article review

8_articles.pdf

Home >Business & Finance homework help >Operations Management homework help >article review

Supply Chain Risk Management Approaches Under Different Conditions of Risk Ila Manuj1, Terry L. Esper2, and Theodore P. Stank3

1University of North Texas 2Sam M. Walton College of Business at the University of Arkansas 3University of Tennessee

A recent Deloitte study of 600 Supply Chain and C-Level executives revealed that 45% felt that their supply chain risk management pro-grams were only somewhat effective or not effective at all, while a mere 33% used risk management approaches to proactively and strate- gically manage supply chain risk based on conditions in their operating environment. Using a two-method approach, the research summarized in this paper investigates the effectiveness of different supply chain risk management approaches by examining how performance varies when these approaches are applied under different risk conditions. The results counter prevailing knowledge regarding the appropriate use of such widely acknowledged risk management approaches as postponement and speculation, and highlight the dangers of functionally isolated decision making. The results lend credence to increasing calls for interdisciplinary research to address broad-based supply and demand chain problems, and support the need to utilize performance metrics such as net profit to accurately assess supply chain decisions.

Keywords: supply chain risk management; global supply chains; risk management approaches; simulation

INTRODUCTION

Effective supply chain risk management seeks to control unex- pected outcomes by systematically implementing appropriate approaches to managing and/or mitigating risk (J€uttner et al. 2003; Norrman and Jansson 2004; J€uttner 2005). Supply chain risk management is of utmost importance to senior managers given the potential dire consequences of risk occurrences. For example, Boeing, Cisco and Pfizer each encountered unexpected losses and/or expenses of greater than $2 billion due to ineffec- tive supply chain risk management decisions (Hult et al. 2010). Despite the potential negative outcomes, a recent Deloitte study of 600 Supply Chain and C-Level executives revealed that 45% felt that their supply chain risk management programs were only somewhat effective or not effective at all, and a mere 33% used risk management approaches to proactively and strategically manage supply chain risk based on conditions in their operating environment (Deloitte Development LLC 2013).

The purpose of this research, therefore, is to investigate the effectiveness of different supply chain risk management approaches by examining how performance varies when approaches are applied under different conditions of supply chain risk. Grounded in systems design theory (SDT), the research uses a two-method approach. First, a conceptualization of key supply chain risk management opportunities is developed based upon results of field-generated research to portray the relationships between the use of different supply chain risk approaches under differing conditions of supply chain risk. Next, computer simula- tion modeling is used to precisely observe the impact of the use of four different supply chain risk approaches (hedging, assum-

ing, postponement, and speculation) under varying conditions of supply chain risk on overall net profit. This research makes two important contributions:

1. We enhance understanding of the relationships between sup- ply chain risk management approaches and performance under different supply chain risk conditions.

2. We provide broad insights on supply chain risk decision mak- ing to inform future research and practice.

The following sections present the results of the grounded the- ory research and computer simulation modeling, providing a brief review of relevant literature as appropriate to support the qualitative results and develop research hypotheses. Results of post hoc analyses and interviews conducted to obtain practitioner insights of the results are then reported, followed by a summary of the research and managerial implications.

STUDY 1: INDUCTIVELY FRAMING THE RESEARCH

As we began the study, anecdotal evidence suggested that firms were struggling to design effective risk management systems (we elaborate on systems design preceding hypothesis development). Specifically, as firms seemed to lack insight into how to align risk strategies with risk scenarios, we conducted in-depth inter- views with 14 supply chain managers from nine different manu- facturing firms. The number and content of in-depth interviews was based on the concept of “theoretical sampling” (Glaser and Strauss 1967; Mello and Flint 2009). The initial participant sam- ple was selected based upon experience with the phenomena, job profile and responsibility, and willingness to participate in the research. Additional participants were selected as the interviews progressed to enable further exploration of new categories and concepts that emerged. Interviews continued until “theoretical saturation” was reached. The participants came from a variety of

Corresponding author: Ila Manuj, Department of Marketing and Logistics, University of North Texas, 1155 Union Circle #311160, Denton, TX 76203-5017, USA; E-mail: [email protected]

Journal of Business Logistics, 2014, 35(3): 241–258 © Council of Supply Chain Management Professionals

different roles within multiple industries, with most having over 10 years of experience. Appendix 1 provides demographic details about the interviewees.

The interview protocol included the use of broad open-ended questions followed by focused and directed questions as concepts emerged within and between successive interviews (Strauss and Corbin 1998; Mello and Flint 2009; Randall et al. 2010; Manuj et al. 2014). The interviews, which were conducted over a six- month period, lasted from 40 and 60 min; interviews were audio- taped and transcribed verbatim. ATLASti was used for coding the transcripts and standard qualitative techniques were followed to develop core categories of strategic approaches, supply and demand risk, and performance outcomes. Categories that emerged from the grounded theory research were then compared to existing research to provide a literature-based grounding for the contextualized concepts found in the field. Table 1 provides sample quotes that support the categorization of supply chain risk management approaches, risk categories, and appropriate performance metrics.

Study 1 results

Managers in the qualitative study emphasized the importance of risk arising from the external environment that was out of their direct control. In particular, managers were most concerned with the performance implications of the strategic choice of risk man- agement approaches that are inappropriate considering the char- acteristics of the risk environment. A review of the academic literature suggests the same theme, as the appropriateness of sup- ply chain risk management approaches based on differing condi- tions of supply and demand risk environments has been identified as an issue that needs further and immediate attention (J€uttner et al. 2003; Wagner and Bode 2008; Schoenherr 2009).

Over 40 supply chain environmental risk conditions were identified. These risk conditions could be grouped into two pre- dominant categories. The first category, supply-side risk, is the risk associated with the availability of raw materials or subcom- ponents from upstream suppliers that affect the ability of the focal firm to meet customer demand within anticipated cost and delivery time requirements (Zsidisin 2003; Manuj and Mentzer 2008a,b). The second risk category is demand-side risk associ- ated with availability of finished product to meet customer demand within anticipated cost and time requirements (Zsidisin 2003; Manuj and Mentzer 2008b). Many events related to secu- rity, policy, competitive, and resource risk eventually manifest themselves as supply or demand risk.

Interview managers indicated that the appropriate risk manage- ment approach involves a decision to minimize or to take on risk. For supply-side risk, the approaches included hedging or assuming. Demand-side risk approaches included postponement or speculation. These approaches are defined below:

1. Hedging is designed to balance exposure to supply-side risk through a globally dispersed portfolio of suppliers and facili- ties such that a single event (like currency fluctuations or nat- ural disasters) does not affect all the entities at the same time and/or with the same magnitude (Carter and Vickery 1989; Bartmess and Cerny 1993).

2. Assuming strategy is designed to internalize supply-side risk through vertical integration of supply to focus resources and exploit economies of scale (Wernerfelt and Karnani 1987).

3. Postponement defers the actual commitment of resources by delaying manufacturing and/or logistics operations to manage risk in demand uncertainty by maintaining flexibility and delaying incurred costs (Bucklin 1965; Wong et al. 2009). It must be noted that while postponement includes both form and time, we focus on form postponement as the qualitative interviews revealed a greater use of mass customization and agile manufacturing as the desired means to improve coordi- nation between supply and demand (Yang et al. 2004; Boone et al. 2007).

4. Speculation involves maintaining an inventory of finished products instead of component parts, basing manufacturing and logistics decisions on anticipation of customer demand, committing resources in advance but reducing unit costs through economies of scale, experience curves, etc. (Bucklin 1965; Miller 1992).

The qualitative research also revealed significant complexities associated with applying the most appropriate approach to effec- tively manage or mitigate supply chain risk. The most important of these was the notion that functional metrics are not sufficient to assess the effectiveness of global supply chain risk manage- ment strategies. Rather, the findings suggest that an overall met- ric such as total profit is preferred as it takes into account multiple performance elements including revenue, operating costs (production, warehousing, transportation, and inventory), and penalty costs associated with risk events. Such a comprehensive and holistic measure was viewed as preferred, but difficult to execute in practice. For example, when discussing total profit as a key measure, participants noted an excessive focus on unit pur- chasing cost, with one respondent commenting, “The import of that product was (supposed to be) very profitable, but a lot of profitability got wiped out.” Another suggested that, “you can see pretty quickly that your leverage on a price increase is far better than your leverage on a cost reduction. (But) we do so much work on that cost reduction.” Similarly, a singular focus on revenue was also cited as a challenge to profitability. For example, one manager noted, “we’ve done a great job growing our revenues, but we’ve not done a great job growing margins.” Despite respondents’ desire to use a broader metric like profit, many stated that obtaining such information was a challenge. For example, one manager complained that, “I don’t have any quan- tifiable facts or figures that I can give you” on profit because he could not gather the data.

