Professional Documents
Culture Documents
On The Benefits of CPFR and VMI A Comparative PDF
On The Benefits of CPFR and VMI A Comparative PDF
Abstract
This paper aims to help managers of a supply chain to determine an appropriate level of collaboration according to their
specific business conditions. For this purpose, a comprehensive simulation model representing two popular supply chain
initiatives, that are collaborative planning, forecasting and replenishment (CPFR) and vendor-managed inventory (VMI),
is constructed. In addition, a traditionally managed supply chain (TSS) is also included in the model as a benchmark. The
results indicate that benefits of CPFR are always higher than VMI. However, we also realize that under certain conditions,
the gap between the performances of CPFR and VMI does not rationalize the additional resources required for CPFR.
Especially, when the lead time is short and/or when available manufacturing capacity is tight, a careful consideration has
to be given on the selection of an appropriate collaboration mode.
r 2008 Elsevier B.V. All rights reserved.
0925-5273/$ - see front matter r 2008 Elsevier B.V. All rights reserved.
doi:10.1016/j.ijpe.2007.10.021
ARTICLE IN PRESS
576 K. Sari / Int. J. Production Economics 113 (2008) 575–586
a larger variance than sales to the buyer (Lee et al., VMI offers a competitive advantage for retailers
1997a, b). In return, high inventory levels and poor because it results in higher product availability and
customer service rates are typical symptoms of the service level as well as lower inventory monitoring
bullwhip effect (Metters, 1997; Chopra and Meindl, and ordering cost (Waller et al., 1999; Achabal et al.,
2001, p. 1363). Today, SCM researchers indicate 2000). For vendors, on the other hand, it results in
that elimination the bullwhip effect plays a vital role reduced bullwhip effect (Lee et al., 1997b; Disney
for supply chain enterprises to gain competitive and Towill, 2003a, b) and better utilization of
advantage. manufacturing capacity (Waller et al., 1999), as well
Most of the researchers focusing on remedies for as better synchronization of replenishment planning
coping with the bullwhip effect dictate that sharing (Waller et al., 1999; C- etinkaya and Lee, 2000).
retail-level information (i.e. point of sales (pos) While many benefits have been identified in the
data) between supply chain members is a prerequi- literature, there are also a number of challenges that
site for elimination of the bullwhip effect, see e.g. may exist in practice and that can potentially reduce
Lee et al. (1997a), Chen et al. (2000a, b), McCullen the benefits obtained from VMI or lead to failures in
and Towill (2002), Dejonckheere et al. (2004), VMI programs. For instance, Spartan Stores, a
Ouyang (2006) and Li et al. (2006). Nevertheless, grocery chain, shut down its VMI effort about 1
retailers, most of the time, do not desire to engage in year after due in part VMI vendors’ inability to
information sharing because it provides ignorable deal with product promotions (Simchi-Levi et al.,
levels of benefits for them, see e.g. Lee et al. (2000), 2003, p. 161). Similarly, Kmart cut a substantial
Yu et al. (2001, 2002), Zhao et al. (2002a, b). amount of VMI contracts because Kmart is not
Therefore, this requires upstream members (e.g. satisfied with the forecasting ability of VMI vendors
suppliers or manufacturers) to offer incentives for (Fiddis, 1997). Consequently, many studies have been
retailers in return for information sharing. Vendor- carried out to investigate the effectiveness of VMI
managed inventory (VMI) and collaborative plan- programs under different conditions. For instance,
ning, forecasting and replenishment (CPFR) are the Kuk (2004) empirically tested the acclaimed benefits
partnership programs primarily developed to en- of VMI programs in electronics industry. Similarly,
courage retailers to share information, see e.g. Lee Sari (2007) used a simulation model to evaluate the
et al. (1997b) and Disney and Towill (2003a, b). benefits of VMI under different market conditions.
