Logistic and Operation Management

Edited by Prof. Dr. Thorsten Blecker,
Prof. Dr. George Q. Huang and Prof. Dr. Fabrizio Salvador
9
Operations and Technology Management
Global Logistics Management
Sustainability, Quality, Risks
Wolfgang Kersten / Thorsten Blecker /
Heike Flämig (Eds.)
erich schmidt verlag
ES
Extract, for more details visit ESV.info/978 3 503 11228 9
Global Logistics Management
Sustainability, Quality, Risks
Edited by
Wolfgang Kersten, Thorsten Blecker,
and Heike Flämig
With Contributions by
Adel Al-Mansi, Hatem Aldarrat, Dagoberto Alves de Almeida,
Nasrin Asgari, Deepak Barman, José Arnaldo Barra Montevechi,
Helmut Baumgarten, Ronivaldo Belan, Leticia Biescas Altelarrea,
Thorsten Blecker, Michael Bohn, Anna Corinna Cagliano, Maria Caridi,
Inga-Lena Darkow, Debadyuti Das, Kateryna Daschkovska,
Ralf Elbert, Heike Flämig, Christian W. Flotzinger, David Francas,
Gabriela Garcia-Lopez, Alexander Goudz, Hans-Dietrich Haasis,
Tobias Held, Hannelore Hofmann-Prokopczyk, Susanne Hohmann,
Tong Huan, George Q. Huang, Yun Huang, Bernd Kaluza,
Wolfgang Kersten, Matthias Klumpp, Jan Koch, Herbert Kotzab,
Antje Krey, Charles Lau, Ma Lin, Matthias Lorenz, Abolfazl Mirzazadeh,
Enzo Morosini Frazzon, Bernd Noche, Wolfgang Ortner,
Manja Ostertag, Margherita Pero, Carlo Rafele, Fathi Rhoma,
Daniel Rief, Hubert B. Schemitsch, Peter Schentler,

Bernd Scholz-Reiter, Jörg Schweiger, Kianoush Siamardi,
Andrea Sianesi, Gyaneshwar Singh Kushwaha, Frank Straube,
Morteza Tolouei, Martin Tschandl, Hans G. Unseld,
Andreas Wieland, Herwig Winkler, Heiko Wöhner,
Reza Zanjirani Farahani, Stephan Zelewski, Hassan Zoghi
ERICH SCHMIDT VERLAG
Extract, for more details visit ESV.info/978 3 503 11228 9
ISBN 978 3 503 11228 9
ISSN 1863-3390
All rights reserved
© Erich Schmidt Verlag GmbH & Co., Berlin 2008
www.ESV.info
This paper fulfi lls the requirements of the
Frankfurter Forderungen of Die Deutsche Bibliothek
and the Gesellschaft für das Buch concerning the paper permanence
and meets the tight regulations of American National Standard
Ansi/Niso Z 39.48-1992 as well as ISO 9706
Printing and Binding: Hubert & Co, Göttingen
Bibliographic information published by Die Deutsche Bibliothek
Die Deutsche Bibliothek lists this publication in the Deutsche Nationalbibliografi e;
detailed bibliographic data are available in the Internet at .
For further information concerning this title please follow this link:
ESV.info/978 3 503 11228 9
from: Kersten/Blecker/Flämig (Eds.), Global Logistics Management.
© Erich Schmidt Verlag GmbH & Co., Berlin 2008


V
Preface
Changes in business environment due to globalization and short product life cycles

have increased the intensity of competition in many markets. Therefore companies
collaborate closely with their suppliers and customers using Supply Chain Man-
agement concepts to optimize the inter-organizational flow of material, information
and capital through their networks. Originating in business logistics management,
Supply Chain Management approaches were enlarged when the synchronization of
manufacturing operations, common product design and standardized information
technologies were integrated. Companies like Dell, Wal-Mart or Benetton gain
their competitive advantage by the application of these Supply Chain Management
techniques. New approaches are developed in the actual academic discussion, like
common sustainability, quality and risk management in supply chains.
With this volume we want to contribute to the holistic global management of
logistics networks. Therefore we structure the book in five main areas of interest.
The first topic highlighted is globalization. Although there already has been a dis-
cussion in science and practice for years, many aspects of these topic are still not
evaluated yet. The authors deal with the consequences of globalization for local lo-
gistics structures as well as with management concepts to control international net-
works. The second area of interest covered is sustainability and the underlying
question how supply chains can become ecologically sustainable. In respect to fu-
ture generations we are convinced that this topic is of tremendous importance for
the management of logistics networks. Two rightly popular topics in science and
practice build the joined third focus of this volume – risk and quality management
in supply chains. Because both concepts seem to be related regarding to the meas-
ures they use as well as to their main approaches. The contributions of our forth
main area show, how products and product related issues interfere with the design
of logistics networks. Finally we present contributions which deal with the per-
formance of supply chain networks.
We would like to thank the authors for their excellent contributions which ad-
vance the logistics research progress. Without their support and hard work, this vo-
lume would not be possible. Additional thanks are due to the publishing company,
the Erich Schmidt Verlag, especially to Dr. Joachim Schmidt for the possibility to

publish this volume and his valuable cooperation. This book would not exist with-
out a good organization and preparation. We would like to thank Philipp Hohrath,
Thomas Will, and Jan Koch for their efforts to prepare, structure, and finish this
book.

