25.

Supply Contracts
The basic unit of analysis when studying logistics is the enterprise. When
studying supply chain management, however, the basic unit is the inter-
enterprise commercial transaction: buying and selling.

To achieve optimal supply chain profits, both seller and buyer along a supply
chain should operate with the same goals and incentives. In reality, of course,
each actor optimizes its own interest, leading, in many cases, to lower profits for
the chain as a whole. For example, a retailer facing uncertain demand would
conduct its own risk/reward trade-off in making the ordering decision, rather than
act in the channel’s interest. This observation has lead companies to many
schemes designed to try to operate in a coordinated fashion, thereby achieving
global optimum. These include collaborative processes such as Vendor-
Managed-Inventory (VMI), Collaborative Planning, Forecasting and
Replenishment (CPFR) and many others (see chapter __).

In this section we use the single period decision process to illustrate how a seller
(say, a manufacturer or a supplier), can restructure the supply contract it offers a
buyer (say, a retailer), so the retailer would have an incentive, while placing an
order, to act in the chain’s global interest. Interestingly, not only can the supplier
improve its own expected profit, but the right supply contract can improve the
retailer’s own expected profit as well (!). The basic framework here is of a retailer
who is facing an uncertain demand. Lead-time is long relative to the selling
season, and therefore there is no opportunity to replenish once the season has
begun. The elements of the analytical framework here are similar to the
newsboy’s problem discussed in Chapter __. The retailer orders one SKU from a
supplier, and it does so in order to maximize its own expected profit. Once the
order is placed, demand is realized and both the retailer and the supplier can
calculate their profits or losses.

Just like with the newsboy problem, the risks addressed here are those
associated with supply/demand mismatch. If supply exceeds demand, items
remaining at the end of the season can only be sold by the retailer at a low
salvage value (which may be even negative value due to disposal cost). When
supply falls short of demand, sales are lost, representing not only forgone profits
but also loss of customer goodwill.

25.1. A Wholesale price Contract

The simplest scenario is when the supplier sells to the retailer the amount that
the retailer ordered, Q, at a fixed wholesale price, $W per unit, determined by the
supplier. The items cost the supplier $C per unit to make (or procure) and the
retailer can sell each one at $R per unit. Demand is a random variable, X, whose

25.1
density, f(x), and cumulative distribution, F(x), are known. Unsold items can be
salvaged by the retailer at a price of $S per unit.

This is a familiar case from the retailer’s point of view – it is facing a “newsboy”
problem. If the realized demand x, is more than the amount ordered (x > Q) the
retailer‘s revenue will be Q•R dollars and if x ≤ Q the retailer’s revenue will be
Q•R dollars. In the latter case there will be (Q-x) items left over and the retailer
can salvage them at a revenue of (Q-x)•S. Based on considerations identical to
those expressed Eq. [10.5], the expected profit for the retailer is:

E [Re tailer Profit ] = R iE [Sales ] + S iE [Unsold ] − W iQ [25.1a]

∞ Q Q
E [Retailer Profit ] = R iQi ∫
x =Q
f ( x )dx +R i ∫
x =0
x if ( x )dx + S i ∫ (Q − x )if ( x )dx − W iQ
x =0
[25.1b]

The supplier’s profit function is trivial with a wholesale contract:

E [Supplier Profit ] = W iQ − C iQ [25.2a]

The total channel’s profit is the sum of the retailer’s and the supplier’s profits

To illustrate these relationships, consider a retailer who is ordering a designer

dress for a given selling season. The dress will sell for R=$200/unit and the
supplier will charge a wholesale price of W=$135/unit. The supplier’s costs are
C=$50/unit and the retailer can salvage unsold inventory at S=$10/unit. Demand
is expected to
follow a Normal Fig 25.1 Expected Profits with Wholesale Price
$150,000
distribution with Retailer
mean µ = 800 Supplier
units and a $100,000
Total
standard deviation
Expected Profit

of σ = 150 units. $50,000

Fig. 25.1 depicts $0

the profit of the 200 400 600 800 1000 1200 1400 1600
retailer, supplier -$50,000
Order Quantity

and the channel

as a whole as a -$100,000
function of the
amount ordered
by the retailer.

