Bahinipati - 2012 - Vertical Collaboration in The Semiconductor Industry

Computers & Industrial Engineering 62 (2012) 504526
Contents lists available at SciVerse ScienceDirect
Computers & Industrial Engineering

journal homepage: www.elsevier.com/locate/caie
Vertical collaboration in the semiconductor industry: A decision framework

for supply chain relationships q
Bikram K. Bahinipati a,, S.G. Deshmukh b
a
b
Quantitative Methods and Operations Management Group, Indian Institute of Management Kozhikode, Kozhikode, India
Department of Mechanical Engineering, Indian Institute of Technology Delhi, New Delhi 110 016, India
a r t i c l e
i n f o
Article history:
Received 1 December 2010
Received in revised form 2 August 2011
Accepted 31 October 2011
Available online 18 November 2011
Keywords:
Vertical collaboration
Buyersupplier relationships
e-Market
Collaboration scheme
a b s t r a c t
Vertical collaboration problem focuses on integrating and modeling the decision problems of the suppliers and buyers together with the market intermediary by identifying the inefciencies in the traditional
marketplace and aligning the incentives of members in the e-marketplace. The present work develops
and solves real life e-marketplace models for complex buyerssuppliers procurement problems by estimating the order quantities in the collaborated supply chain. The newsvendor framework considers
demand to be independent of the selling price as is generally the case in the semiconductor industry supply chain dealing with techno-savvy customers. The vertical collaboration process would be more effective if the length of the planning horizon and order size is considered as a negotiation parameter between
the buyer and supplier. It is observed that the suppliers expected prot function increases with the buyers ordering quantity, which is important in characterizing the general structure of the collaboration
scheme of the supply chain.
2011 Elsevier Ltd. All rights reserved.
1. Introduction
Due to globalization and sustained demand growth, the semiconductor industries are focusing more to sustain their customer
base and to enhance revenue opportunities; so, they must manage
successive technological innovations effectively (Cao & Zhang,
2011; Fabbe-Costes, Roussat, & Colin, 2011). The most ideal approach is to introduce high margin innovative products at the right
time through optimal use of the resources and realignment of the
supply chain members to adjust to this trend (Ku, Gurumurthy, &
Kao, 2007). Due to rapid evolution and change of technology, the
future semiconductor manufacturers must work collaboratively
with its extended supply chain to bring about enhanced synchronization of procurement business functions (Wu, Erkoc, & Karabuk,
2005; Flynn, Huo, & Zhao, 2010).
The collaborative activities include information sharing, joint
relationship effort, and dedicated investments, which lead to
improved customer satisfaction and supply chain performance
(Nyaga, Whipple, & Lynch, 2010). The collaborative relationships,
based on trust and commitment with their supply chain partners,
are critical to achieve efciencies, exibility, and sustainable competitive advantage (Chen, Yen, Rajkumar, & Tomochko, 2011; Cao &
Zhang, 2011; Panayides & Lun, 2009). This collaborative advantage
q
This manuscript was processed by Area Editor Mohamad Y. Jaber.
Corresponding author. Tel.: +91 9567880796.
E-mail address: bikram@iimk.ac.in (B.K. Bahinipati).

0360-8352/$ - see front matter 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.cie.2011.10.017
is the strategic benets gained over competitors in the marketplace

through partner enabled inventory centralization and ordering,
and supply chain partnering, which is the desired synergistic outcome (Cao & Zhang, 2010).
The semiconductor industry sector is characterized by a number of key and unique characteristics from the perspective of product features and the sectors structure, where collaborative
practices are developed in response to the economic pressures
and customer requirement, driving the evolution of the chain
and encourage greater horizontal and vertical coordination (Kapur,
Peters, & Berman, 2003). With the objective of manufacturing and
supplying a specic product or component, or locating new
enterprises, all enterprises in the semiconductor manufacturing
cooperate as synergetic unit to pursue for success (Zhang, Xu, &
Wang, 2004).
The vertical collaboration problem in the semiconductor industry forms the basis of multi-echelon inventory theory, since it considers only two stock keeping locations, i.e., buyer and supplier.
Classical works in multi-echelon inventory theory assume that
these entities cooperate and hence solve their problems using a
centralized approach (Goyal, 1976). This theory suggests integrating and modeling the decision problems of the suppliers and
buyers together with the market intermediary. This is a valid
assumption if the SC entities belong to the same company or
operate under long-term agreements, such as supplier-owned
inventory (SOI) systems (Centikaya & Lee, 2000, 2002). Supply
chain management (SCM) considers both logistics and information
505
B.K. Bahinipati, S.G. Deshmukh / Computers & Industrial Engineering 62 (2012) 504526
ers) are depicted in Fig. 1. As the customer demands are placed at

nished goods level, manufacturing requirements are communicated to nal test, assembly, wafer probe, and to the foundries. Efcient means of communication are necessary to ensure the success
of the process, which can be achieved through the participation of
the supply chain partners in the e-marketplace.
issues as well as dominance relationships between the entities. As

a result, SC studies focus on more than the system wide optimization of inventory problems.
The typical activities which are critical to the management of
vertical collaborative relationships from a procurement perspective are: selecting suppliers, motivating, time of involvement,
assigning physical and human resources, information sharing, synchronizing standards and coordination among supply chain members. So, any member and any relationship within a supply
network are affected by the actions of other members, and thus
have to cope with these to enable successful collaborative relationships (Grover & Saeed, 2007).
The emerging idea is that the greater the extent to which manufacturers engage in e-market enabled transactions with upstream
and downstream members the better the performance (Rosenzweig, 2009). The research reveals that the collaborative e-marketplace is still a relatively new business model (Wu & Chuang, 2010).
It has potential for growth in optimizing supply chain networks,
enabling vertical collaboration between buyers, suppliers and logistic providers (Wang, Potter, Naim, & Beevor, 2011).
Current research in vertical collaboration in semiconductor
industry focuses on inventory management within the rm as well
as throughout the supply chain, to gain cost advantage. This collaborative procurement can be operationalized through the participation of all SC members in a private exchange (e-market) to gain
cost advantage and enhance responsiveness. So, it is imperative
to consider both e-market (centralized) models and traditional
market (decentralized) models for complex supply/replenishment
problems with the objective of obtaining collaborated decentralized solutions for these problems. In order to maintain and improve the competitive power of these industries, it is critical to
select agile, competent and compatible partners quickly and rationally during the formation phase of the collaborative relationship.
Further, an adaptable and reasonable modeling approach is necessary to determine the nature of relationships between buyer and
supplier enterprise (Saen, 2007). This may be possible by developing an analytical framework to generate an evaluation system and
to provide the information regarding the nature or intensity of
buyersupplier relationships. This approach would facilitate decision-making pertaining to vertical collaboration in the semiconductor industry supply chain.
The general relationships between the members of fabless
semiconductor industry supply chain (subcontractors, and custom-
Fabrication
Wafer
Die-bank
2. Motivation and research objectives

The semiconductor industry is a rapidly changing industry with
shortening life cycles, uctuating demand and continuous price
and cost pressures. In order to sustain the dynamic and competitive environment, these enterprises must be exible in the quantity and type of product kept in inventory (Wu et al., 2005). The
business partners share resources to the relationship, establishing
mutual commitment, which reduces opportunistic ambitions and
promotes bilateral governance (Grover & Saeed, 2007). These
structures, implemented through collaborative initiatives, are strategic responses to uncertainty in supply and demand (Zacharia,
Nix, & Lusch, 2011).
The vertical collaboration considered from the procurement
perspective in the semiconductor industries may have market or
industry related constraints to survive in the market, such as resource availability with individual enterprises, and this might contradict with prot maximization objectives. So, these industries
have to satisfy these constraints along with trust related conditions, while working or under a collaborative framework (Panayides & Lun, 2009).
The motivation to order more products/components in semiconductor industry, to achieve economies of scale, is restricted to
a single period in contrast to EOQ-oriented models with innite
planning horizon. The semiconductor industries are characterized
by the fact that increasing the quality of products/components procured/manufactured does not result in xed costs savings in future
periods. The existing literature on buyersupplier coordination assumes that the selling price of items at the buyer is a constant or is
a function of the order quantity. However, in the semiconductor
industry supply chain, with items/components of short product life
cycles, there is always a decline in the economic value of the item
over time, as is evident in PC manufacturing industry. These are
known as permanent markdowns in marketing literature (Nair
& Closs, 2006).
OEM
AT
Customer /
Demand
Fab
1
Probe
1
AT
1
Customer 1
Fab
2
Probe
2
AT
2
Fab
3
Probe
3
AT
3
Customer 3
Fab
4
Probe
4
AT
4
Customer 4
Customer 2
Fig. 1. Relationship between members in a semiconductor supply chain (pull system).
OEM: Original
Equipment
Manufacturer
AT: Assembly/
Testing
Fab:
Fabrication
Probe: Wafer
Probe
506
The researchers should not assume all partnering relationships

in vertical collaboration are strategic and ignore the operational aspects of partnering. So, the current research suggest a continuum
of strategic and operational partnering based on (1) orientation
of the partners (buyer and supplier), and (2) the degree of implementation of partnering between these two independent enterprises. Strategic partnering is a continuous and long-term interenterprise relationship for achieving strategic goals, which delivers
value to the customers and protability to partners. Operational
partnering is an as-needed and short-term relationship for obtaining parity with competitors.
In view of the above, the present work focuses on developing
vertical collaboration models when customer demand is independent of the selling price of the product, which is typical to the
semiconductor industry supply chain dealing with end customers
requiring continuous upgradation of technology to stay competitive in the market, such as big corporate sectors. In this model, a
single-period stochastic buyersupplier problem has been considered, where the buyers (PC manufacturers) successively decrease
their selling price during the planning horizon. It has been assumed that this decrease in selling price is due to the permanent
markdowns. Further, it has been assumed that the end-customers
have different preferences in their willingness to pay. Some customers prefer to buy products immediately after its launching in
the market even if it is costly; some customer prefer to buy it towards the end of the planning horizon when it is cheaper. So, the
demand arrival rate (through point-of-sales data) has been modeled as a pure Poisson process. This collaboration model has been
investigated as a prot maximization problem for this scenario in
semiconductor industry supply chain.
In keeping with the recent trends in supply chain practice, the
objectives of this work are as follows:
(a) Developing a modeling framework and theoretical understanding of vertical collaboration issues in the context of
new initiatives in the supply chain management.
(b) Exploring the conditions under which the vertical collaboration works, and to render insights into contract design and
operational level decision making.
Based on the research objectives, the collaborating enterprises
must develop a viable strategy to determine the relationships to
coordinate inventory management decisions for cost advantage
and ensure the success of the vertical collaborative initiatives.
3. Literature review
Vertical collaboration in buyersupplier network requires that
sensitive information and knowledge may be exchanged to other
parties including competitors through common suppliers (Barratt,
2004). So, the role of e-market is particularly important who act
as intermediaries to maintain the condentiality of information
and knowledge shared by network members (Cheng, Chen, &
Mao, 2010).
In semiconductor industries, an enterprises ability to manage
resources and inventory is the most critical factor for its long-term
success (Cao, Vonderembse, Zhang, & Raghu-Nathan, 2010). These
enterprises must structure their supply chain to respond to demand surge from new product introduction and market upside,
and to absorb short-term decline due to technological change
and market downside (Christopher, Mena, Khan, & Yurt, 2011).
The role of inventory and resource management is critical in these
industries, as the capital equipment cost is high (Bailey & Francis,
2008) and obsolescence rates are high (Gravier & Swartz, 2009).
So, these industries must consider relational contracts between
a enterprises and outsourcing suppliers, where capacity investment must take place before the new product is fully dened (Taylor & Plambeck, 2007).
In view of the above, the literature review on vertical collaboration is presented from three perspectives: (a) supply chain relationships, (b) obsolescence costs, and (c) modeling approaches.
3.1. Perspective of supply chain relationships
The present work has dened vertical collaboration as a relationship in which the buyer and the supplier work together for a common objective by sharing information and resources to solve
problems, improve products, and streamline inventory-related processes. Such collaborative relationships are essential due to the following factors: (a) shorter product life cycles in semiconductor
industry supply chain have enhanced the speed of product innovation (Cao & Zhang, 2011) (b) there is a trend toward mass customization, which compelled the buyers and suppliers to jointly design,
develop and manufacture customized products associated with
high switching costs (Lee, Kim, Hong, & Lee, 2010), and (c) the
increasing need for resource specic investments requires increased
levels of information exchange and trust (Ha, Park, & Cho, 2011). As
the buyers have recognized their interdependence on the suppliers,
they work towards long-term contracts, consensus decision making
and joint problem solving approaches (Kim, Park, Ryoo, & Park,
2010). So, the buyers adopt a supplier policy, which compel the
suppliers to collaborate while protecting the status of each supplier
during the life of a given product (Constantino & Pellegrino, 2010).
While analyzing the vertical supply relationship, it has been
observed that the circulation of information and technology is a
crucial factor for supply chain efciency (Esposito & Passaro,
2009). The collaborative activities, such as information sharing,
joint relationship effort, and dedicated investments lead to trust
and commitment, which in turn lead to improved satisfaction
and performance. The buyers focus more on relationship outcomes,
while suppliers look to safeguard their transaction specic investments through information sharing and joint relationship effort
(Nyaga et al., 2010).
Literature has focused on the perspective of transaction value
by contextualizing inter-rm collaboration in terms of relationship
learning and value co-creation viewed by both the buyers and
suppliers. It has been observed that the global environmental
and inter-organizational conditions inuence learning capabilities,
which in turn inuences relationship value for both supplying and
buying enterprises (Cheung, Myers, & Mentzer, 2010). There is a
positive effect of absorptive capacity, collaborative process competence and level of engagement on the operational and relational
success of a collaboration effort. The collaborative process competence mediates the relationship between absorptive capacity and
collaborative engagement, and positively inuences both operational and relational outcomes (Zacharia et al., 2011).
The protability benets of supply chain relationships are captured predominantly by downstream chain members, whereas
cash cycle benets are realized throughout the supply chain. The
chain members nancial performance varies systematically with
measures of downstream bargaining power, downstream relationship duration, and degree of supply consolidation (Lanier Jr., 2010).
It is supported by the fact that there is a positive relationship between partnership quality and supply chain performance, which
is strengthened in the presence of high demand- and supply-side
risks (Srinivasan, Mukherjee, & Gauret, 2011). The supply chain
relationship quality focuses on communication, cooperation, trust,
adaptation and atmosphere, and this has a signicant positive impact on cooperative strategies, such as relationship persistence,
relationship frequency, and relationship diversity (Su, Song, Li, &
Dang, 2008).
3.2. Perspective of obsolescence cost