Hamburg, August 2008 Wolfgang Kersten, Thorsten Blecker and Heike Flämig
from: Kersten/Blecker/Flämig (Eds.), Global Logistics Management.
© Erich Schmidt Verlag GmbH & Co., Berlin 2008


VII
Table of Content
Preface V
Table of Content VII 
I. Logistics in a Global Context
Transfer Pricing, Taxation and Capacity Planning in International
Manufacturing Networks 3
David Francas
Bringing regional networks back-into global supply chains: Strategies
for logistics service providers as integrators of logistics clusters 21
Hans-Dietrich Haasis and Ralf Elbert
System Dynamics Analysis on the Evolution of Logistics Cluster: An
Empirical Analysis Of Logistics Networks And Nodes In Yangtze River
Delta of China 33
Ma Lin and Tong Huan
Holistic Security in Global Supply Chains - A Strategic Framework 45
Andreas Wieland and Helmut Baumgarten
Logistics in the Context Of Internationalisation – How Chinese and
German Companies Enter Foreign Markets 59
Frank Straube, Michael Bohn, and Daniel Rief

II. Sustainability in Logistics
Future Scenarios 2030: The Contribution of Logistics to a Sustainable
Development 75
Christian W. Flotzinger and Hannelore Hofmann-Prokopczyk
Design of an Environmental Supply Chain Network: A Biosolids Waste
Case Study 87
Adel Al-Mansi, Hatem Aldarrat, Fathi Rhoma, Alexander Goudz,
Matthias Lorenz, and Bernd Noche

from: Kersten/Blecker/Flämig (Eds.), Global Logistics Management.
© Erich Schmidt Verlag GmbH & Co., Berlin 2008
Table of Content

VIII
Biogas Logistics Supply Chain Network Design: A Framework Model 99
Fathi Rhoma, Hatem Aldarrat, Adel Al-Mansi, Alexander Goudz,
and Bernd Noche
Reasons for the Slow Development of Reverse Logistics in the Electronic
Industry of China 113
Charles Lau
III. Risk and Quality Management in Supply Chains and Logistics
Quality Management Impacts on Logistics Networks measured by
Supply Chain Performance Indicators 129
Matthias Klumpp and Manja Ostertag
A Conceptual Approach for Quality Improvements in Project-Oriented
Supply Chain Networks 149
Herwig Winkler, Hubert B. Schemitsch, and Bernd Kaluza
Impact Of Supply Chain Quality Management on Competitive Advantage
and Organizational Performance 171
Deepak Barman, Gyaneshwar Singh Kushwaha, and Debadyuti Das

Supply Chain Risk Management for Robustness in a Logistical
Framework 187
Inga-Lena Darkow and Heiko Wöhner
Container Logistics Networks under Context Diversity 205
Bernd Scholz-Reiter, Kateryna Daschkovska, and Enzo Morosini Frazzon
Risk and Capacity Management in Logistics Networks: The Example
of Global Container Operators 223
Stephan Zelewski, Matthias Klumpp, and Susanne Hohmann
Logistics Services: Theoretical and Empirical Findings on Quality
Perceptions 239
Wolfgang Kersten and Jan Koch
IV. Product Induced Logistics Strategies
Logistic Network Strategies and Product Harmonization – Some Findings
of the Interrelation between Logistics Network Setup and Product
Country Clustering in the Fast Moving Consumer Goods Industry in
Europe 261
Tobias Held
from: Kersten/Blecker/Flämig (Eds.), Global Logistics Management.
© Erich Schmidt Verlag GmbH & Co., Berlin 2008
Table of Content

IX

Optimal Pricing for a Three-Level Supply Chain with Different Channel
Structures Using Game-Theoretic Approach 279
Yun Huang and George Q. Huang
Conceptual Design of a ROI Benefit Assessment for Web-based
Supplier Relationship Management Scenarios 295
Wolfgang Ortner, Jörg Schweiger, and Martin Tschandl
Inbound Supply Chain Management in Mass Customization:

Specific Aspects and Requirements for Efficient Procurement 311
Peter Schentler and Antje Krey
Economic Order Interval under Variable Inflationary Conditions 325
Abolfazl Mirzazadeh
Product Architecture and Supply Chain Design: Impacts on Mix
Flexibility 339
Andrea Sianesi, Margherita Pero, and Maria Caridi
V. Performance and Capacity Management in Supply Chains
A Client-Supplier Relationship Performance Measuring Model Based on
Integrated Engineering 361
Dagoberto Alves de Almeida, Ronivaldo Belan,
and José Arnaldo Barra Montevechi
Airport Integrated Capacity Model 381
Leticia Biescas Altelarrea
Fast Transhipment Equipment and Novel Methods for Rail Cargo 393
Hans Unseld and Herbert Kotzab
Evaluation of Advanced Parking Information Systems at Airports 405
Hassan Zoghi, Kianoush Siamardi, and Morteza Tolouei
Simulation for Logistics Performance Management: Comparing
Different Approaches 423
Anna Corinna Cagliano, and Carlo Rafele
Analysis of on-time delivery measure in a supply chain 443
Nasrin Asgari and Reza Zanjirani Farahani
Development and Implementation of a Six Sigma Readiness Model
in the Food Industry 455
Wolfgang Kersten, Gabriela Garcia-Lopez, and Jan Koch


3
Transfer Pricing, Taxation and Capacity Planning in

International Manufacturing Networks
David Francas
Abstract
This paper considers the impact of corporate taxation and transfer pricing on the
design of a multi-market facility manufacturing network. Using a model based ap-
proach we investigate the network configuration from a cross-plant capacity in-
vestment perspective under demand uncertainty. The results highlight the impor-
tance of creating operational flexibility within the manufacturing network of a
multinational firm, not only to hedge against uncertainty in product demands, but
further to arbitrage taxation by means of profit-shifting activities.

