ARTICLE IN PRESS

Int. J. Production Economics 93–94 (2005) 253–262

www.elsevier.com/locate/dsw

Inventory management in supply chains:

A bargaining problem
Eric Sucky
Department of Supply Chain Management, School of Business and Economics, Goethe-University Frankfurt, Mertonstrasse 17,
D-60054 Frankfurt, Germany

Abstract

This paper is focused on supply chain management from the perspective of inventory management. The coordination
of order and production policies between buyers and suppliers in supply chains is of particular interest. When a buyer of
an item decides independently, he will place orders based on his economic order quantity (EOQ). However, the buyer’s
EOQ may not lead to a favorable policy for the supplier. A cooperative order and production policy can reduce total
cost significantly. Should the buyer have the dominant position to impose his EOQ on the supplier, then consequently
no incentive exists for him to deviate from his EOQ in order to choose a cooperative policy. To induce the buyer to
order in quantities more favorable to the supplier, the supplier could offer a cooperative policy associated by a side
payment to the buyer. The research presented in this paper provides several bargaining models depending on alternative
production policies of the supplier. With these bargaining models the offered cooperative policy and the offered side
payment can be derived.
r 2004 Elsevier B.V. All rights reserved.

Keywords: Inventory management; Joint economic lot size; Buyer–supplier relationships; Bargaining problem

1. Introduction reach goals that cannot be achieved acting

1.1. The problem of coordination: EOQ and ELS
solutions
The term supply chain management refers to
cooperative management of materials and infor-
mation flows between supply chain partners, to
Tel.: +49-69-79828201; fax: +49-69-79828414.
E-mail address: esucky@wiwi.uni-frankfurt.de (E. Sucky).
individually. This paper focuses on the supply
chain from the perspective of inventory manage-
ment.
In contrast to multi-echelon inventory manage-
ment, that coordinates inventories at multiple
locations of one company, a joint inventory
replenishment policy in supply chains involves
coordination among multiple firms (Johnson and
Pyke, 2001, pp. 794–795). Therefore, the coordina-
tion of order policy and production policy between

