreting that its business in 2012 was good Howeverthe target company only is the 15th out of 20 companiesGuided by the business philosophy of developing constantlythis company should boldly improve its production efficiencyby the formation of the alliance

To implement the empirical research the study starts toform virtual alliance and then executes DEA calculation Bycombining the DMU1 with the rest of DMUs the researchgets 39 virtual alliances totally

Here the software of DEA-Solver Pro 50 built by SaitechCompany is used to calculate Super-SBM-I-V model for 39DMUs Table 11 shows the score and ranking results of virtualalliance in 2013

Depending on the results depicted above the researchcan easily compare the efficient frontiers between DMUs andvirtual alliances The changing from original target DMU1


Table4Fo

recaste

dvalues

ofinpu

tsandou

tputso

fallDMUsin2013

and2014

DMUs

Inpu

ts(by10

00000

USdo

llars)

Outpu

ts(by10

00000

USdo

llars)

(I)F

ixed

assets

(I)C

osto

fgoo

dssold

(I)O

peratin

gexpenses

(O)R

evenues

(O)O

peratin

gIncome

(O)R

etainedearning

2013

2014

2013

2014

2013

2014

2013

2014

2013

2014

2013

2014

DMU1

1579125

1871076

13587033

15461311

872406

1033

724

14841231

1692

4408

39421

447578

1757362

2086792

DMU2

197627

218479

1816

859

2038986

7012

872514

1968374

2210

541

8215

410

2947

151314

303501

DMU3

57504

57460

56892

2560509

26563

26402

612042

601475

1699

115402

minus486051

minus474371

DMU4

22332

2179

023282

206

427

7044

8468

238801

212607

641

379

38252

38378

DMU5

2091

124347

228619

232713

8844

8737

244

302

247816

6741

6311

53854

58882

DMU6

28057

2977

1226020

242278

1212

112583

24893

6266

029

10827

11251

69302

7990

9DMU7

2210

2172

17477

17427

10064

10024

28086

28518

2502

5223

minus1803

minus1513

DMU8

4894

5387

34538

34353

1885

1990

37493

37633

1524

1963

3151

4248

DMU9

12612

16394

58852

6198

72909

3410

6593

768877

4454

4114

20837

21211

DMU10

220459

225280

2145607

1966215

84484

8294

42275858

2087859

46434

40414

minus492077

minus459652

DMU11

22565

2273

05175

654443

4676

5728

5110

950117

1919

1200

40674

4116

4DMU12

27735

26843

112742

113568

6735

6806

119396

119260

898

605

8635

8517

DMU13

17621

21065

192296

207290

6655

6758

201690

216438

4014

3707

26655

29556

DMU14

11018

10774

136200

129000

3191

03112

2174303

1644

148226

6753

110667

110578

DMU15

18713

16809

31718

3094

92288

2212

34428

33518

479

410

minus14564

minus17385

DMU16

1576

715090

1844

77

181344

13711

13231

20973

8206

696

1217

913088

5497

910

2306

DMU17

18610

2210

640377

4999

42788

2914

48062

58503

2480

2873

5179

5600

DMU18

10020

1078

178466

98872

4744

4882

8378

910

4486

009

009

9196

9181

DMU19

33552

3379

9627805

62716

92675

626678

673052

67214

225574

33062

minus19

4250

minus18

1281

DMU20

2886

2870

33111

33348

1441

1462

35201

35674

578

841

1135

1116

Sourcecalculatedby

researchers


Table 5 Average MAPE of DMUs

DMUs Average MAPEDMU1 1626DMU2 5663DMU3 2256DMU4 4001DMU5 5021DMU6 0790DMU7 9307DMU8 2180DMU9 9525DMU10 2459DMU11 1369DMU12 7792DMU13 2069DMU14 1124DMU15 14513DMU16 7046DMU17 2089DMU18 15664DMU19 5527DMU20 16422lsquoAverage of all MAPEs 5822

Table 6 Pearson correlation coefficient

Correlation coefficient Degree of correlationgt08 Very high06ndash08 High04ndash06 Medium02ndash04 Lowlt02 Very low

to virtual alliance will clearly indicate the difference Thedifference can be split into two groups Positive results indifference demonstrate the alliance is better than originalDMUsThemore the difference is themore efficient the alliancegets In contrast the negative result means the alliance isworse Tables 13 and 14 respectively present the concreteresult of two groups

DMUsrsquo ranking rose after alliance demonstrates the targetcompany can take advantages from alliance Table 12 showsthat 10 companies includingDMU19DMU5DMU6DMU4DMU14 DMU10 DMU2 DMU16 DMU13 and DMU9have the good characteristic and necessarily match withcandidatesrsquo desire in doing business The virtual companies(DMU19 + DMU1 DMU5 + DMU1 DMU6 + DMU1DMU4 + DMU1) have the highest opportunities to have thebest efficiency in applying strategic alliance business model(scoregt 1)Thus those 4 candidateswill be highly appreciatedin considering strategic alliance In particular DMU19 is thebest potential candidate for strategic alliance because thedifference is the biggest which is 10 (see Table 13) ThereforeDMU19 is the first priority for this strategy The second

group includes the companies in the category of the not-goodalliance partnership

According to data in Table 14 we can see quite clearly that09 companies (DMU20 DMU8 DMU18 DMU7 DMU17DMU3 DMU11 DMU15 and DMU12) get worse afterstrategic alliances (theDMUsrsquo ranking reduced dramatically)Those companieswould not be our choice due to non-benefitsfor Target Company

46 Partner Selection In previous section the study findsthe good alliance partnership based on the position of thetarget company DMU1 In reality we need to analyze thepossibility of alliance partnership against the category of theGood Alliance Partnership (see Table 13)We take the DMUsrsquoranking before alliance and after alliance of the companies inthe category of theGoodAlliance Partnership into considera-tion on their position to find out which companies are willingto cooperate with the target company

As seen clear from Table 15 nine companies would not bewilling to cooperate with the target company The ranking ofthese companies after alliance reduced in comparison withoriginal ones It means the performance of the companiesincludingDMU19DMU5DMU6DMU4DMU14DMU10DMU2 DMU16 and DMU9 is already good they do notneed to make the alliance partnership with the DMU1

Therefore total 9 above companies will have less desire toform an alliance with the target company because coopera-tion might reduce their performance

By reviewing Tables 11 and 13 and checking the perfor-mance before and after the formation of an alliance those fig-ures clearly highlight the combination between DMU13 andthe target DMU1 Before alliance the efficiency of DMU13does not reach the DEA frontier however the ranking ofDMU13 is improved after alliance with DMU1 It meansthe alliance can exhibit the good scenario for productivityimprovement not only for the DMU1 but also for theDMU13In the other words by implementing alliance both DMU1and DMU13might have the chance to manage their resourcemore effectively Hence this research strongly recommendsDMU13 to cooperate with the target company DMU1

Bsed on the findings of case study the researchers wouldlike to put forward recommendations of the strategic alliancepartner selection for the target company DMU1 in order toimprove its operation efficiency The recommendations arepresented clearly tomake readers satisfy the questions ldquowhy isit necessary to follow such recommendationsrdquoThe noticeablecandidates for alliance strategy are the companyDMU19 (thebest efficiency improvement for the target company) and thecompanyDMU13 (the efficiency improvement for both targetDMU1 and partner DMU13)

5 Conclusions

The pace of development in EMS industry is already matureand getting more and more competitive resulting in thehigher and higher demand for electronic devices and revenuemarket has become a battleground Considering of theseissues enterprises always try to enhance their competitive-ness or expand their business scale These missions are


Table 7 Correlation of input and output data in 2009

Fixed assets Cost of goods sold Operating expenses Revenues Operating income Retained earningFixed assets 1 09918877 09948622 09934644 09870813 05567119Cost of goods sold 09918877 1 09883356 09998524 09707767 04694186Operating expenses 09948622 09883356 1 09906894 09908935 05716045Revenues 09934644 09998524 09906894 1 0974622 04824481Operating income 09870813 09707767 09908935 0974622 1 06383985Retained earning 05567119 04694186 05716045 04824481 06383985 1







assigned to enterprisesrsquo managers That is the reason whya decision making approach on strategic alliance of EMSindustry based on Grey theory and DEA is mainly raised inthis study This research focuses on the relationship betweenstrategic alliance and firmsrsquo performance of EMS industryby using GM (1 1) and DEA model The most importantpurpose of this study is to help the target company find theright partners for strategic alliance

Many scholars and experts have already examined therelated subjects of strategic alliance In fact strategic allianceis employed in the era of globalization while the competitive-ness in themarket become fierce it helps firms to reduce risksand creates the mode of penetration But how the strategicalliance leads the firms to be successful is the big challenge

Basing on the realistic data in the past time (period of2009ndash2012) the GM (1 1) model was used to foresee whatwill happen in terms of specific factors fixed assets cost ofgoods sold operating expenses revenue operating incomeand retained earnings Predicting these factors is necessaryfor the enterprises reduce risks and improve the ability ofreacting against different situations immediately Howeverabsolute accuracy is very hard to reach in forecasting there-fore the MAPE is applied to this studies which the averageMAPE values is 5822 The MAPE in this study confirmsthat GM (1 1) provides reliable and acceptable results whichwould be valuable information for firmrsquos management board

In the aspect of DEA model it is based on the resource-based theory And then the study uses the Super-SBM