Keywords: Manufacturing Strategy, Plant Network Configuration, Transfer Pric-
ing, Taxation, Operational Hedging

1 Introduction
A strong motivation for the configuration of an international manufacturing net-
work has been the access to low costs production inputs although the benefit of
centralized production in low-cost countries is often associated with disadvantages
of increased logistics costs. Furthermore, investment decisions of multinational
companies are affected by international tax rate differentials. Exploiting such dif-
ferences by means of cross-country profit shifting can decrease global tax liability.
Finally, companies have to take into account that investment decisions are typically
made long before actual demand is known. Therefore it is of crucial importance
that a supply chain is designed robustly to unknown operating conditions in order
to avoid lost sales or excess capacity. This paper studies a manufacturer’s multi-
market network design problem from a network capacity investment perspective.
Specifically, the optimal network configuration is examined for a firm that pro-
duces two products in order to serve two geographically dispersed markets using a
common component and regionalized final assembly. For this specific setting, four
network strategies are possible with different degrees of flexibility as shown in

Figure 1: Nodes refer to active facilities, i.e. component production plant(s) and fi-
nal product assembly facilities; the connecting arcs represent supply-demand links
either between network sites resulting in component flows or between assembly fa-
from: Kersten/Blecker/Flämig (Eds.), Global Logistics Management.
© Erich Schmidt Verlag GmbH & Co., Berlin 2008
Extract, for more details visit ESV.info/978 3 503 11228 9
David Francas

4
cilities and market areas. Each of the displayed structures corresponds to a generic
plant charter strategy commonly discussed in global manufacturing networks
(Cohen and Lee 1989, Lu and Van Mieghem 2007).


Figure 1: Possible configurations of the international manufacturing network (cf. Lu and Van
Mieghem 2007)

Clearly, shipping costs are higher for a process-focused network than using a mar-
ket-focused configuration, as the common components need to be shipped across
plants. Benefits associated with a market area strategy are reduced shipping costs
and higher responsiveness to market due to advantages of quick delivery, whereas
process-focused networks benefit from economies of scale due to consolidated pro-
duction (Hayes and Wheelwright 1984).
Aside, process-focused networks allow firms to risk-pool the uncertain demand
for the common subassembly from all products and markets. In this regard, the
market area structure describes a dedicated network configuration without any risk-
pooling opportunities, while the hybrid structure is the most general and includes
the concept of ‘chaining’ introduced by Jordan and Graves (1995). This highly
flexible configuration is adapted by many international car manufactures such as
from: Kersten/Blecker/Flämig (Eds.), Global Logistics Management.

© Erich Schmidt Verlag GmbH & Co., Berlin 2008
Extract, for more details visit ESV.info/978 3 503 11228 9
Transfer Pricing, Taxation and Capacity Planning in International Manufacturing Networks

5

VW, Chrysler, and General Motors which operate in an industry exposed to high
market uncertainty (Francas et al. 2007).
Beside these issues which are commonly discussed in operations and logistics
management, we investigate the role of cross-country profit shifting as a further ra-
tional for selecting a certain network configuration. The potential scope of profit-
shifting via an arbitrary use of transfer prices is sizeable because of the ongoing
economic integration and the increasing volume of trade in intermediates (Ernst &
Young 2005). Moreover, these opportunities exist in spite of current regulations
imposed by tax authorities to avoid such behaviour, as has been observed in recent
investigations of Overesch (2006), Weichenrieder (2006), and Huizinga and
Leaven (2007).
Especially in the EU taxation becomes increasingly important since conver-
gence of other factors is extremely high. Due to an enhanced integration of national
markets, disparities of economic and production conditions disappear and differ-
ences in corporate tax rates become even more distinct. Exploitation of these ex-
ogenous market imperfections via production-shifting to low-tax countries thus
represents a viable alternative to multinational enterprises (MNEs) in order to de-
crease their global tax liability (Stöwhase 2003, Overesch 2006).
In this paper, a numerical study is presented to capture the effects of taxation
and demand uncertainty. In particular, we examine the performance of the generic
operations strategies of a market area configuration, onshore and offshore configu-
rations, and a hybrid configuration in between. Our results highlight the importance
of creating operational flexibility within the manufacturing network of a multina-
tional firm, not only to hedge against demand uncertainty, but to further arbitrage

taxation by means of profit-shifting activities.
The remainder of the paper is organized as follows: In Section 2 the role of
transfer pricing within a corporation is introduced. The third section highlights im-
portant factors influencing the configuration of international networks and reviews
the literature on global network design and manufacturing flexibility. The devel-
opment of an adequate model that captures both taxation issues and demand uncer-
tainty in the framework of flexible manufacturing network structures is described in
Section 4. Subsequently, a numerical study is presented to show the effects of taxa-
tion and demand uncertainty (Section 5). The paper concludes in Section 6 provid-
ing a summary of results and managerial insights obtained.
2 Transfer Pricing
Basically, the use of transfer prices serves two major roles. First, in a tax free world
the intention of transfer prices is the efficient allocation of resources within a single
firm, in addition to their role of providing incentives to divisional managers who
are rewarded based on their performance. Hence, the rationale for setting transfer
prices to value internally exchanged commodities originally derives from the moti-
vation of operating autonomous units in a decentralized firm. For an extensive lit-
from: Kersten/Blecker/Flämig (Eds.), Global Logistics Management.
© Erich Schmidt Verlag GmbH & Co., Berlin 2008
Extract, for more details visit ESV.info/978 3 503 11228 9
David Francas

