European Journal of Operational Research 265 (2018) 533–543

Contents lists available at ScienceDirect

European Journal of Operational Research

journal homepage: www.elsevier.com/locate/ejor

Production, Manufacturing and Logistics

Pricing and sourcing strategies for competing retailers in supply

chains under disruption risk
Milan Kumar, Preetam Basu∗, Balram Avittathur
Operations Management Group, Indian Institute of Management Calcutta, Diamond Harbor Road, Kolkata 700104, India

a r t i c l e i n f o a b s t r a c t

Article history: Supply disruption has become a critical concern for businesses around the world. The extant literature
Received 27 June 2016 has dealt mainly with the sourcing decision for a price-taking retailer. In this paper, we study how a
Accepted 10 August 2017
retailer can use pricing decisions along with sourcing strategies under disruption risk while competing
Available online 16 August 2017
against another retailer with a more reliable supply chain. The retailer uses two decision levers namely,
Keywords: price adjustment, and split of order between reliable but expensive supplier and/or cheap but unreli-
Strategic planning able supplier to compete in the end market. Our analyses show that the competitive dynamics is shaped
Supply chain risk management by the cost structure of the players, relative market potential and disruption risk. We find that the re-
Supply disruption tailer focuses on reliable supplies with less price adjustment when it enjoys procurement cost advantage
Sourcing strategies and higher market potential. On the other hand, as the procurement cost advantage and market potential
Pricing decisions shifts to the competitor; the retailer opts for cheaper but risky supplies and relies on drastic price adjust-
ments. These results have important managerial implications and provide critical guidelines for retailers
involved in pricing and sourcing decisions under the threat of supply disruptions.
© 2017 Elsevier B.V. All rights reserved.

1. Introduction orders either with one supplier (single-sourcing) or two suppli-

ers (dual- sourcing). The structure of two suppliers from differ-
Modern supply chains are complex networks that are spread ent geographies is commonly used in supply chain disruption lit-
across the globe where supply disruption risks exist in every link erature (Allon & Van Mieghem, 2010; Ju, Gabor, & van Ommeren,
(Choi, Narasimhan, & Kim, 2012). A recent WEC-Accenture report 2015; Tomlin, 2006). This assumption is used because many firms
(Bhatia, Lane, & Wain, 2013) identifies natural disaster as primary source from low-cost suppliers in developing countries. Sourcing
threat followed by extreme weather, conflict and political unrest, from foreign supplier in a developing country brings cost advan-
terrorism and sudden demand shock as major causes for supply tages but also exposes supply chain to disruptions. Supply disrup-
disruptions. The threat perception has changed over the years as tion is attributed to shipment delays, customs delays, quality prob-
shown in Fig. 1. The report also stresses the importance of resilient lems to name a few (Fang & Shou, 2015). In recent times, firms
supply chains. Resilience is defined as the ability of a supply chain have been adopting sourcing strategies that comprise a mix of sup-
to return to the state it was in before the disruption occurred. Re- plies; a cheaper supplier in a developing country and a more re-
liability and resilience of suppliers are important criteria for long- liable production unit near the end market. This strategy helps to
term success of a retailer competing in end market. Supply disrup- hedge against potential supply disruptions and at the same time
tions often lead to lower return on sales and assets, loss of compet- control costs. In 2009, a Hyundai manufacturing plant in India suf-
itive advantage, market share and goodwill, which in turn impacts fered a prolonged worker’s strike that disrupted overall production.
profitability (Hendricks & Singhal, 2005). Supply chain glitches re- Following this incident, Hyundai decided to move part of their pro-
sult in higher costs and inventories with an immediate effect on duction to Europe. Till 2014, Hyundai was exporting around 45% of
revenues and market share. Supply chain glitches also cause loss the cars manufactured in India to Europe. In 2014, they decided
of shareholders value where smaller firms are affected more than to supply car models to Europe from their plants in Turkey and
the larger firms (Hendricks & Singhal, 2003). Czechoslovakia.1 Hyundai wanted a more stable supply source near
Sourcing has become a crucial strategic decision in mitigating the market (Europe) to compete and improve its presence in Eu-
supply disruption risks. Retailers/Manufacturers often place their rope.

∗ 1
Corresponding author. Gupta S. D. (Aug 08, 2014). As Hyundai changes strategy, India’s status as auto
E-mail addresses: milank12@iimcal.ac.in (M. Kumar), preetamb@iimcal.ac.in (P. export hub in question. Business Standard. Retrieved from: http://www.business-
Basu), balram@iimcal.ac.in (B. Avittathur). standard.com/, accessed on July 28, 2015.

http://dx.doi.org/10.1016/j.ejor.2017.08.019
0377-2217/© 2017 Elsevier B.V. All rights reserved.
534 M. Kumar et al. / European Journal of Operational Research 265 (2018) 533–543

Fig. 1. Causes of supply disruptions.

Source: Supply Chain Risk Radar survey results. World Economic Forum, 2012.

Another phenomenon recently observed is the increase in end- We study two competing retailers engaged in competition in a
market prices because of supply disruptions. In 2004, a devastat- one-period, single-product model. In present times, there are mul-
ing shortage of flu vaccine occurred in the US. Around 46 million tiple instances where two firms dominate market for a product.
doses produced by Chiron, one of the two retailers, had samples Airbus and Boeing in the passenger aircraft market, Intel and AMD
that contained bacterial infection (Yu, Zeng, & Zhao, 2009). The in chips for computing devices, Canon and Nikon in the digital
resulting shortage of drug led to rationing and increase in prices single lens reflex (SLR) camera market are notable examples from
from $60 to $800.2 In another instance, in 2011, the tsunami that different industries. Very often, competing firms use different set
followed a massive earthquake in Japan caused disruption in global of suppliers. In case of duopoly, supply disruption at one of the
production of semiconductors. This calamity led to shortages of firms results in little or no supply in the end-market, thus ben-
parts for Nikon and Canon resulting in increased prices for cameras efitting the competitor and inducing price rise of the product. In
(Fang & Shou, 2015). The price of camera spiraled because of less our model, both the retailers set prices that determine their end-
or no supplies from one of the two suppliers. In both the cases, market demand. In case of disruption, retailers have the option of
the retailers suffered from supply shortages and that resulted in charging a higher price to cover the loss due to lower supplies. In
price rise. The lack of supplies often results in an increase in prices. our model of two competing retailers, one of the retailers has two
This leads to an important research question: how sourcing affects sourcing options; (a) cheaper but unreliable (foreign supplier), and
the competitive dynamics in a system? Retailers often offer substi- (b) reliable but more expensive (domestic supplier) and the sec-
tutable products in the end-market where prices determine con- ond supplier has a reliable supply source. Song and Zipkin (1996),
sumers’ choice. The impact of supply chain glitches on competitors Tomlin (2006), Yang, Aydın, Babich, and Beil (2012) among others,
and pricing as a decision variable have not been explored in detail have assumed no supplies from the foreign supplier in case of dis-
(Hendricks & Singhal, 2003). When faced with lower supplies, re- ruption and we use the same assumption as well. We study the
tailer has an option to increase prices. However, this strategy may competitive dynamics between retailer and the competing retailer,
not be that effective if retailer is competing in the end-market. who is referred as the competitor in this paper henceforth. The
Competitors can take advantage of retailers affected by supply dis- competitor is an integrated firm with a more reliable supply chain.
ruption and capture the market share. In our paper, we look at We determine the cutoff probabilities for optimal sourcing strate-
Bertrand competition between two price-setting retailers and anal- gies for the retailer. We also optimise for the prices to be charged
yse how pricing can be used as an important lever under supply in both the supply disruption and no disruption cases for different
disruptions. We are interested in sourcing and pricing decisions sourcing strategies for both the retailer and the competitor and the
of a retailer exposed to supply disruption risk while competing in split of order for the retailer in case of dual-sourcing. We further
end-market. The competition is further shaped by relative procure- define price-adjustment as a practice by retailer(s) to charge higher
ment cost, market-potential and price-elasticity. More specifically, prices in the end market when one or more retailers receive less
we look at the following research questions: supplies due to supply disruption.
Competitive dynamics is shaped by the interplay between mar-
• What should be the sourcing configuration of a price-setting ket availability, procurement cost and probability of disruption at
retailer under Bertrand competition under different consumer the suppliers’ end. Retailer adjusts prices and uses combination of
price elasticities when supplies are unreliable? cheaper but risky and/or expensive yet reliable sourcing to counter
• How can pricing be used to maximize profit in a competitive supply disruption risks while competing in the end-market. The
scenario when a firm does not receive supplies from one of the broad findings are as follows: on one hand, retailer focuses on re-
suppliers and the competitor has reliable supplies? liable supplies with less price adjustment when the procurement
• How is a reliable supply chain affected by the sourcing struc- cost advantage and higher market potential is with the retailer;
ture of a competing retailer adopting single/dual sourcing on the other, as the procurement cost advantage and market po-
strategies? tential shifts to the competitor, retailer opts for cheaper but risky
supplies and relies on drastic price adjustments. The competitor’s
2
profit function follows a non-monotonic relation with the proba-
Grady D. (Oct 17, 2004). With Few Suppliers of Flu Shots, Shortage Was Long in
Making. The New York Times. Retrieved from: http://www.nytimes.com, accessed on bility of disruption of the retailer’s supplies.
July 28, 2015.
 M. Kumar et al. / European Journal of Operational Research 265 (2018) 533–543 535