High-volume and consumer oriented electronic products and
systems in semiconductor industries are always impacted by the
rapid rate of technology change; so, there is a need to stay on
the leading edge of technology in order to prevent loss of market
share to competitors (Singh & Sandborn, 2006). However, some
product sectors, such as communication systems lag behind the
leading edge because of high cost or long times associated with
new technology adoption (Gravier & Swartz, 2009; Dinesh Kumar
& Saranga, 2010). This phenomenon gives rise to product and component obsolescence. So, this is a major challenge in the semiconductor industries with volatile demand of customer market
segments. Obsolescence of products and components in the semiconductor industry increases the supply chain costs and decreases
the net prot of the collaborative supply chain. When obsolescence
costs are high (i.e., when the product has a low salvage value and a
high retail price as in semiconductor industries) and margins are
low due to short product life cycles, the prot of the buyer decreases in the collaborative supply chain environment (Xiao &
Jin, 2011).
While estimating the total relevant cost of the supply chain, the
obsolescence cost may be estimated based on the most likely disposition value of obsolete inventory (i.e. salvage value of the unsold components and products). The total relevant cost per year
increases with the percentage increase in obsolete demand. Further, as the total relevant cost per year is directly proportional to
the percentage of postponement, it can be stated that the percentage of obsolescence depends primarily on the postponement decisions (Lee, Billington, & Carter, 1993). If the obsolescence costs are
higher, better it is to postpone the nal assembly in semiconductor
industry supply chain. This supports the argument that higher the
risks of product obsolescence, greater are the savings in postponement. The postponement strategy is being used in semiconductor
industries and consumer electronics industry, such as Dell, HP, Xilinx, and Motorola. Manufacturers are nding this strategy increasingly attractive to provide better customer service at reduced cost.
By modularizing the design of power pack and postponing the
assembly of the power pack and manuals with the printer, HP
has reduced the investment in nished goods inventory (Lee
et al., 1993). Postponement is more valuable in a system that faces
returns and in which the salvage values are low, i.e., the return
costs high (Brown, Lee, & Petrakian, 2000).
Reduced supply bases, global sourcing, and reliance on supply
clusters increase a rms vulnerability to parts obsolescence as
the inevitable result of more interconnected and complex supply
chains a disruption to one source of supply affects many buyers
(Craighead, Blackhurst, Rungtusanatham, & Handeld, 2007). The
most common response to diminishing supply sources in the production and operations research has been to develop obsolescence
forecasting methods based on each parts life cycle (Sandborn, Prabhakar, & Ahmad, 2011). The results of technological forecasts are
to minimize the cost of mitigating obsolescence (Singh & Sandborn,
2006). In reality, the fabless semiconductor industry often collaborates with the original equipment manufacturers to raise the salvage value of the products and components by advertising. When
the manufacturer designs its product, it is better to consider the
salvage value of product. This is why the semiconductor industries
attempts to predict or even create a new trend through extensive
advertising campaigns.
3.3. Modeling approaches in literature
Non-collaborative modeling approaches in supply chain literature focuses either on buyer-driven or supplier-driven coordination. For the collaborative modeling approach, this attribute is
507
not applicable. The classical buyersupplier coordination problem

in the literature, is the foundation of the collaboration models in
the present work. An important generalization of the classical
buyersupplier problem in the semiconductor industry supply
chain is when the suppliers (chip manufacturers, electronics component manufacturer and so on) are manufacturers and have nite
production rate. It has been concluded from the existing literature
on supply chain coordination that the type of dispatch policy used
to deliver buyer replenishments is particularly important in this
setting. This is due to the fact that the manufacturing at the supplier enterprises may continue while the buyer enterprises are
being replenished. This may result in very complex inventory load
proles for both the buyer and supplier enterprises depending on
the type of dispatch policy, which impacts the inventory holding
costs. The literature depicts various control policies, which are presented in Table 1.
Typically, in semiconductor industry supply chain, an original
equipment manufacturer (OEM) outsources some of its production
activities to a contract manufacturer, who in turn serves several
customers on the same capacitated production line. It has been
shown that an order release strategy with multiple decision levels
performs signicantly better than an order release strategy with
only one decision level (Boulaksil & Fransoo, 2009). Inventory
models developed in the context of vertical collaboration typically
use newsvendor network settings with single-period and stochastic demand. These models consider inventory investment by independent decision makers in a stationary environment (McCarthyByrne & Mentzer, 2011). So, the prot to the enterprises can be
modeled as a function of the inventory or resource level in the entire supply network. In a newsboy setting, the buyers might benet
from cooperation through coordinated ordering and inventory
pooling. The basic processes consist of placing coordinated orders
by the buyers and use these quantities to satisfy the total demand.
In this way, they can benet from order coordination and perfect
allocation of the ordered quantity to the demands realized.
In the semiconductor industry supply chain, it is assumed that
the buyers are homogeneous, i.e., they have approximately the
same cost and revenue parameters. In this work, review has been
made of the deterministic xed demand rate problem where the
demand rate is independent of the selling price at the buyer enterprises. The deterministic models help to gain insights into the
dynamics of the problem. Further, their solutions can be used as
approximations for stochastic demand problems. However, stochastic problems provide better representations of real life applications in semiconductor industry supply chains. The literature on
the investigation of the stochastic nature of the problem is very
much limited.
The conventional belief has been that long-term buyersupplier
relationships would increase the efciency and service level in the
supply chain. However, the drawback of long-term relationships
might be lack of innovation and benchmarking with the most efcient suppliers at the marketplace. So, the length of the relationship or the planning horizon would be a critical indicator for the
success of the vertical collaboration procurement initiative. Further, the review of literature indicates that trust, collaboration
and resource interdependency are important characteristics of
the e-market oriented governance structures. So, the SCM in an
e-market is an integrative approach for dealing with the planning
and control of materials, components and products moving from
suppliers to end customers (Croom, Romano, & Giannakis, 2000).
In summary, the supply chain models for collaborative inventory management and vertical relationships, presented in the literature, has not paid due attention on the collaborative procurement
in the e-market mode of operation. It is also imperative to explore
the structural and process integration mechanisms used by the
semiconductor industries to form a vertical mode of supply chain
508
Table 1
Perspectives in SC literature supporting vertical collaborative relationships.
Authors
Perspectives
Goyal (1976)
Benets of centralized modeling approach than decentralized modeling approach in terms of supply chain system costs
(referred as the classical buyersupplier problem in the literature)
Analyzed the buyersupplier coordination problem with homogeneous buyers when the price schedule is continuous. This
problem can be treated as equivalent to a single buyer, single supplier problem. This may be due to the fact that the price
schedule that is optimal for one buyer applies to all other buyers
Discounts as a means for the suppliers to encourage the buyers to order more and to increase prots without changing the
buyers cost
Lot-for-lot (LFL) dispatch policy
Identical delivery quantity policy (IDQ) dispatch policy
Deliver-what-is-produced (DWP) dispatch policy
Analyzed the coordination problem under a newsboy problem setting; analyzed both the centralized and decentralized
decision making models of the buyer and the supplier, and suggested conditions for which centralized modeling approach is
signicantly better than the decentralized modeling approach. However, they did not provide any coordination mechanism
Centralized approach can be described as the basic deterministic model in multi-echelon inventory systems. The total
system cost per unit time in the centralized approach is represented as the summation of buyers total cost and the
suppliers total cost per unit time
The suppliers may not always benet from centralization, as it depends on the demand distribution, service levels, and the
level of market search. The suppliers might prefer competing buyers when the market search is high
Proposed the globally optimal dispatch policy for the single buyer, single supplier manufacturing problem. But it has a very
complex structure that may be very difcult to implement
The establishment and maintenance supply chain relationships is a critical element for collaborative initiatives, which
includes mutuality of benets, risks, and prot sharing
The benets of developing long-term relationships through collaboration can include decrease in transaction costs, and
increase in resource sharing, learning and risk sharing.
Compared the buyer-oriented and supplier-oriented supply chains as a solution to the decentralized modeling approach,
where the impact of power structure is an interesting issue
Coordinated ordering and inventory pooling
Lal and Staelin (1984)
Monahan (1984)
Banerjee (1986)
Banerjee and Burton (1994), Lu (1995)
Goyal (1995)
Parlar and Weng (1997)
Silver et al. (1998)
Anupindi and Bassok (1999)

Hill (1999)
Barratt and Oliveira (2001)
Jap (2001)), Cousins (2002))
Ertek and Grifn (2002)
Hartman and Dror (2003, 2005), Slikker
et al. (2005))
Dong and Rudi (2004)), Zhang (2005))
Handeld and Bechtel (2004)
Meca et al. (2004)), van den Heuvel et al.
(2007))
Bagchi, Ha, Skjoett-Larsen, and
Soerensen (2005)
Matopoulos, Vlachopoulou, Manthou,
and Manos (2007)
Bahinipati, Kanda, and Deshmukh
(2009a)
The suppliers are not isolated from the effects of cooperation or competition at the buyers level in the supply chain
Interaction of elements, such as trust, power, and dependence play an inuential role in enterprises decision to collaborate
The buyers benet from reduced ordering and inventory related costs due to coordinated ordering
There is a signicant correlation between the length of the relationship with suppliers and performance measures, such as
total logistics cost, on-time delivery and rate of return. Supply chain managers should continuously benchmark their
suppliers with best practice in the supplier market
Identify how trust, power and dependence interact with each other, and how they affect and determine the intensity of
collaboration, as well as, the selection of the appropriate information-data sharing technologies and techniques
Detailed review of buyersupplier coordination problems, which forms the basis for collaborative procurement
governance or relationship structure (McCarthy-Byrne & Mentzer,

2011). So, the future research must focus on developing analytical
modeling approaches for collaborative ordering and inventory
management in e-market operated supply chains. This approach
may be ideal for operational and tactical decision making to develop strategic relationship among members of supply network.
4. Vertical collaboration in supply chains
The literature has identied two perspectives of buyersupplier
relationships in supply chains: a purchasing and supply perspective, and a transportation and logistics perspective (Tan, 2001).
The objective of the rst perspective is to reduce the supply base
and inventory, and to increase customer satisfaction. The objective
of the second perspective is to reduce transportation costs, reduce
demand uncertainty and provide supply chain visibility. In order to
achieve these goals, collaboration among SC members is critical
(Horvath, 2001; Skjoett-Larsen, Thernoe, & Andresen, 2003).
Classical works in multi-echelon inventory theory assume that
the buyer and the suppliers cooperate and hence solve their problems using a collaborative (centralized) approach. This is a valid
assumption if the supply chain members operate under long-term
agreements, such as supplier-owned inventory (SOI) systems. On
the other hand, supply chain management considers both logistical
and information issues as well as dominance relationships
between the members.
Although external pressure exerted by buyers on suppliers may
encourage the adoption of e-market, it may also result in distrust
that could reduce the effect of actual e-market use. The implemen-
tation of e-markets results in higher perception of risks due to

higher uncertainties, i.e., buyers face the risks of incomplete and
distorted information, whereas suppliers face the possibility that
their price offers will be reduced due to information transparency.
So, supplier distrust and perceived buyer opportunism may cause
supplier dissatisfaction and resistance to the adoption of such emarkets. This underlines the importance of offering collaborative
procurement functions through e-markets, as this can result in a
winwin strategy by eliminating opportunism and price reduction.
The objective of the present work is to explore how rms can
collaborate in purchasing through e-market. Participating in an
e-market is known as collaborative business, irrespective of
whether the SC members pursue business through arms-length
business relationships or through long-term relationships (Barratt
& Rosdahl, 2002). So, e-markets for semiconductor industries are
collaborative initiatives. The researchers are focusing on e-market
supporting portfolios of sourcing strategies, including strategic
supply chain management. Rudberg, Klingenberg, and Kronhamn
(2002) discussed the possibility of supporting supply chain planning using e-markets. Grieger (2003) suggested that there are possibilities for managing multiple supply relationships through emarkets. However, these collaborations have been limited to sharing an e-market infrastructure, but not purchasing. So, the scope of
such collaborative initiatives should be extended to the investment
opportunities in SCM technology and IT infrastructure. Although
there have been a large number of literature from different qualitative perspectives of inter-organizational collaborative relationships, quantitative perspectives of collaboration in the context of
e-market has not been studied systematically.
509
4.1. Classication of vertical collaboration procurement activities
4.2. Impact of e-marketplaces for vertical collaboration
The semiconductor industry supply chain classify vertical collaboration activities as collaborative fulllment, pre-negotiated
purchasing contract execution, product life cycle management,
and supply chain coordination and integration, which can be described as given below:
In SCM, purchasing becomes strategic and tightly coupled

with other functions of supply chain, such as inventory management and product development. Strategic SCM is supported by
information and communication technology (ICT) tools, such as
electronic data interchange (EDI), Internet, and e-market
(Garcia-Dastugue & Lambert, 2003; Lancioni, Smith, & Schau,
2003). Ovalle and Marquez (2003) explored the effectiveness
of various e-collaboration tools to share product information,
customer demand and transaction information. They concluded
that information sharing contributes to faster and more exible
supply chain processes among SC partners, and make the process more responsive to market changes. Some researchers think
that e-markets are only tools that support non-strategic purchasing and short-term relationships (Choudhury & Hartzel,
1998). McLaren, Head, and Yuan (2002) developed a framework
for supply chain activities supported by e-business tools, in
which third party e-markets only support non-strategic product
procurement.
Various methods for synchronizing supply chain information
and processes between organizations are EDI/XML, joining an
e-marketplace, or utilizing shared collaborative SCM systems
(McLaren et al., 2002). There is a lack of suitable frameworks
in literature for analyzing the expected costs and benets of
the supply chain members to determine which approach is most
suitable for an enterprise. In order to optimize the entire supply
network, the enterprises must jointly make supply and demand
decisions that create sustainable value for all members involved.
Collaborative SCM assists in (1) exchanging and integrating
information between suppliers and buyers, (2) tactical decision
making among the partners for collaborative planning, forecasting, distribution and product design (Kumar, 2001), and (3) strategic joint decision making about partnerships and network
design. So, collaborative SCM systems allow enterprises to move
beyond mere operational level information exchange and optimization, and can transform a business and its partners into
more competitive enterprises.
It is inferred from the review of literature that collaborative and
stable relationships have dominated market-oriented and
hierarchical governance structures in purchasing and supply chain
management. Further, if a majority of the purchasing is done
through existing collaborative partners, then that share of the market must be discounted as potential transactional volume (Barratt
& Rosdahl, 2002). This may be one potential reason for the
e-markets focusing on collaborative functionality for procurement
in the semiconductor industry supply chain.
1. Collaborative fulllment in semiconductor industry supply