6
erature review on the managerial use of transfer pricing we refer to Pfaff and Pfeif-
fer (2004), Ewert and Wagenhofer (2005) and Göx and Schiller (2006). In fact, the
managerial use of transfer prices has further been studied in the supply chain con-
tracting literature where, in contrast to the accounting literature, transfers take place
between two independent firms. Cachon (2003) provides a detailed overview on the
supply chain contracting literature.
Second, within a multinational firm, even if decisions are made centrally, transfer

prices need to be established for cross-border trade flows. Consequently, transfer
prices determine the tax burden of each subsidiary located in different countries,
thus affecting the overall tax exposure of the multinational enterprise (Abdallah
1989). A multinational corporation that can adjust the transfer prices for its inter-
mediate products within a certain range can use those as a device for shifting prof-
its to low-tax jurisdictions. This incentive improves the firm’s global after-tax
profit due to decreased tax liabilities (Devereux 2006). Obviously, transfer pricing
is a concern when it comes to minimizing the corporation’s global tax burden (see
Abdallah (1989) for more details on international transfer pricing).
To prevent MNEs from the ‘manipulative’ use of transfer prices and to ensure a
fair allocation of profits most nations have agreed on following the approach set out
in article 9 of the OECD model convention on taxes on income and capital (OECD
2003). The OECD promotes the arm’s length principle for determining transfer
prices for tax purposes, which is defined as the market price that would have been
negotiated between independent firms bargaining at arm’s length. Although the
OECD Guidelines call for the use of the arm’s length principle, in practice, calcu-
lating the arm’s length price can be a challenge. For instance, many intra-firm
transactions are often specialized or intangible in nature so that there is no compa-
rable market (Bartelsman and Beetsma 2003). It is therefore difficult to find the
true value of the internally exchanged good or service. Consequently, companies
can generally identify a range of allowable arm’s length prices for their intermedi-
ates rather than a ‘single’ correct one (OECD 2000). For a detailed overview on ac-
ceptable transfer pricing methods, see Feinschreiber (2004) and Choi and Meek
(2005).
Because of their impact on a global supply chain’s overall performance, arm’s
length transfer prices should not only be a major factor in international financial
management, but rather be incorporated into strategic decisions on global planning
of the internal supply chain. The need for such an investigation is further accentu-
ated when taking into account that in practice the tax effects of transfer pricing are
most often only considered after strategic decisions have been made, as observed

by Cravens (1997).
3 Global Network Design
A crucial component of global supply chain planning is the efficient design of its
multi-plant network. Vidal and Goetschalckx (1997), Cohen and Mallik (1997), and
from: Kersten/Blecker/Flämig (Eds.), Global Logistics Management.
© Erich Schmidt Verlag GmbH & Co., Berlin 2008
Extract, for more details visit ESV.info/978 3 503 11228 9
Transfer Pricing, Taxation and Capacity Planning in International Manufacturing Networks

7

Meixell and Gargeya (2005) present reviews on global supply chain design models.
Some of these models have considered transfer prices either implicitly – see Kou-
velis and Rosenblatt (2001), Martel (2005) – or even explicitly – see Nieckels
(1976), Cohen et al. (1989), Cohen and Lee (1989), Canel and Khumawala (1996)
– as a determinant for minimizing taxes in global supply chains. Vidal and Goet-
schalckx (2001) simultaneously consider endogenously determined transfer prices
and transportation cost allocation to either the selling or buying facility in order to
shift profits to locations with a favourable tax rate.
The complexity in these models is due to numerous network decisions that have
been included and a broader scope based on a large number of commodities, coun-
tries, echelons and time periods. However, a common feature and therefore main
limitation of these ‘network flow models’ is that, given the strategic nature of the
network design problem, the decision framework is deterministic although main
factors such as customer demands or exchange rates are uncertain. Hence, they do
not incorporate the value of operational flexibility under uncertainty which pro-
vides a firm with a profitable response to changing environmental conditions by
creating real options whose value increases with the level of uncertainty (Cohen
and Huchzermeier 1999, Van Mieghem 1998).
Manufacturing flexibility has been considered a promising approach to increase