The remainder of the paper is arranged as follows. Section Most of the papers in disruption literature have looked at sup-
2 provides a brief review of the literature. Section 3 describes the plier competition, and only a few have studied retailers’ com-
analytical models. Sections 4–6 provide a theoretical analysis of the petition. Babich, Burnetas, and Ritchken (2007) discuss the ef-
models. Section 7 provides a detailed numerical analysis and re- fect of competing suppliers in case of one buyer. They derive
sults. In Section 8, detailed discussion and conclusion of the paper that the buyer can extract more profit with an increased corre-
with possible future research directions have been outlined. lation of suppliers’ default risks, and suppliers will extract more
profit if their default risk has a negative correlation. Deo and Cor-
bett (2009) model the supplier decision to enter a market under
2. Literature review Cournot competition between existing suppliers in the face of yield
uncertainty in the production process. Qi, Shi, and Xu (2015) in-
Our present work draws upon three streams of literature: vestigate the effect of supplier’s competition when suppliers are
sourcing strategies, supply disruption risk management and com- competing on pricing and reliability. One of the findings is that
petitive dynamics in supply chains under disruptions. the supplier should go for high wholesale price and reliability. All
There is an extensive literature on sourcing strategies. The lit- the papers mentioned above are mainly about the supplier’s com-
erature on sourcing strategies identifies three types of strategies petition. Tang and Kouvelis (2011) were one of the first to model
based on the number of suppliers from whom the buyer would retailers’ competition in the face of suppliers with varying yield
source, who are selected from among many qualified suppliers: (1) uncertainty. They use yield uncertainty as an exogenous variable.
single sourcing, (2) dual-sourcing and (3) multiple sourcing. Single Further, they use Cournot game setting to analyse the implications
sourcing has several disadvantages. Yu et al., (2009) mention that of dual sourcing and diversification and the implications it has on
dependence on single source exposes the buying firm to greater retailer’s profit. Chen and Guo (2013) looked at retailer competi-
risk. Minner (2003) highlights the importance of multiple sourcing tion in the presence of supply uncertainty. They look at sourcing
to counter the risk of exchange rate volatility, supply disruptions as a more strategic choice when there is a common supplier. Their
due to machine breakdown, labour strikes or political instability. paper mainly focuses on the optimality of different sourcing strate-
Burke, Carrillo, and Vakharia (2007) report that single sourcing is gies namely single and dual sourcing in the presence of yield un-
optimal when the supplier capacity is more than the total demand. certainty for competing retailers. He, Huang, and Yuan (2015) also
The retailer is not able to take advantage of diversification benefits looked at the pricing and ordering decision for competing retail-
when supplier capacity is large. In all other cases, multiple sourc- ers in the presence of supply disruption. They use a common sup-
ing is optimal. We refer the readers to Bozarth, Handfield, and Das plier and spot market for the competing retailer. They compute a
(1998) and Yu et al., 2009 and references therein for a comprehen- reliability threshold value to decide on the sourcing option. Fang
sive review of the evolution of sourcing strategies. and Shou (2015) compare the value of centralization on retailer’s
Supply Chain Risk Management (SCRM) literature and specifi- competition in case of yield uncertainty of supplier. The decision
cally supply disruptions has been categorized into three broad ar- between single sourcing and dual sourcing has been explored in
eas, (1) yield uncertainty i.e., the difference between the order case of a price-taking retailer. We look at these decisions in the
placed and the order received, (2) lead-time uncertainty and (3) face of competing retailers. We use Bertrand game as we model
supplier disruption where supplier will supply either the full or- price as a decision variable and price adjustment strategies to cope
der in case of no disruption and nothing in case of disruption with supply shortages due to disruptions. We further examine how
(Fang & Shou, 2015). Yano and Lee (1995) provide a comprehen- difference in procurement costs impact the competitive dynam-
sive review of the yield uncertainty literature. They look at the ics amongst retailers under supply disruption. To the best of our
split of order between two suppliers in case of varying yield un- knowledge, our paper is one of the first attempts at modeling pric-
certainty and cost differentials. Anupindi and Akella (1993), Ger- ing as a decision variable under the risk of supply disruption in a
chak and Parlar (1990), Parlar and Wang (1993), Deo and Corbett, duopoly where one of the retailers chooses an optimal pricing and
(2009) and Fang and Shou (2015) are some significant research sourcing strategy to compete with another firm that has a more
works where the authors have used yield uncertainty model. The reliable supply source.
supplier disruption model is widely used in the dual sourcing lit-
erature. Parlar and Berkin (1991), Snyder and Shen (2006), Song 3. Model and analyses
and Zipkin (1996), Tomlin (2006), Yang, Aydin, Babich, and Beil
(2009) & Yu et al., 2009 have made significant contribution to this We study a duopoly market in a single period setting, where
stream of research. Gupta, He, and Sethi (2015) studied the im- one of the retailers (R) has two suppliers, one low cost, but unre-
pact of disruption on competing retailers. Their main focus was on liable foreign supplier, and another expensive but reliable domes-
the timing of order and capacity reservations as risk mitigation op- tic supplier. The foreign supplier has a ‘q’ probability of disruption
tions. The industry examples that motivate the research in this pa- and will not supply anything in case of disruption. This assumption
per are closest to the third area of SCRM. Tomlin (2006) discusses implies that orders are received from the foreign supplier in bulk
the difference between risk mitigation and contingency strategies. at the beginning of the selling horizon and not in batches during
Yang et al. (2012) develop a model where supplier reliability is pri- the selling season. Hence, disruption results in no supplies from
vate information and supplier can either choose to opt for backup the foreign supplier. The domestic supplier is reliable and faces
production or pay a penalty. They use mechanism design to select no supply disruption risks. The other competing retailer (referred
which supplier to source from while factoring in the trade-off be- to as competitor, C) is an integrated firm that produces the prod-
tween the expensive backup option and supplier reliability. Wang, uct internally and is not exposed to supply disruptions. This as-
Gilland, and Tomlin (2010) compare the benefits of dual sourcing sumption is important when a retailer attempts to understand the
and process improvement. In case of random capacity, it is more implication of single sourcing vis-a-vis dual-sourcing when com-
beneficial to invest in process improvement than dual sourcing for peting against a reliable supply chain. All parties are assumed to
low cost difference between the suppliers. It is better to invest in be risk-neutral, and there is no information asymmetry. The cost
dual sourcing than process improvement when the difference in of procurement from the different suppliers and the realised state
reliability is high. For random yield, dual sourcing and process im- of supplies from the foreign supplier is known to both the retail-
provement are favoured under high cost and reliability differences ers. Procurement cost for a retailer from a supplier would mean
respectively. the total direct cost per unit incurred by the retailer in acquiring
536 M. Kumar et al. / European Journal of Operational Research 265 (2018) 533–543