VMI, also known as continuous replenishment or Dong and Xu (2002), on the other hand, evaluated
supplier-managed inventory, is one of the most the value of VMI programs both for suppliers and
widely discussed partnering initiatives for encoura- buyers. Most of these studies show that ineffective
ging collaboration and information sharing among usage of retail-level information is one major limita-
trading partners (Angulo et al., 2004). Popularized tion of VMI programs (see e.g. Aviv, 2002; Ovalle
in the late 1980s by Wal-Mart and Procter & and Marquez, 2003; Angulo et al., 2004; Yao et al.,
Gamble (Waller et al., 1999), it was subsequently 2007). That is, since retailers are closer to the
implemented by many other leading companies marketplace, they may have better knowledge about
from different industries, such as Glaxosmithkline customer behaviors, products and marketplace.
(Danese, 2004), Electrolux Italia (De Toni and However, in most, if not all, VMI programs, this
Zamolo, 2005), Nestle and Tesco (Watson, 2005), unique knowledge of the retailers cannot be joined
Boeing and Alcoa (Micheau, 2005), etc. It is a into inventory decisions. This is because in a typical
supply chain initiative where the vendor decides on VMI program, retailers are excluded from demand
the appropriate inventory levels of each of the forecasting process. Indeed, in a VMI system, the
products and the appropriate inventory policies to responsibilities of the retailers are noting more than
maintain those levels. The retailer provides the sharing sales and inventory data.
vendor with access to its real-time inventory level. CPFR, on the other hand, can solve majority of
In this partnership program, the retailer may set the problems that are encountered in adaptation of
certain service level and/or self-space requirements, VMI because it requires all members of a supply
which are then taken into consideration by the chain to jointly develop demand forecasts, produc-
vendor. That is, in a VMI system, the retailer’s role tion and purchasing plans, and inventory replenish-
shifts from managing inventory to simply renting ments (Aviv, 2002). It is a business practice that
retailing space (Simchi-Levi et al., 2003, p. 154; combines the intelligence of multiple trading part-
Mishra and Raghunathan, 2004). ners in the planning and fulfilment of customer
ARTICLE IN PRESS
K. Sari / Int. J. Production Economics 113 (2008) 575–586 577
demand (CPFR Workgroup, 2002). CPFR adds Raghunathan (1999), Aviv (2001, 2002, 2007),
value to the supply chain in the form of reduced Ovalle and Marquez (2003) and Disney et al.
inventory and increased customer service level by (2004) represent most of the developments in this
achieving better match of demand and supply area. Our paper is different from these previous
(Foote and Krishnamurthi, 2001; Aghazadeh, studies in three ways. First, most of the models have
2003; Aichlmayr, 2003; Fliedner, 2003). Nonethe- tended to mainly analytical with some very restric-
less, successful implementation of CPFR is not an tive assumptions (e.g. two-stage supply chain,
easy task. It requires more intensive organizational normally distributed or correlated market demands)
resources than VMI as well as mutual trust of for the sake of mathematical tractability (e.g.
multiple trading partners (Barratt and Oliveria, Raghunathan, 1999; Aviv, 2001, 2002, 2007; Disney
2001; Fliedner, 2003). Furthermore, dramatic et al., 2004). Second, some of the models have been
changes are also required in usual ways of doing developed so far are only concentrated on forecast-
business for CPFR implementation. ing part of CPFR (e.g. Aviv, 2001, 2007). Third, as
Consequently, examination of the previous litera- far as we know, none of the models has explored
ture reveals the fact that adaptation of higher levels CPFR and VMI comparatively in capacitated
of collaboration among members of a supply chain multi-stage supply chains under both stationary
creates greater benefits for the supply chain. On the and non-stationary customer demands (e.g. Ovalle
other hand, we also see that development and and Marquez, 2003) as we have done in this paper.
operational costs of a highly integrated collabora- Therefore, this paper contributes to the current
tion is also higher. That is, while CPFR eliminates literature by extending the results of previous
most of the problems encountered in VMI pro- research studies in a way that managers in a supply
grams; investment and operation costs of CPFR are chain enterprise can determine an appropriate level
substantially higher along with greater implementa- of collaboration for their supply chains.