The retailer will maximize its expected profit by ordering QR* , where:

25.2
 ⎛ R −W ⎞
QR* = F −1 ⎜ ⎟. [25.3]
⎝ R −S ⎠

In this equation, F-1(•) is the inverse cumulative probability function of the

demand. For our example, [(R-W)/(R-S)] = [(200-135)/(200-10)] = 0.342 and the
optimal order quantity can be calculated using Excel as:

QR* = NORMINV (0.342,800,150) = 739 items [25.4]

The profits in this case are going to be allocated as follows (see also Fig. 25.1):

Table 25.1 Profits with Optimal Retail Order

Retailer Opt Order: 739 items
Retailer Profit: $41,533
Supplier Profit $62,814
Total Profit: $104,347

The optimal order quantity for the retailer is, of course, infinity. Given that there
are no capacity constraints and the costs do not change with the order size, the
supplier’s profit will grow linearly with the order size (see Eq. [25.2]).

For the channel as a whole, however, we can ignore the wholesale price
transaction since it is internal: it is a cost to the retailer and revenue to the
supplier. For the channel as a whole, the profit function will be given by the sum
of Eq. [25.1] and Eq. [25.2], yielding

E [Channel Profit ] = R iE [Sales ] + S iE [Unsold ] − C iQ [25.5]

This equation is identical to Eq. [25.1] but with the wholesale cost, W replaced by
the supplier’s cost, C. The optimal order size from the total channel’s point of
view, QT* , is:

⎛ R −C ⎞
QT* = F −1 ⎜ ⎟ [25.6]
⎝R −S ⎠

Since C < W (otherwise the supplier will have no profit), it will always be the case
that QT* ≥ QR* . In other words, the retailer will order less than the optimal amount
from the channel point of view. In the example discussed here, [(R-C)/(R-S)] =
[(200-135)/(200-50)] = 0.433, and

QT* = NORMINV (0.342,800,150) = 921 items [25.7]

25.3
If the retailer were to order QT* = 921 items, the resulting profit allocation would
be:
Table 25.2 Optimal Channel Order
Channel Optimal Order: 921 items
Retailer Profit: $33,516
Supplier Profit $78,259
Total Profit: $111,774

As expected, the order size from the total channel point of view is larger than the
retailer’s optimal order. The total channel profit is also larger ($111,774 vs.
$104,347). Unfortunately, such increased channel profit is not likely to be
realized since the retailer’s expected profit when ordering more, is lower than
before ($33,516 vs. $41,533). Thus, even though the supplier’s profit is
substantially higher, it cannot realize it since the retailer who places the order,
has no incentive to reduce its own profits.

In order to induce the retailer to Fig. 25.2 Retiler's Optimal Order

order more, the supplier can 1000
reduce the wholesale price that it 800
charges the retailer. Using the
Order Size

600
same numerical example, Figure
25.2 depicts the retailer’s optimal 400
order size as a function of the 200
wholesale price. The order size
goes to zero for a wholesale price 0
50 75 100 125 150 175 200
of $200, since at that point the
Wholesale Price
wholesale price is the same as
the retail price and the retailer
has no chance to make any profits. At a wholesale price of $50 (which is equal to
the supplier’s cost), the retailer will order the channel-optimal quantity, 921 items.
This is clear since when W = C the retailer’s order problem is identical to the
channel’s ordering problem.

To understand this further, Fig. 25.3 depicts the expected profits for all parties as
the wholesale price changes. For each given wholesale price the retailer
orders QR* , the optimal amount from its own point of view, and the figure depicts
the resulting expected profits.