chain represents the lowest level of collaboration after the business partners negotiate a deal, which includes payment, delivery, order tracking, and after-sales maintenance. The emarkets support negotiation of contracts and ensures fulllment of these activities. After the auction is over, both the parties (buyer and supplier) continue to cooperate until the
products are delivered and payments are cleared. The e-markets may offer order tracking, requiring suppliers to cooperate
temporarily in order to share certain information with buyers.
2. Pre-negotiated suppliers in contract execution in e-market
helps in transferring transaction data between partners, such
as purchase orders, delivery notices and invoices. These e-markets (private exchanges) are only accessible by a specic buyer
and its recognized suppliers. Through these facilities, the business partners can use the e-market infrastructure to manage
and automate their business processes, while eliminating open
market negotiation.
3. Product life cycle management focuses mostly on the development stage of the product, which may include project management, document management, and collaborative R&D activities.
The common characteristics of the semiconductor manufacturing industries are the large number of parties involved in the
product development. The e-markets for these industries connect suppliers at different echelons of the supply chain by providing business solutions including collaborative engineering,
document management, and computer aided design (CAD) conversions. In the semiconductor industry, numerous component,
subsystem and system designs must be generated by chip and
component manufacturers, and their suppliers when developing new products. These designs and related documentation
may be generated in different formats by different software
applications on different systems. So, the conversion and transmission of this design information among the business partners
are important.
4. Supply chain coordination and integration deals with the transportation and logistics aspects of the semiconductor industry
supply chain. The purpose of this activity is to eliminate inefciencies in the supply chain. The most popular semiconductor
industry supply chain coordination functionality is supplier
owned inventory (SOI) and its variants. Collaborative planning,
forecasting and replenishment (CPFR) is an advanced SOI service used in this industry, which is used to ensure enough quantity to meet buyer demand, while minimizing inventory costs
(Holweg, Disney, Holstrom, & Smaros, 2005). In order to achieve
this goal, business partners agree to mutual business objectives
and measures, develop joint sales and operational plans, and
collaborate to generate and update sales forecasts and replenishment plans.
Most of the semiconductor manufacturing industries use collaborative fulllment, supply chain coordination and integration
activities. Collaborative procurements in e-markets try to build
connectivity between the business partners, which can support
pre-negotiated contract execution, supply chain coordination and
product life cycle management. Among all these activities, prenegotiated contract execution and supply chain coordination are
seen in e-markets.
4.3. Framework for e-market supported vertical collaboration models

Supply chain collaboration in e-market is aimed at improving
the total expected system prots in a decentralized structure
and to bring them closer to those of a centralized structure (Tsay,
Nahmias, & Agrawal, 2000). There is scope for applications of
operations research (OR) models in internet-enabled supply
chains and e-markets (Sodhi, 2001). The basis for developing
frameworks for such models in neutral e-markets, buyer-oriented
e-markets and supplier-oriented e-markets is presented in this
sub-section.
Assume PC be the total expected system prots in the collaborative structure (neutral e-market) and PD be the total expected system prot in the non-collaborative structure (buyer-oriented/
supplier-oriented e-market). The optimal values of PC and PD can
be estimated by using the solutions to the following models in
the e-market:
510
M1 : Neutrale market Model NEM

Maximize expected value Buyer Profit Supplier Profit
Subject to
Buyer constraints
Supplier constraints
4. PCS PSD P PDB P CB : The suppliers gain from the collaborative

structure is at least the same as the buyers loss from the
non-collaborative solution, which is obvious from observations
(2) and (3) above.
4.3.2. Supplier-oriented e-markets
M2 : Buyer orientede market Model BEM

Buyers ProblemBP1
Maximize expected value Buyer Profit
Subject to
Buyer constraints
Suppliers Problem SP1

Maximize expected value Supplier Profit
Subject to
M3 : Supplier orientede market Model SEM
Suppliers ProblemSP2
Maximize expected value Supplier Profit
Subject to
Buyers ProblemBP2
Maximize expected value Buyer Profit
Subject to
Buyer constraints
The objective function value of M1 gives the value of PC. Further, dene P CB and PCS as the expected values of the buyer and the supplier
prots resulting from M1. So, PC PCB P CS .
In the buyer-oriented or supplier-oriented e-markets models,
the sub-problems can be solved sequentially. The buyer may solve
his sub-problem rst, if he dominates the marketplace and vice
versa. In the e-marketplace with supplier competition, the buyeroriented model has greater dominance, and vice versa. The rst
sub-problem (BP1) can be executed by the buyer and its solution
would determine the expected value of the buyers prots in
buyer-oriented e-marketplace (P DB ). Similarly, the objective function value of the second sub-problem (SP1) would determine the
suppliers expected prots in the buyer-oriented e-marketplace
(PDS ). So, P D P DB P DS .
Since the neutral e-market model (NEM) attempts to maximize
the expected value of system prots, its objective function value
would provide an upper bound on the total expected prots of the
buyersupplier system in the e-market. Hence, in the e-marketplace, PD < PC. So, the neutral e-market can be used as a benchmark,
and the gap between PD and PC can be considered as a potential or
motivation to improve the expected prots of the buyer and supplier in the buyer-oriented and supplier-oriented e-markets. The
following observations can be made about buyer-oriented and
supplier-oriented e-markets from these market models:
4.3.1. Buyer-oriented e-markets
1. If PD < PC: The non-collaborative structure is no better than the
collaborative structure as far as system prots are concerned.
2. If P DB P PCB : The buyers expected prots in the non-collaborative
structure are at least the same as in collaborative structure, as
the buyers feasible solution for his decision variables attains
its maximum due to his dominance in the marketplace.
3. If P CS P P DS : The suppliers expected prots in the collaborative
structure are at least the same as in the non-collaborative
structure.
1. If PDS P PCS : The suppliers expected prots in the non-collaborative structure are at least the same as in collaborative structure,
as the suppliers feasible solution for his decision variables
attains its maximum due to his dominance in the marketplace.
2. If PCB P PDB : The buyers expected prots in the collaborative
structure are at least the same as in the non-collaborative
structure.
3. PCB PDB P P DS P CS : The buyers gain from the collaborative
structure is at least the same as the suppliers loss from the
non-collaborative solution, which is obvious from the observations (2) and (3) above.
These observations are crucial for supply chain collaboration in
e-marketplaces, because one enterprises gain from the collaborative initiative is more than the other enterprises loss. In other
words, the suppliers gain from using the collaborative initiative
can be used to compensate the buyers losses under such
initiative as well as to increase the suppliers prots under the
non-collaborative scenario. This requires that the enterprises in
the non-collaborative mode of operation should be engaged in a
collaborative relationship in such a way that it would result in
the same outcome for the decision variables as in a collaborative
initiative. This would result in a mutual acceptable strategy for
fair and equitable sharing of the revenue. The revenue sharing
among the enterprises can be done by means of any supply chain
contract, such as xed payments between the enterprises of the
supply chain, quantity discounts, rebates, return policies, or a
combination of all these. It can be negotiable between these
enterprises of the supply chain, or can be imposed by one
enterprise to control the behavior of the other enterprise
(Bahinipati, Kanda, & Deshmukh, 2009b). These strategies for
generating evenue/prots of the collaborative structure using
non-collaborative strategies in e-marketplaces are called collaboration schemes.
5. Research methodology for e-market supported vertical
collaboration model
In this section, an attempt has been made to develop an analytical collaboration model when the demand for the product/component is independent of the selling price of the product/component.
The problem has been approached by a newsboy setting due to the
following reasons:
(a) It is widely applicable for components with short product
life cycles, such as consumer electronics, software/hardware
products in semiconductor industry supply chain.
(b) The newsboy problem provides the solution for the last period of the corresponding multi-period stochastic problem.
The costs and related parameters used during the development
of analytical model for the vertical collaboration problem have
been dened and presented in Table 2.
5.1. Assumptions and characterization of marketplaces
The following assumptions have been made while analyzing the
problem on hand.
Table 2
Notations in the vertical collaboration model.
511
Table 3
Characterization of marketplaces from the perspective of vertical collaboration.
Notation
Description
Traditional marketplace [no collaboration (NC)]
MU
TU
SKU
HB
Monetary unit
Time unit
Stock keeping unit
Inventory holding cost at the buyer (per unit per unit time)
(expressed in MU)
Length of the planning horizon or cycle time of the component/
product (TU)
Selling price of the component at T = 0 (to end customer)
(expressed in MU)
Depreciation rate in the economical value of the component, such
that G HT > 0
End customer price = G HT (expressed in MU)
Arrival time of the ith demand from the customer
Number of demand arrivals during [0, T] = n
Fixed replenishment cost of the buyer (MU)
Fixed replenishment cost of the supplier (MU)
Manufacturing cost of the component by the supplier (MU)
Salvage cost or cost of obsolescence of the component (of the
unsold components at the end of the period) (MU)
Penalty per unit due to lost sale (MU)
Purchase cost per unit by the buyer from supplier (MU)
Ordering quality of the buyer (at the beginning of the planning
horizon) (SKU)
Optimal ordering quantity by the buyer in the collaborated
scenario (SKU)
Optimal ordering quantity by the buyer in the traditional
marketplace (SKU)
Optimal ordering quantity by the buyer in the electronic
marketplace (SKU)
Expected value of the buyers prot (as a function of the order
quantity) (MU)
Expected value of the suppliers prot (as a function of the order
quantity) (MU)
Expected prot of the collaborated system between buyers and
supplier (MU)
Expected value of the revenue of the buyer (component wise)
(MU)
Expected value of the holding cost of the buyer (MU)
Expected salvage value or cost of obsolescence of the buyer (of the
components) (MU)
Expected value of procurement cost of the buyer (MU)
Expected value of the replenishment cost of the buyer (MU)
Expected value of the lost sales at the buyer (MU)
Buyer takes decision about the optimal ordering quantity, for which its
expected prot is maximum.
Enterprise optimization focus or decentralized decision making
No information sharing between supply chain members
Less formal relationship with no or lower degree of commitment
Authority lies with individual enterprises
No resource, risk or prot sharing between buyer and supplier
Electronic marketplace [complete collaboration (CC)]
T
G
H
R(t)
Ai
X
KB
KS
P
V
B
C
Q
QC
QTM
QEM
PB(Q)
PS(Q)
PC(Q)
E[R]
E[HC]
E[SV]
E[PC]
E[RC]
E[LS]
1. A single short life-cycle product involving a single time-period

is considered for analysis.
2. There is no initial on-hand inventory.
3. The demand arrival during the planning horizon is a random
variable, which follows a certain probability distribution.
4. The replenishment costs of the buyers and suppliers, and the
holding cost of the buyers are xed.
As outlined earlier, the major objective of the analytical modeling is to maximize the expected prot of the collaborated system.
While developing the analytical framework, perspectives of procurement in semiconductor manufacturing industries may be considered. As such, in the semiconductor industries, the products/
components are often substitutable, i.e., the buyers demand may
be satised by another product. Typically, the buyers commit
themselves to a large extent in terms of the number of units to order. So, the only decision to be made is about the value of Q, the
number of units to order, so that these can be placed into inventory
at the beginning of the period. The suppliers have one or more
opportunities for replenishment after the initial order is placed
by the buyer.
It may also be noted that the sales of the nal products are usually inuenced by promotional activities. This is done to reduce the
obsolescence cost in the supply chain, which is very common in
semiconductor industries. It is possible to dispose of any units of
Both buyer and supplier take decision (joint decision making) about the
optimal ordering quantity, for which the expected prot of the collaborated system is maximum.
Planning, division of roles and information and communication channels
are clearly dened.
Supply chain optimization focus
Complete information sharing to e-market intermediary (private
exchange)
Joint authority structure and joint resource sharing
Hybrid marketplace [PARTIAL COLLABORATION (PC)]
Buyer takes the decision about the optimal ordering quantity, for which
the expected prot of the decentralized supply chain system is
maximum.
Information sharing as and when required
Resource allocation, risk and prot sharing is done on mutual basis
Authority lies with individual enterprises
the product remaining at the end of the planning horizon (by

receiving a salvage value for the units). So, the perspective of price
declines and product obsolescence in the semiconductor industry
can be considered as a parameter contributing to the buyers prot
model and can be estimated as the salvage cost of the unsold components and products in a collaborative supply chain setting.
The research has also characterized the marketplace as traditional marketplace (TM), hybrid marketplace (HY) and e-marketplace (EM), for the development of market models and their
analysis. Traditional marketplace (TM) has been dened as a market space in which buyers and suppliers are engaged in independent decision making, and in which the optimal policy is
determined by their individual minimum expected cost; the nature
of buyersupplier relationship in such a case has been dened as
no collaboration (NC). Electronic marketplace (EM) has been dened as a market space in which the e-market intermediary has an
access to the information of both buyers and suppliers and decides
the optimal policy based on the minimum joint total expected cost
of the centralized supply chain; the nature of buyersupplier relationship in such a case has been dened as complete collaboration (CC). Hybrid marketplace (HM) has been dened as a
market space in which the buyer is engaged in independent decision making with the help of an e-market and adopts a policy
based on the minimum joint total expected cost of the decentralized supply chain; the nature of buyersupplier relationship in
such a case has been dened as partial collaboration (PC). The
complete characterization of the three market spaces dened in
the present work has been depicted in Table 3.
Attempt has also been made to develop algorithms for traditional market model and e-market model. It has been operated
for several simulation runs to demonstrate decision policies for
vertical collaboration at the operational and tactical level. Further,
it has been attempted to compare the objective function values in
the traditional marketplace and e-marketplace models for given
sets of parameter values to establish the nature of relationships between the buyer and supplier [NC or PC or CC]. This model has
been extended through simulation to facilitate decision making.
The research framework for this model has been depicted in Fig. 2.
512
PB Q ERevenue Salvage cost EHolding cost
5.2. Development of an analytical collaboration model