tion difficulties. Indeed, these difficulties might Unlike many prior analytical studies which have
explain why many of the CPFR programs have very restrictive assumptions for the sake of math-
not moved beyond a limited number of product ematical tractability (e.g. Mishra and Raghunathan,
categories or a small set of trading partners (see e.g. 2004; Lee and Chu, 2005; Yao et al., 2007), we have
Baird, 2003; Program may build CPFR momentum, used a simulation model in this study to investigate
2005). Therefore, this trade-off between benefits and the benefits of CPFR and VMI under more realistic
costs of supply chain collaborations creates an circumstances. The simulation approach has been
urgent need for SCM practitioners to determine the used extensively in the literature for analyzing
right collaboration level for their supply chains. supply chain systems (e.g. Waller et al., 1999; Zhao
Today, many SCM practitioners try to determine et al., 2002a, b; Angulo et al., 2004; Lau et al., 2004;
the appropriate level of collaboration for their Sari, 2007; Zhang and Zhang, 2007). In this study,
supply chains. Here, the following two questions we considered a four-stage supply chain, which
play a critical role in determining the right consists of four echelons: a manufacturing plant, a
collaboration level: warehouse, a distributor and a retailer. The plant
has limited manufacturing capacity and produces a
Does it is required to invest in CPFR if an earlier single product. Each enterprise replenishes its
supply chain initiative such as VMI, had already inventory from its immediate upstream enterprise.
been adopted? In other words, does the gap The remainder of this study is organized as
between the performances of CPFR and VMI follows. Section 2 clarifies the methodology and
compensate the cost of investing in CPFR? development of the simulation model. Setting of
Which factors are influential in answering the experimental design is identified in Section 3,
question described above? Do capacity of the followed by simulation output analysis in Section 4.
manufacturing facility, lead times, or uncertainty Conclusions are presented in Section 5.
in customer demand influence the desire for
CPFR? 2. The simulation model
To the best of our knowledge, there have been At the initial stages of this research, we intended
very few research studies aiming to explore these to use Microsoft Excel in constructing the simula-
questions. That is, a few research studies e.g. tion model; however, research conducted by Keeling
ARTICLE IN PRESS
578 K. Sari / Int. J. Production Economics 113 (2008) 575–586
and Pavur (2004) indicated that it might possible for no information is shared between members, up-
errors to occur in the random numbers generated by stream stages are unaware of actual demand
Microsoft Excel. Therefore, in order to eliminate information at the market place. That is, while
this potential problem, we have used Crystal Ball, creating demand forecasts and inventory plans,
an Excel add-in published by Decisioneering. It is a supply chain members use only replenishment
popular risk analysis and forecasting program that orders placed by their immediate downstream
uses Monte Carlo simulation in a spreadsheet member. Therefore, each member of the supply
environment. chain replenishes his own inventory by following an
According to different situations of information installation-based (R, S) policy. Under an installa-
sharing and ordering information coordination, the tion-based (R, S) policy, each member considers his
partnership between supply chain members can be local inventory position. The sequence of events
described as one of the following integration levels followed by a supply chain member under TSS is
(see Fig. 1). The first structure is a supply chain outlined as follows:
operated under traditional ways of doing business
(TSS) and the second structure is a supply chain (i) The member receives the delivery from its
model operated under a VMI program. Finally, the immediate upstream member, which was or-
third structure is the supply chain operated under a dered L periods ago (the lead time is L periods).