Table 11 Efficiency and ranking before strategic alliances

Rank DMUs Scores1 DMU14 396491552 DMU11 165927663 DMU7 162890514 DMU9 154331915 DMU16 134850306 DMU2 134804627 DMU20 124571838 DMU19 121827989 DMU5 1212078510 DMU4 1205571711 DMU6 1149148712 DMU10 1131755913 DMU8 1066124814 DMU17 1058372815 DMU1 116 DMU13 0952748317 DMU3 0857819518 DMU18 0798776719 DMU15 0719228220 DMU12 06572566

model to evaluate the real firms individually and measurethe operation performance of virtual decision making unitsfor strategic alliances The proposed methodology can easilyidentify and compare the efficiency before and after alliances

In this research DMU1 one of the EMS companies isemployed to test whether the strategic alliance benefits existsif DMU1 has alliances with other companies in the sameindustry and give the firms suggestions and the directionof improvement In Section 4 the study finds out thatthe following companies DMU19 DMU5 DMU6 DMU4DMU14 DMU10 DMU2 DMU16 DMU13 and DMU9 arethe good candidates for DMU1 to have strategic alliances inwhich the DMU19 DMU5 DMU6 and DMU4 are stronglyrecommended In addition the research also indicates apossible ideal alliance partners for DMU1 That is DMU13Strategic alliance is not only good for theDMU1 but also goodfor the DMU13 as well

Among the 39 combinations the study finds that somecompaniesrsquo efficiency improved however some of themdecreasedwhich indicates that strategic alliancesmaybe copewith many risks Firms which have the better efficiency maylead to poor efficiency when the partners of strategic allianceshave failed On the contrary firms which originally havepoor efficiency may also have the chance to increase theirown efficiency once they have chosen the right strategicalliance partner Therefore we can summarize that strategicalliances not always can help firms increase their efficiencyand performance Blindly proceeding strategic alliances maycause the company lose their overall competitiveness Thecrucial thing in implementing strategic alliance is that enter-prise should consider different aspects to get benefits Once

Table 12 Efficiency and ranking after strategic alliances

Rank DMUs Scores1 DMU14 38119338782 DMU11 16592765663 DMU7 16289051284 DMU9 15433190545 DMU16 13485030056 DMU2 12875511297 DMU20 12457183148 DMU19 12182797559 DMU5 121207848310 DMU4 120557170211 DMU6 114914870312 DMU10 111970089613 DMU8 106612477314 DMU17 105837279815 DMU1 + DMU19 100483594316 DMU1 + DMU5 10024484617 DMU1 + DMU6 100165363418 DMU1 + DMU4 100078896719 DMU1 + DMU14 119 DMU1 + DMU10 119 DMU1 + DMU2 122 DMU1 + DMU16 099951937623 DMU1 + DMU13 099768238324 DMU1 + DMU9 099658695525 DMU1 09963702426 DMU1 + DMU20 099619459127 DMU1 + DMU8 099614893928 DMU1 + DMU18 099551144229 DMU1 + DMU7 099542409430 DMU1 + DMU17 099447456831 DMU1 + DMU3 09940230432 DMU1 + DMU11 099339314933 DMU1 + DMU15 099133272334 DMU1 + DMU12 099132806335 DMU13 095274830436 DMU3 085781953237 DMU18 07987766838 DMU15 071922819839 DMU12 065725664

strategic alliance is considered and treated carefully andproperly firmrsquos operation efficiency is surely raised

This is a new studying method in both academic researchand practical applications by combining Grey theory andSupermdashSBM model The proposed method of this researchnot only forecasts some important business factors for EMSindustry but also provides an accurate and appropriate evalu-ation of the industry at current situationThat could be usefulinformation helping EMS enterprisesrsquo top managers to haveeffective decision making for business strategy (includingalliance strategy) in the future The result after strategic


Table 13 The good alliance partnership

Virtual alliance Target DMU1ranking (1)

Virtual allianceranking (2)

Difference(1) minus (2)

DMU1 + DMU19 25 15 10DMU1 + DMU5 25 16 9DMU1 + DMU6 25 17 8DMU1 + DMU4 25 18 7DMU1 + DMU14 25 19 6DMU1 + DMU10 25 19 6DMU1 + DMU2 25 19 6DMU1 + DMU16 25 22 3DMU1 + DMU13 25 23 2DMU1 + DMU9 25 24 1

Table 14 The unqualified alliance partnership




DMU1 + DMU20 25 26 minus1DMU1 + DMU8 25 27 minus2DMU1 + DMU18 25 28 minus3DMU1 + DMU7 25 29 minus4DMU1 + DMU17 25 30 minus5DMU1 + DMU3 25 31 minus6DMU1 + DMU11 25 32 minus7DMU1 + DMU15 25 33 minus8DMU1 + DMU12 25 34 minus9

Table 15 The impossible alliance partners

DMUs Rank before alliance Rank afteralliance

DMU19 8 15DMU5 9 16DMU6 11 17DMU4 10 18DMU14 1 19DMU10 12 19DMU2 6 19DMU16 5 22DMU9 4 24

alliance provides a meaningful reference to help many otherindustriesrsquo manager in finding the future candidates ofstrategic alliance

Based on the results of this study the researchers concludeclearly alliances model and methodology which we bringup can also apply to the other industries to evaluate thestrategic alliances partnersrsquo selection to enhance the overallcompetitiveness and to avoid the wrong strategic alliances

Although the paper shows that GM (1 1) is a very flexibleand easy model of use to predict what would happen in thefuture of business and DEA is an efficient tool to measure

the firmsrsquo performance we still cannot deny some restrictionsabout these two methods for further studies Because theinformation of some nonlisted companies is difficult toobtain the study was conducted only with the data of 20companies in Top 50 EMS providers The selection of inputand output variables seemnot completely to reflect the overallEMS industry Therefore the limited number of DMU andinput-output variables could leave the issues open for debate

In actual alliances or union the enterprises may havedifferent considerations such as the industry expansiontechnology acquisition and market development As long aswe can properly adjust the input and output factors throughthe method applied and the process established in this studywe can still get results with the reference value

Sources from Websites

Bloomberg Business Week available at httpwwwbusiness-weekcom Bloomberg news available at httpwwwbloom-bergcom Chartered Institute of Management Accoun-tants ldquoGlobal Manufacturing reportrdquo ldquoAugust 2010rdquo availableat httpwwwcimaglobalcom Saitech Companyrsquos websitehttpwwwsaitech-inccomindexasp

Conflict of Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

References

[1] S Gullander and A Larsson ldquoOutsourcing and locationrdquoin Proceedings of the Conference on New Tracks on SwedishEconomic Research Moole Sweden 2000

[2] R J Mockler ldquoMultinational strategic alliances a managerrsquosperspectiverdquo Strategic Change vol 6 no 7 pp 391ndash405 1997

[3] A Parkhe ldquoInterirm diversity organizational learning andlongevity in global strategic alliancesrdquo Journal of InternationalBusiness Studies vol 22 no 4 pp 579ndash601 1991

[4] S H Chan J W Kensinger A J Keown and J D Martin ldquoDostrategic alliances create valuerdquo Journal of Financial Economicsvol 46 no 2 pp 199ndash221 1997

[5] Y Wei Factors influencing the sucess of virtual cooperationwithin DutchmdashChinese strategic alliances [PhD dissertation]University of Twente 2007

[6] Y Wang and Y P Liu ldquoResearch on the evaluation andselection of partner in logistics strategic alliancerdquo in Proceedingsof theInternational Conference on Management Science andEngineering (ICMSE rsquo07) pp 947ndash952 August 2007

[7] J-G Oh ldquoGlobal strategic alliances in the telecommunicationsindustryrdquo Telecommunications Policy vol 20 no 9 pp 713ndash7201996

[8] K Dittrich G Duysters and A-P de Man ldquoStrategic repo-sitioning by means of alliance networks the case of IBMrdquoResearch Policy vol 36 no 10 pp 1496ndash1511 2007

[9] D Ju-Long ldquoControl problems of grey systemsrdquo Systems andControl Letters vol 1 no 5 pp 288ndash294 1982

[10] J L Deng ldquoIntroduction to grey system theoryrdquoThe Journal ofGrey System vol 1 no 1 pp 1ndash24 1989


[11] A Charnes W W Cooper and E Rhodes ldquoMeasuring theefficiency of decision making unitsrdquo European Journal of Oper-ational Research vol 2 no 6 pp 429ndash444 1978

[12] W Chandraprakaikul and A Suebpongsakorn ldquoEvaluationof logistics companies using data envelopment analysisrdquo inProceedings of the 4th IEEE International SymposiumonLogisticsand Industrial Informatics (LINDI rsquo12) pp 81ndash86 SmoleniceSlovakia September 2012

[13] X Zhao L Li and X-S Zhang ldquoAnalysis of operating efficiencyof Chinese Coal Mining industryrdquo in Proceedings of the IEEE18th International Conference on Industrial Engineering andEngineering Management (IEampEM rsquo11) vol Part 2 pp 889ndash893Changchun China September 2011

[14] E Zanboori M Rostamy-Malkhalifeh G R Jahanshahloo andN Shoja ldquoCalculating super efficiency of DMUs for rankingunits in data envelopment analysis based on SBM modelrdquo TheScientific World Journal vol 2014 Article ID 382390 7 pages2014

[15] C-N Wang and Y-R Lee ldquoGM (1 1) and DEA applicationmodel for strategic alliance in taiwanrdquo in Proceedings of the8th International Conference on Intelligent Systems Design andApplications (ISDA rsquo08) pp 454ndash459 Kaohsiung City TaiwanNovember 2008

