Journal of Industrial and Management Optimization

Vol. 19, No. 2, February 2023, pp. 1350–1366

doi:10.3934/jimo.2021232

OPTIMAL RECYCLING PRICE STRATEGY OF CLOTHING

ENTERPRISES BASED ON CLOSED-LOOP SUPPLY CHAIN

Huaqing Cao
College of textile science and engineering
Zhejiang Sci-Tech University, Hangzhou 310000, China

Xiaofen Ji∗
School of International Education
Zhejiang Sci-Tech University, Hangzhou 310000, China

(Communicated by Stefan Wolfgang Pickl)

Abstract. More and more garment enterprises begin to pay attention to the
importance of recycling, take the corresponding recycling strategy to recy-
cle garment products and remanufacture, forming a closed-loop supply chain
(CLSC). In reality, recycling is a complex system, the recycling strategy of
clothing brands will not only affect the reverse channel of closed-loop supply
chain, but also affect the consumer demand of forward channel, and then affect
the profit of supply chain. In order to solve this problem, we propose a CLSC
composed of a manufacturer, a retailer and a collector, establish three differ-
ent Stackelberg leadership models, and derive the optimal recycling strategy.
Our results show that consumers’ sensitivity to the recycling price will affect
the optimal decision of supply chain members. The increase of the recycling
market is not always beneficial to the profits of supply chain members. By
comparing the profits of the three models, it is found that the retailer leader-
ship model is the most effective scenario of CLCS. The results of this paper
provide a reference for garment enterprises to formulate recycling strategies.

1. Introduction. Textile and clothing industry is of great significance to the econ-

omy in terms of trade, employment, investment and income all over the world. How-
ever, the characteristics of this industry are that on the one hand, the manufacturer
has overcapacity, on the other hand, it has caused great losses due to the “aban-
donment” culture. This shows that it is necessary to recover textiles[17]. Recycling
and reuse have become a hot topic in the textile and clothing industry. According
to the report of circular fashion: China new textile economic outlook issued by
China Textile Industry Federation, under the influence of “fast fashion”, the global
clothing utilization rate has decreased by 36% from 2000 to 2015, and more than
50% of fast fashion clothing will be discarded within one year[34]. Although the
original textile production has made significant technological progress in environ-
mental protection, low recovery rate and consumer awareness have an impact on the

2020 Mathematics Subject Classification. Primary: 90B50, 91A10; Secondary: 90C05.

Key words and phrases. Closed-loop supply chain, recycling, price strategy, power structure,
channel leadership.
The research is supported by Zhejiang Soft Science Foundation Project Grant No.2020C35038.
∗ Corresponding author: Xiaofen Ji.

1350
 OPTIMAL RECYCLING PRICE STRATEGY BASED ON CLOSED-LOOP SUPPLY CHAIN
1351

sustainability level of the textile industry[14]. The reuse and recycling of textiles
can be seen as a way to obtain social and economic benefits and to promote the
economic development of a country[8]. Reuse and recycling have generally reduced
the impact on the environment, mainly because of the reduction in demand for pri-
mary resources. Compared with incineration and landfill, textile reuse and recycling
generally reduce the impact on the environment, and reuse is more beneficial than
recycling[25]. Recycling textiles, even reuse of textiles, can reduce the production
of new textiles from raw materials, thereby reducing the use of water, energy and
chemicals in the production chain[9]. By establishing recovery units, some of the
waste generated in the textile and textile fields can be transferred to the recovery
unit where fibers are recovered[21]. Combining with the view of “slow fashion”,
Freudenreich proposed a framework for efficiency-oriented business offering, aiming
to reduce the total amount of clothing produced, used and discarded[11].
On the other hand, clothing reuse is a kind of consumer disposal behavior that
has not been fully exploited, but it has an important impact on society. The survey
shows that 75% of consumers think sustainability is extremely important; More than
one-third of consumers are willing to choose brands with practices in environmental
and social improvement, even if it may not be their preferred brand[34]. From the
perspective of customer led logic, clothing reuse is a kind of consumer value creation
behavior[7]. The results of another study show that people have moderate inter-
est in various clothing related product service systems, and the highest interest in
clothing recycling, exchange and consultation, indicating that there is a certain po-
tential for the development of clothing recycling market[1].Nowadays, major brands
are more and more willing to launch products containing recycled fiber ingredients.
Some products made of 100% recycled materials have appeared in the market, which
is undoubtedly a step towards circular economy. Many retailers are involved in re-
cycling programs (including UNIQLO, Adidas, PUMA, etc.), and they are often
willing to exchange coupons for returned clothing. Fashion enterprises need to pro-
vide more detailed and transparent sustainable fashion production information to
stimulate sustainable fashion consumption and consumer purchase decisions. Fash-
ion companies should strive to produce sustainable fashion to enhance their image
of social responsibility and attract fashion leaders to buy and wear[28]. Research
shows that second hand retail and product recycling plan are the main methods
of address post-retail responsibility of fashion companies, and recycling seems to
be the most important platform to inform consumers of sustainability and circular
economy[16]. If companies that are competitive in price and quality can create
social environmental certification for their brands, labels and products (such as or-
ganic, recyclable, durable, waste reduction or carbon footprint reduction), Chinese
environmentally conscious consumers will be more willing to buy their products,
recommend them to their relatives and friends, and show higher satisfaction with
the products[4]. It can be seen from the literature and practical experience that the
recycling strategy of supply chain will have an impact on consumers’ purchasing
behavior. Based on this, this paper establishes a closed-loop supply chain com-
posed of the manufacturer, retailer and collector. The manufacturer is responsible
for manufacturing new products and remanufacturing with recycled materials, the
retailer is responsible for selling products, and the collector recycles products in
reverse channel. This paper attempts to study the following three questions:
(1) What is the optimal decision of closed-loop supply chain members under the
1352 HUAQING CAO AND XIAOFEN JI

three channel leaders when considering the impact of the recycling price on demand
and recycling quantity?
(2) What is the impact of different channel leaders on the optimal decision?
(3) What is the impact of different market parameters, such as consumer sen-
sitivity to retail price and the recycling price, on the optimal decision of supply
chain?
The rest of this paper is arranged as follows. The second section is literature
review. The third section establishes a basic profit model of closed-loop supply
chain. The fourth section analyzes three kinds of CLCS optimal decisions under
Stackelberg leadership. The fifth section compares the optimal decisions under the
three modes, analyzes the influence of different market parameters on the optimal
decisions. The sixth section presents the results, discusses the significance of the
research and puts forward the future research direction

2. Literature review. This section deals with the relevant literature considering
two different streams of research: closed-loop supply chain and Stackelberg leaders.