CPFR program. If the member is the plant, L is the production
In all three supply chain structures, an (R, S) lead time.
inventory control policy is used for replenishment (ii) The member observes the order placed by its
decisions. Here, R indicates the review interval and immediate downstream member. If the member
S indicates the order-up-to level. R is chosen as 1 is the retailer, the order is the market demand.
week. Order-up-to level, however, is updated at the (iii) The member fulfils the customer orders (plus
beginning of each week to reflect changes in demand backorders if there are any) by on-hand
patterns. inventory, and any unfulfilled customer orders
Under TSS, each member strives to develop local are backordered. The member analyzes the
strategies for optimizing his own organization historical replenishment orders placed by its
without considering the impact of his strategies on immediate downstream member for forecast-
the performance of other members. Moreover, since ing. Based on this demand forecast, the member
updates its order-up-to point. If the member is the same way as in TSS. Here, in order to compute
the retailer, historical market demand data are the echelon order-up-to levels of the retailer and the
analyzed. The order-up-to point of the member distributor, the heuristic developed by Shang and
at stage k, Sk, estimated from the observed Song (2003) is used. Again, in this supply chain, the
demand is as follows (Nahmias, 1997, p. 278): exponential smoothing method is used to update the
order-up-to level at each week.
1 bk
Sk ¼ F k (1) Finally, under CPFR, inventory levels, pos data,
bk þ hk
promotion plans, sales forecasts and all other
where Fk (.) is the distribution function of the information that may be influential on the market
demand realized by the member at stage k. demand are shared between supply chain members
Similarly, bk and hk are backorder and holding (Fig. 1c). Consequently, a single joint demand
costs of the member at stage k, respectively. forecast is created by the contribution of each
Here, parameters of the demand distribution, member. Here, there is no doubt that demand
Fk (.), are updated at the beginning of each forecasts created with the joint contributions of all
week by using the exponential smoothing supply chain members are more accurate than the
method (see e.g. Nahmias, 1997, p. 74) to ones created by the individual organizations (e.g.
reflect changes in demand patterns. demand forecasts created by the supply chains
(iv) The member decides how many units to order operated under TSS or VMI). Indeed, it is very
from its immediate upstream member. The possible that the parameters of the demand dis-
quantity of the order is equal to the difference tribution can be predicted under a CPFR program.
between the order-up-to level and inventory Therefore, in the simulation model, it is assumed
position. If the member is the plant, a produc- that distribution parameters of the market demand
tion order is placed. Here, the plant, because of are predicted under CPFR by contribution of each
its limited manufacturing capacity, cannot member. This assumption does not mean that at the
always produce enough to bring its inventory end of the collaborative forecasting process, the
position up to the updated value of S. In these members can know exactly what the customer
cases, the plant makes full capacity production demand is, but rather they can know what the
by backordering the remaining requirement. parameters of the underlying demand distribution
This modification of order-up-to policy for the are (i.e. if the customer demand is normally
case of limited production capacity provides an distributed, mean and standard deviation of the
optimal solution for uncertain demands (see customer demand is known only). Indeed, this
e.g. Gavirneni et al., 1999; Federgruen and assumption makes it sure that the promise of CPFR
Zipkin, 1986a, b). is realized. That is, the large amounts of informa-
tion available with CPFR are effectively used to
Under VMI, on the other hand, the retailer minimize the uncertainty along the supply chain. Of
provides the distributor with access to its real-time course, in practice, as the model of Disney et al.
inventory level as well as its pos data (Fig. 1b). In (2004) also indicates, it might possible that bulk of
return, the distributor takes the responsibility of information available with CPFR result in confu-
managing the inventories at the retailer. That is, sion of supply chain managers, which leading low
under VMI, the distributor does not only need to levels of supply chain performance. Therefore, the
take its own inventories into account while making benefits of CPFR obtained in this study are valid
inventory plans, but also the inventories of the only if CPFR is properly implemented. Moreover,
retailer. Therefore, under this structure, the dis- under this collaboration mode, an echelon-based
tributor follows an echelon-based policy in his (R, S) inventory policy is used for entire supply
replenishment planning. Under the echelon-based chain. That is, inventory positions and inventory
policy, the distributor looks at its own inventory costs of all four supply chain members are taken
position plus the inventory position of the retailer, into account in replenishment decisions.
instead of his local inventory position only. For a The cost structures for the supply chain members
discussion of installation and echelon policies, see in the simulation model are assumed to be as
Clark and Scarf (1960), Axsäter and Rosling (1993). follows; the unit backorder costs per week for
All other echelons of the supply chain (the plant and the plant, the warehouse, the distributor and
the warehouse), on the other hand, are operated in the retailer are $5, $11, $18 and $25, respectively.