25.4
As the figure
shows, it would be Fig. 6.3 Effect of Wholesale Price
best for the $140,000
supplier to set a $120,000
wholesale price of
$180. At this price $100,000
the retailer will

Expected profits
$80,000
order 612 items Retailer Profit
and the supplier’s $60,000 Supplier Profit
profit will be Channel Profit
$40,000
$79,560. The
retailer’s profit at $20,000
this order size,
$0
however, is small
30 55 80 105 130 155 180
(the retailer -$20,000
retains at this Wholesale Price
point only 12% of -$40,000
the channel’s total
profit). As the wholesale price decreases the retailer will order more, and its profit
(as well as the channel’s profit) will grow. The channel’s expected profit will be
maximized for W = 50 (which is exactly the point where W = C), where the order
size will equal QT* = 921 items (see Eq. [25.7]). Unfortunately, at this point, the
supplier’s profit will be zero since it will be selling each item for exactly its cost.

The channel’s expected profit will always be maximized when the supplier
charges the retailer a wholesale price equal to its costs. To see this note that in
order for the channel’s optimal order size to equal what the retailer would order
(i.e., QR* = QT* ) we need that:

⎛ R −W ⎞ ⎛ R −C ⎞ [25.8]
F −1 ⎜ ⎟ = F −1 ⎜ ⎟,
⎝ R −S ⎠ ⎝ R −S ⎠

leading to the condition: W = C for overall optimality. Thus the supplier cannot
use the wholesale price in order to coordinate the channel, i.e., to induce the
retailer to order the optimal amount from the total channel point of view.

25.2. Buyback Contracts

As an incentive for the retailer to order more and move towards channel
coordination, the supplier can offer to buy back all the unsold units at the end of
the season at a price that is higher than the salvage value. With such a buy-back
contract, the retailer is, essentially, getting a higher “salvage” value for unsold
goods, thereby increasing its expected profit at a given wholesale price. With this

25.5
contract the supplier takes on some of the risk of the unsold units from the
retailer.

To see how this type of contract can coordinate the supply chain, assume that
the supplier is offering to buy all the unsold units at the end of the selling season
at a price of $B/unit. Clearly, B < W, otherwise the retailer would profit from
unsold inventory, and as discussed above, B > S.

The underlying assumption here is that the supplier actually buys back the
unsold goods and salvages them. In reality, the supplier may simply pay the
retailer ($B-$S)/unit for each unsold item and the retailer will dispose of the
items. In many settings, however, the supplier actually takes the units back. This
happens, for example, in cases in which the supplier has a higher salvage value
(when it can re-manufacture or re-distribute the unsold units), or in cases when
the supplier needs to protect its brand against being sold in discount stores who
buy excess inventory and sell it at low prices.

Buyback contracts are common in the book publishing and in the periodicals and
newspaper business. A somewhat similar type of contract is also common in the
consumer electronics business where manufacturers will offer “price support” to
retailers. Such price support includes payments from the manufacturer to the
retailer when the demand for the goods plummets due to the introduction of new
models.

With a buyback contract, the retailer’s expected profit is given by:

E [Re tailer Profit ] = R iE [Sales ] + B iE [Unsold ] − C iQ [25.8]

This equation is identical to Eq. [25.1a] with the only difference being that the
buyback price, B, replaces the salvage value. The supplier’s expected profit
under the buyback contract is:

E [Supplier Profit ] = (W − C )iQ − (B − S ) • E [Unsold ] [25.9]

The supplier buys the unsold inventory from the retailer at $B/item, but it can
presumably still salvage it at a $S/item. Adding the retailer and the supplier’s
expected profit, one gets Eq. [25.5] again as the buyback terms cancel out (they
are internal to the channel).

For the example discussed in this section (R = $200/item, W = $135/item, S =

$10/item and X~N(800,150)), and now with a buyback price of B = $80/item, the
expected profits of the supplier, retailer and the channel as a whole are depicted
in Figure 25.3. Since Eq. [25.8] is identical to Eq. [25.1] but with B replacing S, it
is easy to see that with a buyback contract, the retailer will order, QR* , where:

25.6
 ⎛ R −W ⎞ [25.10]
QR* = F −1 ⎜ ⎟
⎝ R −B ⎠

The critical ratio in this example is (200-135)/(200-80) = 0.54 and thus QR* = 816
items. The channel-optimal
Fig 6.3 Expected Profits with Buyback
order is still QT* = 921 items $120,000
as before.
$100,000