2.1. Closed-loop supply chain. The manufacturer produces clothing products,

sells them to consumers through retailers, and then returns to the manufactur-
ers through recycling channels, forming a closed-loop supply chain. Many scholars
have done research on closed-loop supply chain. As an early scholar of CLSC,
Savaskan compared three CLSC recycling modes: Manufacturer recycling, retailer
recycling and third-party recycling. They found that when members of the sup-
ply chain are closer to consumers, it is the most effective undertaker of product
collection activity. In addition, they designed a coordination mechanism to make
the CLSC profit of decentralized decision-making the same as that of centralized
decision-making[26]. Furthermore, Savaskan studies two competing retailer recy-
cling channels in CLSC[27]. The uncertainty of procurement, processing and market
in closed-loop supply chain greatly increases the complexity of remanufacturing and
reduces the process efficiency. Peng analyzed the causes of uncertainty in different
stages, and determined appropriate methods to quantify the impact of uncertainty
on the production process, so as to ensure the sustainable development of closed-
loop supply chain and clean production[22]. Mukhopadhyay uses modularity in
product design to solve the return policy based on build to order product, estab-
lishes a profit maximization model, and obtains the optimal strategy of recovery
rate and modularity level under certain market reaction parameters[20]. Liu con-
structs an analysis model with uncertainty of demand and income, and studies the
optimal strategy of three-dimensional decision-making with pricing, consumer re-
turn rate and modularization level under the mean square error formula[18]. Choi’s
research is divided into two cases: risk neutral and risk averse mass customization
companies. By comparing the two cases, they reveal the impact of the risk aversion
level of service providers on the optimal return service charge policy. Finally, the
optimal condition of zero return service charge (free return, full refund) is obtained.
These conditions provide free return service for some fashion companies in real life,
but not for every customer[5]. Chen studied a three-level closed-loop supply chain
composed of a manufacturer, a retailer and a third-party logistics supplier, led by
a retailer. The results show that whether it is forward logistics or reverse logis-
tics, the higher the manufacturer’s share in logistics cost, the lower the profit of
each member[3]. Wang draws on variational inequalities to model the closed-loop
 OPTIMAL RECYCLING PRICE STRATEGY BASED ON CLOSED-LOOP SUPPLY CHAIN
1353

supply chain network. Research find that it is necessary to regulate a medium col-
lection rate and a certain minimum recovery rate. The impact of collection rate
and recovery rate on manufacturers are greater than that on retailers. Consumers
can benefit from the increase of the recovery rate as well as the collection rate[29].
Chen extended the closed-loop supply chain to a newsvendor model, which simply
includes supply uncertainty and external government influence. It is found that the
return uncertain lower both the manufacturers’ profits and retail price, where as
its effects on the governments optimal subsequences and the manufacturers optimal
return effects are indicative[2]. Another study considers how the growth factor of
consumer returns affects order quantity, profit and supply chain coordination. By
modeling and analyzing the wholesale price contract and buy back contract between
manufacturer and retailer with random demand and external given retail price, two
decision strategies are compared. One is to consider the consumer return when cal-
culating the decision variables of manufacturer and retailer, the other is to ignore
the consumer return in the optimization model[24].
2.2. Stackelberg leader. With large retailers such as Nike, UNIQLO and SHEIN
becoming channel leaders, the structure of closed-loop supply chain is no longer a
single manufacturer. Choi studies a supply chain composed of retailers, collectors
and manufacturers, and investigates the supply chain performance under differ-
ent channel leadership. Through systematic comparison, the author finds that the
retailer led model provides the most effective customer satisfaction. In addition,
the author finds that the efficiency of remanufacturing system is closely related
to the proximity of supply chain agents to the market. Finally, it is concluded
that the mode dominated by recyclers is not the most effective mode for collecting
products[6]. Edirisinghe studies the influence of channel power on the stability of
supply chain when multiple suppliers sell substitutable products through the same
retailer. It is found that neither manufacturer leader nor retailer leader is a sta-
ble supply chain structure, but the structure of equally distributing power among
agents provides the best stability and performance[10]. Wang studied the responsi-
bility sharing between manufacturers and recyclers in the recycling process of waste
electronic and electrical appliances, and found that from the collection rate improve-
ment perspective, the collector led is a better channel structure of CLSC compared
with centralized and manufacturer led channel structure[30]. Gao established a
CLSC centralized and decentralized game model of one manufacturer and one re-
tailer, to investigate optimal decisions of collection effort, sales effort and pricing
under different channel power structures, they found that the best power struc-
ture for a CLSC varieties with the market demand, as influenced by the collection
effort[12]. Ranjbar takes a TV manufacturer as an example to study a three-tier
closed-loop supply chain composed of manufacturer, retailer and third-party collec-
tor. The main purpose is to evaluate the optimal pricing and collection decision
under two competitive recycling channels (including retailer collection and third-
party collection). The test results from two aspects of environment and consumer
welfare show that the decentralized model of retailer leadership is often the most
effective scenario in the closed-loop supply chain. Considering the overall bene-
fits of the supply chain between the decentralized models, the retailer leadership
model is the best and closest model to the centralized model[23]. Wu studied a
CLSC composed of a single manufacturer and a single retailer. They have three
recycling channels: manufacturer, retailer and the third party. They studied CLSC
performance under different channel leadership and recycling channel combinations.
1354 HUAQING CAO AND XIAOFEN JI