competitiveness in a dynamic marketplace. The optimal use and design of flexibil-
ity to deal with demand uncertainty has been examined by a number of researchers,
of which the most related to the current paper are briefly reviewed. Following the
lead of Fine and Freund (1990) on investment in mix-flexibility, Van Mieghem
(1998) considers a two-product firm that has the option to invest in product-
dedicated resources and/or in a flexible resource being able to produce either prod-
uct. The author formulates the mix-flexible manufacturing capacity investment de-
cision as a two-stage stochastic program with its first-stage capacity decision under
uncertainty and its second-stage production decision after demand realization. Van
Mieghem (1998) proposes that optimal capacity levels of product-dedicated and
mix-flexible resources should be set so as to balance underage and overage costs. It
is argued that by purposely unbalancing the capacity vector, i.e. having capacity in
excess in some resources compared to optimal capacity in a deterministic setting,
firms can alleviate the risk associated with demand uncertainty (risk-pooling).
Typically, this strategy incurs higher costs for acquiring the flexible capacity as op-
posed to the dedicated one, yet is sufficiently inexpensive to be a viable alternative
(Van Mieghem 1998).
Van Mieghem and Rudi (2002) introduce a new class of models, called ‘news-
vendor networks’ that can be used as effective frameworks to study a variety of
stochastic investment decisions while capturing additional features like commonal-
ity, flexible resources, and transhipments. Operational hedging strategies can be
viewed as real options whose value is driven through either timing, i.e. pooling of
safety capacity and postponement of operational decisions (e.g. commonality,
from: Kersten/Blecker/Flämig (Eds.), Global Logistics Management.
© Erich Schmidt Verlag GmbH & Co., Berlin 2008
Extract, for more details visit ESV.info/978 3 503 11228 9
David Francas

8
flexible resources), or scope through providing a set of alternatives instead of a sin-

gle choice (e.g. transhipments), if not both.
Using the newsvendor networks approach, Kulkarni et al. (2004, 2005) and Lu
and Van Mieghem (2007) study global network design problems. The specific sce-
nario illustrated by Kulkarni et al. (2005) is that of a two-plant network assembling
two products with one common and one dedicated component. The common com-
ponent might be either produced in both plants (product-focused network structure)
or in one of the plants. Kulkarni et al. (
2004) determine the better of two given
network configurations for a multi-plant network with commonality. Kul-
karni
et al. (2004, 2005) find that risk-pooling advantages of the process-focused
network structure might be substantial under a wide range of conditions, even in
the absence of scale economies.

Lu and Van Mieghem (2007) present a model to study the sourcing and loca-
tion decisions of commonality in two-market facility networks from a network ca-
pacity investment perspective, resulting in four possible network configurations.
One of the major distinctions that separate their model from the ones presented by
Kulkarni et al. (2004, 2005) is that the authors let the optimal network strategy
emerge from optimization. The authors’ objective is to demonstrate how uncer-
tainty may change the network strategy relative to the intuitive deterministic case.
Under demand certainty, offshoring is optimal when lower manufacturing costs
outweigh higher transportation costs for transhipments; otherwise a market area
strategy is adopted. With ex-ante known market demand, a hybrid structure using
cross-border transhipments will never emerge from optimization. In contrast, in a
stochastic setting, the hybrid structure can become optimal.
Jordan and Graves (1995) introduce the chaining principle as efficient guideline
to configure mix-flexible manufacturing networks. They highlight the fact that an
intelligently designed system in which most plants will only need to produce two
different products approximates the benefit of total flexibility, where each plant can

manufacture all products.
4 Methodology and Model Development
We developed a stochastic optimization model that explicitly considers corporate
taxation in the design of a multi-market facility network and verify the network
configuration from a cross-plant capacity investment perspective under random
demand (Francas et al. 2008). The modelling approach incorporates strategic ca-
pacity investment decisions and tactical resource allocation, transfer pricing, and
transportation cost allocation decisions. The decision process is modelled as a two-
stage stochastic program with recourse.
In the context of a globally integrated two-level, two-market supply chain in
which a common component is used for the local assembly of two regionalized fi-
nal products the network design problem in an uncertain environment principally
boils down to the following main strategic questions: First, whether the common
from: Kersten/Blecker/Flämig (Eds.), Global Logistics Management.
© Erich Schmidt Verlag GmbH & Co., Berlin 2008
Extract, for more details visit ESV.info/978 3 503 11228 9
Transfer Pricing, Taxation and Capacity Planning in International Manufacturing Networks

9

component should be manufactured in a hub location or in two local production
plants, and second, where to locate component capacity if a centralized strategy is
adopted. In other words, the ex-ante capacity acquisition decision for the common
component involves not only how much capacity to acquire, but also where to ac-
quire it from, when additionally taking into account nation-specific corporate tax
rates, a set of acceptable transfer prices for the common component, and the possi-
bility of allocation of transportation costs among network sites. Similar to Vidal
and Goetschalckx (2001), the transfer prices do not include transportation costs,
which are modelled separately. The MNE has to decide on the capacity portfolio
based on the probability distribution of the random demands for final products

thereby determining its global network structure.
The MNE’s overall objective is to coordinate strategic investment decisions in
order to maximize its firm value, which highly depends on the uncertain global op-
erating profit after taxes. Therefore, tactical ex-post decisions need to be antici-
pated for all possible demand realizations. Subsequently, state-dependent global
operating profits are weighted by their individual probability in order to determine
a robust design solution for any demand realization. The corresponding two-stage
decision process is shown in Figure 2.
In the second stage, tactical decisions take place after uncertainty has been re-
solved, i.e. final product demands are considered to be deterministic and entirely
known. In this phase the firm determines the recourse variables in a way that the
global after-tax profit is maximized given earlier specified capacity levels and de-
mand realizations. The recourse variables correspond to three contingent activity
vectors that are production quantities as well as flow-related transfer prices and
transportation cost proportions to be paid by either network site. Depending on the
first-stage decisions, the firm determines production quantities as well as is able to
shift income to tax favourable locations by dynamically adjusting transfer prices
and transportation cost proportions in response to realized demands. Hence, mana-
gerial flexibility is exploited by postponing tactical decisions until market demands
are known.