While analyzing this problem, certain other real-time considerations from the perspective of semiconductor industries are made,
based on the feedback from the focus group supply managers.
These are as follows:
1. One component is required to produce one nal product.
This is due to the fact that each nal product requires one
critical component from its supplier before nal assembly or
processing. The buyer may collaborate with the supplier for
the design, development and manufacturing of this critical
component.
2. The demand arrival process (from the end customer) is a pure
Poisson process (k = Arrival rate). This may be due to the fact
that the distribution of number of demand arrivals during a
specic time period is independent of the time and is independent of the number of arrivals in earlier time intervals. So, the
number of demand arrivals in non-overlapping intervals is statistically independent, which is typical to a demand distribution
in the semiconductor industry supply chain dealing with short
life cycle products.
3. The price of the nal product or component depends on the
time of the life cycle or planning horizon at which it is delivered
or sold to the customer or buyer, i.e., End customer price,
R(t) = G HT (MU). This is due to the fact that the price of the
product or component decreases over time, which is typical to
the semiconductor industry supply chain.
4. The end customer demand is independent of the price of the
product or component. This is due the fact that the major buyers or customers of the semiconductor industry supply chain
are very much technology dependent and they upgrade their
technology continuously. As these buyers want to remain competitive in the marketplace, they cannot wait for the price of the
product or component to fall which is typical during its life
cycle. So, the demand for the products or components is independent of the price.
5. It is important to note that most single-period stochastic
demand problems do not model inventory holding costs. This
is due to the inherent assumption that the period is so short
that these costs can be ignored, or they charge the inventory
holding costs to the end-of-period items (i.e. obsolete or unsold
items) by modifying the salvage value. In this sense, the analysis in semiconductor industry is more realistic, because of the
consideration of an inventory holding cost for each time unit
that a component item stays in the inventory.
6. The semiconductor industries, characterized by volatile demand
of customer market segments and long times associated with
new technology adoption, are associated with the risk of component and product obsolescence. So, while estimating the prot
function of the supply chain, the obsolescence cost is estimated
based on the most likely disposition value of excess obsolete
inventory (i.e. salvage value of the unsold components and
products).
The expected prot of the collaborated system can be estimated as the individual contribution of members to the supply
chain.
PC Q PB Q PS Q
The expected value of the buyers prot depends on revenue earned,

and other total relevant costs, such as inventory holding cost, obsolescence cost, shortage cost, procurement cost from supplier and
xed replenishment cost. The expected prot function of the buyer
can be represented by
Shortage cost Purchase cost

Replenishment cost
The expected value of the supplier prot (for a given variable Q) depends on the difference of manufacturing cost and selling price of
the components, as well as the xed replenishment cost of the supplier. The expected prot function of the supplier can be represented as
PS Q C PQ K S
So, the collaboration model is developed with Q as the decision

variable.
The next step is to estimate the cost components of the buyers
expected prot in Equation (5). As the demand arrival is a random
process, the expected values of revenue, holding cost, obsolescence
cost or salvage cost and the cost of lost sales are to be estimated as
discrete cases.
(a) The expected value of the revenue generated by the buyer in
the discrete case is given by the following expression (Silver,
Pyke, & Peterson, 1998; Papoulis & Pillai, 2002).
ER
1
X
ERjX n:Prob:X n ERjX 0Prob:X 0
n0
Q
X
1
X
ERjX nProb:X n
ERjX nProb:X n
nQ 1
n1
After analysis, simplication and rearrangement of the terms in the

above expression, the expected value of revenue generated by the
buyer can be estimated. This has been presented in the following
Expression (7).
ER 0
"
"
Q
X
n1
1
X
nQ 1
Gn
HnT
2
#
Prob:X n
HTQ Q 1
GQ
2n 1
#
Prob:X n
(a) The expected value of inventory holding cost of the buyer in

the discrete case is represented in the following expression.
EHC
1
X
EHCjX nProb:X n EHCjX 0Prob:X 0
n0
Q
X
EHCjX nProb:X n
n1
1
X
EHCjX n:Prob:X n
nQ1
After analysis, simplication and rearrangement of the terms, the

expected value of the inventory holding cost for the entire cycle
length to be incurred by the buyer can be estimated. This has been
presented in the following Expression (8).
EHC QHB TPr:X 0

"
"
Q
X
n1
nTHB
QTHB
2
#
1
X
HB Q Q 1T
Pr:X n
2n 1
nQ1
#
Pr:X n
8
(c) The expected value of the obsolescence cost of unsold components at the buyer is estimated based on the most likely
disposition value of the obsolete inventory or the salvage
value of the unsold components and products (at the end
of the period). This expected cost of the buyer in the discrete
case is given by the following expression:
ESV
1
X
"
ESVjX nProb:X n
n0
Q
X
#
VQ n Pr:X n
n0
(d) The expected cost of lost sale or goodwill of the buyer in the
discrete case is estimated as the penalty cost for the unsatised demands that arrive after the sale of Q units. The
expected value of this cost is represented in the following
expression:
"
#
1
1
X
X
ELS
ELSjX nProb:X n
Bn Q Pr:X n 10
n0
nQ 1
(e) The expected value of the procurement cost and replenishment costs are the xed costs incurred by the buyer. These
expected cost gures are represented in the following
expression:
EPC ERC CQ K B
11
The expected value of the buyers prot can be estimated by

considering all the transaction cost components represented by
the expressions (7)(11) and incorporating them in the buyer
prot function developed earlier in expression (5). After analysis,
simplication and rearrangement of the cost components, the
expected value of the buyers prot for the entire cycle length
can be estimated. The expected value of the buyer prot function
is depicted in the following expression:
"

#
1
X
HnT
HTQ Q 1
Pr:X n
Pr:X n
P B Q Gn
GQ
2
2n 1
nQ 1
"
#
Q
X
VQ n Pr:X n VQ nPr:X 0
n1
"
#
1
X
nTHB
QTHB
QHB TPr:X 0
Pr:X n
2
"
#n1
1
X
HB Q Q 1T
Pr:X n

2n 1
nQ 1
"
#
1
X

Bn Q Pr:X n CQ K B
513
The expressions for PB (Q) and PC (Q) depict the expected prot
function of the buyer in the traditional market and the prot
function of the collaborative system in e-market respectively.
These expressions must be solved simultaneously for establishing
the ordering quantity for the traditional market and e-market,
for a given T value. The unique optimal ordering quantities
QTM and QEM value necessitates the above prot functions to be
concave.
5.3. Test for concavity of the expected prot function PC (Q) for
collaborative relationships
It is imperative to establish unique optimum values of the
ordering quantities (Q) of the buyer for both traditional market
and e-market, which makes PB(Q) and PC(Q) maximum, i.e., to
establish the concavity of the expected value of the prot
functions.
Let DPC Q PC Q 1 PC Q
D2 PC Q DP C Q 1 DPC Q
If D2PC(Q) < 0, then PC(Q) is a concave function of Q.
The nal expression after simplication and rearrangement of
terms can be represented by Expression (14).

HTQ 1
D2 PC Q V G B
Pr:X Q 1
Q 2

1
X
HTQ 1 HB TQ 1
PrX n

n1
n1
nQ 2
14
As the salvage cost is less than penalty cost due to loss sales and
goodwill, V < B ) V - B < 0. Further, G HT P 0 ) G HT QQ 1
P
2
6 0. So, the rst expression in Equation (14) is
0 ) G HT QQ 1
2
nQ 1
CQ K B QV HB TPr:X 0

Q
X
HnT
HB nT
VQ n QTHB
Pr:X n
Gn
2
2
n1

1
X
HTQ Q 1 HB QQ 1T

Bn Q Pr:X n
GQ
2n 1
2n 1
nQ 1
negative, and hence D2PC(Q) < 0. This implies that the expected
prot function of the collaborated system is a concave function.
So, there is unique value of the decision variable Q, which makes
PC(Q) maximum, i.e., there is a unique QEM for e-market model.
Similarly, it can be proved that PB(Q) is concave and has a unique
QTM for traditional market model.
Let QC = Value of the decision variable for maximum PC(Q). So,
QC must satisfy the following inequalities:
12
PC Q C 1 PC Q C < 0
15
The expected prot of the collaborated system can be estimated by

considering the individual prot contribution of members to the
supply chain (both buyer and supplier) depicted in expression (12)
and expression (6) respectively. These prot functions are incorporated in Expression (4) to compute the expected prot of the collaborated system. This has been presented in the following expression:
PC Q C 1 PC Q C < 0
16
P C Q P B Q P S Q CQ K B Q V HB TPr:X 0

Q
X
HnT
HB nT
Gn
VQ n QTHB
Pr:X n
2
2
n1

1
X
HTQ Q 1 HB Q Q 1T
GQ

Bn Q Pr:X n
2n 1
2n 1
nQ 1
C PQ K S PQ K B K S Q V HB TPr:X 0

Q
X
HnT
HB nT
Gn
VQ n QTHB
Pr:X n
2
2
n1

1
X
HTQ Q 1 HB Q Q 1T
GQ

Bn Q Pr:X n
2n 1
2n 1
nQ 1
13
This system of inequality means that QC is the smallest value of Q

which satises PC(Q + 1) PC(Q) < 0 ) DPC(Q) < 0
But; DPC Q P V HB TPr:X 6 Q

1
X
H HB TQ 1
B Pr:X n
G
n1
nQ1
17
As the demand arrival process from the end customer is assumed to

follow a Poisson process, the probability of occurrence of n arrivals
in [0, T] is given by
Pr:X n
ekT kTn
n!
18
So, the expression (17) can be simplied by incorporating the probability value presented in expression (18).
514
DPC Q P V HB TPr:X 6 Q G Bf1 Pr:X 6 Q g

1
X
ekT kTn
H HB TQ 1
n 1n!
nQ1
In the traditional marketplace, when there is no collaboration between buyer and supplier, QTM is the smallest value of Q that
satises
P G B V HB T G BPr:X 6 Q
1
H HB Q 1T X
ekT kTn1

kT
n 1!
nQ 1
C G B V HB T G BPr:X 6 Q
1
H HB Q 1 X
ekT kTn

k
n!
nQ 2
LetgQ V HB T G BPr:X 6 Q
22
H HB Q 1
Pr:X P Q 2:
k
Then, from expression (19), QTM is the smallest value of Q for which
C + G + B + g(Q) < 0.
H HB Q 1
Pr:X P Q 2 < 0
k
H HB Q 1
Pr:X P Q 2
k
IfQ EM < Q TM ; then C G B gQ EM P 0:
So, the expressions (15) can be represented by the following expression (19).
But; P 6 C: This implies that P G B P C G B:

So; P G B gQ EM P C G B gQ EM P 0:
This implies that PC Q 1 PC Q

H HB Q 1
Pr:X P Q 2
k
19
This violates the expression (21). So, it is essentially inferred that

QEM P QTM, which encourages the buyer to order more in the collaborative environment than in the traditional procurement scenario.
With similar analysis, the expression (16) can be represented by the

following expression:
6. Development of algorithms for vertical collaboration

problem
PC Q 1 PC Q P G B V HB T G BPr:X
The analytical model for the vertical collaboration problem is

the foundation for the developing insights in the real-time procurement scenario in the semiconductor industry supply chain
dealing with short life-cycle components and products. So, the next
step is to develop algorithms for solving the traditional market
model and e-market models to generate the scheme for this collaborative initiative, and to gain research insights. The model has been
operated for several simulation runs to demonstrate decision policies for vertical collaboration at the operational and tactical level,
and to develop managerial implications for the semiconductor
industry supply chain planners. A ow chart depicting the appropriate marketplace dealing with various components/products
(Fig. 3) would offer structural and quantitative insight into the
interplay between vertical collaboration and joint replenishment
decision-making. The algorithm or the logical ow of simulation
for the traditional market model (no collaboration) and e-market
model (complete collaboration) is presented in the next section.
The vertical collaboration models for the buyersupplier procurement system [in traditional market (TM), hybrid market (HY),
and e-market (EM)] have only two controlling and variable parameters: (1) cycle time or planning horizon (T), and (2) ordering quantity (Q). It has been shown that the optimal ordering quantity
depends on the cycle time. This situation has complicated the estimation of expected prot of the SC system for short life cycle products. The expected prot for the SC system depends on these
parameters along with some other xed parameters, such as HB,
G, H, V, C, KB, KS, P, B, and k. These parameters were initialized
and loaded to the simulator to get values of the variable parameters for maximum expected prot of the SC. With the help of the
optimization tool, developed through a C program compiled by
using GNU compiler collection (GCC), this model has generated
the optimal ordering quantity (for e-market, traditional market
and hybrid market), and optimal planning horizon (T) for each
component/product. This will assist the SC planners to get a trigger
for the ordering quantity associated with each component/product,
and to decide the nature of the collaborative relationships for the
short life cycle products, whose cost changes with time. These
models, while operating in traditional and e-markets, would
facilitate individual enterprises to negotiate based on prot sharing (Bahinipati, 2009).
6 Q 1
H HB Q
Pr:X P Q 1
k
20
The value the optimal ordering quantity for the collaborative model
can be determined by solving these relations for various T values to
establish optimal (Q, T) combination for supply chain relationships.
5.4. Comparison between traditional marketplace and e-marketplace
In a centralized supply chain (e-marketplace), information is
open to all echelons and information ow is smooth, and there is
a central planner who makes all the decisions. Hence, optimal
decisions which maximize the system-wide supply chain prot
can be made. On the other hand, in a decentralized supply chain
(traditional marketplace), the entities are separate, and often each
considers its own prot only. In order to improve the supply chain
efciency, the supply chain entities can collaborate by setting an
appropriate contract. This contract which results in decisions by
individual parties that maximize the prot of the entire supply
chain and leaves each member of the chain satised is called a collaboration contract. This work has presented a two parameter relational contract in the form C(Q, T), where Q is the optimal ordering
quantity of the buyer to the supplier and T is the optimal length of
the planning horizon until which the buyer decides to be in the
market for traditional market, hybrid market and e-market. This
contract will help to determine the nature of relationships between
the members based on the transaction cost, i.e., based on the total
expected prot of the stakeholders or the entire SC. This contract
will ensure successful procurement and operational business
practices for long-term relationship in terms of resource sharing,
knowledge sharing, information sharing, and prot and risk
sharing.
In the collaborated e-marketplace, where there is complete collaboration between buyer and supplier, QEM is the smallest value of
Q that satises
P G B V HB T G BPr:X
6 Q
H HB Q 1
Pr:X P Q 2 < 0
k
21
515
Fig. 2. Flow chart depicting the overall logic of the proposed vertical collaboration scheme.
6.1. Simulation results and discussion