ARTICLE IN PRESS
580 K. Sari / Int. J. Production Economics 113 (2008) 575–586
The unit inventory costs per week for the plant, the
warehouse, the distributor and the retailer are $0.25,
$0.50, $0.75 and $1.00, respectively.
technique, 15 replications for each combination of appropriate for our model because MANOVA
the independent variables are conducted. considers the correlations between the dependent
variables in the experimental design, see Hair et al.
3. Experimental design (1998, p. 331). Selected MANOVA results are
presented in Table 2.
Four independent factors are considered in the MANOVA results in Table 2 show that at 5%
experimental design. These are; type of supply significance level, SCTYPE has significant impacts
chain (SCTYPE), available production capacity of on both performance factors, which indicates that
plant (CAP), uncertainty in customer demand (DV) CPFR and VMI have substantial influences on the
and replenishment lead times (L). The number of performance of the supply chain. The performance
levels of these variables and their values are listed in of each type of supply chain is presented in Table 3.
Table 1. Examination of Table 3 reveals that the reduction
Factor SCTYPE refers to the way the supply in total supply chain cost derived from CPFR
chain is operated. Specifically, this factor indicates significantly higher than the reduction derived from
whether the supply chain is operated under TSS, VMI. For example, while CPFR provides 33.90%
VMI, or CPFR. The factor CAP is expressed as the cost savings, VMI provides 17.34% on the average.
ratio of the plant’s total capacity to the total market Similarly, the results also indicate that CPFR
demand. Total capacity of the plant is distributed provides higher level of increase in the customer
to each week, equally. The factor L denotes the service level than VMI does. For example, while
replenishment lead times between each member of CPFR leads to an increase of 3.84% in customer
the supply chain. Finally, the factor DV indicates service level on the average, VMI results in 1.54%
the level of uncertainty seen in market demand. increase in customer service level. Therefore, these
Two factors are used as dependent variables in results lead us to conclude that CPFR produces
the experimental design in order to evaluate benefits substantially higher benefits than VMI in terms of
of CPFR and VMI. These factors are total cost for total supply chain cost and customer service level.
the entire supply chain (TSC) and customer service Actually, these findings are simple and intuitively
level of the retailer (CSL). TSC is the sum of the expected for us, so we will not concentrate on them
inventory holding costs of all members in the supply further. Instead, we will concentrate on how the
chain and backorder cost of the retailer. Here, we performance increase gained from CPFR and VMI
include the backorder cost of the retailer only, change in parallel to changes in specific conditions
because all other backorder costs are internal costs of supply chains. For this purpose, performance of
within the entire supply chain and they are not CPFR and VMI under various capacity levels
actually incurred. Factor CSL is the percentage of (CAP), demand uncertainty (DV) and lead times
customer demand satisfied by the retailer through (L) are produced in Fig. 3.
the available inventory.
4.1. Impact of manufacturing capacity (CAP) on the
supply chain collaboration
4. Simulation output analysis
MANOVA results in Table 2 shows that at 5%
The output from the simulation experiments are
significance level, the interaction effect between
analyzed using MANOVA procedure of the SPSS.
CAP and SCTYPE has significant impacts on both
MANOVA analysis is chosen because it is more
dependent variables. This means that manufactur-
ing capacity has a significant influence on the
Table 1
Independent factors of the experimental design performance of CPFR and VMI for all performance
measures.
Independent factors Levels Examination of Fig. 3 reveals that both supply
1 2 3
chain initiatives better off operating in the environ-
ments where larger manufacturing capacities are
SCTYPE TSS VMI CPFR available. That is, we see that contribution of CPFR
CAP 1.10 1.30 1.50 and VMI is in its lowest level when the plant has its
DV SDV MDV HDV
L 1 4
smallest manufacturing capacity (i.e. CAP ¼ 1.10).