The results of the qualitative research provided an interesting glimpse into the complexities of choosing the appropriate supply chain risk management approach to optimize performance given different supply- and demand-side risk characteristics. Supply chain risk management research and practice could benefit from quantitative research that specifically investigates the relation- ships among the strategic approaches and risk categories identi- fied in Study 1. Such an investigation requires a method that allows for the precise observation of the performance implica- tions of simultaneous and holistic combinations of different stra- tegic risk approaches and environmental conditions. Thus, we

242 I. Manuj et al.

adopted a theory-grounded hypothesis testing approach employ- ing computer simulation modeling for Study 2.

STUDY 2: SIMULATION MODEL

Supply chain system design is one persistent role for supply chain decision makers. Through the effective management and design of supply chains systems, managers answer questions related to the best ways to organize for success, mitigate trade- offs, and overcome resistance to change (Scott and Davis 2006). Thus, SDT is a natural fit for supply chain issues like risk or sus- tainability (Fawcett et al. 2010), particularly when considering the role of the external environment in designing supply chain systems to manage risk and maximize performance (Fawcett et al. 2012).

SDT is a foundational organization theory, focused primarily on the purpose and processes of the organizational unit (Scott and Davis 2006). The core assumptions and arguments of SDT have evolved. Organizations were once viewed as entities focused on basic goal achievement and survival. The more con- temporary perspective conceives organizations as complex sys- tems that engage in a continual process of organizing, adapting, and changing to not just survive, but also to improve perfor- mance. This latter perspective is the “open systems” argument, which primarily emphasizes that organizations inherently depend upon the external environment, and must respond as it evolves (Scott and Davis 2006).

Theoretical frameworks like Contingency Theory (Drazin and Van de Ven 1985) and Resource Dependence Theory (Pfeffer and Salancik 1978), which have proven to be quite popular in supply chain management research, have foundations in the open systems

Table 1: Representative quotes from grounded theory interviews

Concept Representative proof quotes

Supply-Side Risk: Associated with availability of material supply from upstream suppliers that affect the ability of the focal firm to meet customer demand

“The problem with these long supply lines is they’re also highly variable. I mean, it’s not just the mean, it’s the standard deviation of cycle time.”“The problem is that when these suppliers are half a world away from you, they don’t necessarily operate with the same quality and the same safety standards as we adhere to in the US”

Demand-Side Risk: Associated with availability of finished product to meet customer demand within anticipated cost and time requirements

“The company has gotten into severe situations where they have been out of product and can’t respond to the marketplace because of some spike in demand for whatever reason.”“You have no idea what is going to happen next month . . . so you have the huge risk of forecasting incorrectly and it happens over and over.”

Hedging: Balancing exposure to supply-side risk through a globally dispersed portfolio of suppliers and facilities

“I need some flexibility and I can’t have the risk of only being with one. If I absolutely know I’m dependent on you, then I lose some kind of leverage.”“You outsource entire product-line to China, so, here you are suffering all the risks associated with the global outsource. And if RMB strengthens 20% tomorrow, it’s going to wipe out a lot of profitability.”

Assuming: Internalizing supply-side risk through vertical integration of supply to focus resources and exploit economies of scale

“Assuming . . . is a risk/benefit trade-off. We get benefit from moving to China through (lower) costs and we can quantify how much we think we can save by moving product to China.”

Postponement: Delaying operations to manage risks in demand uncertainty by maintaining flexibility and delaying incurred costs

“That factory is flexible enough to operate with a seven-day schedule. On eight, they allow corporate planners to change their schedule any way that’s necessary to react to the orders. You see enormous flexibility in that (cost pressure and fluctuating demand) kind of environment.”

Speculation: Maintaining inventory of finished products based on anticipation of customer demand

“We approach (product) architecture in an integrated fashion today. Greasing the parts and building the product all happen at the same time. (But, we) make sure that you understand what all the risks are and the people are picking up on them and addressing them.”“You are two months away from demand and the additional inventory that you are putting into the system. And, you know, the forecast error multiplies exponentially as you extend the lead time.”

Performance Metric: Includes operating revenues, operating costs, and penalty costs associated with risk events

“We’ve done a great job growing our revenues, but we’ve not done a great job growing margins. . .”“We have to be growing and improving our financial performance, not kind of just barely holding the status quo.”

Supply Chain Risk Management Approaches 243

perspective of SDT. Likewise, by emphasizing the role of environ- mental forces in designing systems as its core, the open systems design perspective is an applicable theoretical foundation for this research. It captures the essence of the issues that emerged from the qualitative research by suggesting that the nature of the exter- nal environment is an important framework through which supply chain systems should be designed. Thus, combinations of hedging, assuming, postponement, or speculation strategies can only prove effective to the extent to which these design strategies align with the characteristics of the external risk environment (Scott and Davis 2006; Fawcett and Waller 2014).

Hypothesis development

Shao and Ji (2009) establish that different approaches to sourcing strategy have different impacts on system performance depending upon supply-side risk conditions. Hedging approaches are partic- ularly preferable when there is a need for high flexibility in sourcing to compensate for uncertain conditions caused by sup- ply chain characteristics such as unreliable supply or unstable manufacturing schedules (Chung et al. 2010). Therefore, for sup- ply chains facing high-supply risk, hedging approaches such as dual or multi-sourcing are likely a more appropriate strategic approach than for supply chains facing low-supply risk. Con- versely, for supply chains facing low-supply risk, the cost of finding and setting up an alternative source of supply may be higher than potential savings resulting from lower risk of supply disruption.

Assuming approaches to supply-side risk may take the form of sourcing from a single supplier, source, or geopolitical area, often employed to obtain lower supply costs (Inderst 2008). However, such an approach will likely not be effective when there is high risk, such as significant variability in performance or opportunism (Berger et al. 2004), because of the risk expo- sure.

Therefore, for high-supply risk (irrespective of the level of demand risk and choice of strategic risk management approach on the demand side):

H1: Supply chains that use a hedging risk management approach under conditions of high-supply risk will yield higher profit than supply chains that use assuming.

For low-supply risk, it is proposed that irrespective of the level of demand risk and risk management strategy on the demand side:

H2: Supply chains that use an assuming risk management approach under conditions of low-supply risk will yield higher profit than supply chains that use hedging.

Similarly, different strategic approaches to downstream supply chain operations have different impacts on system performance depending upon demand-side risk conditions (Perry 1991; Chiou et al. 2002; Yang et al. 2004). Form postponement approaches are appropriate under conditions of high levels of demand varia- tion and uncertainty, demand customization, high component costs, short product life cycle, high levels of product modularity (Perry 1991; Chiou et al. 2002; Yang et al. 2004). Therefore,

supply chains facing high demand-side uncertainty benefit more from form postponement.

By contrast, speculation is appropriate under conditions of low-demand variability and uncertainty, high product popularity, low inventory holding costs, and larger market share (Bailey and Rabinovich 2005). Supply chains facing low-demand variability and uncertainty are better suited to achieve the benefits of specu- lation (LeBlanc et al. 2009).

Therefore, for high-demand risk (irrespective of the level of risk and choice of strategic risk management approach on the supply side):

H3: Supply chains that use a postponement approach under conditions of high-demand risk will yield higher profit than supply chains that use speculation.

For low-demand risk, it is proposed that irrespective of the level of risk and strategy on the supply side:

H4: Supply chains that use a speculation approach under conditions of low-demand risk will yield higher profit than supply chains that use postponement.

Applying the open SDT concept of differential performance under different environmental conditions (Scott and Davis 2006), it can be argued that a supply chain system that is designed with risk management strategic approach combinations that are appropriate for the prevailing characteristics in both supply and demand risk conditions will perform better than sup- ply chains that adopt risk management approaches that misalign with either supply- or demand-side risk (Lee 2002). Building on the base hypotheses presented above that posit the impact of the performance implications of individual risk management strategic approaches, the following hypotheses posit the expected relationships when both supply- and demand-side risk are considered:

H5: Supply chains that use an assuming risk management approach on the supply side and a speculation approach on the demand side under conditions of low-supply risk and low-demand risk will yield higher profit than supply chains that use other combinations of risk management approaches under those risk conditions.

H6: Supply chains that use an assuming risk management approach on the supply side and a postponement approach on the demand side under conditions of low-sup- ply risk and high-demand risk will yield higher profit than supply chains that use other combinations of risk manage- ment approaches under those risk conditions.

H7: Supply chains that use a hedging risk management approach on the supply side and a speculation approach on the demand side under conditions of high-supply risk and low-demand risk will yield higher profit than supply chains that use other combinations of risk management approaches under those risk conditions.