A buyersupplier procurement system in the traditional
marketplace, hybrid marketplace and e-marketplace (private
exchange) has been considered with various parameter values
(C, V, B, P, KB, KS, HB, G, H, k, and T). In the following, ows of
simulation in these marketplaces for decision making have been
presented.
(A) The logical ow of simulation for the TM scenario can be
outlined as follows:
Step A.1: Determine the expected total prot function of the

buyer PB(Q) in terms of Q and T by using Expression (9).
Step A.2: Solve the above expression to determine the optimum
QTM value for maximum PB(Q) value, where T = 1, 2, 3,. . .
Step A.3: Estimate the PB(Q) value for optimal QTM at a particular
T value.
Step A.4: Tabulate the results for each QTM and T combination.
Step A.5: Determine the QTM and T combination at which PB(Q)
is maximum, which proposes the optimal contract C(Q, T) in the
traditional marketplace.
516
Table 4
Diminishing growth rate of ordering quantity and concavity of prot.
T
QTM
QEM
PTM(QTM)
PEM(QEM)
1
2
3
4
5
6
7
8
9
10
9
17
24
31
37
43
47
49
50
50
11
19
27
35
42
48
54
59
62
63
44.788
120.101
179.983
227.220
259.100
279.820
284.960
277.750
265.170
249.190
47.392
123.202
185.086
232.887
266.908
287.694
296.118
293.637
282.985
268.104
% Increase
in QTM
% Increase
in QEM
88.89
41.18
29.17
19.35
16.21
9.30
4.25
2.04
0
72.73
42.10
29.63
20.00
14.28
12.50
9.26
5.08
1.61
Input data set 1: C = 8, V = 3, B = 2, P = 5, KB = 18, KS = 20, HB = 0.4, G = 18, H = 1.5,

k = 8.
Fig. 3. Flow chart for selection of most appropriate marketplace.
(B) The logical ow of simulation for the HY scenario can be outlined as follows:
Step B.1: Repeat the Steps 14 as in TM scenario.
Step B.2: Estimate the PS(Q) value for each QTM and T combination by using expression (3.3).
Step B.3: Estimate the total prot PTM(QTM) = PB(QTM) + PS(QTM),
and tabulate the results for each QTM and T combination.
Step B.4: Determine the optimal QTM and T combination at
which PTM(QTM) is maximum, which proposes the optimal contract C(Q, T) in the HY scenario with parameters QHY and T.
(C) The logical ow of simulation for the EM scenario can be
outlined as follows:
Step C.1: Determine the expected prot function of the collaborated system PC(Q) in terms of Q and T by using Expression (10).
Step C.2: Solve the above expression to determine the optimal
QEM for maximum PC(Q), where T = 1, 2, 3,. . .
Step C.3: Estimate the PC(Q) value for optimal QEM at a particular
T value.
Step C.4: Tabulate the results for each QEM and T combination.
Step C.5: Determine the optimal QEM and T combination at
which PEM(QEM) is maximum, which proposes the optimal contract C(Q,T) in the EM scenario with parameters QEM and T.
The preliminary simulation runs for different xed parameter
values and order arrival rates (with the use of input data set 1) resulted in the following inferences:
1. Irrespective of TM or EM, and cycle time and demand arrival
rate, the expected prot values show concavity (Table 4). The
expected prots in e-markets and traditional markets shows
concavity while plotted against cycle time for all values of
demand arrival rate.
2. Irrespective of the demand pattern and cycle time, the expected
prot margin for operating in e-markets is higher than that
operating in traditional markets. Further, the expected prot
margin in both the cases increase with higher demand arrival
rate (k).
3. For a given demand pattern, the ordering quantity increases
with cycle time and stabilizes at higher cycle time (Table 4).
Further, each row in the Tables 57 compares the objective

functions values in the traditional marketplace, and e-marketplace
models for a given parameter values. Hence, it can be concluded
from the analysis of these simulation results that the cycle time, instead of ordering quantity, is a critical parameter for negotiation
during vertical collaboration. Further, it can also be inferred that
the maximum expected prot of the SC is dependent on cycle time
provided that the ordering quantity is not highly constrained.
It can be noted that the expected prots of e-marketplace and
traditional marketplace models are dependent on T. Further, the
difference in the optimal system prots [PEM(QEM)PTM(QTM)] increases as T increases, for a given input data set. So, more efcient
collaboration mechanisms can be designed, where T can be considered as a negotiable parameter between the buyer and supplier.
This has been exhibited in the following three examples for effective managerial decision.
Considering a traditional market (TM) with both the ordering
quantity (Q) and the length of the planning horizon (T) as the negotiable parameters, the buyer would want to order 43 SKU, and
would be in the market for T = 6 TU (refer Table 5). In this case,
the buyer would want maximize his own prots [PB
(QTM) = 170.82 MU]. However, the system prot would be maximized in hybrid market (HY) at T = 7 TU and QTM = 47 SKU, in
which case, the buyer would lose [170.82163.96 = 6.86 MU], but
the system would gain [284.96279.82 = 5.14 MU]. The cost savings from this solution (i.e., 5.14 MU) cannot be used to compensate the buyer as well as to increase the expected prots of the
buyer and supplier. Further, the collaborated system (EM) prots
would be maximized at T = 7 TU and QEM = 54 SKU, in which case,
the buyer would lose 16.7 MU, but the system would gain 16.298
MU. The cost savings from this solution also cannot be used to
compensate the buyer as well as to increase the expected prots
of the buyer and the supplier. The managerial decision in such a
case is to opt for traditional marketplace (TM) and there is no collaboration (NC) between the buyer and supplier.
In the second case (Table 6) the buyer would want to order 32
SKU, and would be in the market for T = 6 TU. In this case the buyer
would want to maximize his own prots [PB (QTM) = 120.775 MU].
However, the system prot would be maximized in hybrid market
(HY) at T = 7 TU and QTM = 35 SKU, in which case, the buyer would
lose 4.584 MU, but the system would gain 4.416 MU. The cost savings from this solution (i.e., 4.416 MU) cannot be used to compensate the buyer as well as to increase the expected prots of the
the buyer would lose 11.266 MU, but the system would gain
12.734 MU. The cost savings from this solution can be used to compensate the buyer as well as to increase the expected prots of the
buyer and the supplier. The managerial decision in such a case is to
517
Table 5
Prot comparison for optimal time period for input data set 1.
T
QTM
QEM
PBTM
PTM(QTM)
PBEM
PEM (QEM)
PEM (QEM) PTM(QTM)
1
2
3
4
5
6
7
8
9
10
9
17
24
31
37
43
47
49
50
50
11
19
27
35
42
48
54
59
62
63
37.788
89.101
127.983
154.22
168.10
170.82
163.96
150.75
135.17
119.19
44.788
120.101
179.983
227.22
259.10
279.820
284.96
277.75
265.17
249.19
34.392
86.202
124.086
147.89
160.91
163.69
154.12
136.64
116.99
99.10
47.392
123.202
185.086
232.887
266.908
287.694
296.118
293.637
282.985
268.104
2.604
3.101
5.103
5.667
7.808
7.874
11.158
15.887
17.815
18.914
Input data set 1: C = 8, V = 3, B = 2, P = 5, KB = 18, KS = 20, HB = 0.4, G = 18, H = 1.5, k = 8.
Table 6
T
QTM
QEM
PBTM
PTM(QTM)
PBEM
PEM (QEM)
PEM (QEM) PTM(QTM)
1
2
3
4
5
6
7
8
9
10
7
13
18
23
28
32
35
37
37
37
9
15
21
26
32
36
40
44
46
47
21.870
59.617
88.393
107.865
118.456
120.775
116.191
106.832
95.340
83.381
22.870
78.617
122.393
156.865
182.456
196.775
201.191
197.832
186.340
174.381
17.557
55.994
83.872
104.270
111.540
115.096
109.509
96.062
82.549
68.675
24.557
80.994
126.872
162.270
187.540
203.096
209.509
208.062
200.549
189.674
1.687
2.377
4.479
5.405
5.084
6.321
8.318
10.230
14.209
15.293
Input data set 2: C = 8, V = 3, B = 2, P = 5, KB = 18, KS = 20, HB = 0.4, G = 18, H = 1.5, k = 6.
Table 7
T
QTM
QEM
PBTM
PTM(QTM)
PBEM
PEM (QEM)
PEM (QEM) PTM(QTM)
1
2
3
4
5
6
7
8
9
12
22
31
40
48
55
61
63
63
15
25
35
45
54
62
69
75
79
46.232
100.825
133.827
144.977
135.059
105.668
59.886
3.675
56.108
69.232
163.825
232.827
279.977
302.059
300.668
278.886
230.675
170.892
38.007
94.232
126.084
133.944
122.233
91.868
44.121
18.438
87.173
73.007
169.232
241.084
288.944
313.233
314.868
295.121
256.561
203.826
3.775
5.407
8.257
8.967
11.174
14.200
16.235
25.886
32.934
Input data set 3: C = 10, V = 5, B = 6, P = 6, KB = 20, KS = 25, HB = 0.5, G = 20, H = 2, k = 10.
opt for electronic marketplace (EM) and there is complete collaboration (CC) between the buyer and supplier.
In the third case (Table 7), the buyer would want to order 40
SKU, and would be in the market for T = 4 TU. In this case, the buyer
would want to maximize his own prots [PB (QTM) = 144.977 MU].
However, the system prot would be maximized in hybrid market
(HY) at T = 5 TU and QTM = 48 SKU, in which case, the buyer would
lose 9.911 MU, but the system would gain 22.082 MU. The cost
savings from this solution (i.e., 22.082 MU) can be used to compensate the buyer as well as to increase the expected prots of the
the buyer would lose 53.109 MU, but the system would gain
34.891 MU. The cost savings from this solution cannot be used to
compensate the buyer as well as to increase the expected prots
of the buyer and the supplier. The managerial decision in such a
case is to opt for a hybrid marketplace (HY) and there is partial collaboration (PC) between the buyer and supplier.
The insight from the above analysis is that as T increases, the
collaborative system prot increases. But the buyer loses more in
e-market than in traditional market. When the system cost advantage (in terms of prot) by participating in an e-market can be used
to compensate for the buyer losses, the buyer would participate in
collaborative relationships (CC). When the system cost advantage
by participating in a hybrid market can be used to compensate
for the buyer losses, the buyer would participate in partial collaborative relationships (PC). Otherwise, the buyer and supplier would
be engaged in independent decision making leading to no collaborative relationships (NC).
The analysis of the simulation results for the vertical collaboration problem can also be depicted by various characteristic plots.
The relationship between the expected prot of the buyer (in
TM) and the expected prot of the supply chain systems (in both
HY and EM) with the planning (time) horizon and the optimal
ordering quantities have been depicted in the Figs. 4ac and Figs.
5ac respectively. Further, the relationship between optimal ordering quantity and the planning (time) horizon have been presented
in Fig. 6ac for various scenarios. These proles are indicators of
the performance of the supply chain under the characterization
of various collaborative schemes dened in the present study,
518
Fig. 4a. Relationship between planning horizon and expected prots (data set 1).
Fig. 4b. Relationship between planning horizon and expected prots (data set 2).
Fig. 5a. Relationship between optimal ordering quantity and expected prots (data
set 1).
Fig. 5b. Relationship between optimal ordering quantity and expected prots (data
set 2).
Fig. 4c. Relationship between planning horizon and expected prots (data set 3).
which leads to various managerial implications for the procurement professionals of the semiconductor industry supply chain.
The prot functions are also solved for three sets of data, and in
each set, for different order arrival rates (k). The comprehensive
results with associated managerial decisions are depicted in
Tables 810. Further, some analysis has been made for data sets
3.1, 3.3, and 3.4 depicted in Table 10 regarding the optimal decision for collaboration. For data set 3.1, two managerial decisions
are feasible. For the complete collaboration (CC) case with optimal
T = 9, the buyers loss of 11.91 MU can be compensated from overall
system prot of 12.01 MU. So, the net system prot would be
(12.0111.91 = 0.10 MU). For the partial collaboration (PC) case
with optimal T = 8, the buyers loss of 0.64 MU can be compensated
from overall system prot of 7.36 MU. So, the net system prot
would be (7.360.64 = 6.72 MU). So, the optimal managerial
decisions would be the selection of hybrid marketplaces (HY) with
partial collaboration (PC). Similar analysis can be made for the data
sets 3.3 and 3.4 in Table 9, and the optimal managerial decisions
Fig. 5c. Relationship between optimal ordering quantity and expected prots (data
set 3).
would be the selection of e-marketplaces (EM) with complete collaboration (CC) in both the cases.
In real-world semiconductor industries, shorter T is preferred
by both buyer and supplier. This is due to the following. The product is usually set at a xed price (through price negotiations) well
before the R&D has been able to go through the process. So, with a
xed price and shorter T, the individual gain of the buyer and supplier may be high. This strategy may be suited for faster-to-market
of the product. However, due to short life cycle of the product in
this industry, the buyer would be interested to select the optimum
T, as he is investing on the supplier to develop a strategic relationship. The optimum T is selected in such a manner that the systems
gain can be used to compensate for the loss of the buyer. An incentive scheme would be in place for equitable sharing of excess system gain between buyer and supplier, which is the most attractive
feature of this model. This may lead to a situation in which the
519
competition in the marketplace, the present model can be suitably

modied to consider the aspect of securing sustainable stability.
This is beyond the scope of the present work.
6.1.1. Insights
The analytical framework proposes a number of insights, which
are as follows:
Fig. 6a. Relationship between planning horizon and optimal ordering quantity
(data set 1).
Fig. 6b. Relationship between planning horizon and optimal ordering quantity
(data set 2).
(a) It is observed from the literature that the suppliers gain