The results show that M channel is the most effective recycling channel, and the
retailer-led structure is as good as manufacture-led structure[31]. Zheng established
the centralized, manufacturer-led, and retailer-led CLSC models, and analyze the
equilibrium solutions of channel members and the government. Results show that
the manufacturer exists the market earlier in the retailer-led model with regulation
compared with the manufacturer-led model. The government’s optimal collection
target is the same under manufacturer-led and retailer-led models when the regu-
lation comes into force[33].
2.3. Research gaps. However, these studies did not consider the impact of re-
cycling strategy on consumer demand and recycling quantity, because most of the
previous studies were from the perspective of consumer returns, and did not con-
sider the recycling strategy of supply chain from the perspective of active recycling.
In the clothing closed-loop supply chain, the recycling price not only affects the
recycling quantity, but also affects the demand and the profits of the members of
the supply chain because of the stimulation of the recycling price on the purchase
intention of consumers. On the other hand, most of these studies take the manufac-
turer as the Stackelberg leader, without considering the situation of the retailer and
the collector as the leader. The innovation of this paper is to consider the impact
of the recycling price on consumer demand and recycling quantity. When con-
sumers see the recycling advertisement of the retailer, their purchase intention will
increase. The higher the recycling price, the higher the possibility of consumers to
buy clothing products. Therefore, how to determine the optimal recycling strategy
in a complex closed-loop supply chain is the goal of this paper.

3. Problem description and assumptions.

3.1. Problem description. As shown in Figure 1, we consider a closed-loop sup-
ply chain composed of four members: a manufacturer, a retailer, a collector and
consumer. In the forward channel, the manufacturer produces clothing products
and sells them to the retailer at the wholesale price. The retailer sells clothing
products to the consumer at the retail price. In the reverse channel, the collector is
responsible for recycling products and transferring them to the manufacturer, and
the manufacturer pays the collector a certain transfer fee. Based on Stackelberg
game theory, this paper establishes three CLSC models of different channel leaders,
including manufacturer leader (M-led), retailer leader (R-led) and collector leader
(C-led). Different leaders have different decision-making order in CLSC.

Figure 1. Different power structures of CLSC

 OPTIMAL RECYCLING PRICE STRATEGY BASED ON CLOSED-LOOP SUPPLY CHAIN
1355

3.2. Assumptions. In this paper, we make the following assumptions:

Assumption1. It should be noted that in reality, demand may be stochastic. Ac-
cording to the literature on remanufacturing related topics in production economics
and operation management[26, 6, 23], because our model composed of a manufac-
turer, a retailer and a collector, and the optimality of the strategy depends on the
convexity of the demand function. In order to facilitate analysis, we must use this
deterministic and linear requirement in the model, otherwise we will not have any
closed form solution. Therefore, the demand function can be written as:

D(p) = φ − βp + εb

The recycling quantity function can be written as:

R(b, e) = a + lb

The meanings of all symbols are given in Table 1.

Assumption2. We assume that cm > cr , ensure that the cost of remanufacturing
a recycled product is less than the cost of manufacturing a new one, and the quality
of remanufactured products is the same as that of new products and they are sold
in the same market at the same price. Note that p > w > cm > cr .
Assumption3. We consider a closed-loop supply chain with a single manufacturer,
a single retailer and a single collector in the market, and the CLSC decision is
considered in a single cycle. Each agent is an independent decision-maker, and each
of them intends to maximize their profits.

Table 1. Symbols and definitions

Model parameters
φ Basic demand, and φ > 0
β The sensitivity coefficient of demand to the retail price, and β > 0
ε The sensitivity coefficient of demand to the recycling price, and ε > 0
a Basic recycling quantity, and a > 0
l The sensitivity coefficient of recycling quantity to the recycling price, and l > 0
cm The cost of a new product produced by the manufacturer
cr The cost of remanufacturing a new product using recycled products
∆ ∆ = cm − cr , represents the cost saved per unit product during remanufacturing
Profit function symbol, and ΠM C represents the retailer’s profit under the
Π
manufacturer’s leadership mode
w The wholesale price given by the manufacturer to the retailer
p The retailer’s retail price
b The recycling price provided by the collector to the consumer
The transfer price of a single product provided by the manufacturer
A
to the collector

4. Model analysis. In this section, we study the optimal wholesale price, retail
price, recycling price and transfer price in the closed-loop supply chain under three
kinds of Stackelberg game: Manufacturer leader, retailer leader and collector leader.
1356 HUAQING CAO AND XIAOFEN JI

4.1. The manufacturer as Stackelberg leader. A large number of scholars have

studied the supply chain of the manufacturer as Stackelberg leader, their production
decisions directly affect the strategy of downstream members. In this model, the
manufacturer is Stackelberg leader, the retailer and the collector are followers. Here
we use L represent leader and F represent follower and their profits can be expressed
as follows:
L : M ax ΠM
M = (wM − cm )(φ − βpM + εbM ) + (∆ − AM )(a + lbM ) (1)
(wM ,AM )

F : M ax ΠM
R = (pM − wM )(φ − βpM + εbM ) (2)
(pM )

F : M ax ΠM
C = (AM − bM )(a + lbM ) (3)
(bM )

In order to ensure that the manufacturer’s profit function is concave and has a
unique maximum, we have lemma 4.1.
Lemma 4.1. When 16lβ − ε2 > 0 ,the manufacturer’s profit function is concave
and has a unique maximum.
Proof. All the proofs in this section are given in Appendix.

We use reverse induction to obtain the equilibrium decision of each CLSC mem-
ber. First, the retailer decides the retail price p and the collector decides the
recycling price b. Then, the manufacturer decides the wholesale price w and the
transfer price A.
Proposition 1. By solving the first partial derivatives of ΠM M M
M , ΠR and ΠC , we
can get the optimal strategy of the CLSC under the manufacturer leadership mode:
φε − βεcm + 4lβ∆ − 12lβa
b∗M = (4)
16lβ − ε2