While this channel is not

Expected Profit
$80,000
coordinated yet, this order
quantity is higher than the $60,000 Retailer
one achieved without any Supplier
$40,000 Channel
buyback price, leading to
higher profits for both $20,000
channel participants, as can
be seen from a comparison $0
of Table 25.3 to the results 200 400 600 800 1,000 1,200 1,400 1,600
shown in table 25.1. Order Quantity

Table 25.3. Expected Profits with Buyback (B=$80)

Retailer Opt Order: 815 items

Retailer Profit: $44,858
Supplier Profit $64,573
Total Profit: $109,431

The supplier can set both the wholesale price and the buyback price in this
contract. To induce the retailer to order an amount that will coordinate the
channel, the supplier has to set W and B such that QR* = QT* . This will happen
when:
⎛ R −W ⎞ ⎛ R −C ⎞
⎜ R −B ⎟ = ⎜R −S ⎟
⎝ ⎠ ⎝ ⎠

Or:

⎛R −S ⎞ ⎛ R i(C − S ) ⎞
B=⎜ ⎟ iW − ⎜ ⎟ [25.11]
⎝ R −C ⎠ ⎝ R −C ⎠

Any combination of B and W that fulfills Eq. [25.11] will coordinate this channel.
For the example discussed here, the linear relationship between W and B are
depicted in Figure 25.4. The solid line illustrates the relationships between the
wholesale price and the buyback rate. As the wholesale price increases, the

25.7
supplier needs to increase the buyback rate in order to ensure that the retailer
has the right incentive to order the channel-optimal amount.

For the wholesale price

of $135 in our example, Fig. 25.4. Wholesale Price & Buyback
the optimal buyback Amount that Coordinate the Channel
rate is $118. Such a $200 100%
combination of $180 90%

Retailer Share of the

wholesale price and $160 80%

Buyback Rate
$140 70%
buyback rate will cause
$120 60%
the retailer to order the

Profit
$100 50%
channel-optimal
$80 40%
amount of 921 items,
$60 30%
maximizing the
$40 20%
channel’s total profit. $20 10%
$0 0%
The allocation of profits
50 70 90 110 130 150 170 190
in this case between
the supplier and the Wholesale Price
retailer is given by:

Table 25.4. Expected Profits with Optimal Buyback

Retailer Opt Order: 921 items

Retailer Profit: $48,436
Supplier Profit $63,338
Total Profit: $111,774

Note that the supplier’s profit in this case is lower than his profit in the
uncoordinated case with W=$135 and B=$80 (see Table 6.3). Thus one may
think that the supplier would not push to coordinate the channel. This, however,
is not the case. The supplier should be pushing to coordinate the channel but at
a different wholesale price and buyback rate so that he keeps more of the
channel’ profit. For example, at W=$140, and the (coordinating) buyback rate of
$124 the profit allocation between the retailer and the supplier will be:

Table 25.4. Expected Profits with Optimal Buyback

Retailer Opt Order: 921 items

Retailer Profit: $44,710
Supplier Profit $67,064
Total Profit: $111,774

The various combinations of wholesale price and buyback rate, and the resulting
supplier and retailer profit allocations are summarized in Table 25.5:

25.8
 Table 25.5 Summary of Wholesale and Buyback Effects

W= $135 $135 $135* $140*

B= 0 $80 $118* $124*
Retailer Opt Order: 739 items 815 items 921 items 921 items
Retailer Profit: $41,533 $44,858 $48,436 $44,710
Supplier Profit $62,814 $64,573 $63,338 $67,064
Total Profit: $104,347 $109,431 $111,774 $111,774
*These are coordination pairs

As Figure 25.4 demonstrates, there are many combinations of wholesale prices

and buyback rates that will coordinate the channel. They will all yield maximum
channel profit; the difference between them will be the profit allocation between
the retailer and the supplier. The dashed line in Fig. 25.4 depicts the percentage
of the profit that the retailer retains for each wholesale price, assuming that the
corresponding buyback rate is used to coordinate the channel. As one would
expect, at low wholesale price the retailer’s share of the profit is the highest and it
decreases with increasing wholesale price.