For example, when CAP is 1.10, the reductions in
ARTICLE IN PRESS
582 K. Sari / Int. J. Production Economics 113 (2008) 575–586
Table 2
Selected MANOVA results
in the cost savings derived from VMI. For example, and SCTYPE has significant impacts on both
the reduction in total supply chain cost decreases performance factors. This means that lead times
from 24% to 15% when the uncertainty in the have a significant influence on CPFR and VMI for
customer demand (DV) changes from low level all performance measures.
(LDV) to high level (HDV). On the other hand, Examination of Fig. 3 reveals that CPFR and
same level of increase in the demand uncertainty VMI exhibit different performance levels under
does not produce that much performance reduction different replenishment lead times. That is, while
in CPFR. That is, it results in approximately 3% of the reduction amount in total supply chain cost
reduction in total supply chain cost. In addition to under CPFR substantially increases as the replenish-
this, Fig. 3 also shows that the supply chain ment lead times increase, the supply chain cost
operated under CPFR produces higher customer savings gained from VMI do not change signifi-
service level under all levels of the demand cantly. For example, when the supply chain operated
uncertainty. Similarly, the gap between the custo- under CPFR is considered, it is seen that the
mer service levels of VMI and CPFR is in its highest reduction in total supply chain cost dramatically
level when the uncertainty in market demand is in increases from 20% to 40% as lead times increase
its highest level. For example, we see that, while from 1 to 4 weeks. On the other hand, in case of
CPFR produces 1.52% of higher customer service VMI, it is apparent that any level of increase or
level under low level of demand uncertainty decrease in replenishment lead times does not result
(DV ¼ SDV), it is increased to 2.90% when there in significant changes in the reduction of total supply
is high level of uncertainty in the market demand chain cost. Furthermore, when the customer service
(DV ¼ HDV). level is considered, we see that under all levels of lead
The results obtained here lead us to conclude that times, CPFR produces higher level of fill rate than
compared with a VMI system, the value of CPFR is VMI does. There is, in addition, one further point to
substantially greater under the market conditions where make that the gap between the service levels achieved
demand variability is high. Therefore, SCM practi- by CPFR and VMI substantially increase as the
tioners have to be more eager to implement CPFR replenishment lead times increase.
programs under more volatile market conditions. The results obtained here imply that compared
with a VMI system, the value of CPFR is
4.3. Impact of lead times (L) on the supply chain substantially greater under the conditions where
collaboration replenishment lead times are longer. Therefore,
SCM practitioners have to be more eager to invest
MANOVA results in Table 2 show that at 5% in CPFR instead of VMI in supply chains where
significance level, the interaction effect between L lead times are long.
ARTICLE IN PRESS
584 K. Sari / Int. J. Production Economics 113 (2008) 575–586
Barratt, M., Oliveria, A., 2001. Exploring the experiences of Foote, P.S., Krishnamurthi, M., 2001. Forecasting using data
collaborative planning initiatives. International Journal of warehousing model: Wal-Mart’s experience. The Journal of
Physical Distribution and Logistics Management 31 (4), Business Forecasting Methods and Systems 20 (3), 13–17.
266–289. Gavirneni, S., 2002. Information flows in capacitated supply
Chen, F., Drezner, Z., Ryan, J.K., Simchi-Levi, D., 2000a. chains with fixed ordering costs. Management Science 48 (5),
Quantifying the bullwhip effect in a simple supply chain: The 644–651.
impact of forecasting, lead times and information. Manage- Gavirneni, S., Kapuscinski, R., Tayur, S., 1999. Value of
ment Science 46 (3), 436–443. information in capacitated supply chains. Management
Chen, F., Drezner, Z., Ryan, J.K., Simchi-Levi, D., 2000b. The Science 45 (1), 16–24.
impact of exponential smoothing forecasts on the bullwhip Hair, J., Anderson, R.E., Tatham, R.L., Black, W.C.,
effect. Naval Research Logistics 47 (4), 269–286. 1998. Multivariate Data Analysis, fifth ed. Prentice-Hall,
Chopra, S., Meindl, P., 2001. Supply Chain Management. New Jersey.