H8: Supply chains that use a hedging risk management approach on the supply side and a postponement approach on the demand side under conditions of high- supply risk and high-demand risk will yield higher profit

244 I. Manuj et al.

than supply chains that use other combinations of risk management approaches under those risk conditions.

Study 2 method

The hypotheses were tested via a simulation model developed according to established simulation frameworks (Law and Kelton 1982; Van der Zee and Van der Vorst 2005; Manuj et al. 2009) using Supply Chain Guru by the Llamasoft Corporation (Ann Arbor, MI). Multi-country supply chains were simulated since longer and more complex supply chains add to the sources of risk failures and exponentially increase risk (Schoenherr 2009), thereby providing a more rigorous test of the hypotheses. The design con- sists of two levels each of supply- and demand-side risk, and two approaches each to supply risk strategy, and demand risk strategy (a 2 9 2 9 2 9 2 model) resulting in 16 modeling scenarios. Printers were chosen as the simulated product due to the medium value-to-weight and weight-to-bulk ratios, and demand characteris- tics that are relatively predictable over a one-year horizon but are quite variable on a daily basis. Moreover, the use of postponement (Lee et al. 1993; Feitzinger and Lee 1997) and hedging through multiple sourcing (Inderst 2008) has previously been explored for printer companies such as Hewlett-Packard.

Simulation models are most useful when a limited number of alternatives and variables are considered (Rosenfield et al. 1985). Therefore, we created parsimonious but robust operationaliza- tions of strategic approaches and supply- and demand-side risk conditions. The model simulates a manufacturing firm with both assembly and distribution operations (M/D) based in the United States. The firm sells two types of printers that can either be assembled using a combination of two components or can be purchased from suppliers already assembled. The M/D has two supply base options (one each in the United States and China) and two customers (C1 and C2, both located in the United States). The M/D sells two products – Product A to C1 and Product B to C2. Both products are composed of a common component (Common-Component CC), and each have a unique component (A-Component [AC], unique to Product A; and, B- Component [BC] unique to Product B). Both suppliers—S1 and S2—can supply products A and B, and components AC, BC, and CC. Figure 1 portrays the simulation model layout.

The focal concepts in the model were first identified during the grounded theory research and subsequently refined based upon established criteria in extant literature and consultation with

a focus group of seven senior executives of a global manufactur- ing firm. Table 2 provides a list of the four independent vari- ables, including values, distributions, and sources of data. The qualitative research suggested that most supply risk conditions (such as lead times, capacity constraints, opportunism, etc.) man- ifest themselves as variations in lead time, cost, and quality. The low- (high-) supply risk condition for the Chinese supplier was operationalized as low (high) levels of supplier order processing time variability, cost variability, and quality defects. Lead-time variability is further divided into order processing time variabil- ity, and transportation lead-time variability. Transportation times and variability are the same for both low- and high-risk Chinese suppliers because transit time and variability is a function of transportation companies’ capabilities and not of the suppliers’ capabilities.

Most demand risk such as new product introductions, fads, seasonality, demand amplification etc., eventually manifest them- selves as variations in demand. In the interest of parsimony while retaining the essence of demand risk, the low- (high-) demand risk condition was operationalized as low- (high-) demand variability.

The assuming approach to supply chain risk management strategy was operationalized by using a single Chinese supplier. Hedging was operationalized by using one supplier each in the United States and China. Speculation was operationalized by sourcing finished products from suppliers and postponement by sourcing components from the suppliers and assembling the com- ponents at the M/D based upon actual customer demand.

Net profit, the ultimate dependent variable in the simulation, is measured by the difference between total revenue and total cost for the M/D. Total Revenue is calculated as the product of unit sales volume multiplied by the sale price for each product. Total costs include transportation, inventory, production (assembly), warehousing (picking and packing), penalty (for late delivery), and purchase costs.

The model is triggered by the generation of demand at cus- tomer locations. Demand is distributed normally, truncated at zero, with a mean of 1,000 units per day per customer. Demand generated at customers is transmitted instantaneously to the M/D at a cost of $5. Each order is due in 15 days and late orders incur a penalty of $35/unit. The next step is order receipt and processing at the M/D which includes picking products or com- ponents, assembling components (postponement only), packing, and shipping goods. Figure 2 illustrates the sequence of events in the model.

Figure 1: Simulated supply chain.

Supply Chain Risk Management Approaches 245

T ab

le 2:

D es cr ip ti on

of in de pe nd en t va ri ab le s

S u p p ly -s id e ri sk

R is k fa ct or s

D efi n it io n

D is tr ib u ti on

G lo b al

(l ow

) G lo b al

(h ig h )

U n it ed

S ta te s*

S ou

rc es

of d at a

1. S up pl ie r O rd er

P ro ce ss in g T im

e V ar ia bi li ty

T im

e fr om

or de r

pl ac em

en t to

re pl en is hm

en t at

th e

su pp li er

fa ci li ty

N or m al

(M ea n,

S D )

ra ti on al e:

L ar ge

sa m pl e si ze , w el l-

kn ow

n pr op er ti es

(N ee dh am

an d

E ve rs

19 98 )

N (1 5,

1. 5)

N (1 5,

4. 5)

N (1 0, 1)

In te rv ie w s w it h m an ag er s fo r

bu il di ng

si m ul at io n m od el . B as ed

on re la ti ve

re li ab il it y of

or de r

pr oc es si ng

ti m es . (C om

pu S ys te m s,

T ra ns po rt C o,

H om

eC o,

C om

pu te rW

ar e)

2. C os t

V ar ia bi li ty

in co st

T ri an gu la r (T )

ra ti on al e:

sm al le st ,

la rg es t, an d m os t

li ke ly

va lu es

kn ow

n (H

es se

20 00 )

B as ed

on re la ti ve

co st di ff er en ce s

fr om

m an ag er s co ns ul te d fo r

si m ul at io n m od el

de ve lo pm

en t,

B us in es s W ee k re po rt (E ng ar di o

et al . 20 04 ), B ur ea u of

L ab or

S ta ti st ic s D at a on

in te rn at io na l

w ag e ra te s, R M B

(C hi ne se

cu rr en cy ) tr en ds

w it h re sp ec t to

U S D , an d ex pe ct ed

ri se

in pu rc ha se

co st s fr om

C hi ne se

su pp li er s. (C om

pu S ys te m s)

P ro du ct

A or

P ro du ct

B 15 %

fo r lo w -s up pl y

ri sk ; 45 %

fo r hi gh

su pp ly

ri sk

(M in , M ea n,

M ax )

T (6 0, 64 .5 ,

69 )

T (6 0,

73 .5 ,

87 )

C om

po ne nt

A C

or C om

po ne nt

B C

T (1 5,

16 .1 25 ,

17 .2 5)

T (1 5,

18 .3 75 ,

21 .7 5)

C om

po ne nt

C C

T (3 5,

37 .6 25 ,

40 .2 5)

T (3 5,

42 .8 75 ,

50 .7 5)

3. Q ua li ty

V ar ia bi li ty /Y ie ld

V ar ia bi li ty

in us ab le

pr od uc ts an d

co m po ne nt s re ce iv ed

fr om

su pp li er s

1% de fe ct s fo r

do m es ti c

su pp li er ; 2%

de fe ct s

fo r lo w -r is k C hi na

su pp li er ; 10 %

de fe ct s

fo r hi gh -r is k C hi na

su pp li er

0. 98

0. 90

0. 99

B as ed

on in te rv ie w s w it h m an ag er s

co ns ul te d fo r bu il di ng

si m ul at io n

m od el s. (H

om eC

o, C om

pu S ys te m s)

D em

an d -s id e ri sk

R is k fa ct or s

D efi n it io n

R em

ar k s

L ow

ri sk

H ig h ri sk

S ou

rc es

of d at a

1. V ar ia bi li ty

of de m an d

A ve ra ge

va ri at io n in

da il y de m an d

N or m al

(m ea n,

st an da rd

de vi at io n [S D ])

N (1 ,0 00 , 10 0)

N (1 ,0 00 , 30 0)

A ve ra ge

ba se d on

se co nd ar y da ta

of a le ad in g pr in te r

m an uf ac tu re r. S D

va li da te d in

in te rv ie w s. C V

va lu es

by M en tz er

an d G om

es (1 99 1)

(C om

pu S ys te m s, H om

eC o)

N ot es : *U

.S . va lu es

re m ai n co ns ta nt

th ro ug

ho ut

al l sc en ar io s. C V , co ef fi ie nt

of va ri at io n.

246 I. Manuj et al.

In-depth details concerning the determination of values for the key variables in the simulation model are contained in Appen- dix 2. In addition, the results of model verification and validation procedures established by Law and Kelton (1982) and Sargent (2000) are also included in Appendix 2.

