Supplier Selection And Evaluation. Through Activity-Based Costing Approach

Transcription

1 Suppler Selecton nd Evaluaton Through ctvty-based Costng pproach Bran Korea Logstcs Team Industral Engneerng / Pusan Natonal Unversty uthor : Han Lee (Economc System nalyss Lab., Industral Engneerng, Pusan Natonal Unversty) Thess supervsor : Prof. K.K. Lee (Economc System nalyss Lab., Industral Engneerng, Pusan Natonal Unversty)

2 bstract Purchasng from suppler nvolves many knd of sgnfcant costs for purchaser. Purchasng decsons related to theses purchases nclude the selecton of suppler, order quanttes to be placed and etc. Such decsons are very much complcated due to the fact that varous crtera must be consdered n decson makng process. Ths paper deals wth suppler selecton and determnaton of order quantty by ctvty-based Costng(BC) and Total Cost of Ownershp(TCO) concept, whch explan the cost caused by suppler more accurately than tradtonal cost accountng Suppler Selecton and Evaluaton through ctvty-based Costng pproach

3 Contents bstract Introducton...4 n ntegrated method for suppler selecton.5. Crtera for suppler selecton 5. ctvtes and cost drvers and other costs analyss mxed nteger programmng model Suppler s performance evaluaton..8 4 numercal example..9 5 Conclusons.. 6 Reference....3 BK Logstcs Team, uly 000 3

4 Introducton In most of manufacturng frm the cot of raw materals and component parts consttutes the man cost of a product, such that n some cases t can account for up 70%. In crcumstances the purchasng department can play a key role n cost reducton, and suppler selecton s one of the mportant functons of purchasng management. The Suppler selecton decsons determne how many and whch supplers should be selected as supply sources and how order quanttes should be allocated among the selected supplers. These suppler selecton decsons are complcated due to the fact that varous crtera must be consdered n the decson makng process. Tradtonal suppler selecton and evaluaton methods are all too often based on quoted prce, whch gnore the sgnfcant costs assocated wth orderng, expedtng, recevng, nspectng, and usng purchased parts and materals. number of alternatve approaches for tradtonal suppler selecton methods have been suggested to take these other cost of factors nto account. The wdely used approach s Weghted Pont Plan. Ths approach uses a smple scorng method, whch heavly depend on human udgments. The smplest method s the Categorcal Method, whch assgns ether good(+), neutral(0), or unsatsfactory(-) to each defned crtera for all supplers, then a total rate for each suppler s calculated[tmmerman,983]. The nalytcal Herarchy Process(HP) s other method whch uses parwse comparson and t was appled by Narasmhan(983) Recently Weber and Current(993), knc(993), Lee and Kaslngam(996), Ghodsypour and O Bren(998), and Karpak et al(999) suggested varous mathematcal programmng approaches as mxed nteger programmng and goal programmng, whch used varous suppler performances crtera as qualty, delvery, flexblty and so on. The above methods use varous crtera, but not consder that how much mpact poor performance of suppler and suppled components and materals for the purchaser. In ths paper we use ctvty-based Costng(BC) approach and Total Cost of Ownershp concept to obectvely analyze all costs assocated wth orderng, expedtng, recevng, nspectng, and usng purchased parts. nd we apply a mxed nteger programmng, whch use the result of the above BC analyss, to select approprate supplers and to determne the order quanttes for selected supplers 4 Suppler Selecton and Evaluaton through ctvty-based Costng pproach

5 n Integrated method for suppler selecton Tradtonal suppler selecton and evaluaton methods are all too often based on quoted prce, whch gnore the sgnfcant costs assocated wth orderng, expedtng, recevng, nspectng, and usng purchased parts and materals. These added costs are gnored for many reasons, but manly because of the shortcomngs of tradtonal accountng systems and the performance measures that flow from them[derose, 990]. The costs of acqurng and usng the component can also be sgnfcant. For example estmates that 50 percent of a frm s nonconformance costs are caused by the extra work nvolved n dsposng of reparng, scrappng, or reworkng defectve purchased materals. These nonconformance costs also nclude costly schedule change and downtme caused by nonconformng supples and materals. When all these factors are taken nto account, suppler lnkages have a greater effect on total cost than the manufacturng process tself. Tradtonal manufacturng cost systems, however, track only the purchase prce assocated wth a partcular part number or suppler : they bury the cots of orderng, expedtng, recevng, nspectng, and usng purchased goods n factory overhead accounts or general manufacturng expenses by obscurng these addtonal costs, accountng systems encourage purchasng managers to select the lowest bdders to avod unfavorable purchase-prce varance, even though the lowest bds may no represent the best overall value n terms of cost, qualty, and delvery. t the same tme, management loses valuable nformaton for evaluatng suppler performance. BC s a management accountng methodology that ams to assgn costs to the actvtes that generate the costs. TCO focuses on the true costs assocated wth the entre purchasng cycle, thus TCO consders all costs assocated wth the acquston, usage, mantenance, and follow-up of purchased goods or servce, not ust the purchasng prce. The model presented n ths paper apples the above BC methodology and TCO concept to suggest a better and more obectve method for supplers selecton and evaluaton. Crtera for suppler selecton In ths paper we use both fnancal and non-fnancal crtera. Whle fnancal crteron s the prce, non-fnancal crtera are qualty, delvery, quantty, and servce & communcaton. We use performance measures to valuate the performances of suppler for the crtera n followng Table : Table. Crtera and Performance Measures Crtera Performance Measure Fnancal Prce mount of purchased tem Nonfnancal Qualty Delvery Quantty Servce & Communcaton Number of trouble due to qualty problem Number(Days) of late or early delvery Number of trouble due to quantty problem Servce & communcaton problem BK Logstcs Team, uly 000 5