12lφ − cm ε2 + 3lε∆ − 9aε + 4lβcm

p∗M = (5)
16lβ − ε2

∗ 8lφ − cm ε2 + 2lε∆ − 6aε + 8lβcm

wM = (6)
16lβ − ε2
2lφε − 2lβεcm + 8l2 β∆ − 8lβa − aε2
A∗M = (7)
16l2 β − lε2
Proposition 2. Bringing b∗M , p∗M , wM ∗
and A∗M into (1), (2), (3), we can get that
in M-led model the profits of the manufacturer, retailer and collector are respectively:
2a2 βl + a2 ε2 + 3aβcm εl + 4aβ∆l2 − a∆ε2 l − 3aεlφ + 2β 2 cm 2 l2 −
βcm 2 εl2 + 2βl3 ∆2 + βcm cr εl2 − 4βcm l2 φ + ∆εl2 φ + 2l2 φ2
ΠM
M = (8)
l(16βl − ε2 )
2
β(3aε − 4lφ + 4βcm l − εl∆)
ΠM
R = 2 (9)
(16βl − ε2 )
2
(aε2 − 4aβl − εlφ − 4βl2 ∆ + βcm εl)
ΠM
C = 2 (10)
l(16βl − ε2 )
 OPTIMAL RECYCLING PRICE STRATEGY BASED ON CLOSED-LOOP SUPPLY CHAIN
1357

4.2. The retailer as Stackelberg leader. Nowadays, more and more retailers
have become leaders in the fashion supply chain, such as SHEIN in China and
UNIQLO in Japan. They often assign a large number of orders to various manufac-
turers, and have a dominant position in the supply chain. The wholesale price of the
manufacturer is also affected by the retail price. They can lead the supply chain to
maintain their margins while limiting the profits of their manufacturers. In the re-
tailer leadership model, the retailer is the Stackelberg leader, and the manufacturer
and the collector are the followers:
L : M ax ΠR
R = (pR − wR )(φ − βpR + εbR ) (11)
(pR )

F : M ax ΠR
M = (wR − cm )(φ − βpR + εbR ) + (∆ − AR )(a + lbR ) (12)
(wR ,AR )

F : M ax ΠM
C = (AR − bR )(a + lbR ) (13)
(bR )
We use reverse induction to obtain the equilibrium decision of each CLSC mem-
ber. First, the collector decides the recycling price b, then the manufacturer decides
the wholesale price w and the transfer price A, and finally the retailer decides the
retail price p. Here, we need to note that the decision of the collector does not
include any w, so the profit of the manufacturer can increase with the increase of
the wholesale price. Because p > w, the wholesale price w cannot equal the re-
tail price p. In this case, if the manufacturer maximizes its profit function, then
p = w, the retailer’s profit is zero. Therefore, we use the same method as previous
studies[23, 13, 15, 19, 32], assuming that the manufacturer’s unit marginal profit is
equal to the retailer’s unit marginal profit, that is, w = (cm + p)/2.
Proposition 3. By solving the first partial derivatives of ΠR R R
M , ΠR and ΠC , we can
get the optimal strategy of the CLSC under the retailer leadership mode:
4lφε − 4lβεcm − l∆ε2 + 3aε2 − 48lβa + 16l2 β∆
b∗R = (14)
(8lβ − ε2 )8l
4lφε − 4lβεcm − l∆ε2 − aε2 − 16lβa + 16l2 β∆
A∗R = (15)
(8lβ − ε2 )4l

∗ 4lφ − 2cm ε2 + lε∆ − 3aε + 12lβcm

wR = (16)
16lβ − 2ε2
4lφ − cm ε2 + lε∆ − 3aε + 4lβcm
p∗R = (17)
8lβ − ε2
Proposition 4. Bringing b∗R , p∗R , wR
∗
and A∗R into (11), (12), (13), we can get that
in R-led model the profits of the manufacturer, retailer and collector are respectively:
(3aε−4lφ+4βcm l−εl∆)2
ΠR
M = 16l(8βl−ε2 ) −
(aε2 +16aβl−4εlφ+16βl2 ∆−3ε2 l∆+4βcm εl)(5aε2 −16aβl−4εlφ−16βl2 ∆+ε2 l∆+4βcm εl) (18)
32l(8βl−ε2 )2
2
(3aε − 4lφ + 4βcm l − εl∆)
ΠR
R = (19)
16l(8βl − ε2 )
2
(5aε2 − 16aβl − 4εlφ − 16βl2 ∆ + ε2 l∆ + 4βcm εl)
ΠR
C = 2 (20)
64l(8βl − ε2 )
1358 HUAQING CAO AND XIAOFEN JI

4.3. The collector as Stackelberg leader. Nowadays, in other industries such

as metals(Sims Metal Management) and electronic products(IBM’s Global Asset
Recovery Services), collectors can be channel leaders. With the development of
textile recycling, we believe that textile recycling will be popular, and there will
be large textile collectors in the future. In this model, the collector is Stackelberg
leader, and the manufacturer and the retailer are followers:
L : M ax ΠC
C = (AC − bC )(a + lbC ) (21)
(bC )

F : M ax ΠC
M = (wC − cm )(φ − βpC + εbC ) + (∆ − AC )(a + lbC ) (22)
(wC ,AC )

F : M ax ΠC
R = (pC − wC )(φ − βpC + εbC ) (23)
(pC )

We use reverse induction to obtain the equilibrium decision of each CLSC mem-
ber. First, the retailer decides the retail price p, then the manufacturer decides
the wholesale price w and the transfer price A, and finally the collector decides the
recycling price b. Here, we need to note that the retailer’s decision does not include
any A, so the manufacturer’s profit can increase with the decrease of the transfer
price A. Because A < b, the transfer price A cannot be equal to the recycling
price b. In this case, if the manufacturer maximizes its profit function, then A = b,
resulting in zero profit for the collector. Therefore, we adopt the same method as
in 4.2, assuming that the unit marginal profit of the manufacturer in the reverse
channel is equal to the unit marginal profit of the collector, that is, A = (∆ + b)/2.
Proposition 5. By solving the first partial derivatives of ΠC C C
M , ΠR and ΠC , we can
get the optimal strategy of the CLSC under the collector leadership mode:
6lφ + 3lε∆ − 3aε + 2lβcm
p∗C = (24)
8lβ

3l∆ − a
A∗C = (25)
4l

∗ 2lφ + lε∆ − aε + 2lβcm

wC = (26)
4lβ

l∆ − a
b∗C = (27)
2l
Proposition 6. Bringing b∗C , p∗C , wC
∗
and A∗C into (21), (22), (23), we can get that
in C-led model the profits of the manufacturer, retailer and collector are respectively:
2 2
4βl(a + l∆) + (aε − 2lφ + 2βcm l − εl∆)
ΠC
M = (28)
32βl2