[16] K Tone ldquoA slacks-based measure of efficiency in data envelop-ment analysisrdquo European Journal of Operational Research vol130 no 3 pp 498ndash509 2001

[17] S F Liu and Y Lin An Introduction to Grey Systems IIGSSAcademic Publisher Grove City Pa USA 1998

[18] J DengGreyTheory Basis Huazhong University of Science andTechnology Wuhan China 2002 httpwwwsciencedirectcomsciencearticlepiiS0260877409004580bib13

[19] J W Stevenson Operations Management McGraw-Hill 10thedition 2009

[20] F-Y Lo C-F Chien and J T Lin ldquoA DEA study to evaluatethe relative efficiency and investigate the district reorganizationof the Taiwan Power Companyrdquo IEEE Transactions on PowerSystems vol 16 no 1 pp 170ndash178 2001

[21] J C Pruessner C Kirschbaum G Meinlschmid and D HHellhammer ldquoTwo formulas for computation of the area underthe curve represent measures of total hormone concentrationversus time-dependent changerdquo Psychoneuroendocrinology vol28 no 7 pp 916ndash931 2003

[22] S Knack and P Keefer ldquoInstitutions and economic perfor-mance cross-country tests using alternative institutional mea-suresrdquo Economics amp Politics vol 7 no 3 pp 207ndash227 1995

Submit your manuscripts athttpwwwhindawicom

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

MathematicsJournal of


Mathematical Problems in Engineering

Hindawi Publishing Corporationhttpwwwhindawicom

Differential EquationsInternational Journal of

Volume 2014

Applied MathematicsJournal of


Probability and StatisticsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal of


Mathematical PhysicsAdvances in

Complex AnalysisJournal of


OptimizationJournal of


CombinatoricsHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of


Operations ResearchAdvances in

Journal of


Function Spaces

Abstract and Applied AnalysisHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of Mathematics and Mathematical Sciences


The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014


Algebra

Discrete Dynamics in Nature and Society



Decision SciencesAdvances in

Discrete MathematicsJournal of


Volume 2014 Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Stochastic AnalysisInternational Journal of


0

500000

1000000

1500000

Dec 2011998400 Mar 2012

998400 Jun 2012998400 Sep 2012

998400 Dec 2012

RevenueNet income (MTWD)

Figure 1 Revenue and net income of Hon Hai Precision IndustryCo Ltd in period of 2011-2012 (Source Bloomberg News)

The footprint of HonHai Precision Industry is all over theworld in which China is the operation center the rest of theplants are located in India Malaysia Japan Brazil MexicoHungary Slovakia and Czech Republic and Vietnam Forthe future development planning Foxconn Group expectsto expand its production scale not only by looking forinvestment opportunity in more other territories but also byupgrading and building new factories in countries where ithas already run its business

The purpose of this research is to provide an assess-ment model based on Grey theory GM (1 1) and DataEnvelopment Analysis (DEA) helping the target companymdashFoxconn or Hon Haimdashto make a well-considered decisionin finding the right partners At the same time the studyalso provides the prediction about firmsrsquo business in thefuture which is relevant for them when setting strategies forproduction capacity planning and for investment decisionmaking and whether they should expand their business ininternational market or not The results of the case study canbe the reference of the strategic alliances partner selection forworldwide EMS providers

2 Literature Review

21 Strategic Alliances Mockler [2] considers the agreementsbetween companies (partners) to reach objectives of commoninterest Strategic alliances are among the various optionswhich companies can use to achieve their goals they arebased on cooperation between companies This point ofview was shared by Parkhe [3] ldquoStrategic alliance means thepermanent cooperative agreement between the companiesincluding the inflow and linking of resources and its cooper-ative purpose is finishing their company missions in strategicalliancerdquo These definitions emphasize the importance of thecommon business goals of the companies when carrying outalliance strategy

Chan et al [4] see strategic alliance as a cooperativeagreement between various organizations Its purpose aims atreaching the competitive advantage for enterprises that forma partnership Strategic alliance brings together otherwise

independent firms to share resources in product designproduction marketing or distribution In simple words astrategic alliance is sometimes just referred to as ldquopartner-shiprdquo that offers businesses a chance to join forces for amutually beneficial opportunity and sustained competitiveadvantage [5] These definitions attach importance to com-petitive advantages among enterprises in strategic alliance

Wang and Liu [6] researched the evaluation and selectionof partner in logistics strategic alliance The study establishesthe index system of partners in logistics strategic allianceaccording to SCOR index model to evaluate the partnersobjectively Based on AHP method the researchers importthe TOPSIS method to standardize the evaluation result sothat more objective and effective evaluation result could beachieved and the appropriate partner is indicated finally

Oh [7] focused on global strategic alliances in thetelecommunications industry the author points out theglobal marketplace is demanding an increasingly sophisti-cated seamless worldwide communications network alongwith one inexpensive contract for every service GlobalStrategic Alliances is a way of meeting these needs in thecontext of limited available resources

Dittrich et al [8] aimed to show that alliance networkscan play an important role in facilitating large-scale strate-gic change projects They focus on the particular case ofIBM whose radical redirection from an exploitation strat-egy towards an exploration strategy was realized by majorchanges in its network strategy The findings of this papersuggest that the traditional view of large firms as being slowto adapt may not be valid because alliance networks can beused to overcome inertia

22 Grey SystemTheory andDEA Forecast is the explanationof something that has not yet been previously observedor is unknown In recent years many methods have beenproposed for forecasting especially forecasting in businesssuch as fuzzy theory neural networks and grey predictionGrey system theory developed originally in early 1980s by Ju-Long [9] is an interdisciplinary scientific area This theoryhas become a very popular method to deal with uncertaintyproblems under discrete data and incomplete informationHaving superiority to conventional statistical models greymodels require only a limited amount of data to estimate thebehavior of unknown systems [10]

Model Data Envelopment Analysis (DEA) was firstdescribed in a 1978 paper by Charnes Copper and Rhodes(CCR) [11] In that work the authors described a ldquodataorientedrdquo approach for evaluating the performance of a setof peer entities called Decision Making Units (DMUs) whichconvert multiple inputs into multiple outputs The DMU canbe banks managers shipping companies or what we willevaluate in this paper areas within EMS industry Recentyears had a great variety of application of DEA in both publicand private sectors of different countries

Chandraprakaikul and Suebpongsakorn [12] used dataenvelopment analysis to explore the operation performancesof 55 Thai logistics companies from 2007 to 2010 The resultspoint out the reasons for the inefficient Decision Making







Literature review







Step 10 Summary

Part 2 Prediction

Part 3DEA

Part 4Conclusions


factors








3 Methodology










119894119895) isin 119877119898times119899 and 119884 = (119884




min 120588 =1 minus (1119898)sum

119898

119894=1119904minus1198941199091198940

1 + (1119904)sum119904

119894=1119904minus1198941199101198940

(1)

subject to

1199090= 119883120582 + 119904

minus

1199100= 119884120582 minus 119904

+

120582 ge 0 119904minusge 0 119904

+ge 0

(2)








as follows

min 120575 =(1119898)sum

119898

119894=11199091198941199091198940

(1119904) sum119904

119903=11199101199031199101199030

(3)

subject to

119909 ge

119899

sum119895=1119895 =0

120582119895119909119895

119910 le

119899

sum119895=1119895 =0

120582119895119909119895

119910 ge 1199090 119910 le 119910

0 119910119910 ge 119910

0 120582 ge 0

(4)




119903and 119910+

minus119903by

119910+

119903= max119895=1119899

119910119903119895| 119910119903119895gt 0

119910+

119903= min119895=1119899

119910119903119895| 119910119903119895gt 0

(5)


= 119910+minus119903



1199030is never



119904+119903

119910+minus119903(119910+

119903minus 119910+minus119903) (119910+

119903minus 1199101199030) (6)

(2)

119904+119903

(119910+minus119903)2119861 (119910+

119903minus 1199101199030) (7)


















119883(0)

= (119883(0)

(1) 119883(0)

(2) 119883(0)

(119899))

119899 ge 4

(8)




119883(1)

= (119883(1)

(1) 119883(1)

(2) 119883(1)

(119899))

119899 ge 4

(9)

where

119883(1)

(1) = 119883(0)

(1)

119883(1)

(119896) =

119896

sum119894=1

119883(0)

(119894) 119896 = 1 2 3 119899

(10)


119885(1)

= (119885(1)

(1) 119885(1)

(2) 119885(1)

(119899)) (11)


119885(1)

(119896) =1

2(119883(1)

(119896) + 119883(1)

(119896 minus 1))

119896 = 2 3 119899

(12)



1198891198831 (119896)

119889119896+ 119886119883(1)

(119896) = 119887 (13)



119883(1)

(119896 + 1) = [119883(0)

(1) minus119887

119886] 119890minus119886119896

+119887

119886 (14)


[119886 119887]119879= (119861119879119861)minus1

119861119879119884

119884 =

[[[[[

[

119909(0)

(2)

119909(0) (3)

119909(0)

(119899)

]]]]]

]

119861 =

[[[[[

[

minus119911(1)

(2) 1

minus119911(1) (3) 1

minus119911(1) (119899) 1

]]]]]

]

(15)



119883(0)

= 119883(0)

(1) 119883(0)

(2) 119883(0)

(119899)

where 119883(0)

(1) = 119883(0)

(1)

(16)


119883(0)

(119896 + 1) = [119883(0)

(1) minus119887

119886] 119890minus119886119896

(1 minus 119890119886) (17)