Figure 2: Two-stage decision process

The following assumptions are made for the developed model. Generally, we as-

sume that the decision maker is risk-neutral. In line with the literature on manufac-
turing flexibility, it is assumed that the MNE is a price taker in each final product
market. Localized end-products are thus sold at the prevailing market selling price.
Capacity and Network
Design Decision
(1
st
Stage)
Realized
Final
Product
Demand

Production; Transfer
Prices; Transportation
Cost Allocation
(2
nd
Stage)
from: Kersten/Blecker/Flämig (Eds.), Global Logistics Management.
© Erich Schmidt Verlag GmbH & Co., Berlin 2008
Extract, for more details visit ESV.info/978 3 503 11228 9
David Francas

10
Final product demands are supposed to be random variables with known probabil-
ity distribution. Furthermore, there is no demand for the common component alone.
It is possible that some demand may not be met, which will incur a per unit output
shortage penalty cost. Thus, there is just the lost sales case and no backlogging is
allowed. As a single-period model is suggested, there is no inventory turnover in

the event of excess capacity. The technology matrix is linear and each final product
requires one common component.
For expositional simplicity, costs for acquiring long-term capacity are assumed to
be linear for component production plants and local assembly facilities, and corre-
spond to the equivalent annual costs before taxes (see Santoso et al. 2005). Further,
capacity costs are assumed to be similar across network structures.
1
For this spe-
cific setting, four global network strategies are possible with different degrees of
flexibility, as shown Figure 1. Cross-border transhipments represent substitutable
processing activities to satisfy regional demand. Contrary, in the event of a market
area and a process-focused structure (i.e. offshore or onshore) there is no other
lowest-cost choice for the shipping of the common component. Both network struc-
tures can be replicated by the hybrid configuration. Thus, the market area structure
and both process-focused structures can be interpreted as ‘boundary’ solutions of
the general network design problem (Lu and Van Mieghem 2007).
It is assumed that currency exchange rates are fixed so that all prices and costs
are denominated in a single currency. Equivalently, the Purchasing Power Parity
does not hold in the short term, i.e. movements in both prices and costs cannot be
adjusted via exchange rate fluctuations (Krugman and Obstfeld 2006). Finally, it is
economically justified to produce both localized final products so that investments
in dedicated assembly resources are always positive.
It is taken as given that for tax purposes the MNE has identified a range of ac-
ceptable arm’s length prices represented by an interval with a lower and an upper
bound for each plant producing the intermediate product. Any price in this interval
can be justified to the tax authorities as based on acceptable valuation methods and
supporting sets of data. Note that these statutory requirements on transfer pricing
only apply to international transfer prices, i.e. prices for intermediates transferred
across national borders (Choi and Meek 2005). For simplicity reason, it is assumed
that the MNE relies on just one set of books for its transfer prices so that the identi-

fied range of arm’s length prices for each production plant is also applied to the
component flows within a single country.
All transportation costs are linear functions of the quantities shipped. While the
transportation costs of a common component to be transferred within a single coun-
try account solely for shipping the intermediate good, in the event of international
___________________
1
As the optimal structure should emerge from optimization, economies of scale are not
considered but could be modelled using different capacity costs in the respective configura-
tions or by adding a fixed term to the investment cost function. See Kulkarni et al. (2004) and
Lu and Van Mieghem (2007).
from: Kersten/Blecker/Flämig (Eds.), Global Logistics Management.
© Erich Schmidt Verlag GmbH & Co., Berlin 2008
Extract, for more details visit ESV.info/978 3 503 11228 9
Transfer Pricing, Taxation and Capacity Planning in International Manufacturing Networks

11

intermediate flows transportation costs might additionally include tariffs and duties,
if any.
2
Furthermore, the MNE may use these costs at its global advantage by allo-
cating them arbitrarily between the facilities. In addition, variable production and
assembling costs are also presumed to be linear.
Following Martel (2005), income taxes paid in one country are calculated on
the sum of the net income made by all network sites located in that country, which
represent one subsidiary of the MNE. A reported loss of a network site reduces the
overall realized profit before taxes of the respective subsidiary. Therefore, subsidi-
aries which consist in the presented setting of either one or two network sites (de-
pending on the network structure) are assumed to be legally separate entities and

taxation occurs at the source, that is the tax liability of each subsidiary is deter-
mined by its respective taxable income. In particular, subsidiaries are wholly
owned by the MNE.
3
In case that a subsidiary reports a negative net income, no
taxes are to be paid. Also, according to the model developed in Vidal and Goet-
schalckx (2001), no tax credits (tax losses carried forward) are included in the
analysis due to the single-period nature of the proposed model. Corporate tax rates
are assumed to be constant in each county.
Finally, costs for building up plant capacities are to be considered in the prob-
lems’ second-stage formulation, according to the stochastic programming formula-
tion of Cohen and Huchzermeier (1999). That is because investment costs have a
significant impact on the taxable profit of an MNE realized in a single country
(Kruschwitz 2005).
In order to carry out a detailed analysis of different network configurations the
two-stage stochastic program is effectively solved numerically using a scenario
planning approach based on sampling techniques, labelled as the Sample Average
Approximation (see Francas et al. 2007).
5 Numerical Analysis
In the following, we present numerical results for the developed model that demon-
strate the effect of corporate taxation and transfer pricing on the design of a multi-
market facility network. In order to isolate the effect of taxation on the design of
the global network structure, some of the country-specific economic characteristics
are assumed to be identical. These include among others market prices for final
products, marginal production and assembly costs, and capacity costs for produc-
tion plants and assembly facilities. Random product demand is supposed to be in-
___________________
2
This assumption is employed for two reasons: first, within commercial blocks duties and tar-
iffs are to be neglected, and second, it avoids a possible trade-off between minimization of

taxes vs. duties and tariffs.