Study 2 analysis and results

The results of univariate analysis of variance (ANOVA) are pre- sented in Table 3. ANOVA was conducted to test the null hypothesis that the means of the 16 scenarios are same. This omnibus test is important for establishing that further comparison of cell means are appropriate for hypothesis testing. The resulting F-statistics of the four-way ANOVA model indicate that all main effects and interaction effects (both two-way and four-way) are significant (p < .001 for all model terms). The means for net

profit in all scenarios are presented in Table 4. Overall, in all scenarios except the low supply-low demand risk scenario, assuming-speculation shows the widest range of outcomes. This is because assuming and speculation approaches are risk-taking strategies that leverage low costs. If the lowest risk events occur in a run, then low purchase cost as well as lack of assembly costs lead to very low costs, and therefore, high profit. However, if the highest risk events occur in a run, there is no buffer on the supply side and no inventory on the demand side, which leads to high penalty costs and low revenues.

General Linear Model and Tukey’s W procedure for multiple comparisons of means are used to test hypotheses at the 5% level of significance. Tukey’s W procedure is appropriate as it is best at controlling Type 1 error under conditions of nonnormal popu- lations, unequal variances, but equal sample size for multiple comparisons (Ramsey et al. 2010). Tukey’s W procedure uses the Studentized range distribution that makes it a conservative test as there is a low probability of declaring at least one pair of means significantly different when running multiple comparisons (Ott and Longnecker 2001). Results for testing of the hypotheses are summarized in Table 5. In Figure 3, mean values of total profit are plotted for strategies under conditions of low and high levels of supply and demand risk. Figures 3a–d correspond to hypotheses H1 through H4 respectively. Figure 3(a) shows the mean values of total profit for all strategic approaches under high-supply risk, (b) under low-supply risk, (c) under high- demand risk, and (d) under low-demand risk.

H1 is supported (see Table 5 and Figure 3a). As shown in Figure 3a, under conditions of high-supply risk, the graphs corre- sponding to hedging-postponement and hedging-speculation sce- narios lie above the graphs corresponding to the assuming- postponement and assuming-speculation scenarios. This supports the hypothesis that hedging is a better strategic approach to sourcing than assuming for supply chains facing high-supply risk. This result confirms the commonly held view in theory and practice that it is beneficial to hedge risk by creating options when facing high-risk contingencies.

H2 is partially supported (see Table 5 and Figure 3b). Under conditions of low-supply risk, the graph in Figure 3b correspond- ing to the assuming-speculation scenario lies above the graph

Figure 2: Sequence of events in the simulation.

Table 3: Univariate test results*

Source of variation F-ratio df Sig. (p)

SR 71,320 1 <.001 DR 216.277 1 <.001 SS 186.657 1 <.001 DS 6,007 1 <.001 DR 9 DS 73.233 1 <.001 SR 9 SS 141.084 1 <.001 SR 9 DR 9 SS 9 DS 129.148 9 <.001

Notes: *Dependent variable is total profit. Source of variation column shows main and interaction variables. F-ratio column shows the F-statis- tic, which tests whether the effect of each independent variable is signifi- cant. df column shows the degrees of freedom used to obtain observed significance levels. Sig. column shows the significance level (p-value). Interaction variables are abbreviated. For example, DR 9 DS represents interaction between demand risk and demand strategy. SR 9 DR 9 SS 9 DS represents the interaction between supply risk, demand risk, supply strategy, and demand strategy. Other interactions are represented in a similar format.

Supply Chain Risk Management Approaches 247

corresponding to hedging-speculation scenario, while the graph corresponding to assuming-postponement scenario lies below the graph corresponding to hedging-postponement scenario. This finding suggests that, as hypothesized, assuming yields better performance than hedging under low-supply risk, but only when a speculation is used on the demand side.

H3, that postponement would perform better under high-risk conditions, is not supported (see Table 5 and Figure 3c). The graphs corresponding to the assuming-postponement and hedg- ing-postponement scenarios lie below the graphs corresponding to the assuming-speculation and hedging-speculation scenarios respectively. It may be inferred that in all scenarios, speculation was better than postponement. A closer examination of this pat- tern of results suggests that production costs, particularly the cost of assembling products in the United States, and penalty costs were the primary cost drivers that make postponement less profit- able for all scenarios.

H4 is supported, as findings indicate that speculation is better than postponement in the face of low-demand risk (see Table 5 and Figure 3d). The graphs corresponding to the assuming-spec- ulation and hedging-speculation scenarios lie above the graphs corresponding to assuming-postponement and hedging-postpone- ment scenarios respectively. As aforementioned, all postpone- ment scenarios generally incur higher assembly and penalty costs than assuming scenarios, confirming the well-accepted idea in both theory and practice that it is beneficial to focus resources on fewer initiatives when risk is low.

Figure 4 presents the outcomes of all four combinations of strategic approaches under different supply and demand risk

conditions. All four charts show a combination of demand- and supply-side risk management approaches that is distinctly- superior to other combinations of approaches. This confirms the underlying concept that different combinations produce sig- nificantly different outcomes under similar environmental con- ditions.

H5 is supported (see Table 5 and Figure 4a). Although Fig- ure 4a seems to indicate a lack of significant differentiation of the assuming-speculation combination in low supply and demand risk conditions, a closer inspection of Figure 4a reveals other- wise. Not only does assuming-speculation have a higher mean, but it also demonstrated higher minimum outcome values and a narrower range of outcomes.

H6 is not supported (see Table 5 and Figure 4b), as assuming- postponement proved to not be the best strategy combination under conditions of low-supply and high-demand risk. As Fig- ure 4b shows, assuming-postponement lies to the left of all other strategy combinations. It actually turns out to be the worst com- bination. The best combination for a low supply-high demand risk scenario is assuming-speculation. This finding is similar to findings for H3, that the mere presence of high-demand variabil- ity does not justify the use of a postponement strategy but rather is dependent upon the underlying cost structure in manufactur- ing. This is one of the more interesting findings and is discussed in detail in the following discussion of results and sensitivity analysis section.

H7 is supported (see Table 5 and Figure 4c). Under the condi- tions of high-supply and low-demand risk-hedging-speculation lies to the right of all strategies and clearly outperforms all other

Table 4: Mean net profit for all scenarios

Cell # SR DR SS DS Average revenue Average cost

Profit

Average Min Max Range

1 L L He Po 215,508,034 165,494,485 50,013,549ab 46,195,856 54,451,744 8,255,888 2 L L He Sp 215,422,308 150,428,237 64,994,071ac 60,368,816 68,188,192 7,819,376 3 L L As Po 214,668,509 166,783,287 47,885,222d 46,291,680 48,903,504 2,611,824 4 L L As Sp 214,665,879 148,186,321 66,479,558bcd 63,261,712 70,702,416 7,440,704 5 L H He Po 216,024,669 167,146,943 48,877,726 46,034,704 52,437,824 6,403,120 6 L H He Sp 215,923,115 151,475,954 64,447,162e 61,871,744 68,920,448 7,048,704 7 L H As Po 215,009,184 169,299,532 45,709,652 43,842,048 48,079,856 4,237,808 8 L H As Sp 214,724,404 147,835,642 66,888,762e 63,379,632 71,240,304 7,860,672 9 H L He Po 207,105,325 203,154,722 3,950,603fg �1,218,224 7,953,872 9,172,096 10 H L He Sp 207,025,863 185,899,447 21,126,416fhi 18,080,624 25,264,832 7,184,208 11 H L As Po 197,187,995 201,237,901 �4,049,906ghj �9,107,968 1,566,864 10,674,832 12 H L As Sp 197,346,080 186,101,005 11,245,075ij 6,388,032 18,883,712 12,495,680 13 H H He Po 207,048,047 203,564,846 3,483,201k �289,520 8,322,544 8,612,064 14 H H He Sp 206,534,228 184,692,386 21,841,841l 17,087,696 26,420,816 9,333,120 15 H H As Po 196,878,298 201,867,956 �4,989,658k �9,576,240 889,568 10,465,808 16 H H As Sp 196,912,045 186,206,335 10,705,710l 3,631,280 15,613,216 11,981,936

Notes: Cell means with the same subscripts (i.e., all cells with subscript a, all with subscript b, etc.) were used for hypotheses testing and are signifi- cantly different from each other, when accounting for Type 1 Error with a Tukey’s W adjustment, at the p < .05 level. SR, supply risk; DR, demand risk; SS, supply strategy; DS, demand strategy; L, low; H, high; He, hedging; As, assuming; Po, postponement; Sp, spec- ulation.

248 I. Manuj et al.

strategy combinations. As is evident from both the table and the figure, hedging-speculation also shows the narrowest range of outcomes, even narrower than other combinations that show lower profit.