119883(0)

= (78714 913060 1192280 1309450) (18)


119883(1)

= (78714 17002 289248 420193)

119909(1)

(1) = 119909(0)

(1) = 78714

119909(1)

(2) = 119909(0)

(1) + 119909(0)

(2) = 17002

119909(1)

(3) = 119909(0)

(1) + 119909(0)

(2) + 119909(0)

(3) = 289248

119909(1)

(4) = 119909(0)

(1) + 119909(0)

(2) + 119909(0)

(3) + 119909(0)

(4) = 420193(19)



119911(1)

(2) =1

2(78714 + 17002) = 124367

119911(1)

(3) =1

2(17002 + 289248) = 229634

119911(1)

(4) =1

2(289248 + 420193) = 3547205

(20)



913060 + 119886 times 124367 = 119887

1192280 + 119886 times 229634 = 119887

1309550 + 119886 times 3547205 = 119887

(21)












Let

119861 = [

[

minus124367 1

minus229634 1

minus3547205 1

]

]

120579 = [119886

119887]

119910119873= [

[

91306011922801309550

]

]

(22)


[119886

119887] = 120579 = (119861

119879119861)minus1

119861119879119910119873= [

minus0169642810

737498] (23)


119889119909(1)

119889119905minus 0169642810 times 119909

(1)= 737498 (24)


119883(1)

(119896 + 1) = (119883(0)

(1) minus119887

119886) 119890minus119886119896

+119887

119886

119909(1)

(119896 + 1) = (78714 minus 737498

minus0169642810) 1198900169642810times119896

+737498

minus0169642810

= 5134498 sdot 1198900169642810times119896

minus 434736

(25)


119896 = 0 119883(1)

(1) = 78714

119896 = 1 119883(1)

(2) = 1736414

119896 = 2 119883(1)

(3) = 2861192


119896 = 3 119883(1)

(4) = 4193920

119896 = 4 119883(1)

(5) = 5773045

119896 = 5 119883(1)

(6) = 7644121

(26)


119909(0)

(1) = 119909(1)

(1) = 78714 for the year 2009

119909(0)

(2) = 119909(1)

(2) minus 119909(1)

(1) = 949274

forecasted for 2010

119909(0)

(3) = 119909(1)

(3) minus 119909(1)

(2) = 1124778

forecasted for 2011

119909(0)

(4) = 119909(1)

(4) minus 119909(1)

(3) = 1332728

forecasted for 2012

119909(0)

(5) = 119909(1)

(5) minus 119909(1)

(4) = 1579125

forecasted for 2013

119909(0)

(6) = 119909(1)

(6) minus 119909(1)

(5) = 1871076

forecasted for 2014

(27)



MAPE =1

119899sum


times 100

















Table4Fo

recaste

dvalues

ofinpu

tsandou

tputso

fallDMUsin2013

and2014

DMUs

Inpu

ts(by10

00000

USdo

llars)

Outpu

ts(by10

00000

USdo

llars)

(I)F

ixed

assets

(I)C

osto

fgoo

dssold

(I)O

peratin

gexpenses

(O)R

evenues

(O)O

peratin

gIncome

(O)R

etainedearning

2013

2014

2013

2014

2013

2014

2013

2014

2013

2014

2013

2014

DMU1

1579125

1871076

13587033

15461311

872406

1033

724

14841231

1692

4408

39421

447578

1757362

2086792

DMU2

197627

218479

1816

859

2038986

7012

872514

1968374

2210

541

8215

410

2947

151314

303501

DMU3

57504

57460

56892

2560509

26563

26402

612042

601475

1699

115402

minus486051

minus474371

DMU4

22332

2179

023282

206

427

7044

8468

238801

212607

641

379

38252

38378

DMU5

2091

124347

228619

232713

8844

8737

244

302

247816

6741

6311

53854

58882

DMU6

28057

2977

1226020

242278

1212

112583

24893

6266

029

10827

11251

69302

7990

9DMU7

2210

2172

17477

17427

10064

10024

28086

28518

2502

5223

minus1803

minus1513

DMU8

4894

5387

34538

34353

1885

1990

37493

37633

1524

1963

3151

4248

DMU9

12612

16394

58852

6198

72909

3410

6593

768877

4454

4114

20837

21211

DMU10

220459

225280

2145607

1966215

84484

8294

42275858

2087859

46434

40414

minus492077

minus459652

DMU11

22565

2273

05175

654443

4676

5728

5110

950117

1919

1200

40674

4116

4DMU12

27735

26843

112742

113568

6735

6806

119396

119260

898

605

8635

8517

DMU13

17621

21065

192296

207290

6655

6758

201690

216438

4014

3707

26655

29556

DMU14

11018

10774

136200

129000

3191

03112

2174303

1644

148226

6753

110667

110578

DMU15

18713

16809

31718

3094

92288

2212

34428

33518

479

410

minus14564

minus17385

DMU16

1576

715090

1844

77

181344

13711

13231

20973

8206

696

1217

913088

5497

910

2306

DMU17

18610

2210

640377

4999

42788

2914

48062

58503

2480

2873

5179

5600

DMU18

10020

1078

178466

98872

4744

4882

8378

910

4486

009

009

9196

9181

DMU19

33552

3379

9627805

62716

92675

626678

673052

67214

225574

33062

minus19

4250

minus18

1281

DMU20

2886

2870

33111

33348

1441

1462

35201

35674

578

841

1135

1116

Sourcecalculatedby

researchers















5 Conclusions
















































References































Volume 2014




Journal of











Journal of


Function Spaces






Algebra















Literature review







Step 10 Summary

Part 2 Prediction

Part 3DEA

Part 4Conclusions


factors








3 Methodology










119894119895) isin 119877119898times119899 and 119884 = (119884




min 120588 =1 minus (1119898)sum

119898

119894=1119904minus1198941199091198940

1 + (1119904)sum119904

119894=1119904minus1198941199101198940

(1)

subject to

1199090= 119883120582 + 119904

minus

1199100= 119884120582 minus 119904

+

120582 ge 0 119904minusge 0 119904

+ge 0

(2)








as follows

min 120575 =(1119898)sum

119898

119894=11199091198941199091198940

(1119904) sum119904

119903=11199101199031199101199030

(3)

subject to

119909 ge

119899

sum119895=1119895 =0

120582119895119909119895

119910 le

119899

sum119895=1119895 =0

120582119895119909119895

119910 ge 1199090 119910 le 119910

0 119910119910 ge 119910

0 120582 ge 0

(4)




119903and 119910+

minus119903by

119910+

119903= max119895=1119899

119910119903119895| 119910119903119895gt 0

119910+

119903= min119895=1119899

119910119903119895| 119910119903119895gt 0

(5)


= 119910+minus119903



1199030is never



119904+119903

119910+minus119903(119910+

119903minus 119910+minus119903) (119910+

119903minus 1199101199030) (6)

(2)

119904+119903

(119910+minus119903)2119861 (119910+

119903minus 1199101199030) (7)


















119883(0)

= (119883(0)

(1) 119883(0)

(2) 119883(0)

(119899))

119899 ge 4

(8)




119883(1)

= (119883(1)

(1) 119883(1)

(2) 119883(1)

(119899))

119899 ge 4

(9)

where

119883(1)

(1) = 119883(0)

(1)

119883(1)

(119896) =

119896

sum119894=1

119883(0)

(119894) 119896 = 1 2 3 119899

(10)


119885(1)

= (119885(1)

(1) 119885(1)

(2) 119885(1)

(119899)) (11)


119885(1)

(119896) =1

2(119883(1)

(119896) + 119883(1)

(119896 minus 1))

119896 = 2 3 119899

(12)



1198891198831 (119896)

119889119896+ 119886119883(1)

(119896) = 119887 (13)



119883(1)

(119896 + 1) = [119883(0)

(1) minus119887

119886] 119890minus119886119896

+119887

119886 (14)


[119886 119887]119879= (119861119879119861)minus1

119861119879119884

119884 =

[[[[[

[

119909(0)

(2)

119909(0) (3)

119909(0)

(119899)

]]]]]

]

119861 =

[[[[[

[

minus119911(1)

(2) 1

minus119911(1) (3) 1

minus119911(1) (119899) 1

]]]]]

]

(15)



119883(0)

= 119883(0)

(1) 119883(0)

(2) 119883(0)

(119899)

where 119883(0)

(1) = 119883(0)

(1)

(16)


119883(0)

(119896 + 1) = [119883(0)

(1) minus119887

119886] 119890minus119886119896

(1 minus 119890119886) (17)










119883(0)

= (78714 913060 1192280 1309450) (18)


119883(1)

= (78714 17002 289248 420193)

119909(1)

(1) = 119909(0)

(1) = 78714

119909(1)

(2) = 119909(0)

(1) + 119909(0)

(2) = 17002

119909(1)

(3) = 119909(0)

(1) + 119909(0)

(2) + 119909(0)

(3) = 289248

119909(1)

(4) = 119909(0)

(1) + 119909(0)

(2) + 119909(0)

(3) + 119909(0)

(4) = 420193(19)



119911(1)

(2) =1

2(78714 + 17002) = 124367

119911(1)

(3) =1

2(17002 + 289248) = 229634

119911(1)

(4) =1

2(289248 + 420193) = 3547205

(20)



913060 + 119886 times 124367 = 119887

1192280 + 119886 times 229634 = 119887

1309550 + 119886 times 3547205 = 119887

(21)