Fig. 2. Pictorial representation of the model.

Fig. 3. Timeline of events.

the product. The procurement cost will include the sum of pur- Table 1
Notations and variables.
chase cost, logistics cost, handling and labor costs for acquiring the
product (He et al., 2015; Inderfurth, Kelle, & Kleber, 2013). Also, i
Pj,k Prices charged in
⎧
the competitor can ramp up its supplies if required and cater to ⎨SF − Single sourcing f rom F oreign supplier,
market demand within a very short lead time. The suppliers do i= SD − Single sourcing f or Domestic supplier,
⎩DS − Dual sourcing
not have capacity constraints. The model is depicted pictorially in
by, 
Fig. 2.
R = Retailer,
The timeline of the events is shown in Fig. 3. The retailer, con- j=
C = Competitor
scious of the prices that would be charged by the competitor, de- in cases
 of
cides on the sourcing strategy considering the probability of dis- N = No disr uption scenar io,
k=
ruption of her foreign supplier. On similar lines, the competitor D = Disr uption scenar io
c1 Cost per unit of foreign supplier
also decides on her prices and produces the goods. When the dis-
c2 Cost per unit of domestic supplier
ruption status is realised, both the retailers update their prices and c Per unit cost of manufacturing for the competitor
announce that in the market. In case of disruption, the competitor q Probability of disruption of foreign supplier 0 < q < 1
determines the additional quantity to be supplied in the market αl The split of order to foreign supplier in dual sourcing for case l = 1, 2
and delivers it with an updated price. and 3. Also, 0 ≤ αi ≤ 1, αl = 1 implies SF and αl = 0 implies SD
πl,i j Expected profit in case l for player j for sourcing strategy i where l
We have assumed a linear price-dependent demand structure.
denotes the different cases based on differences in procurement cost
This demand structure has been widely used in academic lit- structure of the retailer and the competitor
erature, notably by Anderson and Bao (2010), Deo and Corbett
(2009) and Dixit (1979) among others.
monopolist, i.e., a1 − b1 c2 > 0. This also implies that the expression
D1 = a1 − b1 p1 + p2 will be positive for c1 as c1 < c2 .
Demand :
D2 = a2 − b2 p2 + p1 In case of dual sourcing, the orders are placed with both the
suppliers, i.e., 0 < αl < 1. Also, the subscript l is used to represent
where, a1 and a2 are the market potential of the retailer and the three different procurement cost structures of the retailer with re-
competitor, respectively, and b1 and b2 are own price elasticities spect to the competitor. We discuss the cases in detail in subse-
of the retailer and the competitor, respectively. Following practice, quent sections. Next, we present the profit functions of both the
we assume cross-price elasticity to be one. We also assume own- players.
price elasticity is greater than cross-price elasticity, i.e., b1 , b2 > 1
as used in the papers by Anderson and Bao (2010), Biswas, Avit- 3.1. Expected profit functions under different sourcing strategies
tathur, and Chatterjee (2016). Prices set by the retailer and the
competitor are denoted by p1 and p2 , respectively. The retailer has three different strategies at her disposal. She
It is interesting to study, the sourcing configuration of a retailer can go for single sourcing from either the foreign supplier or
when she is competing against a more reliable supply chain. The the domestic supplier, and dual sourcing. The profit functions of
problem analysis is from a game theoretic perspective to under- the retailer and the competitor under the three different sourcing
stand the dynamics of the pricing decision under supply disruption strategies are described below:
in a duopoly setting. Please refer to Table 1 for all the notations
and variables used in the analytical models. 3.1.1. Single sourcing from foreign supplier
We also assume that the demand will be positive in case the In case of single sourcing from the foreign supplier, the retailer
retailer was offering a product at price c1 and c2 in case it was a will not be able to deliver anything to the market if disruption
 M. Kumar et al. / European Journal of Operational Research 265 (2018) 533–543 537

occurs at the foreign supplier. The expected profit function of the

retailer, in this case, will be:

I. πl,R
SF
= (1 − q )(PR,N
SF
− c1 )(a1 − b1 PR,N
SF SF
+ PC,N )
The expected profit function of the competitor will be:

π SF
l,C = (1 − q )(PC,N
SF
− c )(a2 − b2 PC,N
SF SF
+ PR,N )
+q(PC,D
SF
− c )(a2 − b2 PC,D
SF SF
+ PR,D )

3.1.2. Single sourcing from domestic supplier

In case of single sourcing from the domestic supplier, the re-
tailer gets assured supplies and her profit function will be:
 SD  
II. πl,R
SD
= PR,N SD
− c2 a1 − b1 PR,N SD
+ PC,N
The profit function of the competitor will be:

π SD
l,C = (PC,N
SD
− c )(a2 − b2 PC,N
SD SD
+ PR,N )
Fig. 4. Probability of disruption vs. split of order to foreign supplier.
3.1.3. Dual sourcing
In this case, the retailer places orders with both the suppliers.
In the event of a disruption, she will only receive the order she procuring from the foreign supplier, (ii) the competitor’s procure-
placed with the domestic supplier. Hence, in case of disruption, ment cost lies between the retailer’s cost of procuring from the
supplies will be constrained by the fraction of the order that was foreign supplier and the domestic supplier and (iii) the competi-
placed with only the domestic supplier. Also, prices charged in case tor’s procurement cost is more than the retailer’s cost of procuring
of disruption and no disruption will be different and will be based from the domestic supplier.
on the supplies received in each of the cases. The first scenario leads to the competitor being the Stackel-
The objective function of the retailer will be: berg leader. Both the retailers go for a simultaneous move game
 DS   in the second case and in the third case the retailer is the Stackel-
πI,R
DS
= (1 − q ) PR,N − α1 c1 − (1 − α1 )c2 a1 − b1 PR,N
DS
+ PC,N
DS

 DS   berg leader. In subsequent sections, a detailed analysis of the three

+ q PR,D − c2 a1 − b1 PR,D + PC,D
DS DS cases mentioned above has been provided.
III.
s.t., 4. Competitor as Stackelberg leader
   