As shown in Table 25.5, both the retailer and the supplier have higher profits
when moving from no buyback price to a coordinating buyback price. In that
example, the supplier had to change both the wholesale price and the buyback
rate in order to settle on an allocation that gives him a high profit, while also
giving the retailer a higher profit as compared to the situation in which there was
no buyback.

Both the retailer and the suppliers can be better off by using a wholesale price
and buyback rate pair that coordinates the channel simply because there are
more profits to allocate in this case. The negotiations between them should focus
not on the value of a single price but on the full picture – in this case the pair of
prices. A strong supplier can, of course, set the coordination parameters such
that the retailer has a certain required profit which is higher than the retailer’s
profit without

When the buyback rate is chosen appropriately to coordinate the channel, the
expected profit functions for both the retailer and the channel will show a
maximum profit for the same order size, as shown in Figure 25.4.

25.9
 Fig 25.4 Expected Profits with Coordinating Buyback Rate
$120,000

$100,000

$80,000

Expected Profit
$60,000

$40,000
Retailer
Supplier
$20,000
Channel

$0
200 400 600 800 1,000 1,200 1,400 1,600
Order Quantity

Setting a buyback price is not the only mechanism, however, that can be used to
coordinate a supply chain. Another mechanism involves revenue sharing
between the supplier and the retailer.

25.3. Revenue Sharing

With revenue sharing contracts, the supplier creates an incentive for the retailer
to order more by charging a very low wholesale price. As shown later in this
section, this price has to be below the supplier’s own costs (!). In return, the
retailer shares with the supplier a percentage of the sales revenue.

Let δ be the fraction of the sales revenue that the retailer keeps. Its profit in this
case will be:

E [Re tailer Profit ] = δ iR iE [Sales ] + S iE [Unsold ] − W iQ [25.12a]

The supplier’s profits are going to be:

E [Supplier Profit ] = (W − C )iQ + (1 − δ )iR iE [Sales] [25.12b]

The sum of these profits is the channel’s expected profit. It will be identical to Eq.
[25.5] since the revenue sharing mechanism is internal to the channel and
therefore cancels out. Thus, the channel-optimal order quantity and the
maximum attainable expected profits are still the same as before.

25.10
Consider the
example used Fig 25.6 revenue Sharing Example
$120,000
throughout this
section but this $100,000 Retailer

time with a Supplier

Total Chain
$80,000
wholesale price

Profits
of W = $35 and $60,000
a retailer $40,000
revenue share
of δ = 0.40. The $20,000

expected profit $0
functions will be

0
00

00

00
0

0
00

20

40

60
20

30

40

50

60

70

80

90
as shown in

11

13

15
1,

1,

1,

1,
Figure 25.5. Order Quantity

The retailer order quantity in this case would be 855 items and the total channel
profit would increase from the base case to $110,897.

Eq. [25.12a] is identical to Eq. [25.1] with the only difference being that the
retailer’s revenue, R, is replaced by the fractional revenue, δ⋅R. Thus, the optimal
retailer’s order is given by:

⎛ δ iR − W ⎞
QR* = F −1 ⎜ ⎟ [25.13]
⎝ δ iR − S ⎠

The supplier has to set the wholesale price and the revenue share in order to
cause the retailer to order the optimal amount. To get QR* = QT* we need that:

⎛ δ iR − W ⎞ ⎛ R − C ⎞
⎜ δ iR − S ⎟ = ⎜ R − S ⎟ [25.14]
⎝ ⎠ ⎝ ⎠

This equation leads to the linear relationships between δ and W as follows:

(R − C ) S i(R − C )
δ =Wi − [25.15]
R i(C − S ) R i(C − S )

These relationships are depicted in Figure Fig 25.6 Optmal Revenue Sharing

25.5 for the numerical example discussed 100%

throughout this section. Note that the 80%

highest wholesale price that the supplier

can use is W = $50, which is equivalent to 60%

its cost, C. If the wholesale price is higher 40%

than C, the retailer will not order the

20%
channel-coordinating amount even if it
0%
10 15 20 25 30 35 40 45 50

Whol e sa l e P r i c e

25.11
gets all the revenue, as shown in the case of a wholesale contract.