Let

119861 = [

[

minus124367 1

minus229634 1

minus3547205 1

]

]

120579 = [119886

119887]

119910119873= [

[

91306011922801309550

]

]

(22)


[119886

119887] = 120579 = (119861

119879119861)minus1

119861119879119910119873= [

minus0169642810

737498] (23)


119889119909(1)

119889119905minus 0169642810 times 119909

(1)= 737498 (24)


119883(1)

(119896 + 1) = (119883(0)

(1) minus119887

119886) 119890minus119886119896

+119887

119886

119909(1)

(119896 + 1) = (78714 minus 737498

minus0169642810) 1198900169642810times119896

+737498

minus0169642810

= 5134498 sdot 1198900169642810times119896

minus 434736

(25)


119896 = 0 119883(1)

(1) = 78714

119896 = 1 119883(1)

(2) = 1736414

119896 = 2 119883(1)

(3) = 2861192


119896 = 3 119883(1)

(4) = 4193920

119896 = 4 119883(1)

(5) = 5773045

119896 = 5 119883(1)

(6) = 7644121

(26)


119909(0)

(1) = 119909(1)

(1) = 78714 for the year 2009

119909(0)

(2) = 119909(1)

(2) minus 119909(1)

(1) = 949274

forecasted for 2010

119909(0)

(3) = 119909(1)

(3) minus 119909(1)

(2) = 1124778

forecasted for 2011

119909(0)

(4) = 119909(1)

(4) minus 119909(1)

(3) = 1332728

forecasted for 2012

119909(0)

(5) = 119909(1)

(5) minus 119909(1)

(4) = 1579125

forecasted for 2013

119909(0)

(6) = 119909(1)

(6) minus 119909(1)

(5) = 1871076

forecasted for 2014

(27)



MAPE =1

119899sum


times 100

















Table4Fo

recaste

dvalues

ofinpu

tsandou

tputso

fallDMUsin2013

and2014

DMUs

Inpu

ts(by10

00000

USdo

llars)

Outpu

ts(by10

00000

USdo

llars)

(I)F

ixed

assets

(I)C

osto

fgoo

dssold

(I)O

peratin

gexpenses

(O)R

evenues

(O)O

peratin

gIncome

(O)R

etainedearning

2013

2014

2013

2014

2013

2014

2013

2014

2013

2014

2013

2014

DMU1

1579125

1871076

13587033

15461311

872406

1033

724

14841231

1692

4408

39421

447578

1757362

2086792

DMU2

197627

218479

1816

859

2038986

7012

872514

1968374

2210

541

8215

410

2947

151314

303501

DMU3

57504

57460

56892

2560509

26563

26402

612042

601475

1699

115402

minus486051

minus474371

DMU4

22332

2179

023282

206

427

7044

8468

238801

212607

641

379

38252

38378

DMU5

2091

124347

228619

232713

8844

8737

244

302

247816

6741

6311

53854

58882

DMU6

28057

2977

1226020

242278

1212

112583

24893

6266

029

10827

11251

69302

7990

9DMU7

2210

2172

17477

17427

10064

10024

28086

28518

2502

5223

minus1803

minus1513

DMU8

4894

5387

34538

34353

1885

1990

37493

37633

1524

1963

3151

4248

DMU9

12612

16394

58852

6198

72909

3410

6593

768877

4454

4114

20837

21211

DMU10

220459

225280

2145607

1966215

84484

8294

42275858

2087859

46434

40414

minus492077

minus459652

DMU11

22565

2273

05175

654443

4676

5728

5110

950117

1919

1200

40674

4116

4DMU12

27735

26843

112742

113568

6735

6806

119396

119260

898

605

8635

8517

DMU13

17621

21065

192296

207290

6655

6758

201690

216438

4014

3707

26655

29556

DMU14

11018

10774

136200

129000

3191

03112

2174303

1644

148226

6753

110667

110578

DMU15

18713

16809

31718

3094

92288

2212

34428

33518

479

410

minus14564

minus17385

DMU16

1576

715090

1844

77

181344

13711

13231

20973

8206

696

1217

913088

5497

910

2306

DMU17

18610

2210

640377

4999

42788

2914

48062

58503

2480

2873

5179

5600

DMU18

10020

1078

178466

98872

4744

4882

8378

910

4486

009

009

9196

9181

DMU19

33552

3379

9627805

62716

92675

626678

673052

67214

225574

33062

minus19

4250

minus18

1281

DMU20

2886

2870

33111

33348

1441

1462

35201

35674

578

841

1135

1116

Sourcecalculatedby

researchers















5 Conclusions
















































References































Volume 2014




Journal of











Journal of


Function Spaces






Algebra
















119894119895) isin 119877119898times119899 and 119884 = (119884




min 120588 =1 minus (1119898)sum

119898

119894=1119904minus1198941199091198940

1 + (1119904)sum119904

119894=1119904minus1198941199101198940

(1)

subject to

1199090= 119883120582 + 119904

minus

1199100= 119884120582 minus 119904

+

120582 ge 0 119904minusge 0 119904

+ge 0

(2)








as follows

min 120575 =(1119898)sum

119898

119894=11199091198941199091198940

(1119904) sum119904

119903=11199101199031199101199030

(3)

subject to

119909 ge

119899

sum119895=1119895 =0

120582119895119909119895

119910 le

119899

sum119895=1119895 =0

120582119895119909119895

119910 ge 1199090 119910 le 119910

0 119910119910 ge 119910

0 120582 ge 0

(4)




119903and 119910+

minus119903by

119910+

119903= max119895=1119899

119910119903119895| 119910119903119895gt 0

119910+

119903= min119895=1119899

119910119903119895| 119910119903119895gt 0

(5)


= 119910+minus119903



1199030is never



119904+119903

119910+minus119903(119910+

119903minus 119910+minus119903) (119910+

119903minus 1199101199030) (6)

(2)

119904+119903

(119910+minus119903)2119861 (119910+

119903minus 1199101199030) (7)


















119883(0)

= (119883(0)

(1) 119883(0)

(2) 119883(0)

(119899))

119899 ge 4

(8)




119883(1)

= (119883(1)

(1) 119883(1)

(2) 119883(1)

(119899))

119899 ge 4

(9)

where

119883(1)

(1) = 119883(0)

(1)

119883(1)

(119896) =

119896

sum119894=1

119883(0)

(119894) 119896 = 1 2 3 119899

(10)


119885(1)

= (119885(1)

(1) 119885(1)

(2) 119885(1)

(119899)) (11)


119885(1)

(119896) =1

2(119883(1)

(119896) + 119883(1)

(119896 minus 1))

119896 = 2 3 119899

(12)



1198891198831 (119896)

119889119896+ 119886119883(1)

(119896) = 119887 (13)



119883(1)

(119896 + 1) = [119883(0)

(1) minus119887

119886] 119890minus119886119896

+119887

119886 (14)


[119886 119887]119879= (119861119879119861)minus1

119861119879119884

119884 =

[[[[[

[

119909(0)

(2)

119909(0) (3)

119909(0)

(119899)

]]]]]

]

119861 =

[[[[[

[

minus119911(1)

(2) 1

minus119911(1) (3) 1

minus119911(1) (119899) 1

]]]]]

]

(15)



119883(0)

= 119883(0)

(1) 119883(0)

(2) 119883(0)

(119899)

where 119883(0)

(1) = 119883(0)

(1)

(16)


119883(0)

(119896 + 1) = [119883(0)

(1) minus119887

119886] 119890minus119886119896

(1 minus 119890119886) (17)










119883(0)

= (78714 913060 1192280 1309450) (18)


119883(1)

= (78714 17002 289248 420193)

119909(1)

(1) = 119909(0)

(1) = 78714

119909(1)

(2) = 119909(0)

(1) + 119909(0)

(2) = 17002

119909(1)

(3) = 119909(0)

(1) + 119909(0)

(2) + 119909(0)

(3) = 289248

119909(1)

(4) = 119909(0)

(1) + 119909(0)

(2) + 119909(0)

(3) + 119909(0)

(4) = 420193(19)



119911(1)

(2) =1

2(78714 + 17002) = 124367

119911(1)

(3) =1

2(17002 + 289248) = 229634

119911(1)

(4) =1

2(289248 + 420193) = 3547205

(20)



913060 + 119886 times 124367 = 119887

1192280 + 119886 times 229634 = 119887

1309550 + 119886 times 3547205 = 119887

(21)












Let

119861 = [

[

minus124367 1

minus229634 1

minus3547205 1

]

]

120579 = [119886

119887]

119910119873= [

[

91306011922801309550

]

]

(22)


[119886

119887] = 120579 = (119861

119879119861)minus1

119861119879119910119873= [

minus0169642810

737498] (23)


119889119909(1)

119889119905minus 0169642810 times 119909

(1)= 737498 (24)


119883(1)

(119896 + 1) = (119883(0)

(1) minus119887

119886) 119890minus119886119896

+119887

119886

119909(1)

(119896 + 1) = (78714 minus 737498

minus0169642810) 1198900169642810times119896

+737498

minus0169642810

= 5134498 sdot 1198900169642810times119896

minus 434736

(25)


119896 = 0 119883(1)

(1) = 78714

119896 = 1 119883(1)

(2) = 1736414

119896 = 2 119883(1)

(3) = 2861192


119896 = 3 119883(1)

(4) = 4193920

119896 = 4 119883(1)

(5) = 5773045

119896 = 5 119883(1)

(6) = 7644121

(26)


119909(0)

(1) = 119909(1)