H8 is partially supported (see Table 5 and Figure 4d). The results suggest that under conditions of high-supply and high- demand risk, hedging-speculation outperforms all other approach combinations, instead of the hypothesized hedging-postponement. This finding is similar to findings for H3 and H5 discussed ear- lier, in that the presence of high-demand variability does not nec- essarily justify the use of a postponement.

DISCUSSION

As a final step, Study 2 results were shared with four new man- agers (with extensive experience in the areas of global procure- ment, industrial engineering, and simulation modeling) for additional qualitative insights. While some of the modeling results are expected, others contradict standard theory and prac- tice. Hence, this post hoc phase provided contextualization of the modeling results and the clarified the importance of the research

implications. Overall, the findings of this research challenge existing knowledge of the interrelationships and trade-offs preva- lent in choosing supply chain risk management approaches that most effectively impact net profit.

One major implication is that the financial impact of approaches chosen to manage or mitigate supply-side risk, and those chosen to manage or mitigate demand-side risk, are not mutually exclusive. Specifically, a supply-side risk approach that is typically chosen to maximally impact supply-side performance may not be the most effective choice for impacting overall net profit, and vice versa. This concept also resonated with the managers consulted after the final analysis, as they stressed the importance of profit margins as key to managing risks because healthy margins can help absorb systemic and combined up- and down stream risk management trade-offs. For example, model results indicate that supply chains facing low-supply risk do not always benefit from adopting assum- ing approaches, as prevailing wisdom would suggest. Rather, the overall system effectiveness of assuming supply-side risk depends upon the demand strategy; the model shows that when speculation is used on the demand side, assuming on the supply side is a better combination, but when postponement is used on the demand side, hedging is a better choice on the supply side.

Figure 3: Total profit under different levels of supply and demand risk.

Supply Chain Risk Management Approaches 249

Conceptually, this supports the contention that organizations must seek to integrate demand and supply to fully impact overall firm profitability; choices made to optimize functional perfor- mance often suboptimize the overall system (Esper et al. 2010; Stank et al. 2012). This result highlights the continued need to drive integration of strategic decision making across procure- ment, manufacturing, and logistics. Further, for overall system design decisions, our findings support the use of more encom- passing metrics that track both revenue and total costs of owner- ship across the supply chain rather than functional metrics that may mask system suboptimization. In fact, the managers con- sulted as part of the post hoc assessment of results agree, sug- gesting that when broader metrics such as profit margins were used to assess performance, they were “more inclined to forego a sale and lose revenue on it rather than incur high assembly costs and drive down already thin margins.”

Beyond this, managers also elaborated on the metric issue with regard to the political aspects of supply chain risk manage- ment. It was suggested that even with appropriate metrics and analysis, “a part of what actually takes place is what the leader- ship wants.” In doing so, the analytics and consequent recom- mendations provided by these managers may be “overlooked.” For example, although analysis may suggest the appropriateness of an assuming strategy on the supply side, managers indicate that hedging is almost always “preferred because of the flexibil- ity it affords.” Hence, risk management is often an even more complex notion that involves internal firm dynamics and upper management preference.

Another interesting finding relates to the lack of support for the notion that postponement approaches are most effective when

demand variability is high; in fact, the model results indicate that speculation yields greater net profit under both high- and low- demand risk environments. This finding refutes current thinking on postponement and indicates a need for further research that investigates the sensitivities in cost-benefit trade-offs of post- ponement (Yang et al. 2004) and further development of the conditions necessary to effectively leverage postponement when faced with uncertainty (Boone et al. 2007). Managers in the post hoc qualitative group agreed that the viability of postponement strategies was a concern for their firms. They mentioned that they are “moving from sourcing parts to sourcing subassemblies” to reduce assembly costs in the United States and to reduce cus- tomer lead times.

Postponement approaches were supported under certain condi- tions however. Notably, postponement is appropriate in condi- tions of high-demand risk when accompanied by low supply-side risk. To further explore the impact of postponement approaches, post hoc sensitivity analyses were conducted by increasing the coefficient of variation (CV) of order processing variability of the Chinese supplier, increasing the CV of demand variability, increasing capacity, lowering assembly costs, and increasing lead times to customers. The key results from these sensitivity analy- ses show that:

1. Supply risk conditions have a significant impact on the net profit performance implications of the postponement strategic approach even when demand risk conditions suggest that postponement is an appropriate strategy.

2. When demand variability is increased, the gap between the profitability of the postponement and speculation scenarios is

Table 5: Summary of results

Hypothesis Brief statement of Hypothesis Hypothesis in terms of cell #* Results in terms of cell #* Conclusion

H1 Hedging performs better under High-Supply Risks

9>11; 10>12; 13>15; 14>16 9>11; 10>12; 13>15; 14>16 Supported

H2 Assuming performs better under Low-Supply Risks

3>1; 4>2; 7>5, 8>6 8>6; 4>2; 5>7;1>3 Partially supported

H3 Postponement performs better under High-Demand Risks

5>6; 7>8; 13>14; 15>16 5<6; 7<8; 13<14; 15<16 Not supported

H4 Speculation performs better under Low-Demand Risks

2>1; 4>3; 10>9; 12>11 2>1; 4>3; 10>9; 12>11 Supported

H5 Assuming-Speculation performs best under low supply-low demand risks

4>1, 2, and 3 4>1, 2, and 3 Supported

H6 Assuming-Postponement performs best under low supply-high demand risks

7>5, 6, and 8 7<5, 6, and 8 Not supported

H7 Hedging-Speculation performs best under high supply-low demand risks

10>9, 11, and 12 10>9, 11, and 12 Supported

H8 Hedging-Postponement performs best under high supply-high demand risks

13>14, 15, and 16 13>15; 13<14; 13<16 Partially supported

Notes: Supported hypothesis test results (at p < .05 level of significance) in bold, based on Tukey’s W procedure for mean comparison. *Please see Table 4 for cell numbers.

250 I. Manuj et al.

significantly narrowed. A threshold level of demand variabil- ity may be required to justify postponement; this level is much higher than what was identified by the interview man- agers in Study 1.

3. High-demand variability alone may not justify the investment in postponement; however, when combined with lower assem- bly costs, and/or longer lead times to customers, it becomes a useful strategic approach. The well-known example of Dell computer assembly postponement from the mid-1990s to the mid-2000s comes immediately to mind.

CONTRIBUTIONS AND FUTURE RESEARCH DIRECTIONS

The research makes important contributions to theoretical under- standing of supply chain risk management approaches and sug- gests insights for practical applications that may lead to improved profitability. Although the appropriateness of using some supply chain risk management approaches are supported, the research counters prevailing knowledge regarding the appro- priate conditions for choosing postponement and speculation (see e.g., Fisher 1997; Pagh and Cooper 1998). Significantly, it high- lights the dangers of functionally isolated decision making regarding issues that have broad system-wide impact on overall

performance, lending great credence to the increasing calls for interdisciplinary research to address problems that cross demand and supply as well as those that cross functions within the sup- ply chain. Finally, the research supports the need to utilize cross- functional performance metrics such as net profit as dependent variables to assess the efficacy and effectiveness of supply chain decisions.

While the research results cannot be generalized beyond the scope of the computer model, the conclusions outlined above suggest three themes that could have widespread implications for supply chain management, including:

1. The need for managers to ensure that decisions that impact broad system-wide performance, such as supply network deci- sions and manufacturing strategies are made with an eye toward impact on broad organizational performance; supplier- oriented risk management approaches must consider the impact of demand-oriented approaches, and vice versa.

2. The knowledge that postponement can be an effective approach, but only after fully understanding: (1) the total cost and system implications related the importance of a stable and reliable supply base, (2) the level of demand variability, and (3) the cost of finished goods inventory. The threshold level of variability of demand that makes postponement an effective approach may be much higher than what is commonly assumed.

-20 0 20 40 60 80

F re q u en cy

Millions

-20 0 20 40 60 80

F re q u en cy

Millions

-20 0 20 40 60 80

F re q u en cy

Millions

-20 0 20 40 60 80

F re q u en cy

Millions

(soiranecS sksiR dnameD woL-ylppuS woL)a( b) Low Supply-High Demand Risks Scenarios

(soiranecS sksiR dnameD woL-ylppuS hgiH )c( d) High Supply-High Demand Risks Scenarios

Hedging-Postponement Hedging-Speculation Assuming-Postponement Assuming-Speculation Note: Graphs represent the distribution of values for all 28 runs.

Figure 4: Distribution of total profit under different environmental conditions.

Supply Chain Risk Management Approaches 251

3. The importance of using a “market basket” of key perfor- mance indicators to assess the impact of system and operating decisions; functional indicators are important for assessing ex- ecutional success, but can lead to system suboptimization. Overall system metrics such as net profit should drive big sys- tem-level decisions.