3
Taxation upon repatriation will be ignored since it may be postponed for a long period of time
and may effectively not be an issue if foreign direct investments are financed by the foreign
subsidiaries themselves, rather than through equity transfers, see Hartman (1985).
from: Kersten/Blecker/Flämig (Eds.), Global Logistics Management.
© Erich Schmidt Verlag GmbH & Co., Berlin 2008
Extract, for more details visit ESV.info/978 3 503 11228 9
David Francas

12
dependently and identically normal distributed. The parameter values for the stan-
dard case which is varied throughout the numerical studies are shown in Table 1.
According to an acceptable transfer pricing method for an intermediate good,
the lower bound of the transfer price is assumed to be equal to total per unit pro-
duction costs multiplied with a comparable mark-up, which can be interpreted as
profit margin for the supplying production plant (Vidal and Goetschalckx 2001).
However, it is important to note that a properly defined lower bound for the pro-
posed model could in fact only be determined after optimization, as total per unit
production costs depend on established capacity and production levels after de-
mand realization (Choi and Meek 2005). The upper bound on the transfer prices is
set equal to the comparable market price of the intermediate good if it was sold to
an external party (Choi and Meek 2005). In order to produce a range of potential
transfer prices it is assumed that this market price equals €14 for both production
plants.

Parameter Standard value
Final product price perceived at both facilities
Per unit production costs at both plants

Per unit assembling costs at both facilities
Per unit capacity investment costs at all sites
Transportation costs within a single country
Costs for cross-border transhipments
Lower bound for the transfer price at both plants
Upper bound for transfer price at both plants
Tax rate in the high-tax country 1
Tax rate in the low-tax country 2
Mean of local demand perceived at both facilities
Coefficient of variation of both local demands
20
6
4
2
1
2
11*
14
0.4
0.2
100
0.4
*

includes a mark-up of 37.5% over total component production costs
Table 1: Standard test parameters

Transportation costs for subassembly flows within a single country are supposed to
be lower than for those shipped across countries. Alternate transportation cost dif-
ferentials are obtained by changing the value of the latter for both flow directions

simultaneously. Corporate tax rates are presumed to differ among the two coun-
tries. Country 1 will be denoted as the high-tax country with a tax rate of 40%. A
value of 20% is assigned to the low-tax country 2. Any variations of the tax rate
differential are realized by keeping the tax rate of the high-tax country fix and
changing the tax rate of the low-tax country.
We compare the optimal value of the four possible network configurations en-
compassed by the suggested two-market, two-product setting. These values are ob-
tained by prescribing the respective network structures to the model (Lu and Van
Mieghem 2007). By incorporating a predetermined network configuration into the
model formulation the corresponding optimal value for a given parameter configu-
ration is then realized from optimization.
from: Kersten/Blecker/Flämig (Eds.), Global Logistics Management.
© Erich Schmidt Verlag GmbH & Co., Berlin 2008
Extract, for more details visit ESV.info/978 3 503 11228 9
Transfer Pricing, Taxation and Capacity Planning in International Manufacturing Networks

13

In order to compare the performance of these four network configurations, the
relative value of flexibility is used as a benchmark. In this fashion, the relative dif-
ference between the values of the most general hybrid structure (recall that invest-
ment costs are identical) and any other structure defines the relative lost value of
the respective less general network structure.
5.1 Impact of Taxation
First, a special case of the model is analyzed to obtain insights on the impact of tax
differences on the MNE’s network design. It is assumed that capacity decisions can
be made after demand uncertainty has been resolved. Consequently, the overall
network design problem reduces to a deterministic optimization problem. Local
market demand perceived at the assembly facilities is supposed to be equal to the
standard mean value of the stochastic case.

Figure 3: Deterministic network values as a function of the transportation cost differential (left
side) and tax rate differential (right side), respectively.

Figure 3 demonstrates the global after-tax profit achieved by the network configu-
rations (network value) as a function of the tax rate differential and transportation
cost discrepancy, respectively. The figure shows that increasing transportation costs
for transhipments, as measured by an increasing transportation cost differential, en-
hance the attractiveness of the market area structure, but decrease the attractiveness
of any other configuration. Only when supposing zero transportation cost differen-
tial, centralizing component production in the low-tax country would become opti-
mal. However, in between these two extremes a hybrid structure emerges from op-
timization when some critical thresholds are passed (recall that the hybrid structure
corresponds to investments in component capacity on both markets). This result
seems to be in conflict with the statement of Lu and Van Mieghem (2007) that ex-
pensive transhipments, being substitutable processing activities to cope with de-
mand deviations from the operating point, are never optimal in a deterministic set-
ting.
400
500
600
700
800
01234
Transportation Cost
Differential
Network Value
Hybrid Offshore
Onshore Market