0925-5273/$ - see front matter r 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.ijpe.2004.06.025
 ARTICLE IN PRESS
254 E. Sucky / Int. J. Production Economics 93–94 (2005) 253–262
buyers and suppliers in supply chains is of special
interest (Landeros and Lyth, 1989, pp. 146–147).
When the buyer and supplier treat inventory
problems singly under deterministic conditions,
the economic order quantity (EOQ) formula or the
economic lot size (ELS) formula gives an optimal
solution. However, in general, an order policy
based on the EOQ solution is undesirable to the
supplier and likewise, a production and delivery
policy based on the ELS solution is unacceptable
to the buyer (Lu, 1995, p. 312).
1.2. JELS models with equal and unequal sub-
batches
The problem of coordination between the order
policy and the production policy of a buyer and a
supplier has received considerable attention in
recent years. Goyal (2000), Goyal and Gupta
(1989), Joglekar and Tharthare (1990), Thomas
and Griffin (1996), and Sharafali and Co (2000)
give detailed reviews of integrated buyer–supplier
inventory models. A number of authors, including
Goyal (1976, 1988), Banerjee (1986b), Landeros
and Lyth (1989), Chatterjee and Ravi (1991) and
Agrawal and Raju (1996) demonstrate methods to
gain cost savings. They suggest joint economic lot
size (JELS) models where the objective is to
minimize the joint total relevant costs for both
the buyer and the supplier. It is shown that an
integrated inventory replenishment policy is more
desirable than individual optimal policies of the
parties involved. While all models use the accepted
EOQ formula to determine buyer’s individual
optimal order policy, the distinguishing feature is
the assumed production and delivery policy of the
supplier. Goyal (1976) assumes an infinite produc-
tion rate for the supplier. Banerjee (1986a)
generalizes Goyal’s model by integrating a finite
production rate, assuming that the supplier
follows a lot-for-lot policy. Goyal (1988) further
relaxes the lot-for-lot assumption by assuming that
each production batch is dispatched to the buyer
in an integer number of equal sized sub-batches.
Landeros and Lyth (1989) further generalize these
models by incorporating fixed delivery cost asso-
ciated with each shipment to the buyer. However,
these models assume that the whole production
batch must be finished before any shipments from
the batch can take place. Agrawal and Raju (1996)
consider that the supplier may wish to ship a
number of equal sized sub-batches before the
whole production batch is finished. Based on a
much earlier idea set out by Goyal (1977), Goyal
(1995) shows the use of unequal sized sub-batches.
This production and delivery policy involves
successive shipments within a production batch
so that the size of the sub-batches increases
according to a geometric series. The ith sub-batch
size within a production batch will be: (first
shipment size)
(production rate/demand rate)(i1).
Chatterjee and Ravi (1991) present an equivalent
model considering fixed delivery costs associated
with each shipment. Viswanathan (1998) shows
that neither a policy with equal sized sub-batches
nor a policy with unequal sized sub-batches
dominates the other. Hill (1997), however, sug-
gests a more general class of supplier’s production
and delivery policies. The ith sub-batch size is
given by: (first shipment size)y(i1), with 1pyp
(production rate/demand rate). He shows that, at a
fixed transportation cost per shipment, the total
costs are smaller then using equal sized sub-
batches. Another kind of delivery policy is shown
in Goyal and Szendrovits (1986), Goyal and
Nebebe (2000) and Goyal (2000). A certain
number of unequal-sized sub-batches are com-
bined with a number of equal sized sub-batches.
However, one may identify two deficiencies of
delivery policies with unequal sized sub-batches.
First, the capacity of the handling, packing and
shipping equipment must be at least equal to the
largest sub-batch size and hence, becomes under-
utilized for smaller sub-batch sizes, which leads to
idle-capacity costs (Goyal and Szendrovits, 1986,
p. 204). On the other hand, if the size of a sub-
batch exceeds the load-capacity of the transporta-
tion unit more then one shipment per sub-batch
are necessary. Therefore, using unequal sized sub-
batches, it is inappropriate to treat shipment costs
as being independent of the sub-batch size
(Szendrovits, 1978, p. 1018). Second, supply and
receipt of unequal sized sub-batches associated
with order intervals of different length cause a
prohibitive operational planning and control effort
for the supplier and the buyer (Agrawal and Raju,
 ARTICLE IN PRESS
E. Sucky / Int. J. Production Economics 93–94 (2005) 253–262
1996, p. 580). Nevertheless, neither the models
with equal-sized sub-batches nor the models with
unequal-sized sub-batches consider the power
structure in buyer–supplier relationships.
1.3. Models considering the power structure in
buyer–supplier relationships
Should either party be in a position to impose
his EOQ or ELS on the other party, consequently
no incentive exists for this party to choose a
cooperative policy. Essentially, the dominant party
will be at a disadvantage if the JELS solution is
adopted. However, by adopting the JELS solution
the dominant party’s loss is more than offset by
the gain of the weaker party (Banerjee, 1986a, p.
308). Thus, to provide an incentive to choose a
joint policy the weaker party can offer a side
payment to the dominant party. The order and
delivery quantities and the side payment are
determined through a bargaining process between
the parties. The supplier’s problem to influence the
buyer’s order policy by a price discount was
analyzed by previous authors, including Monahan
(1984), Banerjee (1986b), Lee and Rosenblatt
(1986), Joglekar (1988) and Hill (1997). However,
there are very few contributions dealing with the
bargaining process between the buyer and the
supplier. Side payment problems can be consid-
ered as two-person-nonzero-sum games in which
the buyer and the supplier try to maximize their
individual gains.
The research presented in this paper offers
bargaining models, assuming that the buyer has
the bargaining power to enforce his EOQ on the
supplier in case of a break-down in negotiations.
Therefore, the supplier has to offer a side payment
that will certainly guarantee that the buyer accepts
the contract. With these bargaining models opti-
mal contracts can be derived.
2. Individual optimal order and production policies
2.1. Individual optimal order policy of the buyer
The discussion and analysis in this paper is
restricted to the case of a single supplier (P) and
255
single buyer (A) of a specific item. The demand
rate, b [unit/period], for the item is assumed
constant and deterministic. Shortages are not
permitted at the buyer’s end, and the time horizon
over which the item is ordered by (A) and supplied
by (P) is infinite. The lead-time for (A) is zero. If
(A) operates independently, his total relevant cost
per period is given by
b xA
K A ðxA Þ ¼ B þ hA ð1Þ
xA 2
with b being the buyer’s demand rate [unit/period],
xA the buyer’s order quantity per order [unit],
B the buyer’s ordering cost per order [$], and hA
the buyer’s inventory holding cost [$/unit and
period].
The objective of (A) is to minimize his total
relevant cost per period. It is easy to identify that
for an optimal policy every order is received
precisely when the inventory level drops to zero
(Bramel and Simchi-Levi, 1997, pp. 145–147). The
EOQ and the minimum total relevant cost per
period are given by
sffiffiffiffiffiffiffiffiffi
2Bb pffiffiffiffiffiffiffiffiffiffiffiffiffiffi
xA ¼ ; K A ðxnA Þ ¼ 2BbhA : ð2Þ
hA
2.2. Individual optimal production and delivery
policy of the supplier
The supplier (P) manufactures the regarded item
at a finite production rate of d [unit/period], with
d4b, in batches of xPP [unit] and a batch set-up
cost R1 [$]. Each batch is dispatched to (A) in an
integer number of equal sized sub-batches xPT
[unit]: xPP ¼ IxPT , with I 2 IN. When a produc-
tion batch is split up into an integer number of
sub-batches, shipping costs have to be considered
(Landeros and Lyth, 1989, p. 151). With each
shipment a fixed delivery cost, R2 [$], is associated.
The objective for (P) is to determine the optimal
production and delivery policy leading to a
minimum of total relevant cost per period. The
total costs comprise the production set-up cost, the
cost of handling and processing the shipments to
(A) and the inventory holding cost. When solving
the problem it has to be considered whether lot
streaming is allowed or not.
 ARTICLE IN PRESS