(1 − α1 ) a1 − DS
b1 PR,N + DS
PC,N ≥ a1 − DS
b1 PR,D + DS
PC,D
In this case, c < c1 < c2 , c1 = c + β1 , c2 = c + β1 + β2 where
The expected profit function of competitor will be: β1 & β2 > 0.
Since, the competitor is more efficient in terms of the procure-
πl,C
DS
= (1 − q )(PC,N
DS
− c )(a2 − b2 PC,N
DS DS
+ PR,N ) ment cost; it acts as a Stackelberg leader. The retailer will opti-
+q(PC,D
DS
− c )(a2 − b2 PC,D
DS DS
+ PR,D ) mize her profit function considering the prices of the competitor
as given and based on the results obtained by the retailer; the
Proposition 1. The expected profit functions in I, II and III are con- competitor will optimise her profit function. The summary of the
cave in retailer’s prices for given value of competitor’s prices and vice- results can be found in Table 2 below. We have provided all the
versa. proofs in Appendix B.
Proof. See Appendix A. Lemma 1. The expected profit function in case of dual sourcing
For notational simplicity, αl = 1 will denote SF. This helps us in can be optimized jointly for split of order, prices to be charged
establishing the continuity of the value of αl . For low values of q, under normal and disruption cases for q∗1 < q < 1 where q∗1 =
we find SF to be the dominant strategy. At q = q∗ , which we define b1 β2 (4b1 b2 −3 )
denotes the probability of dis-
as the critical probability, the retailer moves from SF to DS. This (2a2 b1 −c−a1 +3b1 c1 +2b1 b2 (2a1 +c )−4b21 b2 c1 )
critical probability q∗ differs based on the difference in the pro- ruption at which the retailer switches to DS from SF.
curement cost structure. We also find that αl is non-increasing in
Proposition 2. The expected profit function of the retailer, when the
the probability of supply disruption. The properties of q∗ has been
competitor is Stackelberg leader, is optimised for the unique prices and
explored in details in the subsequent sections. How the optimal
split of order as mentioned below in Table 2. Also, the optimal value of
sourcing strategy changes with q∗ is represented diagrammatically
dual variable for the constraint in case of dual sourcing of the retailer
in Fig. 4.
is μ∗1 = (1 − q )β2 . Also, υ = a1 − c + 2b1 a2 + b1 c1 + 2b1 b2 c.
The retailer has three sourcing strategies viz. single sourcing
from domestic supplier (SD), single sourcing from foreign supplier In case of DS, the dual variable is equal to the product of the
(SF) and dual sourcing (DS), as mentioned above while compet- probability of no disruption and difference of the cost between the
ing against a retailer with a more reliable supply chain. The com- two suppliers. It is independent of the cost difference between the
petitor may have different cost structure. Under the game theo- foreign supplier and competitor’s procurement cost.
retic dynamics, in a Stackelberg game, cost efficiency determines
Corollary 1. The critical probability q∗1 and its relation to model pa-
who acts as a leader in the decision making process. A firm with
rameters are as follows:
sufficient cost advantage gets a first-mover advantage in a duopoly
(Amir & Stepanova, 2006). In their paper, van Damme and Hurkens i q∗1 is increasing in β1 & β2 and b1 & b2 .
(1999) argued that in a linear duopoly game, the firm with low ii For fixed market size, where, a1 + a2 = M, q∗1 is increasing in a2
cost will emerge as a leader in a Stackelberg game. This results and decreasing in a1 .
in three cases based on the competitor’s procurement cost: (i) the iii For a given market size of the competitor, q∗1 is decreasing in both
competitor’s procurement cost is less than the retailer’s cost of a2 & a1 .
538 M. Kumar et al. / European Journal of Operational Research 265 (2018) 533–543

Table 2
Optimum values of prices and split of order for case I.

(a)
Strategy Retailer

PR,N PR,D α
(a1 +PC,N
SF ∗
+b1 c1 ) SF ∗
a1 +PC,D
SF SF
PR,N = 2b1
SF
PR,D = b1
1
a1 +b1 c1 +PC,N
DS∗
b1 β2 +q(a1 +PC,D
DS∗
+b1 c1 ) b1 β2 −q(PC,D
DS∗
−PC,N
DS∗
)
DS F or q∗1 ≤ q ≤ 1 DS
PR,N = 2b1
DS
PR,D = 2qb1
α1 = q(a1 −b1 c1 +PC,N
DS )

(b)
Strategy Competitor

PC,N PC,D

SF SF ∗
PC,N = 2(2b υb −1) SF ∗
PC,D = (a1 −c+ b1 a2 +b1 b2 c )
2(2b b −1 )
1 2 1 2
b1 β2 +qυ
DS F or q∗1 ≤ q ≤ 1 PC,N = 2(2b υb −1)
DS∗ DS∗
PC,D = 2q(2b1 b2 −1 )
1 2

The above corollary implies that when the retailer is at a severe Table 3
Optimum values of prices and split of order for case II.
cost disadvantage compared to the competitor, the retailer should
go for single sourcing from the foreign supplier as she cannot go (a)
for the expensive domestic supplier. The corollary also brings for- Strategy Competitor
ward an interesting point regarding market potential for each of PC,N PC,D
the players. With increasing market potential for the retailer, the SF ∗ ω SF ∗ a1 −c+a2 b1 +b1 b2 c
SF PC,N = 4b1 b2 −1
PC,D = 2(b1 b2 −1 )
focus is on reliable supplies and hence DS is the chosen option. ω b1 (δ1 +δ2 )+qω
DS F or q∗2 ≤q≤1 DS∗
PC,N = 4b1 b2 −1
DS∗
PC,D = q(4b1 b2 −1 )
When the overall market potential is fixed, an increase in the com-
petitor’s market potential leads to an increase in the critical proba- (b)
bility implying a shift towards SF. This happens because with fixed Strategy Retailer
market size, the retailer needs to supply at a lower price to gain
PR,N PR,D α
market share and hence is forced to go with cheaper foreign sup- SF ∗
SF ∗ λ a1 +PC,D
plier even though it is unreliable. SF PR,N = (4b b −1) SF
PR,D = b1
1
1 2
2b1 b2 (δ1 +δ2 )+qλ q∗2
However, if the overall market increases with increase in the DS F or q∗2 ≤ q ≤ 1 DS∗
PR,N = (4b bλ −1) DS∗
PR,D = q(4b1 b2 −1 )
α2∗ = q
1 2

competitor’s market potential, then the critical probability de-

creases with the competitor’s market potential and in turn the re-
tailer opts for DS. With increasing market potential of the com-
petitor which leads to overall increase in the market, the retailer
moves towards reliable supplies in order to hold on to her market Corollary 2. The critical probability q∗2 and its relation to model pa-
potential. rameters are as follows:

i q∗2 is increasing in δ1 & δ2 and b1 & b2 .

5. Simultaneous-move game
ii For fixed market size, where, a1 + a2 = M, q∗2 is increasing in a2
and decreasing in a1 .
When c1 < c < c2 , c = c1 + δ1 , c2 = c1 + δ1 + δ2 where δ1 , δ2 >
iii For a given market potential of the competitor, q∗2 is decreasing in
0.
both a2 & a1 .
The competitor procures at cost ‘c’ that lies between the cost
incurred by the retailer while procuring from the foreign and do- The above results are on similar lines to the results obtained in
mestic supplier respectively. The retailer can create a supply port- Corollary 1. When there is no significant cost advantage to any of
folio where the cost of procurement can be less than or greater the parties, the retailer goes for single sourcing from the foreign
than ‘c’. This results in a scenario wherein no player has a clear-cut supplier in order to lower her procurement cost as she competes
cost advantage. This scenario will further result in a simultaneous directly on pricing with the competitor. We obtain trends similar
move game between the retailers. We have compiled the results in to what was observed in the previous case with changing market
Table 3. The proof for all the strategies can be found in Appendix availability.
C.