This can be also seen from Eq. [25.15]. Since the retailer’s share of the revenue
has to be δ ≤ 1, this equation means that W ≤ C .

25.3.1. A Case Study

Until 1998, video rental stores used to buy movie tapes from the major studios for
about $65/tape and rent them to customers at $3/tape. This meant that a tape
had to be rented 22 times just to cover its costs, and a lot more in order to
actually return profit. Unfortunately, consumers’ interest in movie rental peaks at
the time of the release and wanes shortly thereafter when other new movies are
released to the rental market. Thus, demand is high for a short time and the
stores cannot satisfy it since the tapes are too expensive for them to buy in high
quantities. This leaves consumers frustrated and both stores and studios losing
income.

In the summer of 1998 Blockbuster Video solved this problem by restructuring its
supply contracts. Instead of paying $65/tape it negotiated to buy each tape for
only $7, but gave the studios 40% of the rental revenues. Note that even though
Blockbuster got only 60% of the $3 rental -- $1.80 per rental, it could cover the
direct costs in just four rentals. Now Blockbuster could justify buying many more
tapes, leading to many more potential rentals and satisfying customers when
demand was high. In fact, the average inventory of tapes of major movie
releases in Blockbuster stores increased from 24 to 124, as compared with an
average inventory of 12 tapes at independent stores.

The results were spectacular. Blockbuster’s market share increased by 5% in the

first year following the establishment of the new supply contracts (this roughly
equaled the share of the number two retailer, Hollywood Studios). The market
share continued to increase from 24% in 1998 to 40% in 2002. The new
arrangement was very profitable for the studios as well, since minting more tapes
costs very little. (The main cost is in creating the content, of course.)

Two outcomes of the use of revenue sharing contracts by Blockbuster and the
studios stand as testament to the success of the scheme. First, it became
common throughout the industry for all the large chains. The second outcome
was that Blockbuster and the studios were sued for unfair trade practices by the
independent retailers, who did not use revenue sharing contracts (mainly since
they were not computerized and the studios had difficulties verifying the number
of units rented).

25.3.2. Equivalency of Contracts

25.12
Intuitively, buyback contracts and revenue sharing contracts are equivalent; for
each combination of wholesale price and buyback rate in a buyback contract,
there is a pair of wholesale price and revenue sharing allocation that will produce
an identical result under a revenue sharing contract. To see this, denote the
wholesale price under buyback with WR, and denote the wholesale price under
revenue sharing by WB.

We have described buyback contracts as consisting of the retailer paying WB for

each unit purchased, and getting B-S for each unit not sold. This is the same as
saying that the retailer pays WB-(B-S) for each unit ordered and an additional
payment of (B-S) for each unit sold.

For revenue sharing contracts, the retailer pays WR for each unit purchased and
an additional (1 − δ )iR for each unit sold. Therefore, buyback and revenue
sharing contracts are similar when:

WR = WB − (B − S ) [25.16a]

and

B −S [25.16b]
(1 − δ )iR = (B − S ) or: δ = 1−
R

In other words, given a wholesale price and a buyback price under a buyback
contract, one can derive a revenue sharing contract with the parameters given in
Eq. [25.16], that will be give the same profits to the retailers and the supplier.

To see this in the case of the example discussed throughout this section
[R=$200, C=$50, S=$10, X~N(800.150)], assume that WB = $135 and B = $80.
In this case, the retailer order will be 816 items, the retailer’s expected profit will
be $44,858 and the supplier’s expected profit will be $64,573. The equivalent
revenue sharing contract can be found by setting:

WR = 135 − (80 − 10) = 65

and

80 − 10
δ = 1− = 0.65
200

Using these values of WR and δ one gets the exact same order quantity and
profits for the retailer and the supplier.