256 E. Sucky / Int. J. Production Economics 93–94 (2005) 253–262

In case of lot streaming, assuming an unin- to the individual optimal shipment quantity
terrupted production run, any shipments can be ðxPP ¼ xPT Þ. (P) follows a lot-for-lot policy, i.e.
made from a production batch before the whole after finishing the production batch he ships the
batch is finished. However, some suppliers cannot whole batch to (A). The individual optimal
accommodate lot streaming because of regula- production and delivery policy and the total
tions, material handling equipment, or production relevant cost per period are given by
restrictions (Silver et al., 1998, p. 657). Without lot sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
streaming the whole production batch must be 2dðR1 þ R2 Þ
xPT ¼ xPP ¼ ;
finished before any shipments can be made from hP
the batch (Hill, 1997, p. 493). The opportunity of rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
P   2ðR1 þ R2 ÞhP
lot streaming affects supplier’s average inventory. K ðI ¼ 1; xPT Þ ¼ b : (6)
According to Goyal (1988) the average inventory d
without lot streaming will be ðxPT =2ÞðIð1 þ b=dÞ  1Þ
[unit]. With lot streaming, Hill (1997) and Agrawal 2.2.2. Individual optimal production and delivery
and Raju (1996) show the average inventory to be policy with lot streaming
ðxPT =2ÞðI  1  Ib=d þ 2b=dÞ [unit]. The supplier’s If (P) operates independently, using the oppor-
inventory holding cost is given by hP [$/unit and tunity of lot streaming, his total relevant cost per
period]. period is given by
b b
2.2.1. Individual optimal production and delivery K P ðI; xPT Þ ¼ R1 þ R2
IxPT xPT
policy without lot streaming  
xPT Ib 2b
If (P) operates independently, without the þ I 1 þ hP : (7)
2 d d
opportunity of lot streaming, his total relevant
cost per period is given by Using lot streaming the optimum value of xPT
b b for a given value of I is given by
K P ðI; xPT Þ ¼ R1 þ R2 sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
IxPT xPT
    2bðR1 =I þ R2 Þ
xPT b xPT ðIÞ ¼ : ð8Þ
þ I 1þ  1 hP : (3) hP ðI  1  Ib=d þ 2b=dÞ
2 d
The objective of (P) is to minimize his total Substituting xPT in (7) leads to
relevant cost per period. For a given value of I the pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
K P ðIÞ ¼ 2bhP ðR1 =I þ R2 ÞðI  1  Ib=d þ 2b=dÞ:
optimum value of xPT is given by
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ð9Þ
2bðR1 =I þ R2 Þ
xPT ðIÞ ¼ : ð4Þ The optimal integer number of shipments is
ðIð1 þ ðb=dÞÞ  1ÞhP
given by the I  that satisfies
Substituting xPT in (3), the total relevant cost K P ðI   1ÞXK P ðI  ÞpK P ðI  þ 1Þ: ð10Þ
per period as a function of the integer number I of
shipments per batch production run is given by Substituting (9) into (10) and appropriately
P
K ðIÞ rearranging terms, (10) is equivalent to
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
      R1 ð2b=d  1Þ
b b R1 I  ðI  þ 1ÞX XI  ðI   1Þ: ð11Þ
¼ 2bhP R1 1   R2 þ R2 I 1   : R2 ð1  b=dÞ
d d I
(5)
As can be seen from (11), the optimal integer
It is easy to see that the optimal integer number number of shipments can only be determined for
I of shipments is I  ¼ 1. Without lot streaming the 2b4d. In case of 2bpd a lot-for-lot policy will be
individual optimal production lot size is identical assumed.
 ARTICLE IN PRESS