There are limitations to the research. Computer simulation enables precision in control/measurement/manipulation of vari- ables, and partially addresses the research goal of existential real- ism or realism of context, but does not provide an ability to generalize to a population of interest (Bienstock 1994). Future research can focus on complementary methods such as surveys to increase the generalizability of the findings. Another limitation

is the finite number of variables and parameters that could be included. The constraints of approach and the resulting number of variables included in the current research opens the door for multiple future explorations that include variable not endogenous to this model, for example, other supply chain risk management or mitigation approaches, different operationalization of concepts such as financial viability of suppliers, intellectual property risk, and infrastructural issues, greater complexity in product lines and network choices, etc. Moreover, it is very likely that firms may engage in the simultaneous application of several strategic approaches investigated in this research, instead of the more sim- plistic scenarios simulated here whereby only one strategic approach was allowed.

APPENDIX 1

Profile of participants

Pseudonyms* Peter/HomeCo Home appliances and consumer goods. Manufacturer and Distributor.

Recently retired after 32 years as Vice President of Supply Chain Strategy; responsible for transportation, warehousing, manufacturing engineering, forecasting, and corporate planning

Rob/HomeCo 10 years of experience, 5 years with the current firm; Vice President of Technology; responsible for product development and brand management

Kevin/HomeCo 34 years of experience, 32 years with the current firm; Director of Global Sourcing Strategy; responsible for procurement, supply chain, and manufacturing strategy

Jake/HomeCo 20 years of experience, 6 years with the current firm; Assistant Treasurer of Risk Management; responsible for insurance procurement and corporate risk management

Ben/HomeCo 22 years of experience, 6 years with the current firm; Global Director for controls; responsible for controls and wiring within appliances

Luke†/HomeCo 4 years of experience with current firm; Sourcing Senior Manager; responsible for global procurement Jeff†/HomeCo 35 years of experience with current firm; Group Vice President, Operations and Supply Chain Pat/HomeCo 34 years of experience with current firm; Senior Manager, Import/Export; responsible for managing import

and exports including customs, third-party logistics providers, and brokers; and mitigating import risk Tim/HealthCo Consumer health products, pharmaceuticals, biologics, and health devices and diagnostics. Manufacturer

and Distributor. 30 years of experience, 27 years with the current firm; Director of Worldwide Customer and Distribution Services; responsible for global fulfillment and returns processing

Tyler/ElectriCo Electronic and electrical components. Manufacturer and Distributor. 20 years of experience, all with the current firm; Manager, Strategic Supply Management; responsible for strategic sourcing and planning

Bill/ComputerWare Information technology and services. Manufacturer and Distributor. 20 years of experience, 9 years with the current firm; Director of Fulfillment; responsible for contract management, cost containment and process improvement, inventory optimization, risk management, and change management

Tony/MachineCo Machinery. Manufacturer and Dealer. 27 years of experience, 10 years with the current firm; Director of Global Supply Chain Strategy; responsible for client account management, supply chain strategy for the entire organization

Matt/Builders Inc Building materials. Manufacturer and Distributor. 7 years of experience, 5 years with the current firm; Director of Supply Chain; responsible for procurement, distribution, logistics, and operations planning

James/SuppliesCo Office supplies and equipment. Retailer. 22 years of experience, 3 years with the current firm; Executive Vice President of Supply Chain; responsible for inventory management, transportation, warehousing, real estate strategy and, store development

Continued.

252 I. Manuj et al.

Charles/CompuSystems Computer hardware. Manufacturer and Distributor. 27 years of experience, 4 years with the current firm; Vice President of Supply Chain Integration; responsible for global distribution

Ted/CareCo Personal and health care products. Manufacturer and Distributor. 10 years of experience; Director of Customer Supply Chain Strategy; responsible for reducing inventory and costs and improving availability

Multiple participants Third-party logistics. Logistics Service Provider. Consulted only for data collection for simulation model

Notes: *Company and participant names have been changed to ensure the anonymity of the participants. †Participated only in focus group discussion.

APPENDIX 2

ESTABLISHING VALUES FOR KEY MODEL VARIABLES AND MODEL VERIFICATION AND VALIDATION RESULTS

The values for reorder point (ROP) and replenishment quantity are calculated using the economic-order quantity formulae (Coyle et al. 2003; Williams and Tokar 2008) that incorporate variability in demand and lead time and the expected cost of stock-outs. The calculated value of Q is rounded to the nearest integer that is a multiple of a 40-foot container-load quantity. For the assuming scenario, all orders are assigned to the single Chinese supplier. For the hedging scenario, each replenishment order has an equal probability (0.5) of being assigned to either supplier.

For the assuming scenarios, calculation of ROP and Q values is based on the Chinese supplier. For the hedging scenarios, Q is based on both suppliers; ROP is based on the Chinese supplier. An in-stock availability rate of 84% (standard deviation = 1) is used for assuming and 99.87% (standard deviation = 3) for postpone- ment to provide the postponement scenario with the flexibility to react to high-demand variability. Inventory value of products and components is assessed at average purchase cost. Inventory carry- ing cost is set to 21%, the average cost of carrying inventory over the past six years (Wilson 2007–2012).

The overall manifestation of quality problems for managers consulted for setting up the simulation model was in the form of “usable items received” or “yield rate.” Every item was

given a specific rate of being defective based upon values estab- lished in the qualitative interviews. Order processing time and variability at the suppler is modeled using the First-In, First-Out priority. The supplier has no capacity constraints, there are no backorders, and every order is filled complete. For cost, the product purchase price from the Chinese supplier is set to $60/ unit. Typically, the purchase cost of electronic products and components is around 20%–30% cheaper in China (Engardio et al. 2004). The variability of purchase cost of products and components from the Chinese supplier is set to a high of 15% for low-risk scenarios and a high of 45% for high-risk scenar- ios. The low value of 15% was arrived at by extrapolating the Chinese wage rate increase over the past six years due to the gradual strengthening of Chinese currency. The high value of 45% was based on trends in price increases of raw materials and components that go onto electronic products, cost of labor shortages, and oil price increases.

Transportation of product to the M/D is the final step of the model. From the Chinese supplier, the goods are sent to the Hong Kong port using domestic transportation where they are loaded onto a ship. The ship travels from the Hong Kong port to the U.S. Los Angeles port. At the port, goods are cleared through customs and loaded onto trucks, which transport the goods to the M/D. At the domestic supplier, the goods are shipped to the M/D in full truckloads. Key values used to operationalize model variables are provided in Table A2a. The total basic cost for each strategy is summarized in Tables A2b and A2c. The model was verified and validated using procedures established by Law and Kelton (1982) and Sargent (2000). These procedures and results are included as Table A2d.

Table A2a: Values for important model variables

Cost/Value Time Policy/Remarks Data Source/Rationale/

Justification

1. Order processing costs and constraints

1a. Speculation Cost to pick and pack is $10/unit

Capacity is 130% of daily capacity, i.e., 1,300 units per day maximum; 1 shift/7 days a week/365 days a year

Data from a major third-party logistic (3PL) (TransportCo)

Continued.

APPENDIX 1 Profile of participants

Supply Chain Risk Management Approaches 253

Table A2a: (Continued)

Cost/Value Time Policy/Remarks Data Source/Rationale/

Justification

1b. Postponement Cost to pick, assemble, and pack is $20/unit

Capacity is 130% of daily capacity, i.e., 1,300 units per day maximum; 1 shift/7 days a week/365 days a year

Data from a major 3PL (TransportCo)

2. Transportation Cost and Time

2a. Chinese supplier ship complete order to HK Port

0 1 day Transportation cost included in per container charge from China port to U.S. port

Data from interviews

At Hong Kong Port 0 T(4,5,6) Port costs included in per container charge from China port to U.S. port

Data from interviews

HK Port to LA Port $3,000 per container T(13, 15, 20) $3,000/container includes the cost from China supplier through LA port including all taxes, charges, and other duties

Data from interviews Published ship sailing schedules of major ocean carriers

At LA Port 0 T(3, 4, 5) Port costs cost included in per container charge from China port to U.S. port

Data from interviews

From LA Port to Manufacturer/Distributor

$3,000 per TL T(4, 5, 6) Cost quote from trucking company; times validated in interviews

2b. U.S. supplier from Supplier to Manufacturer/Distributor

$3,000 per TL T(4, 5, 6) Cost quote from trucking company; times validated in interviews

3. Shipment to customers 3a. Cost $10/unit LTL transportation Cost quote from a leading

3PL company’s website 3b. Late orders $35/unit $35 is penalty cost for each unit

delivered late to the customer Penalty cost validated in interviews (SuppliesCo)

3c. Transit time 3 days Data from interviews 4. Selling price $150/ unit Calculated based on

secondary data on gross margins for a major printer manufacturer

Notes: LTL, less-than-truckload; T, triangular.