25.13
25.4. Option Contracts

Another way to coordinate a channel is to go into an options contract. One of the

simplest forms of such a contract is when the retailer is allowed to buy Q call
options at a price of $W/item. The supplier is assumed to produce/procure the
ordered item at a cost of $C/item. As demand materializes, the retailer can
exercise the options and buy each item at a price of $P/item.

In this simple version of this model the retailer is assumed to not be able to buy
more then the ordered amount even at a higher price and the supplier actually
makes the amount ordered and cannot make any more available.

The retailer’s expected profit in this case will be:

E [Re tailer Profit ] = (R − P )iE [Sales ] − W iQ [25.17a]

The supplier’s expected profits will be:

E [Supplier Profit ] = (W − C )iQ + P iE [Sales ] + S iE [Unsold ] [25.17b]

In this contract the retailer has no “overage” risk since if demand is less than the
number of call options it bought, the retailer will not exercise its option for the
extra items and buy exactly according to demand. Thus it has no use for any
salvage activities. It has some exposure, though since it had to buy the options in
the first place and his profit has to cover at least that amount; it will need to sell
enough items (at a margin of (R - P) each) to cover the cost of the options, Q⋅W.
The supplier is assumed to produce/acquire the items and therefore he can
salvage them at a value S.

Note that the option price, W, has to be lower than the supplier’s costs. This is
because for a price higher than the supplier’s cost we know from the discussion
of wholesale
Fig. 25.7. Optimal Number of Call Options
contracts that the
channel cannot be $120,000
coordinated. Any
$100,000
option price higher
than the supplier’s $80,000
cost would mean that
Profits

the retailer will end $60,000 Retailer

up paying at least Supplier

$40,000
that even for unsold Total
(or in this case un- $20,000
exercised) items.
Thus, for example, $0
200
300
400
500
600
700
800
900
1,000
1100
1,200
1300
1,400
1500
1,600

for an option price of

$25 and an exercise Number of calls

25.14
price of $120, the example used throughout this section [R=$200, C=$50, S=$10,
X~N(800,150)], the expected profit functions are depicted in Figure 25.6.

The optimal retailer order can be found by inspection: Compare Eq. [25.17] to
[25.1] and note that in [25.17] (R-P) replaces R, and S is zero, leading to:

⎛ R − P −W ⎞
QR* = F −1 ⎜ ⎟ [25.18]
⎝ R −P ⎠

To make sure that QR* = QT* , the wholesale price and option price should fall along
the line:
R −S [25.19]
P = R −W i
C −S

This line is depicted in

Fig. 25.8. Option Price and Excercise Price
Fig.25.8.
$200
$180
$160
$140
Excercise Price

Option contracts are $120

sometimes used when $100
reserving manufacturing $80
capacity. A buyer may reserve $60
Q units of capacity for $40
$W/units and pay an $20
additional $P for each unit $0
used. $0 $5 $10 $15 $20 $25 $30 $35 $40 $45
Option Price

25.4.1. Equivalency of Contracts

An options contract is identical to a buyback contract and a revenue sharing

contract with the right choice of parameters. The argument is very similar to the
one used to show the equivalence between buyback and revenue sharing
contracts. In an options contract the buyer pays $WO for each unit reserved and
an additional $P for each unit actually purchased. Thus, given a buyback contract
with parameters (WB, B), an equivalent options contract can characterized by:

WO = WB − (B − S ) and P = B −S [6.20a]

Given a revenue sharing contract with parameters (WR, δ), the equivalent option
contract would be:

[6.20b]

25.15
 WO = WR and P = (1 − δ )iR

25.5. Summary

This chapter uses the single ordering framework to show that by allocating risk
between seller and buyer, the supply chain profit can be increased. Furthermore,
the profit of each of the channel members can increase as well. Consequently,
this is a true win/win situation. The relationships can be struck such that when
each channel member is acting to maximize his own profits, the contract is such
that both of them are better off. Rather than playing a zero-sum game they are
both acting on their own behalf and at the same time increase the channel’s total
profits.