E. Sucky / Int. J. Production Economics 93–94 (2005) 253–262 257

3. Integrated order and production policies quantity xG ¼ xA ¼ xPT [unit] can be derived as
follows (Agrawal and Raju, 1996, p. 582):
(P) delivers equal-sized sub-batches of xPT [unit].  
b R1
(A) orders quantities of xA [unit]. So, in the JELS K G ðI G ; xG Þ ¼ Bþ þ R2
xG IG
solution the sub-batch size corresponds to the   
order quantity, i.e. xA ¼ xPT . The joint total xG I G b 2b
þ IG  1  þ hP þ hA : (15)
relevant cost per period for a joint order and 2 d d
delivery quantity xG ¼ xA ¼ xPT without lot
streaming is given by (Goyal, 1988, p. 237; Equivalent to the case without lot streaming,
Landeros and Lyth, 1989, p. 154) presented above, the joint economic order and
  production policy for both (A) and (P) is given by
b R1
K G ðI G ; xG Þ ¼ Bþ þ R2
xG IG xA ¼ xPT ¼ xG
    sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
xG b
þ I G hP 1 þ  hP þ hA : 2bðB þ R1 =I G þ R2 Þ
2 d ¼ ;
ðI G  1  I G b=d þ 2b=dÞhP þ hA

(12)
xPP ¼ I G xG ; (16)
The joint economic order and delivery policy is
given by
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi where the joint optimal integer number of ship-
2bðB þ ðR1 =I G Þ þ R2 Þ ments is given by the I  that satisfies
xA ¼ xPT ¼ xG ¼ ;
I G hP ð1 þ b=dÞ  hP þ hA R1 ðhA þ hP ð2b=d  1ÞÞ
I G ðI G  1Þp pI G ðI G þ 1Þ:
xPP ¼ I G xG ; (13) hP ðB þ R2 Þð1  b=dÞ
where the joint optimal integer number of ship- ð17Þ
ments is given by the I G that satisfies
Fig. 2 shows the inventory cycles for (A) and (P)
R1 ðhA  hP Þ
I G ðI G þ 1ÞX   XI G ðI G  1Þ: with lot streaming.
ðB þ R2 ÞhP 1 þ b=d However, for xA axPT , obviously the most
ð14Þ common case, the optimal joint order and delivery
Fig. 1 shows the inventory cycles for (A) and (P) quantity xG is situated in the interval between the
without lot streaming. individual optimal solutions, i.e. xG 2 xA ; xPT ½.
If (P) uses lot streaming, the joint total relevant For any joint policy xG 2 xA ; xPT ½ the buyer’s
cost per period for any joint order and delivery
Buyer(A)
Buyer (A)
Inventory

xA
xA
Inventory

Time
Time
Supplier (P)
Supplier (P)
xPP
Inventory

Inventory

xPT

Time Time

Fig. 1. Inventory cycles without lot streaming (I ¼ 2). Fig. 2. Inventory cycles with lot streaming (I ¼ 3).
 ARTICLE IN PRESS

