Transcript Impact of Returns on Supply Chain Coordination

Impact of Returns on Supply
Chain Coordination
Ana Muriel
Department of Mechanical and Industrial Engineering,
University of Massachusetts
In collaboration with Rocio Ruiz-Benitez
Outline
Motivation
Model
Analysis
Computational Study
Conclusions
Motivation
The value of commercial product returns now exceeds
$100 billion annually in the US (Stock, Speck and Shear
(2002))
Commercial product returns: Products returned for any reason
within 90 days of purchase.
Hewlett Packard recently estimated the cost of consumer
returns for North America exceeded 2% of their total outbound
sales revenue.
Reason
% of returns
Returns ~ 6% of sales
Defective
20%
Could not install
27.5%
Performance
40%
Convenience
12.5%
Ferguson, Guide and Souza (2005)
Motivation
Policy of most US retailers:
Full returns no question asked!!
Return rates: 6% to 15% (Dekker and Van der Laan
(2003))
Mail order companies and e-tailers: as high as
35%
Largely ignored in supply chain coordination and
contracts literature
Most research on consumer returns concerns inventory
policies, production planning and reverse logistics
(Fleischmann and Kuik (2003), Kiesmuller (2003))
Literature Review
Wood (2001), “Remote Purchase Environments: The influence of Return
Policy Leniency on Two-Stage Decision Processes”, Journal of Marketing
Research 38, 157-169.
Dekker and Van der Laan (2003), “Inventory control in reverse logistics”,
chapter in Business Aspects of Closed-Loop Supply Chains, V.D. Guide Jr., L.N.
Van Wassenhove, editors. Carnegie Mellon University Press, Pittsburgh, PA
Fleischmann M. and Kuik R. (2003), “On optimal inventory control with
independent stochastic items returns”, European Journal of Operational
Research 151, 25-37
Kiesmuller, G.P. (2003), “Optimal control of a one product recovery system
with leadtimes”, International journal of Production Economics 81-82, 333-
340
Ferguson, Guide and Souza (2005), “Supply Chain Coordination for False
Failure Returns”, working paper. Georgia Institute of Technology.
Souza, Guide, van Wassenhove and Blackburn (2005), “Time Value of
Commercial Product Returns”, working paper. University of Maryland.
Research Questions:
What is the profit impact of incorporating consumer
returns in our decision models?
Centralized system
Decentralized system
How does it affect retail prices and quantities
ordered?
How does this depend on
the magnitude of logistics costs?
the relative share between retailer and manufacturer?
the proportion of product that is returned?
Classical Model
Two-echelon supply chain
cQ
Sales S = min(y,Q)
wQ
Manufacturer
Retailer
rS
s(Q-S)
Stochastic and price dependent
Manufacturer’s decision variables:
Retailer’s decision variables:
Single replenishment opportunity
demand y
wholesale price w
repurchase price s
order quantity Q
selling price r
Returns Model
vR
l1R
l2R
wQ
cQ
Manufacturer
s(Q-S)
wR
rS
Retailer
rR
A percentage of sales is returned Returns R = aS
Manufacturer’s returns logistics cost:
l1
Retailer’s returns handling cost:
l2
This costs include inspection, shipping, sorting, repackaging,
remanufacturing, disposal
Average salvage value of returned item v
Costs Associated with Returns
System costs:
=r-v+l
Manufacturer costs
1 = w - v + l1
Retailer costs
2 = r – w + l2
Demand Distribution
y = stochastic and price dependent demand faced by
the retailer:
y=xD(r)
x= positive r. v. with mean 1 and density function f()
D(r) = expected demand quantity, decreasing in
retail price
Demand density function
1
g ( y; r ) 
D( r )
 y 

f 
 D( r ) 
Profit Functions and Optimal Decision
Variables:
Centralized System
C = rS – cQ – R
 r  a  c 
Q  D(r ) F 

r

a


*
C
1
Decentralized System
T = R +  M
Retailer
R = rS +s(Q-S)– wQ – 2R
 r   2a  w 
Q  D(r ) F 