6 . ctvtes and cost drvers and other costs analyss In ths secton we analyze actvtes, cost drvers, and other costs caused by the poor performance of suppler for the above crtera. For example, a delvered part that does not conform wth qualty standards causes a producton stop, return or rework and so on. The nferor part used should be replaced by the suppler when delverng the next order. The results of ths analyss are gven n Table. But, It s possble that each purchaser has dfferent actvtes, cost drvers, and other costs. For ths reason we suggest the most general actvtes, cost drvers and other costs Table. ctvtes, Cost Drvers and Costs caused by Poor Performances of Suppler Performance Crtera Qualty Problem Delvery Late Early Quantty Problem Servce & Communcaton Problem ctvty / Cost Drver -Return or rework -Reschedulng or plannng -Reorderng -Repackagng -Re-recevng -ddtonal nspecton -Producton stoppng(set-up) -Dsposng of scrap -Follow-up or expedtng actvty -Reschedulng or plannng -Premum transportaton -Producton stoppng -Holdng & dmnstratve actvty -ddtonal recepton -ddtonal set-up -Reschedulng and plannng -ddtonal nspecton -Reorderng -Informaton exchangeablty Costs Caused by poor performance -Cost of carryng excess nventory due to ther unrelablty -Cost of tran suppler n qualty methods -Lost sales -Lost sales -Cost of carryng extra nventory -Inventory holdng cost -Lost sales -Cost of delays due to slow response to problem.3 mxed nteger programmng model Based on the above actvty analyss, the suppler selecton problem can be formulated as the followng mxed nteger programmng model. Mn z = s. t P X + Y F I 5 I K = = n= = = k= X Q = n nk + Y 5 I M n= = m= = F n D nm nm + 5 I n= = = for =,,3,..., I (.) Y for =,,3,..., I (.) = C n Y + = Z S 6 Suppler Selecton and Evaluaton through ctvty-based Costng pproach

7 X V for =,,3,..., I, =,,3,..., (.3) X for =,,3,..., I, =,,3,...,, M s a arbtrary large number (.4) I = MY Y MZ for =,,3,..., (.5) Y w for =,,3,..., I (.6) = Z W for =,,3,..., (.7) X 0 for =,,3,..., I, =,,3,..., (.8) Y = 0 or for =,,3,..., I, =,,3,..., (.9) Z = 0 or for =,,3,..., (.0), : s tem ndex ( =,,3..., I ) and s suppler ndex ( =,,3..., ) n : Crtera ndex (:qualty, : late delvery, 3:early delvery, 4:quantty, 5:servce) k,m : k & m are ctvty, ndex ( k =,,3,... K ),( m =,,3..., M ) : Prce of tem from suppler P F n Q, V X Y Z nk nm D nm C n S w W : The quantty of performance measure of crteron n for the tem from suppler : Q s demand for the tem, V s Supply capacty of suppler for the tem I : Order quantty of suppler for the tem : =, f suppler s selected for supplyng tem, 0, otherwse : = f suppler s used at all, 0 otherwse : The cost of actvty k caused by performance n of suppler for the tem (Superscrpt means actvty s an actvty that uses performance measure as a cost drver) : The cost of actvty m caused by performance n of suppler for the tem (Superscrpt means actvty s an actvty that does not use performance measure but others as a cost drver) : The cost drver of actvty m caused by performance n of suppler for the tem : The other cost caused by performance n of suppler for the tem : Suppler sustanng cost : The maxmum number of suppler for tem : The maxmum number of suppler The obectve s to mnmze the sum of purchasng cost and addtonal actvtes costs caused by suppler, other costs, and suppler sustanng cost. ddtonal actvtes costs consst of two costs term. The frst s costs of actvtes use performance measure as cost drver and the second s not. Suppler sustanng costs are assocated wth a gven suppler that are ndependent of the quantty and nclude mantanng fles on suppler characterstcs and performance, and perodc evaluatons of suppler performance. Equaton (.) represents the demand constrant for each tem. Equaton (.) requres that all tems be suppled. The capacty of tems from varous supplers s modeled n equaton (.3). Equaton (.4) represent a suppler s selected before orders are placed, (.5) requre tems to be suppled only BK Logstcs Team, uly 000 7