Table A2b: Basic cost for each strategy: cost sheet for finished products

Cost/unit Product A or B from China Source of data

Product A or B from the United States Source of data

Purchase Price 60.00 Product cost from a major 3PL 80.00 Imported products are approximately 25% cheaper

China to United States by Ship

2.50 Cost per container = $3,000 Number of printers in a container = 1,200

Inbound Receiving at Port*

0.00 Included in end-to-end shipping cost above NA

Continued.

254 I. Manuj et al.

Table A2b: (Continued)

Cost/unit Product A or B from China Source of data

Product A or B from the United States Source of data

Port/U.S. supplier to M/D

2.50 Cost per TL = $3,000 Number of printers in a container = 1,200

Inbound Receiving at M/D, processing, picking and packing

10.00 Cost from a leading 3PL 10.00 Cost from a leading 3PL

Ship 10.00 Cost from a leading 3PL 10.00 Cost from a leading 3PL Total 85.00 102.50 Average cost per unit in Hedging strategy

93.75

Average cost per unit in Assuming strategy

85.00

Notes: 3PL, third-party logistic; M/D, manufacturing firm with both assembly and distribution operations; TL, truck load.

Table A2c: Basic cost for each strategy: cost sheet for components assembled at M/D

Cost/unit

Components imported and assembled in the United States Source of data

Components bought in the United States and assembled Source of data

Purchase Price of A or B Component

15.00 Cost from a major 3PL 20.00 Imported products are approximately 25% cheaper

Purchase Price of C Component

35.00 50.00 Imported products are approximately 25% cheaper

China to United States by Ship (A or B)

0.61 Cost per container = $3,000 Number of components in a container = 4,880

China to United States by Ship (C)

2.26 Cost per container = $3,000 Number of components in a container = 1,330

Inbound Receiving at Port* Included in end-to-end shipping cost above

Port/U.S. supplier to M/D Component A or B

0.61 Cost per container = $3,000 Number of A or B components in a container = 4,880

Port/U.S. supplier to M/D Component C

2.26 Cost per container = $3,000 Number of C components in a container = 1,330

Inbound Receiving at M/D, processing, picking and packing

10.00 Cost from a leading 3PL Cost from a leading 3PL

Assembly 20.00 Cost from a leading 3PL 20.00 Cost from a leading 3PL Ship 10.00 Cost from a leading 3PL 10.00 Cost from a leading 3PL Total 94.73 102.87 Average cost per unit in Hedging strategy

98.31

Average cost per unit in Assuming strategy

94.73

Notes: 3PL, third-party logistic; M/D, manufacturing firm with both assembly and distribution operations.

Supply Chain Risk Management Approaches 255

Table A2d: Verification and validation of simulation model

Model element Relevance Ways to address the element

Model verification

Ensures that the computerized model and its implementation are correct (Sargent 2000)

Programming experts in supply chains modeling were consulted to verify the basic model structure. The uniformity and variability of the model’s random number generators (e.g., transportation times) were validated by comparing outputs with manually calculated solutions. Simulation results for short pilot runs of simple cases for the complete model were compared with manual calculations. The model was built in stages where each submodel was verified by replacing stochastic elements with deterministic elements and gradually integrating these sub- models into the main model.

Model validation

Ensures that that the computerized model possesses a satisfactory range of accuracy consistent with the intended application of the model (Sargent 2000)

Subject matter experts (SMEs), including several academic scholars and practitioners, were regularly consulted in the conceptual development of model. A structured walk-through of the model and a review of the simulation results for reasonableness with a separate set of SMEs. Sensitivity analyses were performed to see which model factors had the greatest impact on the performance measures.

Sample size determination

Ensures the balance between absolute precision and external validity

The relative-precision procedure (Law and Kelton 1982; Bienstock 1996) was used for sample size determination. Procedure resulted in 28 runs for each of the 16 scenarios at 5% relative-precision level. Run length was set to two years (730 days) as it is the typical contract duration.

Model stability Minimizes the effect of initial conditions on the model

Warm up period was set to 60 days as by then scenarios stabilized in terms of direction profit, stability in penalty costs of late deliveries, and stable order fill rates. Initial inventory at the M/D was set to the ROP level.

Notes: M/D, manufacturing firm with bothassembly and distribution operations; ROP, reorder point.

REFERENCES

Bailey, J.P., and Rabinovich, E. 2005. “Internet Book Retailing and Supply Chain Management: An Analytical Study of Inventory Location Speculation and Postponement.” Transportation Research: Part E 41(3):159–77.

Bartmess, A., and Cerny, K. 1993. “Building Competitive Advantage Through a Global Network of Capabilities.” California Management Review 35(2):78–103.

Berger, P.D., Gerstenfeld, A., and Zeng, A.Z. 2004. “How Many Suppliers Are Best? A Decision-Analysis Approach.” Omega (Oxford) 32(1):9–15.

Bienstock, C.C. 1994. The Effect of Outsourcing and Situational Characteristics on Physical Distribution Transportation Efficiency. Dissertation at Virginia Polytechnic Institute and State University, Blacksburg, VA.

Bienstock, C.C. 1996. “Sample Size Determination in Logistics Simulations.” International Journal of Physical Distribution and Logistics Management 26(2):43–50.

Boone, C.A., Craighead, C.W., and Hanna, J.B. 2007. “Postponement: An Evolving Supply Chain Concept.” International Journal of Physical Distribution and Logistics Management 37(8):594–611.

Bucklin, L.P. 1965. “Postponement, Speculation and the Structure of Distribution Channels.” Journal of Marketing Research 2(1):26–31.

Carter, J.R., and Vickery, S.K. 1989. “Currency Exchange Rates: Their Impact on Global Sourcing.” Journal of Purchasing and Materials Management 25(3):19–25.

Chiou, J.-S., Wu, L.-Y., and Hsu, J.C. 2002. “The Adoption of Form Postponement Strategy in a Global Logistics System: The Case of Taiwanese Information Technology Industry.” Journal of Business Logistics 23(1):107–24.

Chung, W., Talluri, S., and Narasimhan, R. 2010. “Flexibility or Cost Saving? Sourcing Decisions With Two Suppliers.” Decision Sciences 41(3):623–50.

Coyle, J.J., Bardi, L.B., and Langley, C.J., Jr. 2003. The Management of Business Logistics. 7th ed. Ohio: South- Western.

Deloitte Development LLC. 2013. “The Ripple Effect: How Manufacturing and Retail Executives View the Growing Challenge of Supply Chain Risk.” http://deloitte.wsj.com/cfo/ files/2013/02/the_ripple_effect_supply_chain.pdf

Drazin, R., and Van de Ven, A.H. 1985. “Alternative Forms of Fit in Contingency Theory.” Administrative Science Quarterly 30(4):514–39.

Engardio, P., Roberts, D., and Bremner, B. 2004. The China Price. Business Week. http://www.businessweek.com/magazine/content/ 05_22/b3935151.htm?chan=search

Esper, T., Ellinger, A., Stank, T.P., Flint, D., and Moon, M. 2010. “Demand and Supply Integration: A Conceptual Framework of Value Creation Through Knowledge

256 I. Manuj et al.

Management.” Journal of Academy of Marketing Science 38(1):5–18.

Fawcett, S.E., Fawcett, A.M., Watson, B.J., and Magnan, G.M. 2012. “Peeking Inside the Black Box–Toward an Understanding of Supply Chain Dynamics.” Journal of Supply Chain Management 48(1):44–72.

Fawcett, S.E., and Waller, M.A. 2014. “Supply Chain Game Changers–Mega, Nano, and Virtual Trends–And Forces That Impede Supply Chain Design (i.e., Building a Winning Team).” Journal of Business Logistics doi: 10.1111/jbl.12058

Fawcett, S.E., Waller, M.A., and Fawcett, A.M. 2010. “Elaborating a Dynamic Systems Theory to Understand Collaborative Inventory Successes and Failures.” International Journal of Logistics Management 21(3):510–27.

Feitzinger, E., and Lee, H.L. 1997. “Mass Customization at Hewlett-Packard: The Power of Postponement.” Harvard Business Review 75(1):116–21.

Fisher, M.L. 1997. “What is the Right Supply Chain for Your Product?” Harvard Business Review 75(2):105–16.

Glaser, B.G., and Strauss, A.L. 1967. The Discovery of Grounded Theory: Strategies for Qualitative Research. New York: Aldine De Gruyter.

Hesse, R. 2000. “Triangle Distribution: Mathematica Link for Excel.” Decision Line 31(3):12–14.

Hult, G.T.M., Craighead, C.W., and Ketchen, D.J., Jr. 2010. “Risk Uncertainty and Supply Chain Decisions: A Real Options Perspective.” Decision Sciences 41(3):435–58.