400
500
600
700
800
0% 10% 20% 30% 40%
Tax Rate
Differential
Hybrid Offshore
Onshore Market
from: Kersten/Blecker/Flämig (Eds.), Global Logistics Management.
© Erich Schmidt Verlag GmbH & Co., Berlin 2008
Extract, for more details visit ESV.info/978 3 503 11228 9
David Francas

14
All other things being equal, in the absence of taxes demand would always be
satisfied by the less expensive alternative that is supplying components from the
production plant located within the same country due to lower transportation costs.
However, in the presence of taxes a non-homogenous domain is created in which a
realized profit in one country may not have the same value because of different tax
rates. An arbitrary use of transfer pricing and transportation cost allocation enables
the MNE to (partly) shift a profit realized in the high-tax country to the low-tax one
when relying on component flows that pass national borders. Thus, cross-border
trade flows, even though more expensive, might become a valuable option to sat-
isfy demand, as it might be possible to reduce the profit of the high-tax country as
close as possible to zero. In resorting to a combination of substitutable processing
activities, a hybrid structure might be capable of effectively exploiting the tax rate
differential while simultaneously minimizing transportation costs compared to the
network configuration of component centralization in the low-tax country. Subse-

quently, a combination of component flows within a single country and cross-
border transhipments might be the cost minimizing option.
If such a combination cannot be found, a market area structure becomes opti-
mal when transportation cost savings through local sourcing outweigh possible tax
savings via profit-shifting activities. Otherwise, in the event of less expensive tran-
shipments, it is more likely to satisfy the optimality condition of offshore produc-
tion. In contrast, onshore production – although being a process-focused network
configuration with cross-border trade flows too – is dominated by the market area
structure. This is due to the fact that at most all profit
4
realized from satisfying de-
mand in the low-tax country could be transferred to that country, while all profit
from satisfying local demand remains in the high-tax country. Hence, this is com-
parable to the profit allocation obtained when applying a market area structure,
though at the expense of higher transportation costs (when assuming a transporta-
tion cost differential greater than zero).
It should be noted that these critical thresholds do not only depend on the rela-
tive magnitude of transportation cost and tax rate differentials, but further on the
values of several exogenous parameters. Ceteris paribus, higher market prices and
lower capacity and production costs for assembly render an offshore configuration
optimal. However, with increased component production or capacity costs one
would rather focus on each market separately, thus resorting to the market area
network structure. These results are reasonable when considering that lower prices
and higher costs reduce the trade flows’ per unit net income (including costs for
capacity investments) thereby lowering the impact of profit-shifting activities
among subsidiaries.
___________________
4
This would be the case when a lower bound on transfer prices equal to total component pro-
duction costs is identified so that in equating optimal transfer price to this lower bound the

production plant of the high-tax country would earn zero profit for supplying the assembling
facility of the low-tax country.
from: Kersten/Blecker/Flämig (Eds.), Global Logistics Management.
© Erich Schmidt Verlag GmbH & Co., Berlin 2008
Extract, for more details visit ESV.info/978 3 503 11228 9
Transfer Pricing, Taxation and Capacity Planning in International Manufacturing Networks

15

Summarizing, this first study shows that strategic planners of a firm operating
in a global environment should consider the joint effects of transportation costs and
taxation, while taking into account that a clear dominance of one configuration
strategy over another is not evident from the relative level of transportation cost
and tax rate differentials. In fact, the MNE must trade off operating costs for trans-
ferring common components and benefits derived from exploiting differentials in
corporate tax rates. It should be pointed out that even under demand certainty the
presence or addition of cross-border transhipments (hybrid structure) might in-
crease optimal network value significantly and consequently strengthen the com-
petitive position of the MNE. These findings underscore the importance of analyz-
ing the network design problem in the more complex setting of stochastic demand.
5.2 Impact of Demand Uncertainty
Especially in international scenarios uncertainty becomes more pronounced than in
domestic ones, making the network design decision even more critical. Besides ex-
ploiting arbitrage opportunities offered by taxation differentials, the international
manufacturing network should be designed to be robust to changing operating con-
ditions as capacity investment decisions are typically made before uncertainty has
been resolved. Figure 4 plots the network values as a function of the tax rate differ-
ential and transportation cost differential, respectively.

Figure 4: Stochastic network values as a function of the transportation cost differential (left side)
and tax rate differential (right side), respectively.
300
400
500
600
700
01234
Transportation Cost
Differential
Network Value
Hybrid Offshore
Onshore Market
300
400
500
600
700
0% 10% 20% 30% 40%
Tax Rate
Differential
Hybrid Offshore
Onshore Market
from: Kersten/Blecker/Flämig (Eds.), Global Logistics Management.
© Erich Schmidt Verlag GmbH & Co., Berlin 2008
Extract, for more details visit ESV.info/978 3 503 11228 9
www.ESV.info

69,00

€(D)
^ Nowadays the supply chain management framework
is improved continuously by new elements, such as quality
or risk management. This affects the inter-organizational
management system in a global perspective and the design
of every single node in a supply chain.
This volume, edited by Wolfgang Kersten, Thorsten
Blecker and Heike Flämig provides a valuable insight into:
• latest findings in the area of global logistics
• new concepts for product induced logistics strategies
• quality and risk management for logistics

and supply chains
• the challenge of sustainability
• performance and capacity management.
With this volume you will learn how to optimize the
management of existing networks in different fields of
logistics.
Extract, for more details visit ESV.info/978 3 503 11228 9