258 E. Sucky / Int. J. Production Economics 93–94 (2005) 253–262

total relevant cost will be 4. The bargaining game

b xG
K A ðxG Þ ¼ B þ hA The bargaining game belongs to the class of
xG 2
  two-person-nonzero-sum games in which (A) and
1 xA xG
¼ þ  K A ðxA Þ (18) (P) try to maximize their individual gains. The
2 xG xA game can be described as follows:
with
   (A) has the power to enforce his EOQ on (P) in
1 xA xG
þ 41: ð19Þ case of a break-down in negotiations. The threat
2 xG xA point, realized in the case of a break-down in
If (A) behaves individually rational, he selects his negotiations (Eichberger, 1993, p. 238), is given
individual optimal policy xA . Should (A) possess by ðK A ðxnA Þ; K P ðI A ; xA ÞÞ.
the dominant position to enforce his EOQ on (P),  (P) offers a joint policy x~ G [unit] with an
then no incentive exists for him to choose a joint associated side payment z [$/period]. He makes
policy xG 2 xA ; xPT ½. Essentially, relation (19) a take-it-or-leave-it-offer. In the first stage (P)
shows that (A) will make a loss if the JELS makes an offer, and then, in the second stage,
solution is adopted. Without bargaining, (P) must (A) can either accept or reject. The game is
adapt the buyer’s order policy. For a given immediately terminated after acceptance or
individual order policy xA the optimal integer refusal by (A) (Eichberger, 1993, pp. 55–57).
number of shipments is given by the I A that satisfies  (P) has complete information about the cost
P P function of (A).
K ðxA ; I A  1ÞXK ðxA ; I A ÞpK P ðxA ; I A þ 1Þ:
 The feasible cost combinations are given by
ð20Þ ðK A ðxG Þ  z; K P ðI; xG Þ þ zÞ.
With (20) follows
2bR1 The objective of (P) is to minimize his total
I A ðI A  1Þp pI A ðI A þ 1Þ
ðxA Þ2 hP ð1
þ b=dÞ relevant cost per period. The following non-linear
minimization problem has to be solved:
ðwithout lot streamingÞ
min K P ðI; xG ; zÞ ¼ K P ðI; xG Þ þ z
and ð22Þ
s:t:
2bR1
I A ðI A  1Þp pI A ðI A þ 1Þ
ðxA Þ2 hP ð1
 b=dÞ
K A ðxG Þ  zpK A ðxnA Þ; ð23Þ
ðwith lot streamingÞ: (21)
Nevertheless, from the supplier’s point of view, in xG ; zX0; I 2 IN: ð24Þ
comparison to the individual optimal order policy
of (A), any joint policy xG 2 xA ; xPT ½ leads to lower Condition (23) ensures individual rationality: it
total relevant cost. Therefore, to provide an must be attractive for (A) to accept the offer. For a
incentive to choose a joint policy (P) can offer a given value of I, the optimization problem with
side payment to (A), which compensates the constraints can be transformed to the following
increase in total relevant cost resulting from this unconstrained minimization problem:
policy. A side payment is defined as an additional
monetary transfer between the parties involved that min LðxG ; zÞ ¼ K P ðI; xG Þ þ z
is used as an incentive for deviating from the  lðK A ðxA Þ  K A ðxG Þ þ zÞ: (25)
individual optimal policy (Rubin and Carter, 1990,
p. 22). (P) must determine the offered delivery As the cost functions K P ðI; xG Þ and K A ðxG Þ are
quantity, the offered side payment and his produc- strictly convex in xG , the Karush–Kuhn–Tucker-
tion policy simultaneously. (KKT-) conditions are sufficient to derive an
 ARTICLE IN PRESS

E. Sucky / Int. J. Production Economics 93–94 (2005) 253–262 259

optimal solution: Substituting (32) and (30) into (22), the total
P A relevant cost per period of (P) is given by
qL qK ðI; xG Þ qK ðxG Þ
¼ þl X0; ð26Þ
qxG qxG qxG K P ðIÞ
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
¼ 2bðB þ R1 =I þ R2 ÞðIhP ð1 þ b=dÞ  hP þ hA Þ
 P  pffiffiffiffiffiffiffiffiffiffiffiffiffiffi
qL qK ðI; xG Þ qK A ðxG Þ  2BbhA : (33)
xG ¼ xG þl ¼ 0;
qxG qxG qxG
Minimizing of (33) is achieved by minimizing
ð27Þ
P R1
K^ ðIÞ ¼ ðB þ R2 ÞIhP ð1 þ b=dÞ þ ðhA  hP Þ:
I
qL qL
¼ 1  lX0 and z ¼ zð1  lÞ ¼ 0; ð28Þ ð34Þ
qz qz
The optimal integer number of shipments is
given by the I~ that satisfies
qL
 ¼ K A ðxA Þ  K A ðxG Þ þ zX0 and P P
ql ~ K^ P ðI~ þ 1Þ:
K^ ðI~  1ÞXK^ ðIÞp ð35Þ
lðK A ðxA Þ  K A ðxG Þ þ zÞ ¼ 0: (29)
Substituting (34) into (35) and appropriately
rearranging terms, (35) is equivalent to (14). (P)
For xG axA follows z40. From z40 and (28) offers exactly the joint economic policy (13),
follows l ¼ 1. For l ¼ 1 and (29) follows presented in the preceding section, associated with
z ¼ K A ðxG Þ  K A ðxA Þ a (theoretical) side payment in accordance with (30).
    pffiffiffiffiffiffiffiffiffiffiffiffiffiffi In the case of lot streaming, with (31),
1 xA xG
¼ þ  1 2BbhA : (30) xG ¼ xA ¼ xPT , (1) and (7) follows
2 xG xA sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2bðB þ R1 =I þ R2 Þ
x~ G ðIÞ ¼ : ð36Þ
(P) will compensate exactly the increase in cost ðI  1  Ib=d þ 2b=dÞhP þ hA
of (A), induced by deviating from the individual
optimal order policy. Condition (23) is fulfilled. Using the opportunity of lot streaming, compar-
The theoretical side payment (30) leads to the able to the case without lot streaming, (P) offers
buyer being just indifferent between accepting and the joint economic policy (16), presented in the
rejecting the supplier’s offer. To create a practical preceding section, associated with a side payment
side payment a small extra incentive of d [$/period] in accordance with (30). In both cases, (P) behaves
should be added. If (A) behaves individually like a master planner and selects the joint
rational, he accepts an offer associated with a side economic order and production policies. A funda-
payment zþd [$/period] and realizes mental principle is illustrated here: if the two
K A ðxG Þ  z  doK A ðxA Þ. It remains to be solved, parties negotiate cooperatively with complete
which joint policy x~ G will be offered by (P). From information, a side payment will lead to a joint
l ¼ 1, (27) and xG 40 follows economic order and production policy (Rubin and
Carter, 1990, p. 25). This result will be illustrated
qK P ðI; xG Þ qK A ðxG Þ by a numerical example.
þ ¼ 0: ð31Þ
qxG qxG