Lemma 2. The expected profit function in case of dual sourc- 6. Retailer as Stackelberg leader
ing can be optimized jointly for split of order and prices charged
in normal and disruption cases for q∗2 < q < 1 where, q∗2 = When c1 < c2 < c, c2 = c1 + γ1 , c = c1 + γ1 + γ2 where γ1 , γ2 >
(δ1 +δ2 )(2b1 b2 −1 ) 0.
(a2 +c1 +2a1 b2 +b2 c−2b1 b2 c1 ) , denotes the probability of disruption at Here, the retailer is more efficient in terms of her procurement
which the retailer switches to DS from SF.
cost than the competitor. This cost structure results in the retailer
Proposition 3. The expected profit functions of the retailers in the assuming the role of a Stackelberg leader. We have assumed that
simultaneous move game will be optimised for the unique value of even in the presence of a reliable domestic supplier that is sup-
prices and split of order as mentioned below in Table 3. The optimal plying to the retailer at a price c2 , the competitor is procuring at
value of dual variable for the constraint in dual sourcing is μ2 ∗ = a price higher than that. At times, the production is handled by
(1 − q )(δ1 + δ2 ). the ancillary unit of the parent company. This brings the benefit of
Also, λ = a2 + 2a1 b2 + b2 c + 2b1 b2 c1 reliability and control over the production. The downside may be
the higher cost. Disposing of an existing costlier production plant
ω = a1 + 2 a2 b1 + b1 c1 + 2 b1 b2 c is not easy owing to controls on mass layoffs and opposition from
labour. The parent company, in that case, ends up procuring at a
(δ1 + δ2 )(2b1 b2 − 1 )
q∗2 = higher price even when an alternate cheaper option is available.
( a2 + c1 + 2 a1 b2 + b2 c − 2 b1 b2 c1 )
 M. Kumar et al. / European Journal of Operational Research 265 (2018) 533–543 539

The results have been presented in Table 4 below. The proof of all ii For q∗l < q < 1, πl,R
DS ≥ Max (π SF , π SD ); where l = 1, 2 & 3.
l,R l,R
the results can be found in Appendix C.
Proof. The proof has been given in Appendix D in Online supple-
Lemma 3. The expected profit function in case of dual sourc- ment.
ing can be optimized jointly for split of order, prices charged
under normal and disruption cases for q∗3 < q < 1 where, q∗3 = We have defined critical probabilities of disruption of the for-
γ1 (2b1 b2 −1 ) eign supplier that determine the sourcing configuration for the
(a2 +c1 +2a1 b2 +b2 c−2b1 b2 c1 ) , denotes the probability of disruption at retailer when competitor’s cost falls in three different categories.
which the retailer switches to DS from SF. In the next section, we provide a detailed numerical analysis to
Proposition 4. The expected profit in case of Retailer as SL can be op- compare the analytical results obtained in the three cases outlined
timised for the unique value of prices and split of order as summarised above.
below in Table 4. Dual variable for the constraint in dual sourcing is
μ3 ∗ = (1 − q )γ1 . 7. Numerical analysis
Where τ = a2 − c1 + 2a1 b2 + b2 c + 2b1 b2 c1
In this section, we report the results obtained from detailed nu-
γ ( 2b1 b2 − 1 ) merical analysis. The analysis provides further insights into the an-
q∗3 =
( a2 + c1 + 2 a1 b2 + b2 c − 2 b1 b2 c1 ) alytical results obtained. We have run our models for various input
parameters. However, for expositional brevity, we report the results
for the following dataset. We take market potential a1 = a2 = 1.0
Table 4
Optimum value of prices and split of order for case III.
and b1 = b2 = 1.1. The value of c1 has been taken to be 0.33 and
value of c2 has been taken as 0.66. The value of q has been taken
(a) as 0.25.
Strategy Retailer

PR,N PR,D α 7.1. The Profit Function in case of different values of Competitor’s
SF SF ∗
PR,N τ
= 2(2b b −1) SF ∗
PR,D =
SF ∗
a1 +PC,D
1
Procurement Cost
1 2 b1
γ1 (2b1 b2 −1 )+qτ q∗3
DS F or q∗3 ≤ q ≤ 1 DS∗
PR,N = 2(2b τb −1) DS∗
PR,D = 2q(2b1 b2 −1 )
α3∗ = q
1 2
Here we study the retailer’s and the competitor’s profit under
(b) the three different cost structures that arise due to difference in
Strategy Competitor the procurement cost of the competitor compared with the re-
PC,N PC,D
tailer. The different cost structures have been discussed in detail in
Sections 4–6 (Case I, II and III). In Fig. 5a and b, the value of c2 is
(a2 +b2 c+PR,N
SF ∗
)
SF SF
PC,N = 2b2
SF ∗
PC,D = (a1 −c+ b1 a2 +b1 b2 c )
2(2b b −1 ) 0.66 and 0.50 respectively to understand the trend in the results
1 2
(a2 +b2 c+PR,N
DS∗
) (a2 +b2 c+PR,D
DS∗
)
DS F or q∗3 ≤ q ≤ 1 DS
PC,N = 2b2
DS
PC,D = 2b2
for two values of the competitor’s procurement cost. The maxi-
mum profit made by the retailer (solid line) and the corresponding
aggregate profit of the retailer and competitor (dashed line) across
Corollary 3. The critical probability q∗3 and its relation to model pa-
the three strategies, is shown in the graphs below.
rameters are as follows:
The retailer, as well as the competitor, makes the least profit in
i q∗3 is increasing in γ1 , b1 & b2 but it is decreasing in γ2 . case of the simultaneous move game. In Bertrand competition, the
ii For fixed market size, where, a1 + a2 = M, q∗3 is increasing in a2 reaction function is increasing in nature, hence, the retailer (or the
and decreasing in a1 . competitor) makes the maximum profit when it acts as a Stackel-
iii For a given market potential of the competitor q∗3 is decreasing in berg leader, followed by the case when the retailer (or the com-
both a2 & a1 . petitor) acts as Stackelberg follower and the least profit is made
in a simultaneous move game (Gal-Or, 1985; Hamilton & Slutsky,
The results are similar to the ones obtained in the previous two 1990). From the above graphs, we can conclude that the simulta-
cases except for the relationship of q∗3 with γ2 . We find that with neous move game is clearly not advantageous to either the retailer
increasing difference in the cost of procurement from the domes- or the competitor. The retailer makes the maximum profit in Case
tic supplier and the competitor’s procurement cost, dual sourcing III where her procurement cost is lower than that of the competi-
becomes the dominant strategy for a larger range of the values of tor. Therefore, strategically to minimize the region for Case II, the
probability of disruption. With increasing γ2 , it is profitable for the retailer should try to lower her procurement cost from the domes-
retailer to go with dual sourcing and place larger orders with the tic supplier. This is highlighted in Fig. 5b where the region for Case
reliable domestic supplier as the competitor is stuck with procure- II is decreased by lowering the value of c2a from 0.66 to 0.50.
ment costs that are even higher than the retailer’s cost of procur-
ing from the domestic supplier. 7.2. Prices and demand under no-disruption and disruption for
SF strategy is driven by cost minimization objectives and is different values of competitor’s procurement cost
used more often as a sourcing strategy. The retailers opt for the
cheaper supplier and it exposes them to supply risk. The other two In this section, we report the analysis of the prices charged by
sourcing strategies namely, Single Domestic (SD) and Dual Sourcing the retailer and the competitor under normal as well as disrup-
(DS) are driven by risk mitigation objectives. We use the expected tion and the total supplies in the market under different scenar-
profit in case of SF for the retailer as a benchmark and compare ios given by the cost structures that arise due to the difference
the subsequent profit change in case of other sourcing strategies. in the procurement cost of the competitor compared with the re-
The next proposition outlines the optimality of sourcing strategies tailer. In Fig. 6a, we analyze the prices charged by the retailer and
for a given value of probability of disruption. the competitor under no supply disruption. The retailer uses dual
sourcing in all the three cases. In Case I, the cost differential be-
Proposition 5. The SF is dominant strategy over SD for q less than
tween the competitor’s and the retailer’s procurement cost is not
q∗ . The dual sourcing is dominant strategy for the value of q above
that high. As the cost differential decreases, with increasing pro-
q∗ . SD is a dominant strategy only at q = 1.
curement cost of the competitor, the retailer starts procuring more
i For 0 ≤ q ≤ q∗l , πl,R
SF ≥ π SD
l,R
from the domestic supplier and reduces her prices to capture the
540 M. Kumar et al. / European Journal of Operational Research 265 (2018) 533–543