8 from the selected supplers. The maxmum number of supplers to be employed for each tem s represented by equaton (.6). Equaton (..7) represents the maxmum number of supplers to be employed. The non-negatvty and ntegralty restrctons are represented by equaton (.8), (.9) and (.0). 8 Suppler Selecton and Evaluaton through ctvty-based Costng pproach

9 3 Suppler s performance evaluaton In the prevous sectons we use varous factors to select approprate supplers and order quanttes. Because the values of these factors are estmated to compute the degree of mpacts from poor performances of suppler, the values of these factors are dfferent from the actual values. In ths secton we analyze the varances between the estmated and the actual value. There are several advantages to analyze the varances wth BC and TCO concept, not only for the purchaser but also for the suppler. For the purchaser the BC and TCO allow to dentfy the relatve mportance of the dfferent cost components and to gve the useful nformaton about cost reducton. That s, the purchaser may attempt to nfluence estmated unts of cost drvers by reducng or elmnatng some of the actvtes. For the suppler t provdes an obectve ndcaton about the degree of affecton that caused by her own poor performances for the varous crtera, and the mportance of the dfferent crtera nvolved n the supplyng process. For example, n Table 3 we present Cost Drver Rate varance and Cost Drver Frequency varance of qualty performance for tem. The Cost Drver Rate varance refers to factors that allow the purchaser to mprove ts effcency by reducng ts cost drver rates. The purchaser s responsble for ths varance. The Cost Drver frequency varance refers to the dfference actual and estmated use of cost drvers (performance measure) for whch the suppler s responsble snce he can potentally affect the cost drvers. Table 3. The Varances of qualty performance for tem Cost Drver Rate Varance : CD (CDR- SCDR) T F ( ) + F D ( k k k k m T m m m Cost Drver Frequency Varance : SCDR (CD-SCD) T k ( F F ) + ( F D F D m m k m T m ( s actual and S s Estmated standard, CD s a number of cost drver, CDR s cost drver rate) * T s a set of supplers for tem ) ) Together wth the above varance analyss, we can evaluate the suppler s performance by calculatng the rate of supplers for each crteron, whch can be used to reflect suppler s performance n next purchasng decsons. Suppler Ratng Index(SRI) for the suppler performance s calculated as follow : (ctualactvty cost - Estmated standardactvty cost) SRI = + Estmated standardactvty cost for suppler Then we can modfy the obectve functon as follow consderng ths SRI : P X + Y F I 5 I K = = n= = = k= n nk SRI + n 5 I M n= = m= = Y F D nm nm SRI + n 5 I n= = = C n Y + = Z S BK Logstcs Team, uly 000 9

10 4 numercal example n example problem wth solutons s presented n ths secton to llustrate the model for suppler selecton. The demand for the three tems and the sellng prc es and capactes of supplers for the three tems, and the suppler sustanng cost for the four supplers are gven n Table 4. The suppler performance assocated wth the crtera for the three tems are gven n Table 5. Table 7 represents actvtes costs. The results are summarzed n Table 8 and the varances are summarzed n Table 9. Table 4. Prce, Capacty, Demand for Item & Suppler Sustanng Cost Suppler Suppler Suppler 3 Suppler 4 Total Demand Item Item Item 3 Prce($) Capacty Prce($) Capacty Prce($) Capacty Sustanng Cost($) Table 5. Estmated Standard(ctual) Suppler Performance for Item (unt: Frequency) Suppler Suppler Suppler 3 Suppler 4 Performance Item Item Item Item Item3 Item Item3 Item Item3 F 3(5*) 9(0) 8(5) 4 5() 8 4 5(5) () F 7(5) (8) () 6 6(5) 3 4 7(8) 3(5) F 3 8(5) 7(7) 7(6) 9 5(7) 8 6(4) 4(5) F 4 5() 4(3) 3(3) 7 4(6) 9 9 3(3) 0(0) F 5 (5) 5(3) 4(6) 6 6(3) 3 6 () 3(3) (* s actual) (F : qualty problem, F :late delvery, F 3 :early delvery, F 4 :quantty problem, F 5 :servce problem) Table 6. Other Cost caused by Poor Performance (unt:$) Item Item Item 3 C C C Suppler Selecton and Evaluaton through ctvty-based Costng pproach