2
(aε − 2lφ + 2βcm l − εl∆)
ΠC
R = (29)
64βl2
2
(a + l∆)
ΠC
C = (30)
8l
 OPTIMAL RECYCLING PRICE STRATEGY BASED ON CLOSED-LOOP SUPPLY CHAIN
1359

5. Model comparison and sensitivity analysis. In this section, in order to

compare the results of different models and illustrate the sensitivity of CLSC mem-
bers’ optimal decisions to different market parameters, we use numerical analysis
to simulate the results of the previous section. Data is collected through inter-
views with relevant experts and studying existing data of a garment enterprise in
Hangzhou. The market parameters are as follows:
φ = 1000; β = 5; ε = 5; a = 10; l = 10; cm = 100; cr = 50.
5.1. Comparison among M-led, R-led and C-led models. In section 4, we
analyze the optimal decision of CLCS under three leadership models and the profits
of each supply chain member. Further, in this section, using the results of numerical
analysis, we will compare CLCS decisions and profits under different supply chain
leadership structures and different game scenarios.
Corollary 1. (a) The optimal wholesale prices w under M-led, R-led and C-led
models satisfy the following relationship: wC > wM > wR . (b) The optimal retail
prices p under M-led, R-led and C-led models satisfy the following relationship:
pC > pM > pR .
Obviously, the wholesale price is smallest in R-led model, because the retailer
has a strong incentive to minimize the wholesale price. Interestingly, the optimal
wholesale price in C-led model is larger than that in M-led model. This is because
when the collector becomes channel leader, there is a “repeated double marginal
effect”, that is, between the collector and the manufacturer, the manufacturer and
the retailer, which makes the wholesale price and retail price double increase. The
wholesale price has a direct impact on the retail price. In C-led model, the wholesale
price is largest, so the retail price is also largest.
Corollary 2. The optimal recycling prices b under M-led, R-led and C-led models
satisfy the following relationship: bC > bR > bM .
The recycling price is largest when the collector is the channel leader, because
the collector has enough motivation to increase the recycle quantity by increasing
the recycling price to increase profit. However, in proposal 5.4, we will see that this
approach is not always beneficial to the collector. The recycling price is smallest in
M-led model.
Corollary 3. Let 16lβ(l∆ + a) − 4ε2 (l∆ − a) = A, 16lβ(l∆ + a) − ε2 (3l∆ − 5a) =
B. The optimal transfer prices A under M-led, R-led and C-led models satisfy
the following relationships: when 8lε(φ − βcm ) < A, AM < AR < AC ; when
8lε(φ−βcm ) > B, AC < AM < AR ; when A < 8lε(φ−βcm ) < B, AM < AC < AR .
If the manufacturer assigns a larger transfer price, the collector will collect more
products, but at the same time, the manufacturer’s remanufacturing cost will also
increase. Therefore, the manufacturer faces tradeoff of the transfer price. Under
different leadership models, the transfer price changes with the change of market
parameters. When the purchase demand is strong and the recycling demand is
weak, that is, φ and ε decrease and β, a and l increase, the transfer price of C-
led model is the largest. When the recycling demand is strong and the purchase
demand is weak, the transfer price of R-led model is largest.
Corollary 4. The profits of manufacturer, retailer and collector in M-led, R-led
and C-led models satisfy the following relationships: (a) ΠR M C
M < ΠM < ΠM ; (b)
1360 HUAQING CAO AND XIAOFEN JI

√ √ √
ΠMR <
C R
√ΠR < ΠR ; (c) Let (12 +√ 2)16lβ(l∆ +
2
√ a) − ε [(1 + 2 2)l∆ − (5 + 2)a] = C,
(1 + 2)16lβ(l∆ + a) − ε [ 2l∆ − (4 + 2)a] = D, the profits of the collector
satisfy the following relationships: when 4lε(βcm − φ) < C, ΠM R C
C < ΠC < ΠC ; when
4lε(φ−βcm ) > D, ΠC M R M C
C < ΠC < ΠC ; when C < 8lε(φ−βcm ) < D, ΠC < ΠC < ΠC .
R

Interestingly, the manufacturer’s profit is largest in C-led model. This is because

the recycling price is largest in C-led model, so the sales quantity and recycling
quantity are also the largest. Obviously, the retailer’s profit is largest under R-led
model. The collector’s profit varies with market parameters. Similar to the transfer
price, when the purchase demand is strong and the recycling demand is weak, that
is, φ and ε decrease and β, a and l increase, the collector has the largest profit in
C-led model. On the contrary, if the recycling demand is strong and the purchase
demand is weak, the collector’s profit is largest in M-led model.

5.2. Sensitivity analysis. In this section, we analyze the changes of optimal deci-
sion and profit with market size (φ), demand sensitivity to retail price (β), demand
sensitivity to recycling price (ε), basic recycling quantity (a) and recycling quantity
sensitivity to recycling price (l) under the three models.
5.2.1 The influence of market size on optimal decision. As shown in Figure
2, the horizontal axis is the market size φ, ranging from 800 to 1000, and the
vertical axis is the wholesale price w, retail price p, recycling price b, transfer price
A, manufacturer’s profit, retailer’s profit, collector’s profit and total supply chain
profit, respectively. When the market size increases, due to the high purchase
intention of consumers, the supply chain does not need to reduce the retail price to
increase the sales, on the contrary, it can increase its profit by increasing the retail
price. In M-led model and R-led model, w, p, b, A will increase. The profits of
manufacturer and retailer also increase with the increase of market size. In C-led
model, w and p increase, but the basic sales have no effect on b and A in reverse
channel. Therefore, in C-led model, retailer’s profit is not affected.

Figure 2. The influence of market size on optimal decision

5.2.2 The influence of demand sensitivity to retail price on the optimal decision.
As shown in Figure 3, the horizontal axis is demand sensitivity to retail price,
ranging from 5 to 8, and the vertical axis is the wholesale price w, retail price p,
recycling price b, transfer price A, manufacturer’s profit, retailer’s profit, collector’s
profit and total supply chain profit, respectively. When the demand sensitivity to
 OPTIMAL RECYCLING PRICE STRATEGY BASED ON CLOSED-LOOP SUPPLY CHAIN
1361

retail price increases, the supply chain has to reduce the retail price to increase the
sales quantity. In this case, w and p in M-led model and R-led model will decrease.
Due to the transmission of supply chain cost, A and b in reverse channel will also
decrease. The profits of manufacturer and retailer also decrease with the increase
of β. In C-led model, w and p decrease. However, because the collector is leader,
β has no effect on b and A in reverse channel. Therefore, the profit of collector in
C-led model is not affected.