Fig. 5. Competitor’s procurement cost vs. retailer’s profit and total profit.

Fig. 6. Competitor’s procurement cost vs. prices (6a and b) and total supplies (6c).

competitor’s market share. The competitor, in turn, is also forced creases the order quantities placed with the domestic supplier to
to charge a lower price. However, the decrease in prices for the re- take advantage of the reliable sourcing option even if the cost is
tailer is much steeper than the competitor as the competitor still higher as the competitor loses its cost advantage. In this scenario,
enjoys a cost advantage in this case. the cost of procurement for both the competitor as well as the re-
In Case II, the cost differential between the retailer and the tailer increases. Hence, the retailer must also increase her prices to
competitor is negligible, since the procurement cost of the com- maintain profit levels.
petitor lies between the foreign and domestic procurement cost of In Case III, the procurement cost for the competitor is maxi-
the retailer. Now, as the procurement cost of the competitor in- mum vis-à-vis Case I and II. Therefore, in this scenario, the retailer
creases (in Case II), the competitor is forced to charge a higher increases the share of its supplies from the domestic supplier. This
price. In this case, the competitor cannot lower the price and in- makes the retailer’s supplies more reliable but enhances the cost
crease her profit because the cost advantage that she had over the of procurement. Hence, in this case, both the prices charged by the
retailer in Case I is not available here. The retailer, in turn, in- competitor as well as the retailer increases significantly.
 M. Kumar et al. / European Journal of Operational Research 265 (2018) 533–543 541

Similar price movements are observed in case of prices under advantage, she goes for reliable supplies. Although, her sourc-
supply disruption. Here, the prices charged by the competitor is ing cost increases but the reliability in supplies leads to higher
considerably lower in all the cases than that of the retailer be- supplies in the market and this combined with the higher mar-
cause the competitor enjoys a more reliable supply chain and is ket potential results in higher profit than the competitor. Also,
not facing any supply disruptions. Any disruption at the retailer’s the relative profit of the retailer increase across the diagonal as
end results in an advantageous outcome for the competitor who high probability of disruption is countered by growth in market
in turn charges a much higher price than what it charges under potential.
no disruption but is still in a position to charge a price lower than
that of the retailer. 7.4. Price adjustment and sourcing strategy
In Fig. 6c, we analyse the expected supplies in the market. From
Fig. 6a and b, we find that the prices are decreasing in Case I In this section, we study the ‘price adjustment’ done by the re-
whereas they increase in both Case II and III. This results in in- tailer when disruption happens. We define ‘price adjustment’ as
crease in expected supplies in Case I with decrease in prices. In the percentage change in the disruption prices over the normal
Case II and III, as prices increase, the expected supplies in the mar- prices. We also analyze the split of order that the retailer places
ket decrease. Therefore, from the consumers’ point of view, it is with the foreign supplier under different relative market potential
most beneficial when the competitor’s procurement cost is in the of the retailer and different values of probability of disruption cat-
neighbourhood of the retailer’s cost of procuring from the foreign egorized by high (H) and low (L). We study the above decisions
supplier. Here, the overall procurement costs are the least, leading under market scenarios marked by low (L), and high (H) market
to lowest market prices and maximum overall supplies. potential of the retailer compared to that of the competitor and
cost advantages available either to the competitor (C, Case I), or
7.3. Relative market potential vs. relative profit the retailer (R, Case III) or nobody in particular (No Adv, Case II).
The results are summarized in Table E1 (Appendix E).
Here, we analyze the expected profit earned by the retailer over The retailer adjusts the disruption prices drastically when the
that of the competitor under different values of relative market po- cost advantage is with the competitor and the probability of dis-
tential (RMP) of the retailer (compared to that of the competitor) ruption is low. When the competitor has cost advantage, the re-
and different levels of probability of disruption categorized by high tailer charges a low price and at low probability of disruption, she
(H), medium (M) and low (L). The results provide important in- sources mainly from the foreign supplier. Hence the retailer’s sup-
sights into the profits earned by the retailer at different probabili- plies to the market are adversely affected if disruption occurs and
ties of disruption and varying levels of market potential. We have she is forced to increase the disruption prices significantly. At high
used the following values of c = 0.165, 0.495 & 0.825 for the three probability of disruption, the retailer sources mainly from the reli-
cases. The numerical result has been summarised in Table 5 below. able domestic supplier and hence even under disruption her sup-
The increase in relative market potential results in improve- plies to the end-market remain relatively unaffected. Therefore, the
ment of the profit for the retailer but the advantage available to retailer adjusts the disruption prices marginally. The above phe-
the retailer decreases with increase in the probability of disrup- nomenon is higher in case of lower market potential for the re-
tion. Interestingly, the retailer makes a higher profit than the com- tailer.
petitor even under high probability of supply disruption when its The retailer orders less from the foreign supplier and goes for
relative market potential is high and it has no significant cost reliable supplies with improving cost advantage, higher market
disadvantage compared to the competitor. When the retailer en- potential and lower probability of disruption. When there is no
joys higher relative market potential even when there is no cost cost advantage, the total profit for both the players decrease and
competition brings down the market prices. Hence, the retailer
goes for cheaper and risky foreign supplier and this explains the
Table 5 increase in the split of order.
Expected profit.
From the above analysis, we conclude that the price adjustment
(a) and the optimal sourcing strategy are interplay of three factors:
Case I RMP market potential, relative cost advantage and probability of disrup-
L H tion. Next, we club the results of Tables E1a and E1b and focus
πR πc πR πc on the cases where either of the players enjoys cost advantage.
Our findings highlight broad strategic implications on how a re-
Prob. of disruption L 0.083 0.474 0.172 0.273
H 0.065 0.458 0.153 0.257
tailer can use price adjustment and sourcing structure to compete
against another player with reliable supply chain under different
(b) market potentials and cost structures at varying levels of probabil-
Case II RMP ity of disruption. The findings are presented in Table 6.
L H Table 6 summarises the broad pricing and sourcing strategy
that can be employed by the retailer for different market condi-
πR πc πR πc
tions, cost structures and various levels of probability of disrup-
Prob. of disruption L 0.086 0.291 0.194 0.145 tion. When the competitor enjoys a significant cost advantage, it
H 0.062 0.276 0.170 0.131
is advantageous for retailer to go for cheap and risky supplies and
(c)
counter supply disruptions with very high price adjustments. With
Case III RMP increasing market potential and/or higher probability of disruption
retailer should move towards more reliable supplies and rely less
L H
on price adjustments. Nevertheless, as the retailer starts enjoying
πR πc πR πc procurement cost advantage, the focus should be on gaining mar-
Prob. of disruption L 0.155 0.214 0.306 0.109 ket share. When the retailer has less market potential, she does not
H 0.130 0.205 0.281 0.100 have pricing power and hence should go for cheaper supplies and
π R – expected profit of retailer, π c – expected profit of competi- relatively high price adjustment in case of disruptions. However,
tor. with an increase in market potential, the retailer should go for
542 M. Kumar et al. / European Journal of Operational Research 265 (2018) 533–543