Inderst, R. 2008. “Single Sourcing Versus Multiple Sourcing.” RAND Journal of Economics (Blackwell Publishing Limited) 39(1):199–213.

J€uttner, U. 2005. “Supply Chain Risk Management.” International Journal of Logistics Management 16(1):120–41.

J€uttner, U., Peck, H., and Christopher, M. 2003. “Supply Chain Risk Management: Outlining an Agenda for Future Research.” International Journal of Logistics: Research and Applications 6(4):197–210.

Law, A.M., and Kelton, W.D. 1982. Simulation Modeling and Analysis. 4th ed. New York: McGraw-Hill, Inc.

LeBlanc, L.J., Hill, J.A., Harder, J., and Greenwell, G.W. 2009. “Modeling Uncertain Forecast Accuracy in Supply Chains With Postponement.” Journal of Business Logistics 30(1):19–31.

Lee, H.L. 2002. “Aligning Supply Chain Strategies With Product Uncertainties.” California Management Review 44(3):105–19.

Lee, H.L., Billington, C., and Carter, B. 1993. “Hewlett-Packard Gains Control of Inventory and Service Through Design for Localization.” Interfaces 23(4):1–11.

Manuj, I., and Mentzer, J.T. 2008a. “Global Supply Chain Risk Management.” Journal of Business Logistics 29(1):133–55.

Manuj, I., and Mentzer, J.T. 2008b. “Global Supply Chain Risk Management Strategies.” International Journal of Physical Distribution and Logistics Management 38(3):192–223.

Manuj, I., Mentzer, J.T., and Bowers, M.R. 2009. “Improving the Rigor of Discrete-Event Simulation in Logistics and Supply Chain Research.” International Journal of Physical Distribution and Logistics Management 39(3):172–201.

Manuj, I., Omar, A., and Pohlen, T. 2014. “Inter-Organizational Learning in Supply Chains: A Focus on Logistics Service Providers and Their Customers.” Journal of Business Logistics. 35(2):103–20.

Mello, J., and Flint, D.J. 2009. “A Refined View of Grounded Theory and Its Application to Logistics Research.” Journal of Business Logistics 30(1):107–25.

Mentzer, J.T., and Gomes, R. 1991. “The Strategic Planning Model: A PC-Based Dynamic, Stochastic, Simulation DSS Generator for Managerial Planning.” Journal of Business Logistics 12(2):193–219.

Miller, K.D. 1992. “A Framework for Integrated Risk Management in International Business.” Journal of International Business Studies 23(2):311–31.

Needham, P.M., and Evers, P.T. 1998. “The Influence of Individual Cost Factors on the Use of Emergency Transshipments.” Transportation Research: Part E 34 (2):149–60.

Norrman, A., and Jansson, U. 2004. “Ericsson’s Proactive Supply Chain Risk Management Approach After a Serious Sub-Supplier Accident.” International Journal of Physical Distribution and Logistics Management 34(5):434–56.

Ott, R.L., and Longnecker, M. 2001. An Introduction to Statistical Methods and Data Analysis. Pacific Grove, CA: Duxbury.

Pagh, J.D., and Cooper, M.C. 1998. “Supply Chain Postponement and Speculation Strategies: How to Choose the Right Strategy.” Journal of Business Logistics 19(2):13–33.

Perry, J.H. 1991. “Emerging Economic and Technological Futures: Implications for Design and Management of Logistics Systems in the 1990s.” Journal of Business Logistics 12(2):1–16.

Pfeffer, J., and Salancik, G.R. 1978. The External Control of Organizations: A Resource Dependence Perspective. New York: Harper and Row.

Ramsey, P.H., Ramsey, P.P., and Barrera, K. 2010. “Choosing the Best Pairwise Comparisons of Means From Non-Normal Populations, With Unequal Variances, but Equal Sample Sizes.” Journal of Statistical Computation and Simulation 80 (6):595–608.

Randall, W.S., Pohlen, T.L., and Hanna, J.B. 2010. “Evolving a Theory of Performance-Based Logistics Using Insights From Service Dominant Logic.” Journal of Business Logistics 31 (2):35–61.

Rosenfield, D.B., Copacino, W.C., and Payne, E.C. 1985. “Logistics Planning and Evaluation When Using ‘What-If’ Simulation.” Journal of Business Logistics 6(2):89–119.

Sargent, R.G. 2000. Verification, Validation, and Accreditation of Simulation Models. Proceedings of the Winter Simulation Conference. San Diego, CA: Society for Computer Simulation International.

Schoenherr, T. 2009. “Logistics and Supply Chain Management Applications Within a Global Context: An Overview.” Journal of Business Logistics 30(2):1–25.

Scott, W.R., and Davis, G. 2006. Organizations and Organizing: Rational, Natural, and Open Systems Perspectives. Englewood Cliffs, NJ: Prentice-Hall.

Shao, X.-F., and Ji, J.-H. 2009. “Effects of Sourcing Structure on Performance in a Multiple-Product Assemble-to-Order Supply Chain.” European Journal of Operational Research 192(3):981–1000.

Stank, T.P., Esper, T., Autry, C., and Crook, R. 2012. “Creating Relevant Value Through Demand and Supply Integration.” Journal of Business Logistics 33(2):167–72.

Supply Chain Risk Management Approaches 257

Strauss, A., and Corbin, J. 1998. Basics of Qualitative Research: Grounded Theory Procedures and Techniques, 2nd ed. Thousand Oaks, CA: Sage Publications Inc.

Van der Zee, D.J., and Van der Vorst, J.G.A.J. 2005. “A Modeling Framework for Supply Chain Simulation: Opportunities for Improved Decision Making.” Decision Sciences 36(1):65–95.

Wagner, S.M., and Bode, C. 2008. “An Empirical Examination of Supply Chain Performance Along Several Dimensions of Risk.” Journal of Business Logistics 29(1):307–25.

Wernerfelt, B., and Karnani, A. 1987. “Competitive Strategy Under Uncertainty.” Strategic Management Journal 8(2):187– 94.

Williams, B.D., and Tokar, T. 2008. “A Review of Inventory Management Research in Major Logistics Journals Themes and Future Directions.” International Journal of Logistics Management 19(2):212–32.

Wilson, R. 2007–2012. 18th through 23rd Annual State of Logistics Reports. Washington, DC: Council of Supply Chain Management Professionals.

Wong, H., Wikner, J., and Naim, M. 2009. “Analysis of Form Postponement Based on Optimal Positioning of the Differentiation Point and Stocking Decisions.” International Journal of Production Research 47(5):1201–24.

Yang, B., Burns, N.D., and Backhouse, C.J. 2004. “Postponement: A Review and an Integrated Framework.” International Journal of Operations and Production Management 24(5):468–87.

Zsidisin, G.A. 2003. “A Grounded Definition of Supply Risk.” Journal of Purchasing and Supply Management 9(5/6):217–24.

SHORT BIOGRAPHIES

Ila Manuj (PhD University of Tennessee) is an Associate Pro- fessor of Logistics at the Department of Marketing and Logistics

at the University of North Texas. Dr. Manuj’s areas of interest are risk and complexity management and interorganizational learning in global supply chains. She has published in the Jour- nal of Business Logistics, The International Journal of Logistics Management, International Journal of Physical Distribution and Logistics Management, and Transportation Journal, and authored book chapters.

Terry L. Esper (PhD University of Arkansas) is the Oren Harris Chair in Logistics and an Associate Professor of Supply Chain Management in the Sam M. Walton College of Business at the University of Arkansas. His primary research interests include supply chain management strategy, behavioral supply chain dynamics, supply chain orientation, retail supply chain marketing, and supply chain integration. Dr. Esper’s research has been published in the Journal of Business Logistics, Journal of Retailing, Journal of the Academy of Marketing Science, the International Journal of Physical Distribution and Logistics Management, and Supply Chain Management: An International Journal, among other outlets. He is also co-author of the book The Definitive Guide to Inventory Management.

Theodore P. Stank (PhD University of Georgia) is the Harry and Vivienne Bruce Chair of Business Excellence and Professor of Logistics and SCM in the College of Business Administration at the University of Tennessee at Knoxville. His research focuses on the strategic implications and performance benefits associated with logistics and supply chain management best practices. He is author of over 100 articles in academic and professional journals including Journal of Business Logistics, Journal of Operations Management, Management Science, Supply Chain Management Review, and Journal of the Academy of Marketing Science. He is also co-author of the books Global Supply Chains: Evaluating Regions on an EPIC Framework, 21st Century Logistics: Making Supply Chain Integration a Reality, and co-editor of Handbook of Global Supply Chain Management.

258 I. Manuj et al.

Copyright of Journal of Business Logistics is the property of Wiley-Blackwell and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.