Figure 3. The influence of the demand sensitivity to retail price

on optimal decision

5.2.3 The influence of demand sensitivity to recycling price on the optimal deci-
sion. As shown in Figure 4, the horizontal axis is the demand sensitivity to recycling
price, ranging from 2 to 5, and the vertical axis is the wholesale price w, retail price
p, recycling price b, transfer price A, manufacturer’s profit, retailer’s profit, collec-
tor’s profit and total supply chain profit, respectively. When the demand sensitivity
to recycling price increases, the higher the recycling price is, although the recycling
cost increases, the increase of sales quantity can make the supply chain gain more
profits. Therefore, w, p, b and A will increase in M-led model and R-led model.
The profits of manufacturer and retailer also increase with the increase of ε. In
C-led model, w and p increase as well. However, because the collector is leader, the
impact of demand sensitivity to recycling price on the supply chain at the purchase
stage cannot be transferred to the recycling price and transfer price in the reverse
channel. Therefore, the profit of collector in C-led model is not affected.
5.2.4 The influence of basic recycling quantity on optimal decision. As shown
in Figure 5, the horizontal axis is the basic recycling quantity, ranging from 10 to
50, and the vertical axis is the wholesale price w, retail price p, recycling price b,
transfer price A, manufacturer’s profit, retailer’s profit, collector’s profit and total
supply chain profit, respectively. When the basic recycling quantity increases, the
manufacturer will reduce the transfer price to increase the recycling profit, and the
collector’s recycling price will be reduced accordingly. As the demand is affected by
the recycling price, p and w also decrease. Therefore, the profits of manufacturer
and retailer are reduced. Benefiting from the increase of basic recycling quantity,
the collector’s profit increases.
5.2.5 The influence of the recycling quantity sensitivity to recycling price on
the optimal decision. As shown in Figure 6, the horizontal axis is the recycling
quantity sensitivity to recycling price, ranging from 5 to 8, and the vertical axis is the
1362 HUAQING CAO AND XIAOFEN JI

Figure 4. The influence of the demand sensitivity to recycling

price on optimal decision

Figure 5. The influence of basic recycling quantity on optimal decision

wholesale price w, retail price p, recycling price b, transfer price A, manufacturer’s

profit, retailer’s profit, collector’s profit and total supply chain profit, respectively.
In M-led model and R-led model, similar to the change of a, all decisions decrease
with the increase of l. Benefiting from the significant increase of recycling quantity,
the profits of the manufacturer and collector increase with the increase of l, while
the retailer do not participate in recycling activities, and the profit decrease with
the increase of l. In C-led model, the situation is different. The collector become
channel leader and makes decisions first. When l increases, the recycling quantity
is more significantly affected by the recycling price. If the recycling price increases
slightly, the unit recycle profit decreases slightly, but the recycling quantity increase
significantly, and the collector’s total profit increases. The decisions of all members
in C-led model are affected by b, so A, p and w increase, and the profits of supply
chain members increase with l. By comparing figure 6, we can find the impact of l
on w and p is not significant, while in the reverse channel, the impact of l on b and
A is significant.

6. Conclusions. This paper discusses the optimal strategy of different members

in the closed-loop supply chain when the recycling price affects the consumer de-
mand under three leader models. Closed-loop supply chain is a complex system,
 OPTIMAL RECYCLING PRICE STRATEGY BASED ON CLOSED-LOOP SUPPLY CHAIN
1363

Figure 6. The influence of recycling quantity sensitivity to recy-

cling price on optimal decision

each parameter will affect the optimal strategy. Based on Stackelberg game theory,
we discuss the optimal strategy of three different models: Manufacturer leadership,
retailer leadership and collector leadership. For each case, we derive the optimal
wholesale price, retail price, recycling price and transfer price by using the reverse
induction method. Furthermore, through numerical analysis, we compare the op-
timal strategies of the three models and study sensitivity analysis on the optimal
strategy, and explain the influence of different parameters on the strategy. In terms
of consumer value, environmental impact and profit, our main findings are as fol-
lows:
• In the purchase stage, the larger the sensitivity of consumer demand to the
recycling price, the larger the sales volume, and the larger the profits of the
manufacturer and retailer. This shows the importance of the retailer develop-
ing recycling advertising and improving consumers’ awareness of environmen-
tal protection. Although the retailer do not participate in recycling activities,
they can still benefit from it;
• In the recycling stage, when the recycling market increases, it is not always
beneficial to the manufacturer and retailer. When the basic recycling quantity
increases, the collector will reduce the recycling price to increase the unit
recycle profit, which leads to the reduction of market demand in the positive
channel. In this case, the retailer should make consumers understand the
recycling strategy through advertising, improve the sensitivity of consumer
demand to recycling price, and increase sales, otherwise the profits of the
retailer will decrease with the increase of recycling market;
• When the retailer becomes leader, supply chain efficiency is higher and mem-
bers can get larger profits. This is an interesting finding. Most of the literature
assumes that the manufacturer are the leader, which seems natural because
the manufacturer are involved in manufacturing and remanufacturing at the
same time. However, as an upstream member of the supply chain, the man-
ufacturer often can not consider the information of consumers when making
decisions, which may be the reason for the higher efficiency of the supply chain
under the retailer leadership model;
• Although the recycling price is the highest when the collector is the leader,
the retail price is also the highest. When the retailer is the channel leader,
1364 HUAQING CAO AND XIAOFEN JI

the difference between the retail price and the recycling price is the smallest,
and consumers can get the greatest welfare. This is because the retailer is
the member of the supply chain closest to consumers. This result is similar
to previous studies[6, 23];
Although this paper obtains the optimal strategy under different channel lead-
ership models, there are still many problems worth exploring in the future. First
of all, future research can explore the change of forward and reverse channel sup-
ply chain when there are multiple retailers competing. Secondly, when the retailer
or the manufacturer also participate in recycling activities, how does the optimal
strategy of supply chain change. Finally, it can study the performance of consumers
of different clothing brands (such as luxury and fast fashion) in recycling, so as to
help the clothing supply chain to make better recycling strategy.