The literature on supply contracts is vast. Pasternak analyzed buyback and

revenue sharing contracts and demonstrated many of the results shown in this
chapter. Cachon (2002) provided a very comprehensive summary of the
literature which includes many other contracts that can be used to coordinate the
channel. Examples include (i) quantity-flexibility contracts in which the supplier
gives the retailer a full rebate on a portion of the order quantity. Tsay (1999)
mentions that such contracts are frequently used by electronics firms and PC
manufactures to form agreements with their component suppliers and (ii) sales-
rebate contract, in which the supplier gives the retailer a rebate for every unit
sold above a certain level. Such contracts are used in the computer hardware
and the automotive industry. (see, for example, Taylor, 2001).

The basic model has been extended to account for elastics demand, multiple
buyers, multiple sellers, continuous replenishment and in many other directions.
As mentioned above, however, all these contract forms demonstrate that the
concept of win/win can modeled and be quantified.

25.6 References

1. Cachon, Gerrard, P. (2002) “Supply Chain Coordination with Contracts”, 2nd draft,
invited to be published in the Handbooks in Operations Research and
Management Science: Supply Chain Management, September 2002
2. Pasternack, Barry Alan (1985) “Optimal Pricing and Return Policies for
Perishable Commodities”, Marketing Science, Vol. 4, No. 2, Spring 1985
3. Tsay, Andy A. (1999) “The Quantity Flexibility Contract and Supplier-Customer
Incentives”, Management Science, Vol. 45, No. 10, October 1999
4.

25.16
 25.7 Appendix

To see the equivalency between buyback and revenue sharing contracts,

consider the retailer expected profits under both. (Let the wholesale price under
the buyback contract be denoted by WB, to distinguish it from the wholesale price
under the revenue sharing contract, WR.)

Under buyback:

∞ Q Q
E [Retailer Profit ]Buyback = R iQ i ∫
x =Q
f ( x )dx +R i ∫
x =0
x if ( x )dx + B i ∫ (Q − x )if ( x )dx − W
x =0
B iQ

Under revenue sharing:

∞ Q Q
E [Retailer Profit ]Re v .Sharing = δ iR iQ i ∫
x =Q
f ( x )dx +δ iR i ∫
x =0
x if ( x )dx + S i ∫ (Q − x )if ( x )dx − W
x =0
R iQ

To see when these are equivalent, set them equal to each other, getting:

E [Retailer Profit ]Buyback − E [Retailer Profit ]Re v .Sharing = 0

∞ Q Q
(1 − δ )iR iQ i ∫
x =Q
f ( x )dx +(1 − δ )iR i ∫
x =0
x if ( x )dx + (B − S )i
x =0
∫ (Q − x )if ( x )dx − (W B + WR )iQ = 0

⎛ Q
⎞ Q Q Q
(1 − δ )iR iQ i⎜ 1 − ∫ f ( x )dx ⎟ + (1 − δ )iR i ∫ x if ( x )dx + (B − S )iQ i ∫ f ( x )dx − (B − S )i ∫ x if ( x )dx − (WB + WR )iQ = 0
⎝ x =0 ⎠ x =0 x =0 x =0
Q Q

((1 − δ )iR + (WR − WB ))iQ + ( −(1 − δ )iR + (B − S ))iQi ∫ f ( x )dx − ( −(1 − δ )iR + (B − S ) )i ∫ x if ( x )dx = 0
x =0 x =0

⎛ Q Q
⎞
((1 − δ )iR + (WR − WB ))iQ + ( −(1 − δ )iR + (B − S ))i⎜ Qi ∫ f ( x )dx − ∫ x if ( x )dx ⎟ = 0
⎝ x =0 x =0 ⎠

Since this is true for all values of Q, both the term multiplying Q and the term
multiplying the integrals in the last parentheses should be zero. Thus:

(1 − δ )iR + (WR − WB ) = 0 and: −(1 − δ )iR + (B − S ) = 0 ,

leading to Eq. [16].

25.17