(1) = 78714 for the year 2009

119909(0)

(2) = 119909(1)

(2) minus 119909(1)

(1) = 949274

forecasted for 2010

119909(0)

(3) = 119909(1)

(3) minus 119909(1)

(2) = 1124778

forecasted for 2011

119909(0)

(4) = 119909(1)

(4) minus 119909(1)

(3) = 1332728

forecasted for 2012

119909(0)

(5) = 119909(1)

(5) minus 119909(1)

(4) = 1579125

forecasted for 2013

119909(0)

(6) = 119909(1)

(6) minus 119909(1)

(5) = 1871076

forecasted for 2014

(27)



MAPE =1

119899sum


times 100

















Table4Fo

recaste

dvalues

ofinpu

tsandou

tputso

fallDMUsin2013

and2014

DMUs

Inpu

ts(by10

00000

USdo

llars)

Outpu

ts(by10

00000

USdo

llars)

(I)F

ixed

assets

(I)C

osto

fgoo

dssold

(I)O

peratin

gexpenses

(O)R

evenues

(O)O

peratin

gIncome

(O)R

etainedearning

2013

2014

2013

2014

2013

2014

2013

2014

2013

2014

2013

2014

DMU1

1579125

1871076

13587033

15461311

872406

1033

724

14841231

1692

4408

39421

447578

1757362

2086792

DMU2

197627

218479

1816

859

2038986

7012

872514

1968374

2210

541

8215

410

2947

151314

303501

DMU3

57504

57460

56892

2560509

26563

26402

612042

601475

1699

115402

minus486051

minus474371

DMU4

22332

2179

023282

206

427

7044

8468

238801

212607

641

379

38252

38378

DMU5

2091

124347

228619

232713

8844

8737

244

302

247816

6741

6311

53854

58882

DMU6

28057

2977

1226020

242278

1212

112583

24893

6266

029

10827

11251

69302

7990

9DMU7

2210

2172

17477

17427

10064

10024

28086

28518

2502

5223

minus1803

minus1513

DMU8

4894

5387

34538

34353

1885

1990

37493

37633

1524

1963

3151

4248

DMU9

12612

16394

58852

6198

72909

3410

6593

768877

4454

4114

20837

21211

DMU10

220459

225280

2145607

1966215

84484

8294

42275858

2087859

46434

40414

minus492077

minus459652

DMU11

22565

2273

05175

654443

4676

5728

5110

950117

1919

1200

40674

4116

4DMU12

27735

26843

112742

113568

6735

6806

119396

119260

898

605

8635

8517

DMU13

17621

21065

192296

207290

6655

6758

201690

216438

4014

3707

26655

29556

DMU14

11018

10774

136200

129000

3191

03112

2174303

1644

148226

6753

110667

110578

DMU15

18713

16809

31718

3094

92288

2212

34428

33518

479

410

minus14564

minus17385

DMU16

1576

715090

1844

77

181344

13711

13231

20973

8206

696

1217

913088

5497

910

2306

DMU17

18610

2210

640377

4999

42788

2914

48062

58503

2480

2873

5179

5600

DMU18

10020

1078

178466

98872

4744

4882

8378

910

4486

009

009

9196

9181

DMU19

33552

3379

9627805

62716

92675

626678

673052

67214

225574

33062

minus19

4250

minus18

1281

DMU20

2886

2870

33111

33348

1441

1462

35201

35674

578

841

1135

1116

Sourcecalculatedby

researchers















5 Conclusions
















































References































Volume 2014




Journal of











Journal of


Function Spaces






Algebra

























119883(0)

= (119883(0)

(1) 119883(0)

(2) 119883(0)

(119899))

119899 ge 4

(8)




119883(1)

= (119883(1)

(1) 119883(1)

(2) 119883(1)

(119899))

119899 ge 4

(9)

where

119883(1)

(1) = 119883(0)

(1)

119883(1)

(119896) =

119896

sum119894=1

119883(0)

(119894) 119896 = 1 2 3 119899

(10)


119885(1)

= (119885(1)

(1) 119885(1)

(2) 119885(1)

(119899)) (11)


119885(1)

(119896) =1

2(119883(1)

(119896) + 119883(1)

(119896 minus 1))

119896 = 2 3 119899

(12)



1198891198831 (119896)

119889119896+ 119886119883(1)

(119896) = 119887 (13)



119883(1)

(119896 + 1) = [119883(0)

(1) minus119887

119886] 119890minus119886119896

+119887

119886 (14)


[119886 119887]119879= (119861119879119861)minus1

119861119879119884

119884 =

[[[[[

[

119909(0)

(2)

119909(0) (3)

119909(0)

(119899)

]]]]]

]

119861 =

[[[[[

[

minus119911(1)

(2) 1

minus119911(1) (3) 1

minus119911(1) (119899) 1

]]]]]

]

(15)



119883(0)

= 119883(0)

(1) 119883(0)

(2) 119883(0)

(119899)

where 119883(0)

(1) = 119883(0)

(1)

(16)


119883(0)

(119896 + 1) = [119883(0)

(1) minus119887

119886] 119890minus119886119896

(1 minus 119890119886) (17)










119883(0)

= (78714 913060 1192280 1309450) (18)


119883(1)

= (78714 17002 289248 420193)

119909(1)

(1) = 119909(0)

(1) = 78714

119909(1)

(2) = 119909(0)

(1) + 119909(0)

(2) = 17002

119909(1)

(3) = 119909(0)

(1) + 119909(0)

(2) + 119909(0)

(3) = 289248

119909(1)

(4) = 119909(0)

(1) + 119909(0)

(2) + 119909(0)

(3) + 119909(0)

(4) = 420193(19)



119911(1)

(2) =1

2(78714 + 17002) = 124367

119911(1)

(3) =1

2(17002 + 289248) = 229634

119911(1)

(4) =1

2(289248 + 420193) = 3547205

(20)



913060 + 119886 times 124367 = 119887

1192280 + 119886 times 229634 = 119887

1309550 + 119886 times 3547205 = 119887

(21)












Let

119861 = [

[

minus124367 1

minus229634 1

minus3547205 1

]

]

120579 = [119886

119887]

119910119873= [

[

91306011922801309550

]

]

(22)


[119886

119887] = 120579 = (119861

119879119861)minus1

119861119879119910119873= [

minus0169642810

737498] (23)


119889119909(1)

119889119905minus 0169642810 times 119909

(1)= 737498 (24)


119883(1)

(119896 + 1) = (119883(0)

(1) minus119887

119886) 119890minus119886119896

+119887

119886

119909(1)

(119896 + 1) = (78714 minus 737498

minus0169642810) 1198900169642810times119896

+737498

minus0169642810

= 5134498 sdot 1198900169642810times119896

minus 434736

(25)


119896 = 0 119883(1)

(1) = 78714

119896 = 1 119883(1)

(2) = 1736414

119896 = 2 119883(1)

(3) = 2861192


119896 = 3 119883(1)

(4) = 4193920

119896 = 4 119883(1)

(5) = 5773045

119896 = 5 119883(1)

(6) = 7644121

(26)


119909(0)

(1) = 119909(1)

(1) = 78714 for the year 2009

119909(0)

(2) = 119909(1)

(2) minus 119909(1)

(1) = 949274

forecasted for 2010

119909(0)

(3) = 119909(1)

(3) minus 119909(1)

(2) = 1124778

forecasted for 2011

119909(0)

(4) = 119909(1)

(4) minus 119909(1)

(3) = 1332728

forecasted for 2012

119909(0)

(5) = 119909(1)

(5) minus 119909(1)

(4) = 1579125

forecasted for 2013

119909(0)

(6) = 119909(1)

(6) minus 119909(1)

(5) = 1871076

forecasted for 2014

(27)



MAPE =1

119899sum


times 100

















Table4Fo

recaste

dvalues

ofinpu

tsandou

tputso

fallDMUsin2013

and2014

DMUs

Inpu

ts(by10

00000

USdo

llars)

Outpu

ts(by10

00000

USdo

llars)

(I)F

ixed

assets

(I)C

osto

fgoo

dssold

(I)O

peratin

gexpenses

(O)R

evenues

(O)O

peratin

gIncome

(O)R

etainedearning

2013

2014

2013

2014

2013

2014

2013

2014

2013

2014

2013

2014

DMU1

1579125

1871076

13587033

15461311

872406

1033

724

14841231

1692

4408

39421

447578

1757362

2086792

DMU2

197627

218479

1816

859

2038986

7012

872514

1968374

2210

541

8215

410

2947

151314

303501

DMU3

57504

57460

56892

2560509

26563

26402

612042

601475

1699

115402

minus486051

minus474371

DMU4

22332

2179

023282

206

427

7044

8468

238801

212607

641

379

38252

38378

DMU5

2091

124347

228619

232713

8844

8737

244

302

247816

6741

6311

53854

58882

DMU6

28057

2977

1226020

242278

1212

112583

24893

6266

029

10827

11251

69302

7990

9DMU7

2210

2172

17477

17427

10064

10024

28086

28518

2502

5223

minus1803

minus1513

DMU8

4894

5387

34538

34353

1885

1990

37493

37633

1524

1963

3151

4248

DMU9

12612

16394

58852

6198

72909

3410

6593

768877

4454

4114

20837

21211

DMU10

220459

225280

2145607

1966215

84484

8294

42275858

2087859

46434

40414

minus492077

minus459652

DMU11

22565

2273

05175

654443

4676

5728

5110

950117

1919

1200

40674

4116

4DMU12

27735

26843

112742

113568

6735

6806

119396

119260

898

605

8635

8517

DMU13

17621

21065

192296

207290

6655

6758

201690

216438

4014

3707

26655

29556

DMU14

11018

10774

136200

129000

3191

03112

2174303

1644

148226

6753

110667

110578

DMU15

18713

16809

31718

3094

92288

2212

34428

33518

479

410

minus14564

minus17385

DMU16

1576

715090

1844

77

181344

13711

13231

20973

8206

696

1217

913088

5497

910

2306

DMU17

18610

2210

640377

4999

42788

2914

48062

58503

2480

2873

5179

5600

DMU18

10020

1078

178466

98872

4744

4882

8378

910

4486

009

009

9196

9181

DMU19

33552

3379

9627805

62716

92675

626678

673052

67214

225574

33062

minus19

4250

minus18

1281

DMU20

2886

2870

33111

33348

1441

1462

35201

35674

578

841

1135

1116

Sourcecalculatedby

researchers















5 Conclusions
















































References































Volume 2014




Journal of











Journal of


Function Spaces






Algebra











1198891198831 (119896)