Table 6
Retailer’s strategies.

Cost Adv w competitor Cost Adv w retailer

RPM RPM

L H L H

Prob. of disruption L "VH" price adj & “VH" cheap "H" price adj & "H" cheap and "H" price adj & "H" cheap and "H" price adj & "H" cheap and
and risky sourcing risky sourcing risky sourcing risky sourcing
H "L" price adj & "H" expensive "L" price adj & "H" expensive "L" price adj & "H" expensive "VL" price adj & "VH" expensive
and reliable sourcing and reliable sourcing and reliable sourcing and reliable sourcing
VH – very high, H – high, L – low and VL – very low.

reliable supplies. With increasing probability of supply disruption, As own-cross price elasticity of the retailer or the competi-
the shift towards reliable supplies is more pronounced. The above tor increases, the retailer moves to cheaper supplies even at the
findings provide critical managerial insights for a retailer compet- cost of higher risk since she cannot charge higher prices to off-
ing against another player with a reliable supply chain. set the increased cost of sourcing from the reliable but expensive
domestic supplier. In case of fixed market, as the share of market
potential increases for the retailer, she moves for stable supplies
8. Discussion and conclusion
and the order placed with the reliable domestic supplier increases.
On the contrary, when the market potential of the competitor in-
In this paper, we have studied the effect of competition on
creases, the retailer prioritises cheaper and riskier supplies. In case
sourcing strategy of a retailer sourcing from costly but reliable
of expanding market, the retailer moves for consistent supplies and
domestic supplier and cheap but unreliable foreign supplier. We
larger orders are placed with reliable supplier. Based on the above
included the power dynamics that pans out in the system be-
findings we recommend that in growing markets, larger emphasis
cause of cheaper procurement cost. We prove the concavity of the
should be placed on increasing market share by relying on stable
profit functions in three strategies and then optimize it for differ-
supplies.
ent power structures.
The profit of the retailer is decreasing in probability of disrup-
We define supply chain disruption as a low frequency-high im-
tion. Also, the profit of the competitor increases and then it starts
pact event that results in severance of one or more nodes of the
decreasing. Total profit of both the retailers combined follows a
supply chain leading to unavailability of services or goods. Events
similar trend and increases till the critical probability and then it
such as socio-political instability, civil unrest, natural hazards, ter-
starts decreasing. The important insight is that unreliable supplies
rorist attack, and epidemics can be classified under catastrophic
of even one retailer decrease the overall profit of system.
risk (Kleindorfer & Saad, 2005). In our model, the foreign supplier
The joint profit of both the players is maximum when competi-
has a ‘q’ probability of meeting such catastrophic events and will
tor with reliable supplies has lesser procurement cost. The total
not supply anything in case of such events. Tang and Nurmaya
profit is least when both the players are involved in a simultane-
Musa (2011), Wagner and Bode (2008), Jüttner (2005) and refer-
ous move game. One crucial insight is that it is not only beneficial
ences therein provide a comprehensive list of supply disruption
for the retailer to have lesser procurement cost but also the differ-
risks that affect firms. Natural disasters result in transportation de-
ence between the costs of supply from both the suppliers should
lays, closure of ports, and closure of production facilities and so on.
be less.
Many times, news agencies and weather channels assign probabil-
The prices are decreasing in procurement cost of the competitor
ities or report the chance of occurrence of natural calamities; the
till the competitor’s cost of procurement reaches the cost of pro-
news agencies assign a probability in case of any looming disas-
curement of the retailer from foreign supplier and then it starts
ter that may affect a particular region. In terms of socio-economic
increasing. On the other hand, the supplies are increasing in pro-
and political disruptions, multiple groups and organisations such
curement cost of the competitor till the cost of procurement of the
as Euromoney.com that publishes report on country-specific risk,
competitor reaches that of the retailer’s cost of procurement from
The Economist publishes risk associated with countries, The PRS
the foreign supplier and then it starts decreasing. Hence, the con-
Group publishes monthly report on country-specific risk, and the
sumers are benefited the most when the competitor’s procurement
World Bank publishes the ease of doing business that can be used
cost is in the neighbourhood of the cost of procurement of the re-
as a proxy for reflection of uncertainty.3 The retailers need to draw
tailer from the foreign supplier which in turn means overall the
a list of risks that affect their business with a foreign supplier
procurement cost should be less.
and can calculate the probability of supply disruption (q) based on
The retailer’s profit is decreasing in the probability of disrup-
these reports and by compounding the impact of events that affect
tion. In the case of competitor, it is decreasing till the point when
them and their suppliers.
the retailer moves from SF to DS. This implies that higher proba-
We were able to calculate the critical probabilities of disrup-
bility of disruption of the retailer’s foreign supplier has an inverse
tion that determines the sourcing structure that will be used by
effect on the competitor’s profit.
the retailer when competing with a competitor with more reliable
Finally, we conclude that the competitive dynamics in our mod-
supplies. We further jointly optimise the prices and split of order
els is shaped by the interplay of three important factors; market
for the retailer in case of dual sourcing.
potential, relative cost advantage and probability of disruption and
the retailer can use combination of price adjustments and sourcing
3
Euromoney.com publishes a report on ranking of countries for political risk,
strategies as critical decision levers to maximize her profits. The
economic performance/projections, structural assessment, debt indicators, credit procurement cost advantage and higher market potential is with
Ratings, access to bank finance and access to capital markets. the retailer then she should aim at gaining market share by opt-
The Economist publishes a list assigning risk to each country. ing for reliable supplies and relying less on price adjustments to
The Political Risk Services Group publishes a monthly report on country specific po-
counter supply disruptions. On the other hand, as the procurement
litical, financial and economic risk.
The World Bank under the domain name doingbussiness.org/rankings provides the cost advantage and market potential shifts to the competitor, the
list of ease of doing business for each country on multiple parameters. retailer should opt for cheaper but risky supplies and rely on high
 M. Kumar et al. / European Journal of Operational Research 265 (2018) 533–543 543