11 C C Table 7. Estmated Standard(ctual) ctvtes Costs caused by Poor Performance of Suppler for Item Suppler Suppler Suppler 3 Suppler 4 ctvtes Item Item Item Item Item 3 Item Item 3 Item Item 3 k m k m 3k 3m 4k 4m 5k k= 0(8*) 6() 7() 9 7(7) 6(6) 0(0) 4() 4(4) 9(0) 9 8(8) 8 3 () 9(6) 3 5(9) 5(5) (0) 8(7) 4 8 (0) 7(7) m= 9(9) 3(3) (4) 9 () 8 9 7(5) 5(0) 6(9) (0) 5(8) 6 5(5) 6 8 () 9(9) k= 3(5) 7(7) (0) 7 () 0 6 7(7) 9() (4) (5) 4(9) 0 (9) 8 9 () 0(0) m= () 3(3) (0) 5 7(7) 7 8 3(3) 7(9) 9(0) 7() 9(0) 8 5(8) 8 7 8(8) 4(6) k= (0) 0(0) () 8 9(9) 7 3 () 3(3) 9(0) 7(7) 7(3) 4 (0) 3 6(30) 0(5) m= 6(5) 3(0) 4(3) 6 4(4) 6 9 (8) 0(0) 5() 5(5) (8) 3 0(0) 5() 0(5) k= (0) 9(9) 8(0) 9 () 7 7 5(9) 3() 5(0) 3(3) 7() 8 4(4) 7 5 4(4) 3(3) m= 8(0) 8(8) 5(7) 4 () 8 9 8(7) () 4(3) (0) (5) 3(3) 5 0(0) 7(7) k= 7(0) 0(3) 3(5) 7 (7) 9 3 7(6) 5(5) (0) 6(6) 8(5) 7 3(3) 4 7 7(7) 0(4) 5m () 6(6) 5(5) 7 7(7) () 7(7) (unt:$) (* s actual actvty cost) Table 8. Results of Example Problem Item Item Item 3 Suppler Suppler Suppler 3 Suppler BK Logstcs Team, uly 000

12 Table 9. The Varances (unt : $) Item Item Item 3 CDR Var. CDF Var. CDR Var. CDF Var. CDR Var. CDF Var. Qualty 30 U 75 F 85 F 3 U 48 F 86 F Delvery Late 77 U 6 F 8 F 9 F 35 U 33 U Early 74 F 0 F 33 F 48 F 43 U 5 U Quantty 30 U 56 F 3 F 94 F 8 U 4 U Servce F 7 U 8 U 04 F 30 U 53 F Total Var. 6 U 396 F F 45 F 34 U 5 F 34 F 536 F 9 U (F: Favorable, U: Unfavorable) Table 0. The SRI of Suppler for Item Suppler Suppler Suppler 4 Performances I I I I3 I I3 F F F F F In Table 9, because of better performances of supplers the total varance s favorable for tem and because of the efforts of cost reducton of purchaser and better performances of supplers the total varance s favorable for tem and cost drver rate varance of tem 3 s unfavorable by reasons of neffcency of purchaser. Table 0 represents the SRI of each suppler for tem. Suppler Selecton and Evaluaton through ctvty-based Costng pproach

13 5 Conclusons In ths paper we proposed an ctvty-based Costng approach and Total Cost of Ownershp concept to obectvely analyze all costs assocated wth suppler and suppler components and materals. nd we apply a mxed nteger programmng, whch use the result of the above analyss, to select approprate supplers and to determne order quanttes for selected suppler. The proposed method wll help to make better suppler selecton decson resultng n decreased overall costs. Own approach wll gve the purchaser useful nformaton for possble cost reducton and gve an obectve ndcaton about the degree of affecton that caused by own poor performances for the varous crtera. BK Logstcs Team, uly 000 3

14 6 References knc, U. (993), Selectng a set of vendors n a manufacturng envronment, ournal of Operatons Management,, pp.07- Ghodsypour, S.H, and O Bren, C. (998), decson support system for suppler selecton usng an ntegrated analytc herarchy process and lnear programmng, Internatonal ournal of Producton Economcs, 56-57, pp.99- Karpak, B. and et al. (999), Mult-Obectve Decson-Makng In Suppler Selecton: n pplcaton of Vsual Interactve Goal Programmng, The ournal of ppled Busness Research, 5,, pp.57-7 Kaslngam, R.G., and Lee, C.P (996), Selecton of vendors- mxed nteger programmng approach, Computers & Industral Engneerng, 3,/, pp Narasmhan, R.(983), n analytcal approach to suppler selecton, ournal of Purchasng and Supply Management, wnter, pp. 7-3 Tmmerman, E.(986), n approach to vendor performance evaluaton, ournal of Purchasng and Supply Management, wnter, pp. -8 Weber, C.., and Current,.R (993), multobectve approach to vendor selecton, European ournal of Operatonal Research, 68, pp Suppler Selecton and Evaluaton through ctvty-based Costng pproach