119889119896+ 119886119883(1)

(119896) = 119887 (13)



119883(1)

(119896 + 1) = [119883(0)

(1) minus119887

119886] 119890minus119886119896

+119887

119886 (14)


[119886 119887]119879= (119861119879119861)minus1

119861119879119884

119884 =

[[[[[

[

119909(0)

(2)

119909(0) (3)

119909(0)

(119899)

]]]]]

]

119861 =

[[[[[

[

minus119911(1)

(2) 1

minus119911(1) (3) 1

minus119911(1) (119899) 1

]]]]]

]

(15)



119883(0)

= 119883(0)

(1) 119883(0)

(2) 119883(0)

(119899)

where 119883(0)

(1) = 119883(0)

(1)

(16)


119883(0)

(119896 + 1) = [119883(0)

(1) minus119887

119886] 119890minus119886119896

(1 minus 119890119886) (17)










119883(0)

= (78714 913060 1192280 1309450) (18)


119883(1)

= (78714 17002 289248 420193)

119909(1)

(1) = 119909(0)

(1) = 78714

119909(1)

(2) = 119909(0)

(1) + 119909(0)

(2) = 17002

119909(1)

(3) = 119909(0)

(1) + 119909(0)

(2) + 119909(0)

(3) = 289248

119909(1)

(4) = 119909(0)

(1) + 119909(0)

(2) + 119909(0)

(3) + 119909(0)

(4) = 420193(19)



119911(1)

(2) =1

2(78714 + 17002) = 124367

119911(1)

(3) =1

2(17002 + 289248) = 229634

119911(1)

(4) =1

2(289248 + 420193) = 3547205

(20)



913060 + 119886 times 124367 = 119887

1192280 + 119886 times 229634 = 119887

1309550 + 119886 times 3547205 = 119887

(21)












Let

119861 = [

[

minus124367 1

minus229634 1

minus3547205 1

]

]

120579 = [119886

119887]

119910119873= [

[

91306011922801309550

]

]

(22)


[119886

119887] = 120579 = (119861

119879119861)minus1

119861119879119910119873= [

minus0169642810

737498] (23)


119889119909(1)

119889119905minus 0169642810 times 119909

(1)= 737498 (24)


119883(1)

(119896 + 1) = (119883(0)

(1) minus119887

119886) 119890minus119886119896

+119887

119886

119909(1)

(119896 + 1) = (78714 minus 737498

minus0169642810) 1198900169642810times119896

+737498

minus0169642810

= 5134498 sdot 1198900169642810times119896

minus 434736

(25)


119896 = 0 119883(1)

(1) = 78714

119896 = 1 119883(1)

(2) = 1736414

119896 = 2 119883(1)

(3) = 2861192


119896 = 3 119883(1)

(4) = 4193920

119896 = 4 119883(1)

(5) = 5773045

119896 = 5 119883(1)

(6) = 7644121

(26)


119909(0)

(1) = 119909(1)

(1) = 78714 for the year 2009

119909(0)

(2) = 119909(1)

(2) minus 119909(1)

(1) = 949274

forecasted for 2010

119909(0)

(3) = 119909(1)

(3) minus 119909(1)

(2) = 1124778

forecasted for 2011

119909(0)

(4) = 119909(1)

(4) minus 119909(1)

(3) = 1332728

forecasted for 2012

119909(0)

(5) = 119909(1)

(5) minus 119909(1)

(4) = 1579125

forecasted for 2013

119909(0)

(6) = 119909(1)

(6) minus 119909(1)

(5) = 1871076

forecasted for 2014

(27)



MAPE =1

119899sum


times 100

















Table4Fo

recaste

dvalues

ofinpu

tsandou

tputso

fallDMUsin2013

and2014

DMUs

Inpu

ts(by10

00000

USdo

llars)

Outpu

ts(by10

00000

USdo

llars)

(I)F

ixed

assets

(I)C

osto

fgoo

dssold

(I)O

peratin

gexpenses

(O)R

evenues

(O)O

peratin

gIncome

(O)R

etainedearning

2013

2014

2013

2014

2013

2014

2013

2014

2013

2014

2013

2014

DMU1

1579125

1871076

13587033

15461311

872406

1033

724

14841231

1692

4408

39421

447578

1757362

2086792

DMU2

197627

218479

1816

859

2038986

7012

872514

1968374

2210

541

8215

410

2947

151314

303501

DMU3

57504

57460

56892

2560509

26563

26402

612042

601475

1699

115402

minus486051

minus474371

DMU4

22332

2179

023282

206

427

7044

8468

238801

212607

641

379

38252

38378

DMU5

2091

124347

228619

232713

8844

8737

244

302

247816

6741

6311

53854

58882

DMU6

28057

2977

1226020

242278

1212

112583

24893

6266

029

10827

11251

69302

7990

9DMU7

2210

2172

17477

17427

10064

10024

28086

28518

2502

5223

minus1803

minus1513

DMU8

4894

5387

34538

34353

1885

1990

37493

37633

1524

1963

3151

4248

DMU9

12612

16394

58852

6198

72909

3410

6593

768877

4454

4114

20837

21211

DMU10

220459

225280

2145607

1966215

84484

8294

42275858

2087859

46434

40414

minus492077

minus459652

DMU11

22565

2273

05175

654443

4676

5728

5110

950117

1919

1200

40674

4116

4DMU12

27735

26843

112742

113568

6735

6806

119396

119260

898

605

8635

8517

DMU13

17621

21065

192296

207290

6655

6758

201690

216438

4014

3707

26655

29556

DMU14

11018

10774

136200

129000

3191

03112

2174303

1644

148226

6753

110667

110578

DMU15

18713

16809

31718

3094

92288

2212

34428

33518

479

410

minus14564

minus17385

DMU16

1576

715090

1844

77

181344

13711

13231

20973

8206

696

1217

913088

5497

910

2306

DMU17

18610

2210

640377

4999

42788

2914

48062

58503

2480

2873

5179

5600

DMU18

10020

1078

178466

98872

4744

4882

8378

910

4486

009

009

9196

9181

DMU19

33552

3379

9627805

62716

92675

626678

673052

67214

225574

33062

minus19

4250

minus18

1281

DMU20

2886

2870

33111

33348

1441

1462

35201

35674

578

841

1135

1116

Sourcecalculatedby

researchers















5 Conclusions
















































References































Volume 2014




Journal of











Journal of


Function Spaces






Algebra



















Let

119861 = [

[

minus124367 1

minus229634 1

minus3547205 1

]

]

120579 = [119886

119887]

119910119873= [

[

91306011922801309550

]

]

(22)


[119886

119887] = 120579 = (119861

119879119861)minus1

119861119879119910119873= [

minus0169642810

737498] (23)


119889119909(1)

119889119905minus 0169642810 times 119909

(1)= 737498 (24)


119883(1)

(119896 + 1) = (119883(0)

(1) minus119887

119886) 119890minus119886119896

+119887

119886

119909(1)

(119896 + 1) = (78714 minus 737498

minus0169642810) 1198900169642810times119896

+737498

minus0169642810

= 5134498 sdot 1198900169642810times119896

minus 434736

(25)


119896 = 0 119883(1)

(1) = 78714

119896 = 1 119883(1)

(2) = 1736414

119896 = 2 119883(1)

(3) = 2861192


119896 = 3 119883(1)

(4) = 4193920

119896 = 4 119883(1)

(5) = 5773045

119896 = 5 119883(1)

(6) = 7644121

(26)


119909(0)

(1) = 119909(1)

(1) = 78714 for the year 2009

119909(0)

(2) = 119909(1)

(2) minus 119909(1)

(1) = 949274

forecasted for 2010

119909(0)

(3) = 119909(1)

(3) minus 119909(1)

(2) = 1124778

forecasted for 2011

119909(0)

(4) = 119909(1)

(4) minus 119909(1)

(3) = 1332728

forecasted for 2012

119909(0)

(5) = 119909(1)

(5) minus 119909(1)

(4) = 1579125

forecasted for 2013

119909(0)

(6) = 119909(1)

(6) minus 119909(1)

(5) = 1871076

forecasted for 2014

(27)



MAPE =1

119899sum


times 100

















Table4Fo

recaste

dvalues

ofinpu

tsandou

tputso

fallDMUsin2013

and2014

DMUs

Inpu

ts(by10

00000

USdo

llars)

Outpu

ts(by10

00000

USdo

llars)