from the e-market model solution is at least equal to the
buyers loss from the traditional market model solution. This
is the basis for vertical collaboration. This requires that the
solution of the traditional market model be coordinated in
such a manner that it results in the same outcome for the
decision variables, as does the solution of the e-market
model.
(b) Vertical collaboration problem focuses on improving the
total expected SC system prots in a traditional market
model and minimizes the gap with the e-market model.
In the traditional market model, the sub-problems are
solved sequentially. Whether the buyer or the supplier
solves his sub-problem rst depends on the party, which
dominates the systems.
As the e-market model maximizes the expected system
prots, its objective function value is an upper bound
on the total expected prots of collaborated buyersupplier system.
(c) The e-market model can be used as a benchmark and the gap
between the prots in the traditional market model and emarket model can considered as an inducement to improve
the solution of traditional market model.
(d) The e-market model can present a centralized approach to
maximize prots (collaboration scheme), and a mutually
agreeable way of sharing prots. The prot sharing can be
negotiable between the parties or forced by one SC member
to inuence the behavior of the other member.
6.1.2. Managerial Implications
The analysis leads to a number of managerial insights, which
are as follows:
Fig. 6c. Relationship between planning horizon and optimal ordering quantity
(data set 3).
OEM may revise the price of the product to increase the market
share.
In summary, the semiconductor industries has to set incentives
to create a situation, in which all the supply chain members has
self-interest to secure the sustainable stability of the entire network. On one hand, these incentives must be of monetary nature
to create a short-term winwin situation (higher prots). On the
other hand, the incentives have to be of non-pecuniary nature to
create unique long-term relationship proposition, which cannot
be imitated easily by competitors (Gulati, Lawrence, & Puranam,
2005). This might result in higher prots or joint growth opportunities in the future. The collaboration in supply chain network relies on condence and understanding. These characteristics have to
develop over time to achieve a superior joint solution of the problem. As the present work has not considered the perspective of
The suppliers expected prot function is increasing with the

buyers ordering quantity Q. This is important in characterizing
the general structure of the mechanism for collaborated supply
chain.
It is always advantageous for the supplier to receive larger order
from the buyer to achieve vertical collaboration. In fact, the
foundries/contract manufacturers are producing bigger wafers
to generate a large number of chips to minimize the system
cost, and to achieve vertical collaboration. This is due to the fact
that a discount scheme that offers a unit discount for order sizes
greater than or equal to QEM can be used by the contract manufacturers (suppliers) to inuence the ordering behavior of the
buyers.
The collaboration effort would be more effective, if the length of
the planning horizon or the time until a new product is
launched in the market (T) is considered as a variable to negotiate a contract between buyer and supplier. This time is an
important component of the pricing strategy in semiconductor
industry supply chain.
The present work has focused on demand which is independent
on the selling price. However, when the product attains maturity in its lifecycle, the demand is dependent on the selling
price, which in turn depends on the time. So, time dependent
incentives may be offered by the supplier to change the ordering behavior of the buyer.
520
Table 8
Managerial decision-making concerning collaboration for input data set 1.
SET 1
Optimal Q (SKU)
PB(Q) MU
PSYS(Q) MU
Optimal time
period T (TU)
Buyers loss
(DPB) MU
System prot
(DPSYS) MU
1.1 Order arrival rate k = 6 SKU

TM
24
HY
29
EM
38
72.450
66.532
38.159
143.450
157.532
165.159
4
5
6
5.918
34.291
14.082
21.709
1.2 Order arrival rate k = 8 SKU

TM
32
HY
44
EM
50
108.472
77.469
64.817
211.472
228.469
239.817
4
6
6
31.003
43.655
16.997
28.345
1.3. Order arrival rate k = 10 SKU

TM
40
HY
48
EM
62
144.977
135.059
91.868
279.977
302.059
314.868
4
5
6
9.918
53.109
22.082
34.891

TM
48
HY
58
EM
74
181.822
169.807
119.199
348.822
376.807
390.199
4
5
6
12.015
62.623
27.985
41.377

TM
60
HY
72
EM
92
237.554
222.342
160.362
452.554
485.342
503.362
4
5
6
15.212
77.192
32.788
50.808
Managerial
decision
PC
T=5
NC
T=4
PC
T=5
PC
T=5
PC
T=5
Input data: C = 10, V = 5, B = 6, P = 6, KB = 20, KS = 25, HB = 0.5, G = 20, H = 2.
Table 9
SET 2
Optimal Q (SKU)
PB(Q) MU
PSYS(Q) MU
Optimal time
period T (TU)
Buyers loss
(DPB) MU
System prot
(DPSYS) MU

TM
32
HY
35
EM
40
120.775
116.191
109.509
196.775
201.191
209.509
6
7
7
4.584
11.266
4.416
12.734

TM
43
HY
47
EM
54
170.82
163.96
154.12
279.82
284.96
296.118
6
7
7
6.86
16.7
5.14
16.298

TM
54
HY
60
EM
68
221.111
211.801
198.935
363.111
371.801
382.935
6
7
7
9.31
22.176
8.69
19.824

TM
65
HY
72
EM
81
271.546
259.718
247.049
446.546
455.716
470.049
6
7
7
11.828
24.497
9.17
23.503

TM
82
HY
82
EM
90
347.244
347.244
335.828
573.244
573.244
585.828
6
6
6
0
11.416
0
12.584
Managerial
decision
CC
T=7
NC
T=6
NC
T=6
NC
T=6
CC
T=6
Input data: C = 8, V = 3, B = 2, P = 5, KB = 18, KS = 20, HB = 0.4, G = 18, H = 1.5.
6.2. Signicance of cycle time (T)

In semiconductor industry, the buyer (OEM) is a product technology leader so that he does not carry an inventory of the component at the end of each planning horizon or product life cycle since
the component is marginally improved after this time period.
Within each contract review period, the buyer makes replenishment. At the beginning of the planning horizon, the buyer rm decides to choose either long-term or short-term suppliers and keeps
the choice of suppliers intact until the end of planning horizon. The
use of these two extreme policies will help us obtain robust analytical results.
The buyer rm, facing dynamics and uncertainty in price and
demand for its major component, must make its purchasing plan.
The model is used to compare the effect of sourcing strategy alternatives for components on the total cost (purchasing, inventory,
shortage and obsolescence costs) incurred by the supply chain
members over the length of its entire planning horizon. Over the
planning horizon, the buyer rm could use either short- or longterm suppliers. Long-term contracts cover at least 1824 months
and include provisions for multiple shipments consistent with
the replenishment policies of the manufacturer. In contrast,
short-term contracts comprise much shorter periods (typically 3
6 months), with provisions for multiple shipments during the contract period. The choice of short-term suppliers and contract prices
are based on a periodic search process initiated by manufacturers.
In particular, it evaluates the cost and benets of short term and
long term contracts for input components.
521
Table 10
SET 3
Optimal Q (SKU)
PB(Q) MU
PSYS(Q) MU
Optimal time
period T (TU)
Buyers loss
(-DPB) MU
System prot
(DPSYS) MU

TM
38
HY
42
EM
50
170.40
169.76
158.41
231.40
238.76
243.41
7
8
9
0.64
11.91
7.36
12.01

TM
52
HY
56
EM
66
234.99
233.53
219.97
323.99
330.53
336.97
7
8
9
1.46
15.02
6.54
12.98

TM
65
HY
71
EM
77
299.902
294.455
291.605
415.902
424.455
430.605
7
8
8
5.447
8.297

TM
78
HY
85
EM
93
364.947
359.577
351.865
505.947
514.577
522.865
7
8
8
5.37
13.082
8.63
16.918

TM
87
HY
97
EM
104
451.549
434.315
427.377
610.549
613.315
620.377
6
7
7
17.234
24.172
2.766
9.828
8.5533
14.703
Managerial decision
CC (PC)
T = 9 (8)
PC
T=8
CC (PC)
T = 8 (8)
CC (PC)
T = 8 (8)
NC
T=6
Input data: C = 6, V = 2, B = 1.5, P = 4, KB = 12, KS = 15, HB = 0.3, G = 16, H = 1.2.
6.2.1. Long-term suppliers

Long-term relationships with a small number of suppliers
have a number of advantages. It can facilitate easier and more
frequent communication and information exchange, more effective monitoring of production processes, economies of scale in
production, elimination of expenses associated with frequent
rebidding for contracts, early supplier involvement (for faster
product development), and mutual cooperation (for order fulllment). Therefore, long-term contracts can lead to reduction in
purchase prices and improvement in delivery lead time performance, over time, as a consequence of mutual efforts by the
buyer and the supplier.
The long-term contract covers the whole planning horizon of T,
including provisions for multiple shipments consistent with the
replenishment policies of the buyer rm. Once a long-term contract is signed, the buyer could purchase as many components as
actual demand requires. When the market price goes down, the
buyer rm has the exibility to purchase the minimum commitment specied in the long-term contract and purchase the rest
from the market. This means that strategically, when the price is
predicted to vary signicantly in the future, as is the case of a semiconductor manufacturing industry, choosing a long-term supplier
(extending over the entire product life cycle of approximately
18 months) could help the buyer rm to hedge against the risk of
incurring high purchasing costs.
Typically, long term contracts allow risk-averse decision makers to hedge against price uncertainty for components by specifying a xed price for the duration of the contract. In order to
motivate the long-term supplier to upgrade the component, when
signing a long-term contract, the buyer rm provides the supplier
a xed relation-specic investment, as an incentive, for the entire
duration of the contract. This investment may include the effort
and cost of performing a detailed technical and nancial audit
of the supplier, technical and nancial support, investment in
equipment, and in some instances, acquisition of a share in the
suppliers equity. Thus, signing a long-term contract for the supply of the component will guarantee stability and reliability in the
light of the fact that the component price can be quite dynamic
and erratic due to market competition and to volatility in supply
and demand.
6.2.2. Short-term suppliers

Short planning horizons tend to favor short-term contracts and
discourage switching to long-term contracts. If sourcing planning
is done for short planning periods, it is less likely for the supply
managers to select long-term contracts. This is applicable to the
highly innovative components or products which have very small
product life cycles. During periods of rapid technological change
in semiconductor industries, reductions in length of planning horizon can be viewed as the means to reduce long-term planning risk
associated with component and product obsolescence. So, emphasizing planning horizon or time period (T) perspective for collaborative relationships may allow for a convergence of planning and
improvisation in new product development that would enable supply chain members to reduce the response time for new product
design to development to market introduction. This would directly
encourage the marketing personnel for close involvement with the
product or component developers and suppliers, and would accelerate time to market, which may be critical for achieving and
maintaining competitive advantage. The short-term contracts usually require enterprises to pay the market price for components.
Thus, they offer a speculative advantage as well as the exibility
to switch to other suppliers.
Supply managers may act optimally by not investing in long
term contracts. This would be possible if the business environment
favors short-term contracts to dominate over long-term contracts.
In the present study on focus group of supply managers from eight
leading semiconductor manufacturing companies, it was indicated
that the average contract length was only 18 months, and even
these were reviewed frequently. Almost 25% of the contracts were
shorter than 6 months, and few were longer than 36 months. The
study also indicated that the process of vendor qualication is
tedious, long (up to 25 weeks) and expensive. The contract
negotiations can sometimes take up to 60 weeks. Thus, the cost
of entering into a long-term contract is perceived to be high.
Further, the supply managers often perceive a great degree of
uncertainty in component prices and technology changes (especially in semiconductor industries, where the product life cycle is
approximately 18 months). These conditions will tend to favor
short-term contracts in semiconductor industries dealing with
short life cycle products and components.
522
If the buyer rm decides to choose a short-term supplier over

the planning horizon, the buyer rm avoids being locked into a
long-term contract and is able to take the advantage of the best
price offered on the spot market in each time period. However,
to nd the best price in each period, the buyer rm has to conduct
a market search to determine the likelihood of the best price on the
spot market that is independent of the demand distribution of the
component and to complete the supplier qualication process before a supplier is chosen. The prices of short-term suppliers are
random variables and depend on changes in demand and supply
in the market, supplier competition, and developments in the
product technology. The buyer rm purchases the exact amount
of realized demand from the short-term supplier after the demand
and the price are realized. Therefore, there must be a trade-off between long-term and short-term suppliers depending on the type
of product or component in semiconductor manufacturing
industry.
6.2.3. The short-term versus long-term supplier strategy
In the present analysis, T is the total number of time periods
(usually in quarters or three months) in the planning horizon with
t as the index for each period, T = 1, 2, ..., T. The planning horizon
for the problem is T strategic review periods. Reasonable values
for T are 1824 months. Each strategic review period consists of
n tactical review periods (e.g., each tactical review period is
3 months long and there are 6 of them in a 18 months strategic review period, T). It may be noted that one Time Unit is the tactical
review period of 3 months. Since all periods in the planning horizon are independent from each other, this work calculated the
optimal ordering quantity (for both traditional market and e-market) for each period independently considering the mean demand
arrival rate (k) with a various sets of xed input data.
The purchasing system is sensitive towards T. This has been
clearly reected in the simulation results. The objective is to nd
the optimal (Q, T) combination at which prot functions are optimal for a no-collaboration (NC), partial collaboration (PC) and complete collaboration (CC) situations. These combinations can be
termed as system optimal (Q, T) combination for a given set of input data. So, the supply managers must critically consider the
selection and implementation of a particular T value in the contract. Often, T is extended beyond the life cycle of the product to
consider the future requirement during phase out period, before
complete withdrawal from the market. Further, the short life cycle
of semiconductor industry is always associated with supplier
competition. This may result in shorter T. However, to gain
competitive advantage, the competing suppliers must resort to
horizontal collaboration to counter the effect of volatile market
demand and obsolescence in the system. This collaborative
network of suppliers would result in an optimal selection of (Q,
T) combination at which the prot function of the supply chain system would be optimum.
The present model quanties the tradeoffs between short-term
supplier and long-term supplier. It is proved through analytical
and quantitative means that long-term contracts may not always
be optimal and developed conditions under which short-term contracts will be preferred for cost advantage. The decision to enter
into long-term contracts is highly dependent on the nature of the
markets and the decision makers risk attitudes. When the market
conditions for components and products are evolving (non-stationary) over the length of the planning horizon, long-term contracts
may not be optimal right at the beginning of the planning horizon
even for risk-averse OEMs. It may be optimal to delay the decision
to switch to the long-term contract until more price realizations
have been observed. Thus, a combination strategy may be best suited for the short-lifecycle components and products in the semiconductor industry supply chain.
6.3. Signicance of vertical collaboration