 r   2a  s 
*
D
1
Manufacturer
M = (w-c)Q – s(Q-S) –1R
Analysis
Objective: Compare the following decision rules
Policy IR:
Ignores customer returns
when optimizing
QIR, rIR, wIR, sIR
Customer returns
considered a posteriori, to
calculate respective profits
Expected profit: IR
Policy CR:
Considers customer
returns when optimizing
QCR, rCR, wCR, sCR
Expected profit: CR
Analysis: Centralized System
Proposition: Under deterministic and price
dependent demand, the optimal retail price
increases and the order quantity decreases
when considering consumer returns. That is,
QCR< QIR and rCR> rIR
Intuitive since the profit margin is reduced
by consumer returns.
Analysis: Centralized System
Theorem: Under stochastic and price
dependent demand we have that
1. For fixed r,
QCR(r)< QIR(r)
2. For fixed Q,
rCR(Q)> rIR(Q)
3. Under mild conditions,
QCR< QIR and rCR> rIR
C1: For all r>
rIR,
C2: For all Q<QIR,
QIR(r)
rIR(Q)
 IR IR
Q (r )
 IR IR
r (Q )
Analysis:
Decentralized System
Corollary: Given w, the retailer’s optimal
decisions satisfy:
1. For fixed r,
QCR(r)< QIR(r)
2. For fixed Q,
rCR(Q)> rIR(Q)
3. Under mild conditions,
QCR< QIR and rCR> rIR
C1: For all r>
rIR,
C2: For all Q<QIR,
QIR(r)
rIR(Q)
 IR IR
Q (r )
 IR IR
r (Q )
Question
Will consumer returns always result in
higher prices and lower quantities in a
decentralized supply chain?
Analysis: System Coordination
Under Buy-Back Contracts
Theorem: Under consumer returns, a policy
that allows for unlimited returns at a partial
credit s will lead to supply chain coordination
for appropriate values of s and w. In
particular,
a
s
1a
(c  l1 )
Allowing no returns is system suboptimal
Extension of Pasternack(1985), demand is not price dependent
Computational Study
Assumptions:
f(x) ~ uniform distribution in [0,2]
Linear demand model
D(r)=b(r-k)
where b<0 and k>0 constants
b=-3, k=5
(Emmons and Gilbert (1998))
Centralized System
Sensitivity Analysis with respect to a
l=1
l=2
l=3
8
6
CR
IR
4
2
0
-2
6%
10%
15%
20%
25%
30%
-4
We observe: QCR < QIR and rCR > rIR
QCR decreases as l increases
Profit difference increases with l and a
35%
a
Decentralized System
Optimal r and Q
CR
IR
Q*
r*
8
7
6
5
4
3
2
1
0
We observe:
QCR < QIR
rCR > rIR
1
1.4
1.8
2.2
2.6
3
3.4
3.8
4.2
4.6
w
Profit functions
8
Manuf.
Retail.
Total
For fixed value of w,
RCR > RIR
6
4
2
0
-2 1 1.2 1.4 1.6 1.8 2 2.2 2.4 2.6 2.8 3 3.2 3.4 3.6 3.8 4 4.2 4.4 4.6 4.8
-4
w
But for optimal w,
RIR > RCR
Profit Functions at optimal w
CR
Manuf.
Retail.
Total
Profit Functions
IR
6
5
4
3
2
1
0
6%
10%
15%
20%
a
Percent Savings
Manufacturer: up to 10%
Retailer:
9% to 66%
Total:
6% to 37%
25%
30%
35%
Sensitivity Analysis
With respect to:
1) Share of logistic cost faced by
retailer ()
2) Percentage of consumer returns (a)
Optimal Q, r and w
4.5
Q*
r*
w*
CR
IR
Under policy IR…
4
QIR, rIR and wIR constant
3.5
3
logistics costs do not intervene in
the decision making process
2.5
2
1.5
1
Under policy CR…
0.5
0
5%
25%
50%
75%
95%

Manufacturer decreases wCR
Manuf.
Retail.
Total
Profit functions
QCR and rCR increase with ;
as incentive for retailer to increase
order quantity
Ends up bearing all logistics cost
3.5
3
2.5
2
1.5
1
0.5
0
5%
25%
50%

75%
95%
If  > 70% => RIR* < RCR*
Retailer's profits
Manufacturer's Profits
3.5
3
2.5
2.5
2
1.5
2
1.5
1
1
0.5
0.5
0
0
5%
5%
25%
50%
75%
95%

25%
50%
-0.5
-1

Total Profits
a=.06
a=.2
a=.35
6
5
4
3
CR
IR
2
1
0
5%
25%
50%

75%
95%
75%
95%
Conclusions
When considering returns …
Centralized system:
1) Lower quantities and higher retail prices
2) Significant profit increase
Decentralized system:
1) Lower quantities and higher retail prices
2) Poor coordination of the supply chain
All members worse off in general
Ignoring returns reduces double marginalization
3) The manufacturer bears the returns logistics costs:
Higher percentage
manufacturer decreases
incurred by retailer
wholesale price to compensate