Acknowledgment. This research is supported by Zhejiang Soft Science Founda-

tion Project Grant No.2020C35038.

Appendix.
Proof of Lemma 4.1. The second-order Hessian matrix of manufacturer’s profit function
with respect to wholesale price w and transfer price A is:
" 2 M
∂ 2 ΠM
# 
∂ ΠM
−β 4ε
M

∂ 2 w2 ∂w∂A
H = ∂ 2 ΠM ∂ 2 ΠM = ε
M M
2 2 4 −l
∂A∂w ∂ p

In order to guarantee the existence of the optimal solution, H must be negative, so

we have 16lβ − ε2 > 0.
Proof of Proposition 1. The profit function of the collector is: ΠM
C = (AM −bM )(a+
∂ΠM
lbM ). By solving the first derivative ∂bM
C
= 0, we can get bM = lAM2l−a . Intro-
ducing bM into the retailer’s profit function, we can get ΠM
R = (pM − wM )(φ −
∂ΠM
βpM + ε lAM2l−a ). By solving the first derivative R
∂pM = 0, we can get pM =
2lφ+ε(lAM −a)+2lβwM
4lβ . Introducing bM and pm into the manufacturer’s profit func-
2lφ+ε(lAM −a)+2lβwM −a
tion, we can get ΠM
M= (wM − cm )(φ − 4l
+ ε lAM
2l
) + (∆ − AM )(a +
lAM −a ∂ΠM ∂ΠM
2
. By solving the partial derivative ∂wM
) M
= 0 and ∂AM M
= 0, we can get
∗ 8lφ−cm ε2 +2lε∆−6aε+8lβcm ∗ 2lφε−2lβεcm +8l2 β∆−8lβa−aε2
wM = 16lβ−ε2 , AM = 16l2 β−lε2 . By introduc-
∗ ∗ ∗ 12lφ−cm ε2 +3lε∆−9aε+4lβcm
ing wM and AM into pM and bM , we can get pM = 16lβ−ε2 ,
b∗M = φε−βεc16lβ−ε
m +4lβ∆−12lβa
2 .

Proof of Proposition 3. The profit function of the collector is: ΠR

C = (AR − bR )(a +
∂ΠR
lbR ). By solving the first derivative ∂bRC = 0, we can get bR = lAR2l−a . Introducing
bR into the manufacturer’s profit function, we can get ΠR M = (wR − cm )((φ − βpR +
ε lAR2l−a ) + (∆ − AR )(a + lAR2−a ). Here because bR does not include any w, we use
the same method as [28, 31-34], let wR = cm +p 2
R
. By solving the partial derivative
∂ΠR ε(pR −cm )−2a+2l∆
M
∂AR = 0, we can get AR = 4l .
Introducing AR into bR , we can get
ε(pR −cm )+2l∆−6a
bR = 8l . Introducing wR and bR into the retailer’s profit function,
2
we can get ΠR = (pR − cm +p
R
2
R
)(φ − βpR + ε (pR −cm )+2εl∆−6εa
8l ). By solving the first
∂ΠR ∗ 4lφ−cm ε2 +lε∆−3aε+4lβcm
derivative ∂pR = 0, we can get pR =
R
8lβ−ε2 . By introducing p∗R
 OPTIMAL RECYCLING PRICE STRATEGY BASED ON CLOSED-LOOP SUPPLY CHAIN
1365

4lφε−4lβεcm −l∆ε2 +3aε2 −48lβa+16l2 β∆

into bR , AR and wR , we can get b∗R = (8lβ−ε2 )8l , A∗R =
4lφε−4lβεcm −l∆ε2 −aε2 −16lβa+16l2 β∆ ∗ 4lφ−2c ε 2
+lε∆−3aε+12lβc
(8lβ−ε2 )4l , wR = m
16lβ−2ε2
m
.

Proof of Proposition 5. The profit function of the retailer is: ΠC

R = (pC − wC )(φ −
∂ΠC
βpC + εbC ). By solving the first derivative ∂pCR = 0, we can get pC = φ+εbC2β+βwC .
Introducing pC into the manufacturer’s profit function, we can get ΠC M = (wC −
cm )((φ − φ+εbCw+βwC + εbC ) + (∆ − AR )(a + lbC ). Here because pC does not include
any A, we use the same method as Proposition 3, let AC = ∆+b 2 . By solving
C

∂ΠC φ+εbC +βcm

the partial derivative M
∂wC = 0, we can get wC = 2β .
Introducing wC
3φ+2εbC +βwC
into pC , we can get pC = 4β . Introducing AC into the collector’s profit
C ∆+bC
function, we can get ΠC = ( 2 − bC )(a + lbC ). By solving the first derivative
∂ΠC ∗ l∆−a ∗
∂bC = 0, we can get bC = 2l . By introducing bC into pC , AC and wC , we can
C

6lφ+3lε∆−3aε+2lβc 2lφ+lε∆−aε+2lβc
get p∗C = 8lβ
m
, A∗C = 3l∆−a ∗
4l , wC = 4lβ
m
.

REFERENCES

[1] C. M. Armstrong, K. Niinimäki, S. Kujala, E. Karell and C. Lang, Sustainable product-service