Prentice-Hall, New Jersey. Keaton, M., 1995. Inventory control under gamma demand and
Clark, A.J., Scarf, H., 1960. Optimal policies for a multi-echelon stochastic lead time. Journal of Business Logistics 16 (1),
inventory problem. Management Science 6 (4), 475–490. 107–131.
CPFR Workgroup, 2002. Collaborative Planning Forecasting Keeling, K.B., Pavur, R.C., 2004. Numerical accuracy issues in
and Replenishment Voluntary Guidelines Voluntary Inter- using excel for simulation studies. Proceeding of the 2004
industry Commerce Standards, available at /http://www.vics. Winter Simulation Conference, 1513–1518.
org/committees/cpfr/voluntary_v2/CPFR_Tabs_061802.pdfS Kuk, G., 2004. Effectiveness of vendor managed inventory in the
(accessed 17 May 2006). electronics industry: Determinants and outcomes. Informa-
- etinkaya, S., Lee, C.Y., 2000. Stock replenishment and
C tion and Management 41 (5), 645–654.
shipment for vendor-managed inventory systems. Manage- Lau, J.S.K., Huang, G.O., Mak, K.L., 2004. Impact of
ment Science 46 (2), 217–232. information sharing on inventory replenishment in divergent
Danese, P., 2004. Beyond vendor managed inventory: The supply chains. International Journal of Production Research
Glaxosmithkline case. Supply Chain Forum: An International 42 (5), 919–941.
Journal 5 (2), 32–39. Lee, C.C., Chu, W.H.J., 2005. Who should control inventory in a
Dejonckheere, J., Disney, S.M., Lambrecht, M.R., Towill, D.R., supply chain. European Journal of Operational Research 164
2004. The impact of information enrichment on the bullwhip (1), 158–172.
effect in supply chains: A control engineering perspective. Lee, H., Padmanabhan, V., Whang, S., 1997a. Information
European Journal of Operational Research 153 (3), 727–750. distortion in a supply chain: The bullwhip effect. Manage-
De Toni, A.F., Zamolo, E., 2005. From a traditional replenish- ment Science 43 (4), 546–558.
ment system to vendor-managed inventory: A case study from Lee, H., Padmanabhan, V., Whang, S., 1997b. The bullwhip
the household electrical appliances sector. International effect in supply chains. Sloan Management Review 38 (3),
Journal of Production Economics 96 (1), 63–79. 93–102.
Disney, S.M., Towill, D.R., 2003a. Vendor-managed inventory Lee, H., So, K.C., Tang, C.S., 2000. The value of information
(VMI) and bullwhip reduction in a two level supply chain. sharing in a two-level supply chain. Management Science 46
International Journal of Operations and Production Manage- (5), 626–664.
ment 23 (6), 625–651. Li, J., Sikora, R., Shaw, M.J., Tan, G.W., 2006. A strategic
Disney, S.M., Towill, D.R., 2003b. The effect of vendor managed analysis of inter organizational information sharing. Decision
inventory dynamics on the bullwhip effect in supply chains. Support Systems 42 (1), 251–266.
International Journal of Production Economics 85 (2), Mason-Jones, R., Towill, D.R., 2000. Coping with uncertainty:
199–215. Reducing ‘‘bullwhip’’ behaviour in global supply chains.
Disney, S.M., Naim, M.M., Potter, A., 2004. Assessing the Supply Chain Forum 1 (1), 40–45.
impact of e-business on supply chain dynamics. International McCullen, P., Towill, D.R., 2002. Diagnosis and reduction of
Journal of Production Economics 89 (2), 109–118. bullwhip in supply chains. Supply Chain Management: An
Dong, Y., Xu, K., 2002. A supply chain model of vendor International Journal 7 (3), 164–179.
managed inventory. Transportation Research Part E: Logis- Metters, R., 1997. Quantifying the bullwhip effect in supply
tics and Transportation Review 38 (2), 75–95. chains. Journal of Operations Management 15 (3),
Federgruen, A., Zipkin, P., 1986a. An inventory model with 89–100.
limited production capacity and uncertain demands I: The Micheau, V.A., 2005. How Boeing and Alcoa implemented a
average-cost criterion. Mathematics of Operations Research successful vendor managed inventory program. Journal of
11 (2), 193–207. Business Forecasting 24 (1), 17–19.