The analytical framework for vertical collaboration has certain
implications for the semiconductor industry, which are as follows:
1. This relationship is critical to determine the positioning of the
industry for long-term survival in terms of supplier involvement for innovative product development and market share.
2. This collaborative approach establishes a perfect balance
between the industries technological development and strategic orientation to facilitate acquisition of resources and core
competencies, and allocation and fulllment of order among
compatible supply chain members.
3. The vertical relationship focuses on developing conducive organizational climate and managerial involvement apart from routine operational and nancial involvement from all the
partnering organizations.
4. The vertical collaborative initiative requires strong internal
capabilities and strategic coherence with partners business
objectives and policies, for strategic outsourcing and long-term
survival.
5. As new technologies increase in complexity due to product
innovation, the enterprises must invest on suppliers and
acquire external capabilities to complement its internal core
competencies.
In order to maximize procurement business output through
investment on resource creation, collaborative initiatives be organized in such a manner that all the SC members should be both
responsible and autonomous. At the same time, the partners must
be united by a cohesive structure. The success of such a structure
requires that the managers be focused to partner rms nancial
and resource involvement and their motivation to acquire desired
core competencies.
7. Perspectives and contributions
The major contribution of the present work is the characterization and analysis of a critical aspect of the supply chain collaboration
based on the vertical relationships among stakeholders, which is
lacking in the existing literature. The present work has attempted
to redene collaboration from the perspective of procurement practices in the semiconductor industry supply chain (Bahinipati, 2009).
Collaboration creates a synergistic business environment which
needs more than partnerships and aligned interests, and where
the enterprises leverage each other on an operational basis so
that the joint performance is better than the individual
performance.
Further, the proposed model contributes an analytical framework and a decision support approach for procurement professionals and practitioners in the semiconductor industry supply chain.
This framework may be the starting point for tackling operational
and tactical issues to develop strategic relationships between
buyer(s)-supplier(s). In this regard, the most signicant contribution of this work to the body of literature is that the collaborative
ordering motivates the buyers for larger orders from the suppliers
covering the entire life cycle of the product. Secondly, the vertical
collaborative relationship facilitates a contract in which the length
of the planning horizon is a negotiable parameter between buyer
and supplier for strategic cost and competitive advantage. Finally,
the analysis presents an alternate denition of the vertical collaborative relationships to the supply chain literature along with their
characteristics from the perspective of procurement in an e-market
mode of operation (Bahinipati, 2009).
Vertical collaboration refers to a hybrid collaboration strategy

based on technology (e-market, and IT), time (cycle time, pull
process), resource productivity (inventory reduction, and facility, resource and equipment utilization) and relationships
(between buyers and suppliers) by sharing, interchanging and
investing in mutual plans, information and resources, and providing mutual visibility to change behavior, where the procurement transactions can be automated through decision making
support of e-market intermediary by signicantly improving
the prot value.
The characteristic features of the relationship based on vertical
collaborative initiatives can be summarized as follows, which
would provide suggestions and guidelines to the practitioners in
the semiconductor industry supply chain.
(a) The collaborative relationships are mutual agreements based
on trust, cooperation, and shared risks and investments.
(b) The e-market enhances transactions capability by identifying and selecting new suppliers, establishing prices for products, components and materials, and aligning procurement
of materials/components from suppliers.
(c) The e-market enhances the collaborative capability by communicating delivery request to suppliers, communicating
manufacturing requirements to suppliers, and managing
and communicating engineering changes.
(d) Use of e-market, internet and other IT infrastructure based
systems enhance the speed of operation to compress the
cycle time.
(e) The enterprises increase customer satisfaction through better visibility of the procurement process and reduction of
errors.
523
(f) The partnering enterprises must have good cost information,

which is critical for selecting manufacturing locations as
well as developing plans for procurement.
(g) This relationship incorporates more than shared information
and a focus on SC cost, such as SC planning strategies and
schemes among partners.
(h) Better management control can be realized due to embedded business schemes in the procurement system by
encouraging outsourcing and allow the enterprises to
engage in e-business to have more control of their operations by sharing customer-specic information.
(i) Cost saving can be realized through less costly and optimized processes for procurement and lower cost indirect
goods by use of e-market.
In summary, the supply chain relationship diagram (Fig. 1) can
be suitably modied by incorporating the proposed vertical collaborative relationships, which is depicted in Fig. 7.
8. Concluding remarks
The present work aims at developing generic analytical modeling frameworks for vertical collaboration initiatives in the semiconductor industry supply chain dealing with short lifecycle
products where the demand is independent of the selling price.
The procurement process has been investigated by characterizing
the market space into three different types, such as traditional
marketplace, hybrid marketplace and e-marketplace. This process
has explored the conditions in which vertical collaboration works,
and has been extended through simulation to comprehensively assess the nature of relationship with individual supply chain mem-
Fig. 7. Proposed relationship between members in a semiconductor supply chain (pull system). (see above-mentioned references for further information.)
524
bers and to evaluate whether such an initiative is really feasible. It

has also investigated the perspectives of information sharing in
collaborative networks and its implementation in the semiconductor industry supply chain. The results of the work indicate that the
proposed approach results in fair and reasonable conclusions. The
outcome of this analysis is to provide an effective method for the
semiconductor industry supply chain members to evaluate the
success of such shared collaborative procurement systems in a
structured and simple manner. So, this research is signicant because it provides a comprehensive vertical collaboration assessment tool for operational and tactical decision making.
The contributions to the literature on supply chain collaboration resulted from this study are: (1) Collaborative efforts among
chain members should be encouraged in order to improve operational and tactical performance. (2) As the SC members have varied
perception about high and low performing collaborative practices,
it may be recommended that the chain members should create a
negotiation mechanism that encourages the vertical collaborative
practices to achieve better performance.
The proposed method may be equally useful in analyzing and
assessing the success of any collaborative procurement in other
similar industries. The applicability of the proposed schemes for
the collaborative initiatives may be different for different industries. However, the proposed schemes may present general frameworks for such initiatives in most industries dealing with short life
cycle products. So, the chain members need to modify and customize the approaches for collaborative practices that suit their unique
environment under which they operate.
As the expectation from the supply chain may be different for
different products and industries, which poses the major limitation
of the present work. Other limitations of this study are as follows:
(1) The study has not considered the competition from the similar
SC members from a global perspective, and (2) The study did not
consider the management structures of collaborative procurement
initiative from a global perspective. However, it seems that only
the threat of global competition can motivate the formation of
large-scale vertical collaboration initiative in industries. These limitations lead to several directions for future research as depicted
below.
(a) From the perspective of vertical collaboration, this research
has drawn conclusion from dyadic relationships only, i.e.,
buyersupplier. So, extending this research focus to more
complex supply chain relationships across the entire chain
would be useful. Further, some quantitative testing of the
model in similar industries is needed with the aim of theoretical replication.
(b) Customer order arrival rate plays an important role in the
vertically collaborated supply chain system. The present
work has assumed the order arrival as a Poisson process.
With the increase in the customer orders, in many situations, the classical Poisson process order arrival may not
be appropriate. In reality, the order arrival rate is typically
non-homogeneous, i.e., the arrival rate typically varies with
time. This occurs due to the customer market in which
demand is dependent on price. Analysis of such situations
in the semiconductor industry supply chain dealing with
short life cycle products would be an important area of
future research.
(c) In reality, most of the suppliers have the right to set their
own selling price, which can directly affect the market
demand. Thus, in a price-endogenous setting, the decision
which maximizes the expected prot should be determined
jointly with respect to the inventory level (ordering quantity) and the selling price. Because of the complexity of the
decision based on both the price and ordering quantity, the
collaboration in the case of the price-endogenous scenario