At first, the case where (P) cannot use lot 5. A numerical example
streaming will be analyzed. From (31),
xG ¼ xA ¼ xPT , (1) and (3), follows 5.1. The bargaining situation
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2bðB þ R1 =I þ R2 Þ
x~ G ðIÞ ¼ : ð32Þ Consider the following information for both (A)
IhP ð1 þ b=dÞ  hP þ hA and (P) displayed in Table 1.
 ARTICLE IN PRESS
260 E. Sucky / Int. J. Production Economics 93–94 (2005) 253–262
Table 1
Numerical example
Buyer (A) Supplier (P)
b=10 000 [unit/period] d=15 000 [unit/period]
B=45 [$] R1 ¼ 360 [$], R2 ¼ 40 [$]
hA ¼ 10 [$/unit and period] hP ¼ 3 [$/unit and period]
Fig. 3. Individual optimal solutions.
The individual optimal policy of (A), i.e. five
orders per period of xA ¼ 300, leads to total
relevant cost of K A ðxA Þ ¼ 3000. The individual
optimal policy of (P) is given by I  ¼ 1,
xPT ¼ xPP ¼ 2000, with K P ðI  ; xPT Þ ¼ 4000 (with-
out lot streaming) or xPT ¼ 894:43, I  ¼ 3,
xPP ¼ 2683:29, with K P ðI  ; xPT Þ ¼ 3577:71 (with
lot streaming), i.e. xA axPT applies. Fig. 3 shows
these individual optimal solutions.
5.2. Bargaining solution without lot streaming
Without lot streaming, Table 2 shows the results
for (A) and (P) in case of
 (A) and (P) choosing the JELS solution,
 (P) adapting his delivery policy to the individual
optimal order policy of (A) and
 (P) offering a joint policy associated by a side
payment.
The individual optimal policy of (P) leads to
K P ðI  ; xPT Þ ¼ 4000. As the buyer’s EOQ of 300
units is the order and delivery quantity, the
supplier’s total relevant cost, according to Eq. (3)
and condition (20), is K P ðI A ; xA Þ ¼ 6883:33. The
absolute cost penalty of (P) is 2883.33 [$/period],
i.e. an increase of 72.08 percent. Adopting the
JELS solution, 558.36 units, would result in a cost
penalty of 597.73 [$/period] on the part of the
buyer, i.e. an increase of 19.92 percent. Therefore,
(A) has no incentive to choose the JELS solution.
But, adopting the JELS solution, (P) saves 988.96
[$/period]. Thus the buyer’s loss is more than
offset by the supplier’s gain. If (P) offers a side
payment of 597.73 [$/period] and persuades (A) to
change his order policy from 300 to 558.36 units,
the increase in total relevant cost of (A) will be
compensated exactly by (P). If (A) accepts the
offer (P) gains 391.23 [$/period].
5.3. Bargaining solution with lot streaming
These results change in case of using lot
streaming on supplier’s side (see Table 3).
The individual optimal policy of (P) leads to
K P ðI  ; xPT Þ ¼ 3577:71. If (A) enforces his EOQ on
(P); his absolute cost penalty is 588.95 [$/period],
i.e. 16.46 percent. Adopting the JELS solution,
398.34 units, would result in a cost penalty of
102.5 [$/period] on the part of the buyer and at the
same time, the supplier saves 261 [$/period]. If (P)
persuades (A) to change his order policy from 300
to 398.34 units, offering a side payment of 102.5
[$/period], (P) gains 158.5 [$/period]. With lot
streaming, (P) can react more flexible on a given
order policy of the buyer. The cost penalty of
adopting the buyer’s individual optimal order
policy, and likewise, the gain induced by the
bargaining solution are smaller than without lot
streaming.
6. Conclusion
In this paper a joint order and production policy
is developed as a bargaining solution assuming
that the buyer has the dominant position. The
paper presents an analytical approach to deter-
mining the terms of a supplier-oriented optimal
side payment scheme. It is shown that the joint
 ARTICLE IN PRESS