price adjustments. The above result has important strategic impli- Gupta, V., He, B., & Sethi, S. P. (2015). Contingent sourcing under supply disrup-
cations for retailers competing in a duopoly under supply disrup- tion and competition. International Journal of Production Research, 53(10), 3006–
3027.
tion risks. Hamilton, J. H., & Slutsky, S. M. (1990). Endogenous timing in duopoly games: Stack-
Most of the papers in dual-sourcing literature look at the sourc- elberg or Cournot equilibria. Games and Economic Behavior, 2(1), 29–46.
ing decision from a price-taking retailer’s perspective. We have He, B., Huang, H., & Yuan, K. (2015). The comparison of two procurement strategies
in the presence of supply disruption. Computers and Industrial Engineering, 85,
included competition and how it affects sourcing and pricing 296–305.
decisions when a competitor has more reliable supplies and can Hendricks, K. B., & Singhal, V. R. (2003). The effect of supply chain glitches on share-
have variable cost of procurement. We have assumed determinis- holder wealth. Journal of Operations Management, 21(5), 501–522.
Hendricks, K. B., & Singhal, V. R. (2005). Association between supply chain glitches
tic demand structure which can be extended to a stochastic setting
and operating performance. Management Science, 51(5), 695–711.
and yield uncertainty at the supplier’s end can be included in a fu- Inderfurth, K., Kelle, P., & Kleber, R. (2013). Dual sourcing using capacity reservation
ture research endeavor. We have not assumed capacity constraints and spot market: Optimal procurement policy and heuristic parameter determi-
nation. European Journal of Operational Research, 225(2), 298–309.
on the supplier. This can be included to build a richer model. Un-
Ju, W., Gabor, A. F., & van Ommeren, J. C. W. (2015). An approximate policy for
certainty in the competitor’s supply chain can be studied in an- a dual-sourcing inventory model with positive lead times and binomial yield.
other interesting future extension. European Journal of Operational Research, 244(2), 490–497.
Jüttner, U. (2005). Supply chain risk management: Understanding the business re-
quirements from a practitioner perspective. The International Journal of Logistics
Supplementary materials Management, 16(1), 120–141.
Kleindorfer, P. R., & Saad, G. H. (2005). Managing disruption risks in supply chains.
Supplementary material associated with this article can be Production and Operations Management, 14(1), 53–68.
Minner, S. (2003). Multiple-supplier inventory models in supply chain management:
found, in the online version, at doi:10.1016/j.ejor.2017.08.019. A review. International Journal of Production Economics, 81-82, 265–279.
Parlar, M., & Berkin, D. (1991). Future supply uncertainty in EOQ models. Naval Re-
References search Logistics, 38(1), 107–121.
Parlar, M., & Wang, D. (1993). Diversification under yield randomness in inventory
Allon, G., & Van Mieghem, J. A. (2010). Global dual sourcing: Tailored base-surge models. European Journal of Operational Research, 66, 52–64.
allocation to near- and offshore production. Management Science, 56(1), 110–124. Qi, L., Shi, J. (Junmin), & Xu, X. (2015). Supplier competition and its impact on firm’s
Amir, R., & Stepanova, A. (2006). Second-mover advantage and price leadership in sourcing strategy. Omega, 55, 91–110.
Bertrand duopoly. Games and Economic Behavior, 55(1), 1–20. Song, J.-S., & Zipkin, P. H. (1996). Inventory control with information about supply
Anderson, E. J., & Bao, Y. (2010). Price competition with integrated and decentralized conditions. Management Science, 42(10), 1409–1419.
supply chains. European Journal of Operational Research, 200(1), 227–234. Snyder, L. V., & Shen, Z.-J. M. (2006). Supply chain managing under the threat of
Anupindi, Ravi, & Akella, Ram (1993). Diversification under supply uncertainty. Man- disruptions. The Bridge (National Academy of Engineering), 36(4), 39.
agement Science, 39(8), 944–963. Tang, O., & Nurmaya Musa, S. (2011). Identifying risk issues and research advance-
Babich, V., Burnetas, A. N., & Ritchken, P. H. (2007). Competition and diversifica- ments in supply chain risk management. International Journal of Production Eco-
tion effects in supply chains with supplier default risk. Manufacturing & Service nomics, 133(1), 25–34.
Operations Management, 9(2), 123–146. Tang, S., & Kouvelis, P. (2011). Supplier diversification strategies in the presence
Bhatia, G., Lane, C., & Wain, A. (2013). Building Resilience in Supply Chains. Re- of yield uncertainty and buyer competition. Manufacturing & Service Operations
tireived from World Economic Forum. www3.weforum.org. Management, 13(4), 439–451.
Biswas, I., Avittathur, B., & Chatterjee, A. K. (2016). Impact of structure, market share Tomlin, B. (2006). On the value of mitigation and contingency strategies for manag-
and information asymmetry on supply contracts for a single supplier multiple ing supply chain disruption risks. Management Science, 52(5), 639–657.
buyer network. European Journal of Operational Research, 253(743), 593–601. van Damme, E., & Hurkens, S. (1999). Endogenous Stackelberg leadership. Games and
Bozarth, C., Handfield, R., & Das, A. (1998). Stages of global sourcing strategy evolu- Economic Behavior, 28(1), 105–129.
tion: An exploratory study. Journal of Operations Management, 16(2–3), 241–255. Wagner, S., & Bode, C. (2008). An empirical examination of supply chain perfor-
Burke, G. J., Carrillo, J. E., & Vakharia, A. J. (2007). Single versus multiple supplier mance along several dimensions of risk. Journal of Business Logistics, 29(1),
sourcing strategies. European Journal of Operational Research, 182(1), 95–112. 307–325.
Chen, J., & Guo, Z. (2013). Strategic Sourcing in the presence of uncertain supply and Wang, Y., Gilland, W., & Tomlin, B. (2010). Mitigating supply risk: Dual sourcing or
retail competition. Production and Operations Management, 23(10), 1748–1760. process improvement? Manufacturing & Service Operations Management, 12(3),
Choi, K., Narasimhan, R., & Kim, S. W. (2012). Postponement strategy for interna- 489–510.
tional transfer of products in a global supply chain: A system dynamics exami- Yang, Z., Aydin, G., Babich, V., & Beil, D. R. (2009). Supply disruptions, asymmet-
nation. Journal of Operations Management, 30(3), 167–179. ric information, and a backup production option. Management Science, 55(2),
Deo, Sarang, & Corbett, Charles J. (2009). Cournot competition under yield uncer- 192–209.
tainty: The case of the U.S. influenza vaccine market. Manufacturing & Service Yang, Z. (Ben), Aydın, G., Babich, V., & Beil, D. R. (2012). Using a dual-sourcing option
Operations Management, 11(4), 563–576. in the presence of asymmetric information about supplier reliability: competi-
Dixit, A. (1979). A model of duopoly suggesting a theory of entry barriers. The Bell tion vs. diversification. Manufacturing & Service Operations Management, 14(2),
Journal of Economics, 10(1), 20–32. 202–217.
Fang, Y., & Shou, B. (2015). Managing supply uncertainty under supply chain Yano, C. A., & Lee, H. L. (1995). Lot sizing with random yields: A review. Operations
Cournot competition. European Journal of Operational Research, 243(1), 156–176. Research, 43(2), 311–334.
Gal-Or, E. (1985). First mover and second mover advantages. International Economic Yu, H., Zeng, A. Z., & Zhao, L. (2009). Single or dual sourcing: Decision-making in
Review, 26(3), 649–653. the presence of supply chain disruption risks. Omega, 37(4), 788–800.
Gerchak, Y., & Parlar, M. (1990). Yield randomness, cost tradeoffs, and diversification
in the EOQ model. Naval Research Logistics, 37(3), 341–354.