has not been fully studied, which creates scope for further
research.
In summary, the present study has investigated the operational
and tactical decision policies for vertical collaboration problems in
semiconductor industry supply chain. These policies would offer
structural and quantitative insight into the interplay between vertical collaboration and joint replenishment decision-making. These
policies may also offer insights into supplier selection problem for
such a collaborative initiative. In reality, a mixed-mode collaborative relationship is highly desirable where the buyer collaborates
with a network of competing supplier for resource and capacity
sharing, which is the scope for future research.
Acknowledgment
The authors are grateful to the reviewers and the area editor for
their valuable comments and suggestions, which led to signicant
improvement in the manuscript.
References
Anupindi, R., & Bassok, Y. (1999). Centralization of stocks: Retailers vs.
manufacturer. Management Science, 45(2), 178191.
Bagchi, P. K., Ha, B. C., Skjoett-Larsen, T., & Soerensen, L. B. (2005). Supply chain
integration: A European survey. The International Journal of Logistics
Management, 16(2), 275294.
Bahinipati, B. K. (2009). A framework for collaborative procurement. Ph. D. Thesis,
IIT Delhi, New Delhi, India.
Bahinipati, B. K., Kanda, A., & Deshmukh, S. G. (2009a). Coordinated supply
management: Review, insights and limitations. International Journal of Logistics:
Research and Applications, 12(6), 407422.
Bahinipati, B. K., Kanda, A., & Deshmukh, S. G. (2009b). Revenue sharing in
semiconductor industry supply chain: Cooperative game theoretic approach.
SADHANAAcademy Proceeding in Engineering Sciences, 34(3), 127.
Bahinipati, B. K., Kanda, A., & Deshmukh, S. G. (2009c). Horizontal collaboration in
semiconductor manufacturing industry supply chain: An evaluation of
collaboration intensity index. Computers and Industrial Engineering, 57(3),
880895.
Bailey, K., & Francis, M. (2008). Managing information ows for improved value
chain performance. International Journal of Production Economics, 111(1), 212.
Banerjee, A. (1986). A joint economic lot size model for a purchaser and vendor.
Decision Science, 17(3), 292311.
Banerjee, A., & Burton, J. S. (1994). Coordinated vs. Independent inventory
replenishment policies for a vendor and multiple buyers. International Journal
of Production Economics, 35(13), 215222.
Barratt, M. (2004). Understanding the meaning of collaboration in the supply chain.
Supply Chain Management: An International Journal, 9(1), 3042.
Barratt, M., & Oliveira, A. (2001). Exploring the experiences of collaborative
planning initiatives. International Journal of Physical Distribution and Logistics
Management, 31(4), 266289.
Barratt, M., & Rosdahl, K. (2002). Exploring business-to-business market sites.
European Journal of Purchasing and Supply Management, 8(2), 111122.
Boulaksil, Y., & Fransoo, J. C. (2009). Order release strategies to control outsourced
operations in a supply chain. International Journal of Production Economics,
119(1), 149160.
Brown, A. O., Lee, H. L., & Petrakian, R. (2000). Xiling improves its semiconductor
supply chain using product and process postponement. Interfaces, 30(4), 6580.
Cao, M., Vonderembse, M. A., Zhang, Q. Y., & Raghu-Nathan, T. S. (2010). Supply
chain collaboration: Conceptualization an instrument development.
International Journal of Production Research, 48(22), 66136635.
Cao, M., & Zhang, Q. (2010). Supply chain collaborative advantage: A rms
perspective. International Journal of Production Economics, 128(1), 358367.
Cao, M., & Zhang, Q. (2011). Supply chain collaboration: Impact on collaborative
advantage and rm performance. Journal of Operations Management, 29(3),
163180.
Centikaya, S., & Lee, C.-Y. (2000). Stock replenishment and shipment scheduling for
vendor managed inventory systems. Management Science, 46(2), 217232.
Centikaya, S., & Lee, C.-Y. (2002). Outbound transportation: Modeling inventory and
cargo capacity. Naval Research Logistics, 49(6), 531556.
Chen, J. V., Yen, D. C., Rajkumar, T. M., & Tomochko, N. A. (2011). The antecedent
factors on trust and commitment in supply chain relationships. Computer
Standards & Interfaces, 33(3), 262270.
Cheng, H. C., Chen, M. C., & Mao, C. K. (2010). The evolutionary process and
collaboration in supply chains. Industrial Management and Data Systems, 110(3),
453474.
Cheung, M.-S., Myers, M. B., & Mentzer, J. T. (2010). Does relationship learning lead
to relationship value? A cross-national supply chain investigation. Journal of
Operations Management, 28(6), 472487.
Choudhury, V., & Hartzel, K. S. (1998). Uses and consequences of electronic markets:
An empirical investigation in aircraft parts industry. MIS Quarterly, 22(4),
471503.
Christopher, M., Mena, C., Khan, O., & Yurt, O. (2011). Approaches to managing
global sourcing risk. Supply Chain Management: An International Journal, 16(2),
6781.
Constantino, N., & Pellegrino, R. (2010). Choosing between single and multiple
sourcing based on supplier default risk: A real options approach. Journal of
Purchasing and Supply Management, 16(1), 2740.
Cousins, P. D. (2002). A conceptual model for managing long-term interorganisational relationships. European Journal of Purchasing and Supply
Management, 8(2), 7182.
Craighead, C. W., Blackhurst, J., Rungtusanatham, M. J., & Handeld, R. B. (2007). The
severity of supply chain disruptions: Design characteristics and mitigation
capabilities. Decision Sciences, 38(1), 131156.
Croom, S., Romano, P., & Giannakis, M. (2000). Supply chain management: An
analytical framework for critical literature review. European Journal of
Dinesh Kumar, U., & Saranga, H. (2010). Optimal selection of obsolescence
mitigation strategies using a restless bandit model. European Journal of
Operational Research, 200(1), 170180.
Dong, L., & Rudi, N. (2004). Who benets from transshipment? Exogenous vs.
endogenous wholesale prices. Management Science, 50(5), 645657.
Ertek, G., & Grifn, P. M. (2002). Supplier and buyer-driven channels in a two-stage
supply chain. IIE Transactions, 34(8), 691700.
Esposito, E., & Passaro, R. (2009). The evolution of supply chain relationships: An
interpretative framework based on the Italian inter-industry experience. Journal
of Purchasing & Supply Management, 15(2), 114126.
Fabbe-Costes, N., Roussat, C., & Colin, J. (2011). Future sustainable supply chains:
What should companies scan? International Journal of Physical Distribution and
Logistics Management, 41(3), 228252.
Flynn, B. B., Huo, B., & Zhao, X. (2010). The impact of supply chain integration on
performance: A contingency and conguration approach. Journal of Operations
Garcia-Dastugue, S. J., & Lambert, D. M. (2003). Internet-enabled coordination in the
supply chain. Industrial Marketing Management, 32(3), 251263.
Goyal, S. K. (1976). An integrated inventory model for a single supplier single
customer problem. International Journal of Production Research, 5(1), 107111.
Goyal, S. K. (1995). A one-vendor multi-buyer integrated inventory model: A
comment. European Journal of Operational Research, 82(1), 209210.
Gravier, M. J., & Swartz, S. M. (2009). The dark side of innovation: Exploring
obsolescence and supply chain evaluation for sustainment-dominated systems.
Journal of High Technology Management Research, 20(2), 87102.
Grieger, M. (2003). Electronic marketplaces: A literature and a call for supply chain
management research. European Journal of Operational Research, 144(2),
280291.
Grover, V., & Saeed, K. A. (2007). Impact of product, market, and relation
characteristics on inter-organizational system integration in manufacturer
supplier dyads. Journal of Management Information System, 23(4), 185216.
Gulati, R., Lawrence, P. R., & Puranam, P. (2005). Adaptation in vertical relationships:
Beyond incentive conicts. Strategic Management Journal, 26(5), 415440.
Ha, B.-C., Park, Y. K., & Cho, S. (2011). Suppliers affective trust and trust in
competency in buyers: Its effect on collaboration and logistics efciency.
International Journal of Operations and Production Management, 31(12), 5677.
Handeld, R. B., & Bechtel, C. (2004). Trust, power, dependence, and economics: Can
SCM research borrow paradigms? International Journal of Integrated Supply
Hartman, B., & Dror, M. (2003). Optimizing centralized inventory operations in a
cooperative game theory setting. IIE Transactions, 35(3), 243257.
Hartman, B., & Dror, M. (2005). Allocation of gains from inventory centralization in
newsvendor environments. IIE Transactions, 37(2), 93107.
Hill, R. M. (1999). The optimal production and shipment policy for the single vendor
single buyer integrated production inventory problem. International Journal of
Production Research, 37(11), 24632475.
Holweg, M., Disney, S., Holstrom, J., & Smaros, J. (2005). Supply Chain Collaboration:
Making Sense of the Strategy Continuum. European Management Journal, 23(2),
170181.
Horvath, L. (2001). Collaboration: The key to value creation in supply chain
management. Supply Chain Management: An International Journal, 6(5), 205207.
Jap, S. D. (2001). Perspectives on joint competitive advantages in buyersupplier
relationships. International Journal of Research in Marketing, 18(12), 1935.
Kapur, V., Peters, J., & Berman, S. (2003). High-tech 2005: The horizontal,
hypercompetitive future. Strategy and Leadership, 31(2), 3447.
Kim, K. K., Park, S., Ryoo, S. Y., & Park, S. S. (2010). Inter-organizational cooperation
in buyersupplier relationships: Both perspectives. Journal of Business Research,
63(8), 863869.
Ku, K.-C., Gurumurthy, C. K., & Kao, H.-P. (2007). Inter-rms collaboration of joint
ventures in IC foundry business. Technovation, 27(5), 296305.
Kumar, K. (2001). Technologies for supporting supply chain management.
Communications of ACM, 44(6), 5861.
Lal, R., & Staelin, R. (1984). An approach for developing an optimal discount pricing
policy. Management Science, 30(12), 15241539.
525
Lancioni, R. A., Smith, M. F., & Schau, H. J. (2003). Strategic internet application
trends in supply chain management. Industrial Marketing Management, 32(3),
211217.
Lanier, D., Jr., Wempe, W. F., & Zacharia, Z. G. (2010). Concentrated supply chain
membership and nancial performance. Chain- and rm-Level perspectives.
Journal of Operations Management, 28(1), 116.
Lee, B.-C., Kim, P.-S., Hong, K.-S., & Lee, I. (2010). Evaluating antecedents and
consequences of supply chain activities: An integrative perspective.
International Journal of Production Research, 48(3), 657682.
Lee, H. L., Billington, C., & Carter, B. (1993). HewlettPackard Gains Control of
Inventory and Service through Design for Localization. Interfaces, 23(4), 111.
Lu, L. (1995). Theory and methodology: A one-vendor multi buyer integrated
inventory model. European Journal of Operational Research, 81(2), 312323.
Matopoulos, A., Vlachopoulou, M., Manthou, V., & Manos, B. (2007). A conceptual
framework for supply chain collaboration: An empirical evidence from the agrifood industry. Supply Chain Management: An International Journal, 12(3),
177186.
McCarthy-Byrne, T. M., & Mentzer, J. T. (2011). Integrating supply chain
infrastructure and process to create joint value. International Journal of
Physical Distribution and Logistics Management, 41(2), 135161.
McLaren, T., Head, M., & Yuan, Y. (2002). Supply chain collaboration alternatives:
Understanding the expected costs and benets. Internet Research: Electronic
Networking Applications and Policy, 12(4), 348364.
Meca, A., Timmer, J., Garcia-Jurado, I., & Borm, P. E. M. (2004). Inventory games.
European Journal of Operational Research, 156(1), 127139.
Monahan, J. P. (1984). A quantity discount pricing model to increase vendor prots.
Management Science, 30(6), 720726.
Nair, A., & Closs, D. J. (2006). An examination of the impact of coordinating supply
chain policies and price markdowns on short lifecycle product retail
performance. International Journal of Production Economics, 102(2), 379392.
Nyaga, G., Whipple, J., & Lynch, D. (2010). Examining supply chain relationships: Do
buyer and supplier perspectives on collaborative relationships differ? Journal of
Operations Management, 28(2), 101114.
Ovalle, O. R., & Marquez, A. C. (2003). The effectiveness of using e-collaboration
tools in the supply chain: An assessment study with system dynamics. Journal of
Panayides, P. M., & Lun, Y. H. V. (2009). The impact of trust on innovativeness and
supply chain performance. International Journal of Production Economics, 122(1),
3546.
Papoulis, A., & Pillai, S. U. (2002). Probability, random variables and stochastic
processes (4th ed.). New Delhi India: Tata McGraw Hill.
Parlar, M., & Weng, Z. K. (1997). Designing a rms coordinated manufacturing and
supply decisions with short product life cycles. Management Science, 43(10),
13291344.
Rosenzweig, E. D. (2009). A contingent view of e-collaboration and performance in
manufacturing. Journal of Operations Management, 27(6), 462478.
Rudberg, M., Klingenberg, N., & Kronhamn, K. (2002). Collaborative supply chain
planning using electronic marketplaces. Integrated Manufacturing Systems,
13(8), 596610.
Saen, R. F. (2007). Supplier selection in the presence of both cardinal and ordinal
data. European Journal of Operational Research, 183(2), 741747.
Sandborn, P., Prabhakar, V., & Ahmad, O. (2011). Forecasting electronic part
procurement lifetimes to enable the management of DMSMS obsolescence.
Microelectronics Reliability, 51(2), 392399.
Silver, E. A., Pyke, D. F., & Peterson, R. (1998). Inventory management and production
planning and scheduling (3rd ed.). New York: John Wiley & Sons.
Singh, P., & Sandborn, P. (2006). Obsolescence driven design refresh planning for
sustainment-dominated systems. The Engineering Economist, 51(2), 115139.
Skjoett-Larsen, T., Thernoe, C., & Andresen, C. (2003). Supply chain collaboration:
Theoretical perspectives and empirical evidence. International Journal of Physical
Distribution and Logistics Management, 33(6), 531549.
Slikker, M., Fransoo, J., & Wouters, M. (2005). Cooperation between multiple newsvendors with transshipments. European Journal of Operational Research, 167(2),
370380.
Sodhi, M. S. (2001). Applications and opportunities for operations research in
internet-enabled supply chains and electronic marketplaces. Interfaces, 31(2),
5667.
Srinivasan, M., Mukherjee, D., & Gauret, A. S. (2011). Buyersupplier partnership
quality and supply chain performance: Moderating role of risks, and
environmental uncertainty. European Management Journal. doi:10.1016/
j.emj.2011.02.004.
Su, Q., Song, Y.-t., Li, Z., & Dang, J.-x. (2008). The impact of supply chain relationship
quality on cooperative strategy. Journal of Purchasing & Supply Management,
14(4), 263272.
Tan, K. C. (2001). A framework of supply chain management literature. European
Journal of Purchasing and Supply Management, 7(1), 3948.
Taylor, T. A., & Plambeck, E. L. (2007). Supply chain relationships and contracts: The
impact of repeated interactions on capacity investment and procurement.
Management Science, 53(10), 15771593.
Tsay, A., Nahmias, S., & Agrawal, N. (2000). Modeling supply chain contracts: A
review. In S. Tayur, R. Ganeshan, & M. Magazine (Eds.), Quantitative models for
supply chain management (pp. 299330). Norwell, MA: Kluwer Academic
Publishers.
van den Heuvel, W., Borm, P. E. M., & Hamers, H. (2007). Economic lot-sizing games.
European Journal of Operational Research, 176(2), 11171130.
526
Wang, Y., Potter, A., Naim, M., & Beevor, D. (2011). A case study exploring drivers
and implications of collaborative electronic logistics marketplaces. Industrial
Marketing Management, 40(4), 612623.
Wu, I.-L., & Chuang, C.-H. (2010). Examining the diffusion of electronic supply chain
management with external antecedents and rm performance: A multi-stage
analysis. Decision Support Systems, 50(1), 103115.
Wu, S. D., Erkoc, M., & Karabuk, S. (2005). Managing capacity in high-tech industry:
A review of literature. The Engineering Economist, 50(2), 125158.
Xiao, Tiaojun, & Jin, Jiao (2011). Coordination of a fashion apparel supply chain
under lead-time dependent demand uncertainty. Production Planning & Control,
22(3), 257268.
Zacharia, Z. G., Nix, N. W., & Lusch, R. F. (2011). Capabilities that enhance outcomes
of an episodic supply chain collaboration. Journal of Operations Management,
29(6), 591603.
Zhang, D. Y., Xu, X. F., & Wang, G. (2004). Process and method of model reuse for
agile virtual enterprise. Computer Integrated Manufacturing Systems, 10(1),
2329.
Zhang, J. (2005). Transshipment and its impact on supply chain members
performance. Management Science, 51(10), 15341539.

Bahinipati - 2012 - Vertical Collaboration in The Semiconductor Industry

Hochgeladen von

Dokumentinformationen

Originaltitel

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

Bahinipati - 2012 - Vertical Collaboration in The Semiconductor Industry

Hochgeladen von

Copyright:

Verfügbare Formate

Computers & Industrial Engineering 62 (2012) 504526

Contents lists available at SciVerse ScienceDirect

Computers & Industrial Engineering

Vertical collaboration in the semiconductor industry: A decision framework

This manuscript was processed by Area Editor Mohamad Y. Jaber.

Corresponding author. Tel.: +91 9567880796.

E-mail address: bikram@iimk.ac.in (B.K. Bahinipati).

is the strategic benets gained over competitors in the marketplace

ers) are depicted in Fig. 1. As the customer demands are placed at

issues as well as dominance relationships between the entities. As

2. Motivation and research objectives

Fig. 1. Relationship between members in a semiconductor supply chain (pull system).

The researchers should not assume all partnering relationships

3.2. Perspective of obsolescence cost

not applicable. The classical buyersupplier coordination problem

Lal and Staelin (1984)

Silver et al. (1998)

Anupindi and Bassok (1999)

governance or relationship structure (McCarthy-Byrne & Mentzer,

tation of e-markets results in higher perception of risks due to

4.1. Classication of vertical collaboration procurement activities

4.2. Impact of e-marketplaces for vertical collaboration

In SCM, purchasing becomes strategic and tightly coupled

1. Collaborative fulllment in semiconductor industry supply

4.3. Framework for e-market supported vertical collaboration models

M1 : Neutrale market Model NEM

4. PCS PSD P PDB P CB : The suppliers gain from the collaborative

M2 : Buyer orientede market Model BEM

Suppliers Problem SP1

Traditional marketplace [no collaboration (NC)]

1. A single short life-cycle product involving a single time-period

the product remaining at the end of the planning horizon (by

PB Q ERevenue Salvage cost EHolding cost

5.2. Development of an analytical collaboration model

The expected value of the buyers prot depends on revenue earned,

Shortage cost Purchase cost

So, the collaboration model is developed with Q as the decision

ERjX n:Prob:X n ERjX 0Prob:X 0

After analysis, simplication and rearrangement of the terms in the

(a) The expected value of inventory holding cost of the buyer in

EHCjX nProb:X n EHCjX 0Prob:X 0

After analysis, simplication and rearrangement of the terms, the

EHC QHB TPr:X 0

The expected value of the buyers prot can be estimated by

The expected prot of the collaborated system can be estimated by

This system of inequality means that QC is the smallest value of Q

But; DPC Q P V HB TPr:X 6 Q

As the demand arrival process from the end customer is assumed to

DPC Q P V HB TPr:X 6 Q G Bf1 Pr:X 6 Q g

IfQ EM < Q TM ; then C G B gQ EM P 0:

But; P 6 C: This implies that P G B P C G B:

This implies that PC Q 1 PC Q

This violates the expression (21). So, it is essentially inferred that

With similar analysis, the expression (16) can be represented by the

6. Development of algorithms for vertical collaboration

The analytical model for the vertical collaboration problem is

6.1. Simulation results and discussion

Step A.1: Determine the expected total prot function of the

Input data set 1: C = 8, V = 3, B = 2, P = 5, KB = 18, KS = 20, HB = 0.4, G = 18, H = 1.5,

Fig. 3. Flow chart for selection of most appropriate marketplace.

Further, each row in the Tables 57 compares the objective

PEM (QEM) PTM(QTM)

Input data set 1: C = 8, V = 3, B = 2, P = 5, KB = 18, KS = 20, HB = 0.4, G = 18, H = 1.5, k = 8.