E. Sucky / Int. J. Production Economics 93–94 (2005) 253–262 261

Table 2
Illustration of the solution without lot streaming

Joint policy Buyer’s cost Supplier’s cost

Buyer’s solution xG ¼ xA ¼ xPT ¼ 300 K A ðxA Þ ¼ 3000 K P ðI A ; xA Þ ¼ 6883:33

I A ¼ 4; xPP ¼ 1200
JELS solution xG ¼ xA ¼ xPT ¼ 558:36 K A ðxG Þ ¼ 3597:73 K P ðI G ; xG Þ ¼ 5894:37
I G ¼ 2; xPP ¼ 1116:72
Bargaining solution x~ G ¼ xA ¼ xPT ¼ 558:36; I~ ¼ 2 K A ðx~ G Þ  z ¼ 3000 ~ x~ G Þ þ z ¼ 6492:1
K P ðI;
xPP ¼ 1116:72; z ¼ 597:73

Table 3
Illustration of the solution with lot streaming

Joint policy Buyer’s cost Supplier’s cost

Buyer’s solution xG ¼ xA ¼ xPT ¼ 300 K A ðxA Þ ¼ 3000 K P ðI A ; xA Þ ¼ 4166:66

I A ¼ 9; xPP ¼ 2700
JELS solution xG ¼ xA ¼ xPT ¼ 398:34 K A ðxG Þ ¼ 3102:5 K P ðI G ; xG Þ ¼ 3905:66
I G ¼ 7; xPP ¼ 2725:38
Bargaining solution x~ G ¼ xA ¼ xPT ¼ 398:34; I~ ¼ 7 K A ðx~ G Þ  z ¼ 3000 ~ x~ G Þ þ z ¼ 4008:16
K P ðI;
x~ PP ¼ 2725:38; z ¼ 102:5