(I)F

ixed

assets

(I)C

osto

fgoo

dssold

(I)O

peratin

gexpenses

(O)R

evenues

(O)O

peratin

gIncome

(O)R

etainedearning

2013

2014

2013

2014

2013

2014

2013

2014

2013

2014

2013

2014

DMU1

1579125

1871076

13587033

15461311

872406

1033

724

14841231

1692

4408

39421

447578

1757362

2086792

DMU2

197627

218479

1816

859

2038986

7012

872514

1968374

2210

541

8215

410

2947

151314

303501

DMU3

57504

57460

56892

2560509

26563

26402

612042

601475

1699

115402

minus486051

minus474371

DMU4

22332

2179

023282

206

427

7044

8468

238801

212607

641

379

38252

38378

DMU5

2091

124347

228619

232713

8844

8737

244

302

247816

6741

6311

53854

58882

DMU6

28057

2977

1226020

242278

1212

112583

24893

6266

029

10827

11251

69302

7990

9DMU7

2210

2172

17477

17427

10064

10024

28086

28518

2502

5223

minus1803

minus1513

DMU8

4894

5387

34538

34353

1885

1990

37493

37633

1524

1963

3151

4248

DMU9

12612

16394

58852

6198

72909

3410

6593

768877

4454

4114

20837

21211

DMU10

220459

225280

2145607

1966215

84484

8294

42275858

2087859

46434

40414

minus492077

minus459652

DMU11

22565

2273

05175

654443

4676

5728

5110

950117

1919

1200

40674

4116

4DMU12

27735

26843

112742

113568

6735

6806

119396

119260

898

605

8635

8517

DMU13

17621

21065

192296

207290

6655

6758

201690

216438

4014

3707

26655

29556

DMU14

11018

10774

136200

129000

3191

03112

2174303

1644

148226

6753

110667

110578

DMU15

18713

16809

31718

3094

92288

2212

34428

33518

479

410

minus14564

minus17385

DMU16

1576

715090

1844

77

181344

13711

13231

20973

8206

696

1217

913088

5497

910

2306

DMU17

18610

2210

640377

4999

42788

2914

48062

58503

2480

2873

5179

5600

DMU18

10020

1078

178466

98872

4744

4882

8378

910

4486

009

009

9196

9181

DMU19

33552

3379

9627805

62716

92675

626678

673052

67214

225574

33062

minus19

4250

minus18

1281

DMU20

2886

2870

33111

33348

1441

1462

35201

35674

578

841

1135

1116

Sourcecalculatedby

researchers















5 Conclusions
















































References































Volume 2014




Journal of











Journal of


Function Spaces






Algebra










119896 = 3 119883(1)

(4) = 4193920

119896 = 4 119883(1)

(5) = 5773045

119896 = 5 119883(1)

(6) = 7644121

(26)


119909(0)

(1) = 119909(1)

(1) = 78714 for the year 2009

119909(0)

(2) = 119909(1)

(2) minus 119909(1)

(1) = 949274

forecasted for 2010

119909(0)

(3) = 119909(1)

(3) minus 119909(1)

(2) = 1124778

forecasted for 2011

119909(0)

(4) = 119909(1)

(4) minus 119909(1)

(3) = 1332728

forecasted for 2012

119909(0)

(5) = 119909(1)

(5) minus 119909(1)

(4) = 1579125

forecasted for 2013

119909(0)

(6) = 119909(1)

(6) minus 119909(1)

(5) = 1871076

forecasted for 2014

(27)



MAPE =1

119899sum


times 100

















Table4Fo

recaste

dvalues

ofinpu

tsandou

tputso

fallDMUsin2013

and2014

DMUs

Inpu

ts(by10

00000

USdo

llars)

Outpu

ts(by10

00000

USdo

llars)

(I)F

ixed

assets

(I)C

osto

fgoo

dssold

(I)O

peratin

gexpenses

(O)R

evenues

(O)O

peratin

gIncome

(O)R

etainedearning

2013

2014

2013

2014

2013

2014

2013

2014

2013

2014

2013

2014

DMU1

1579125

1871076

13587033

15461311

872406

1033

724

14841231

1692

4408

39421

447578

1757362

2086792

DMU2

197627

218479

1816

859

2038986

7012

872514

1968374

2210

541

8215

410

2947

151314

303501

DMU3

57504

57460

56892

2560509

26563

26402

612042

601475

1699

115402

minus486051

minus474371

DMU4

22332

2179

023282

206

427

7044

8468

238801

212607

641

379

38252

38378

DMU5

2091

124347

228619

232713

8844

8737

244

302

247816

6741

6311

53854

58882

DMU6

28057

2977

1226020

242278

1212

112583

24893

6266

029

10827

11251

69302

7990

9DMU7

2210

2172

17477

17427

10064

10024

28086

28518

2502

5223

minus1803

minus1513

DMU8

4894

5387

34538

34353

1885

1990

37493

37633

1524

1963

3151

4248

DMU9

12612

16394

58852

6198

72909

3410

6593

768877

4454

4114

20837

21211

DMU10

220459

225280

2145607

1966215

84484

8294

42275858

2087859

46434

40414

minus492077

minus459652

DMU11

22565

2273

05175

654443

4676

5728

5110

950117

1919

1200

40674

4116

4DMU12

27735

26843

112742

113568

6735

6806

119396

119260

898

605

8635

8517

DMU13

17621

21065

192296

207290

6655

6758

201690

216438

4014

3707

26655

29556

DMU14

11018

10774

136200

129000

3191

03112

2174303

1644

148226

6753

110667

110578

DMU15

18713

16809

31718

3094

92288

2212

34428

33518

479

410

minus14564

minus17385

DMU16

1576

715090

1844

77

181344

13711

13231

20973

8206

696

1217

913088

5497

910

2306

DMU17

18610

2210

640377

4999

42788

2914

48062

58503

2480

2873

5179

5600

DMU18

10020

1078

178466

98872

4744

4882

8378

910

4486

009

009

9196

9181

DMU19

33552

3379

9627805

62716

92675

626678

673052

67214

225574

33062

minus19

4250

minus18

1281

DMU20

2886

2870

33111

33348

1441

1462

35201

35674

578

841

1135

1116

Sourcecalculatedby

researchers















5 Conclusions
















































References































Volume 2014




Journal of











Journal of


Function Spaces






Algebra










Table4Fo

recaste

dvalues

ofinpu

tsandou

tputso

fallDMUsin2013

and2014

DMUs

Inpu

ts(by10

00000

USdo

llars)

Outpu

ts(by10

00000

USdo

llars)

(I)F

ixed

assets

(I)C

osto

fgoo

dssold

(I)O

peratin

gexpenses

(O)R

evenues

(O)O

peratin

gIncome

(O)R

etainedearning

2013

2014

2013

2014

2013

2014

2013

2014

2013

2014

2013

2014

DMU1

1579125

1871076

13587033

15461311

872406

1033

724

14841231

1692

4408

39421

447578

1757362

2086792

DMU2

197627

218479

1816

859

2038986

7012

872514

1968374

2210

541

8215

410

2947

151314

303501

DMU3

57504

57460

56892

2560509

26563

26402

612042

601475

1699

115402

minus486051

minus474371

DMU4

22332

2179

023282

206

427

7044

8468

238801

212607

641

379

38252

38378

DMU5

2091

124347

228619

232713

8844

8737

244

302

247816

6741

6311

53854

58882

DMU6

28057

2977

1226020

242278

1212

112583

24893

6266

029

10827

11251

69302

7990

9DMU7

2210

2172

17477

17427

10064

10024

28086

28518

2502

5223

minus1803

minus1513

DMU8

4894

5387

34538

34353

1885

1990

37493

37633

1524

1963

3151

4248

DMU9

12612

16394

58852

6198

72909

3410

6593

768877

4454

4114

20837

21211

DMU10

220459

225280

2145607

1966215

84484

8294

42275858

2087859

46434

40414

minus492077

minus459652

DMU11

22565

2273

05175

654443

4676

5728

5110

950117

1919

1200

40674

4116

4DMU12

27735

26843

112742

113568

6735

6806

119396

119260

898

605

8635

8517

DMU13

17621

21065

192296

207290

6655

6758

201690

216438

4014

3707

26655

29556

DMU14

11018

10774

136200

129000

3191

03112

2174303

1644

148226

6753

110667

110578

DMU15

18713

16809

31718

3094

92288

2212

34428

33518

479

410

minus14564

minus17385

DMU16

1576

715090

1844

77

181344

13711

13231

20973

8206

696

1217

913088

5497

910

2306

DMU17

18610

2210

640377

4999

42788

2914

48062

58503

2480

2873

5179

5600

DMU18

10020

1078

178466

98872

4744

4882

8378

910

4486

009

009

9196

9181

DMU19

33552

3379

9627805

62716

92675

626678

673052

67214

225574

33062

minus19

4250

minus18

1281

DMU20

2886

2870

33111

33348

1441

1462

35201

35674

578

841

1135

1116

Sourcecalculatedby

researchers















5 Conclusions
















































References































Volume 2014




Journal of











Journal of


Function Spaces






Algebra























5 Conclusions
















































References































Volume 2014




Journal of











Journal of


Function Spaces






Algebra























































References































Volume 2014




Journal of











Journal of


Function Spaces






Algebra





























Volume 2014




Journal of











Journal of


Function Spaces






Algebra









Research Article A Study of the Strategic Alliance for EMS...

Documents

Transcript of Research Article A Study of the Strategic Alliance for EMS...