systems for clothing: Exploring consumer perceptions of consumption alternatives in Finland,
Journal of Cleaner Production, 97 (2015), 30–39.
[2] X. H. Chen, K. Li, F. Q. Wang and X. H. Li, Optimal production, pricing and government
subsidy policies for a closed loop supply chain with uncertain returns, J. Ind. Manag. Optim.,
16 (2020), 1389–1414.
[3] X. H. Chen, P. Xu, J. J. Li, T. Walker and G. Q. Yang, Decision-making in a retailer-led
closed-loop supply chain involving a third-party logistics provider, J. Ind. Manag. Optim.,
2020.
[4] T. Chi, Consumer perceived value of environmentally friendly apparel: An empirical study of
Chinese consumers, The Journal of The Textile Institute, 106 (2015), 1038–1050.
[5] T. M. Choi, Optimal return service charging policy for a fashion mass customization program,
Service Science, 5 (2013), 56–68.
[6] T. M. Choi, Y. Li and L. Xu, Channel leadership, performance and coordination in closed
loop supply chains, International Journal of Production Economics, 146 (2013), 371–380.
[7] J. Cruz-Cárdenas, J. Guadalupe-Lanas and M. Velı́n-Fárez, Consumer value creation through
clothing reuse: A mixed methods approach to determining influential factors, Journal of
Business Research, 101 (2019), 846–853.
[8] S. Cuc and M. Vidovic, Environmental sustainability through clothing recycling, Operations
and Supply Chain Management: An International Journal, (2014), 108–115.
[9] H. Dahlbo, K. Aalto, H. Eskelinen and H. Salmenperä, Increasing textile circulation–
Consequences and requirements, Sustainable Production and Consumption, 9 (2017), 44–57.
[10] N. C. P. Edirisinghe, B. Bichescu and X. Shi, Equilibrium analysis of supply chain structures
under power imbalance, European J. Oper. Res., 214 (2011), 568–578.
[11] B. Freudenreich and S. Schaltegger, Developing sufficiency-oriented offerings for clothing
users: Business approaches to support consumption reduction, Journal of Cleaner Produc-
tion, 247 (2020), 119589.
[12] J. Gao, H. Han, L. Hou and H. Wang, Pricing and effort decisions in a closed-loop supply
chain under different channel power structures, Journal of Cleaner Production, 112 (2016),
2043–2057.
[13] B. C. Giri, A. Chakraborty and T. Maiti, Pricing and return product collection decisions in
a closed-loop supply chain with dual-channel in both forward and reverse logistics, Journal
of Manufacturing Systems, 42 (2017), 104–123.
[14] G. Hole and A. S. Hole, Recycling as the way to greener production: A mini review, Journal
of Cleaner Production, 212 (2019), 910–915.
[15] S. Jørgensen and G. Zaccour, Equilibrium pricing and advertising strategies in a marketing
channel, J. Optim. Theory Appl., 102 (1999), 111–125.
1366 HUAQING CAO AND XIAOFEN JI

[16] K. Kant Hvass, Post-retail responsibility of garments–a fashion industry perspective, Journal
of Fashion Marketing and Management, 18 (2014), 413–430.
[17] W. Leal Filho, D. Ellams, S. Han, D. Tyler, V. J. Boiten, A. Paço, H. Moora and A. L.
Balogun, A review of the socio-economic advantages of textile recycling, Journal of Cleaner
Production, 218 (2019), 10–20.
[18] N. Liu, T. Choi, C. M. Yuen and F. Ng, Optimal pricing, modularity, and return policy
under mass customization, IEEE Transactions on Systems, Man, and Cybernetics - Part A:
Systems and Humans, 42 (2012), 604–614.
[19] T. Maiti and B. C. Giri, A closed loop supply chain under retail price and product quality
dependent demand, Journal of Manufacturing Systems, 37 (2015), 624–637.
[20] S. K. Mukhopadhyay and R. Setoputro, Optimal return policy and modular design for build-
to-order products, Journal of Operations Management, 23 (2005), 496–506.
[21] L. J. R. Nunes, R. Godina, J. C. O. Matias and J. P. S. Catalão, Economic and environmen-
tal benefits of using textile waste for the production of thermal energy, Journal of Cleaner
Production, 171 (2018), 1353–1360.
[22] H. Peng, N. Shen, H. Liao, H. Xue and Q. Wang, Uncertainty factors, methods, and solutions
of closed-loop supply chain–A review for current situation and future prospects, Journal of
Cleaner Production, 254 (2020), 120032.
[23] Y. Ranjbar, H. Sahebi, J. Ashayeri and A. Teymouri, A competitive dual recycling channel in
a three-level closed loop supply chain under different power structures: Pricing and collecting
decisions, Journal of Cleaner Production, 272 (2020), 122623.
[24] R. Ruiz-Benitez and A. Muriel, Consumer returns in a decentralized supply chain, Interna-
tional Journal of Production Economics, 147 (2014), 573–592.
[25] G. Sandin and G. M. Peters, Environmental impact of textile reuse and recycling-A review,
Journal of Cleaner Production, 184 (2018), 353–365.
[26] R. C. Savaskan, S. Bhattacharya and L. N. Van Wassenhove, Closed-loop supply chain models
with product remanufacturing, Management Science, 50 (2004), 239–252.
[27] R. C. Savaskan and L. N. Van Wassenhove, Reverse channel design: The case of competing
retailers, Management Science, 52 (2006), 1–14.
[28] B. Shen, J. H. Zheng, P. S. Chow and K. Y. Chow, Perception of fashion sustainability in
online community, The Journal of The Textile Institute, 105 (2014), 971–979.
[29] W. B. Wang, P. Zhang, J. F. Ding, J. Li, H. Sun and L. Y. He, Closed-loop supply chain
network equilibrium model with retailer-collection under legislation, J. Ind. Manag. Optim.,
15 (2019), 199–219.
[30] W. Wang, Y. Zhang, K. Zhang, T. Bai and J. Shang, Reward–penalty mechanism for closed-
loop supply chains under responsibility-sharing and different power structures, International
Journal of Production Economics, 170 (2015), 178–190.
[31] Z. D. Wu, X. H. Qian, M. Huang, W. K. Ching, H. B. Kuang and X. W. Wang, Channel
leadership and recycling channel in closed-loop supply chain: The case of recycling price by
the recycling party, J. Ind. Manag. Optim., 17 (2020), 3247–3268.
[32] J. Xie and A. Neyret, Co-op advertising and pricing models in manufacturer-retailer supply
chains, Computers & Industrial Engineering, 56 (2009), 1375–1385.
[33] B. R. Zheng and X. P. Hong, Effects of take-back legislation on pricing and coordination in
a closed-loop supply chain, J. Ind. Manag. Optim., 2021.
[34] Make Fashion Circular: Outlook for a New Textile Economy in China, Report of China
National Textile, 2020. Available from: http://sdgstewardship.org/circular/.

Received July 2021; revised October 2021; early access January 2022.
E-mail address: [email protected]
E-mail address: [email protected]