optimal policy with an according side payment
results from negotiations. Adopting the bargaining
solution, the gain of the supplier depends on his
production policy, i.e. if the supplier has the
opportunity to use lot streaming or not. With lot
streaming the supplier is provided with more
flexibility to react on buyer’s individual optimal
order policy. The supplier realizes a lower cost
penalty with lot streaming than without lot
streaming. Nevertheless, in both cases, with or
without lot streaming, the supplier benefits from
the presented bargaining solutions. This raises a
crucial practical issue. In order to determine the
side payment the supplier must obtain the
numerical estimates of buyer’s ordering cost per
order and buyer’s inventory holding cost. While
the buyer’s periodical demand can be anticipated
from the past ordering behavior of the buyer, it is
very difficult to estimate buyer’s holding and
ordering costs. This is true unless the buyer is
willing to disclose the true values of his cost
parameters (Landeros and Lyth, 1989, p. 313).
However, the buyer’s holding cost can be specified
as an implicit function of his ordering cost and his
EOQ and, likewise, the buyer’s ordering cost can
be specified as an implicit function of his holding
cost and his EOQ. The relative insensitivity of z ¼
K A ðxG Þ  K A ðxA Þ to measurement errors permits
the establishment of a ‘‘good’’ side payment.
Therefore, a bargaining solution succeeds if (P)
estimates the cost structure of (A) with sufficient
accuracy.
References
Agrawal, A.K., Raju, D.A., 1996. Improved joint economic lot
size model for a purchaser and a vendor. In: Khan, M.K.,
Wright, C.S., Whalley, R. (Eds.), Advanced Manufacturing
Processes, Systems, and Technologies (AMPST 96),
pp. 579–587.
Banerjee, A., 1986a. A joint economic-lot-size model for
purchaser and vendor. Decision Sciences 17, 292–311.
Banerjee, A., 1986b. Notes on ‘‘A quantity discount pricing
model to increase vendor profits’’. Management Science 32,
1513–1517.
Bramel, J., Simchi-Levi, D., 1997. The Logic of Logistics—
Theory, Algorithm, and Applications for Logistics Manage-
ment. New York.
Chatterjee, A.K., Ravi, R., 1991. Joint economic lot-size model
with delivery in sub-batches. Opsearch 28, 118–124.
 ARTICLE IN PRESS
262 E. Sucky / Int. J. Production Economics 93–94 (2005) 253–262
Eichberger, J., 1993. Game Theory for Economists. Academic
Press, San Diego.
Goyal, S.K., 1976. An integrated inventory model for a single
supplier–single customer problem. International Journal of
Production Research 15, 107–111.
Goyal, S.K., 1977. Determination of optimum production
quantity for a two-stage production system. Operational
Research Quarterly 28, 865–870.
Goyal, S.K., 1988. ‘‘A joint economic-lot-size model for
purchaser and vendor’’: A comment. Decision Sciences 19,
236–241.
Goyal, S.K., 1995. A one-vendor multi-buyer integrated
inventory model: A comment. European Journal of Opera-
tional Research 82, 209–210.
Goyal, S.K., 2000. On improving the single-vendor single-buyer
integrated production inventory model with a generalized
policy. European Journal of Operational Research 125,
429–430.
Goyal, S.K., Gupta, Y.P., 1989. Integrated inventory models:
the buyer–vendor coordination. European Journal of
Operational Research 41, 261–269.
Goyal, S.K., Nebebe, F., 2000. Determination of economic
production—shipment policy for a single-vendor–single-
buyer system. European Journal of Operational Research
121, 175–178.
Goyal, S.K., Szendrovits, A.Z., 1986. A constant lot size model
with equal and unequal sized batch shipments between
production stages. Engineering Costs and Production
Economics 10, 203–210.
Hill, R.M., 1997. The single-vendor single-buyer integrated
production-inventory model with a generalised policy.
European Journal of Operational Research 97, 493–499.
Joglekar, P.N., 1988. Comments on: A quantity discount
pricing model to increase vendor profits. Management
Science 34, 1391–1398.
Joglekar, P.N., Tharthare, S., 1990. The individually respon-
sible and rational decision approach to economic lot size for
one vendor and many purchasers. Decision Sciences 21,
492–506.
Johnson, M.E., Pyke, D.F., 2001. Supply chain management.
In: Gass, S.I., Harris, C.M. (Eds.), Encyclopaedia of
Operations Research and Management Science. Centennial
Edition, Kluwer, Boston, pp. 794–806.
Landeros, R., Lyth, D.M., 1989. Economic-lot-size models for
cooperative interorganizational relationships. Journal of
Business Logistics 10, 146–158.
Lee, H.L., Rosenblatt, M.J., 1986. A generalized quantity
discount pricing model to increase supplier’s profits.
Management Science 32, 1177–1185.
Lu, L., 1995. A one-vendor multi-buyer integrated inventory
model. European Journal of Operational Research 81,
312–323.
Monahan, J.P., 1984. A quantity discount pricing model to
increase vendor profits. Management Science 30, 720–726.
Rubin, P.A., Carter, J.R., 1990. Joint optimality in buyer–sup-
plier negotiations. Journal of Purchasing and Materials
Management (Spring), 20–26.
Sharafali, M., Co, H.C., 2000. Some models for understanding
the cooperation between the supplier and the buyer.
International Journal of Production Research 38,
3425–3449.
Silver, E.A., Pyke, D.F., Peterson, R., 1998. Inventory
Management and Production Planning and Scheduling,
3rd Edition. Wiley, New York.
Szendrovits, A.Z., 1978. A comment on determination of
optimum production quantity for a two-stage production
system. Journal of the Operational Research Society 29,
1017–1020.
Thomas, D.J., Griffin, P.M., 1996. Coordinated supply chain
management. European Journal of Operational Research
94, 1–15.
Viswanathan, S., 1998. Optimal strategy for the integrated
vendor–buyer inventory model. European Journal of
Operational Research 105, 38–42.