Federgruen, A., Zipkin, P., 1986b. An inventory model with Mishra, B.K., Raghunathan, S., 2004. Retailer- vs vendor-
limited production capacity and uncertain demands II: The managed inventory and brand competition. Management
discounted-cost criterion. Mathematics of Operations Re- Science 50 (4), 445–457.
search 11 (2), 208–215. Nahmias, S., 1997. Production and Operations Analysis. Irwin/
Fiddis, C., 1997. Manufacturer–Retailer Relationships in the McGraw-Hill, Homewood, IL.
Food and Drink Industry: Strategies and Tactics in the Battle Ouyang, Y., 2006. The effect of information sharing on supply
for Power. Financial Times Retail & Customer Publications, chain stability and the bullwhip effect. European Journal of
London, UK. Operational Research.
Fliedner, G., 2003. CPFR: An emerging supply chain tool. Ovalle, O.R., Marquez, A.C., 2003. The effectiveness of using
Industrial Management & Data Systems 103 (1), 14–21. e-collaboration tools in the supply chain: An assessment study
ARTICLE IN PRESS
586 K. Sari / Int. J. Production Economics 113 (2008) 575–586
with system dynamics. Journal of Purchasing & Supply Watson, E., 2005. Nestle switches to vendor managed inventory
Management 9 (4), 151–163. with Tesco. Food Manufacture 80 (8), 20–20.
Program may build CPFR momentum, 2005. DSN Retailing Yao, Y., Evers, P.T., Dresner, M.E., 2007. Supply chain
Today 44 (10), 14–14. integration in vendor-managed inventory. Decision Support
Raghunathan, S., 1999. Interorganizational collaborative fore- Systems 43 (2), 663–674.
casting and replenishment systems and supply chain implica- Yu, Z., Yan, H., Cheng, T.C.E., 2001. Benefits of information
tions. Decision Sciences 30 (4), 1053–1071. sharing with supply chain partnerships. Industrial Manage-
Sari, K., 2007. Exploring the benefits of vendor managed ment and Data Systems 101 (3), 114–119.
inventory. International Journal of Physical Distribution Yu, Z., Yan, H., Cheng, T.C.E., 2002. Modelling the benefits of
and Logistics Management 37 (7), 529–545. information sharing-based partnerships in a two-level supply
Shang, K.H., Song, J.S., 2003. Newsvendor bounds and heuristic chain. Journal of Operational Research Society 53 (4), 436–446.
for optimal policies in serial supply chains. Management Zhang, C., Zhang, C., 2007. Design and simulation of demand
Science 49 (5), 618–638. information sharing in a supply chain. Simulation Modelling
Simchi-Levi, D., Zhao, Y., 2003. The value of information Practice and Theory 15 (1), 32–46.
sharing in a two stage supply chain with production Zhao, X., Xie, J., Zhang, W.J., 2002a. The impact of information
capacity constraints. Naval Research Logistics 50 (8), sharing and ordering co-ordination on supply chain perfor-
888–916. mance. Supply Chain Management: An International Journal
Simchi-Levi, D., Kaminsky, P., Simchi-Levi, E., 2003. Designing 7 (1), 24–40.
and Managing the Supply Chain, second ed. McGraw-Hill, IL. Zhao, X., Xie, J., Leung, J., 2002b. The impact of forecasting
Waller, M.A., Johnson, M.E., Davis, T., 1999. Vendor-managed model selection on the value of information sharing in a
inventory in the retail supply chain. Journal of Business supply chain. European Journal of Operational Research 142
Logistics 20 (1), 183–203. (2), 321–344.