w6l1 ShortageCosts ANNOTATED v20 PDF

CTL.
SC1x -Supply Chain & Logistics Fundamentals
Single Period Inventory Models:

Allowing for Stockouts
MIT Center for

Transportation & Logistics
Assumptions: EOQ with Planned Backorders

Demand
n
n
n
Discounts
Constant vs Variable
Known vs Random
Continuous vs Discrete
n
n
Excess Demand
Lead Time
n
n
n
n
Instantaneous
Deterministic vs Stochastic
Internally Replenished
n
n
n
n
n
Independent
Correlated
Indentured
n
n
n
n
One vs Multi vs Multi-Echelon
Capacity / Resources
n
Continuous vs Periodic
Number of Locations
Unlimited
Limited / Constrained
Single Period
Finite Period
Infinite
Number of Items
n
One vs Many
Form of Product
n
n
CTL.SC1x - Supply Chain and Logistics Fundamentals
None
Uniform with time
Non-linear with time
Planning Horizon
Review Time
n
None
All orders are backordered
Lost orders
Substitution
Perishability
Dependence of Items
n
None
All Units vs Incremental vs One Time
Single Stage
Multi-Stage
Lesson: Single Period Inventory Models
EOQ with Planned Backorders

Inventory On Hand
T*
Q*
Time
What will happen to Q* and T* if we allow
for planned backorders at some cost (cs)?
EOQ with Planned Back Orders
Notation
D = Average Demand (units/time)
c = Variable (Purchase) Cost ($/unit)
ct = Fixed Ordering Cost ($/order)
h = Carrying or Holding Charge ($/inventory $/time)
ce = c*h = Excess Holding Cost ($/unit/time)
cs = Shortage Cost ($/unit/time)
Q = Replenishment Order Quantity (units/order)
T = Order Cycle Time (time/order)
N = 1/T = Orders per Time (order/time)
TRC(Q) = Total Relevant Cost ($/time)
TC(Q) = Total Cost ($/time)
Inventory On Hand

T
From similar triangles:
Q-b
Q Qb
b
=
=
T
T1
T2
T2
T1
T1 Q b
=
T
Q
T2 b
=
T Q
Time
! D$
! 1 $! T1 $
! 1 $! T2 $
TRC(Q,b) = ct # & + ce # &# & Q b + cs # &# & b
"Q%
" 2 %" T %
" 2 %" T %
()
! D$
! 1 $! (Q b) $
! 1 $! b $
TRC(Q,b) = ct # & + ce # &#
& Q b + cs # & # & b
"Q%
" 2 %" Q %
" 2 %" Q %
()
!
! D$
# Qb
TRC(Q,b) = ct # & + ce #
# 2Q
"Q%
Single Period Inventory Models
"
Lesson:
$
! b2 $
&
&& + cs # 2Q &
" %
%
Planned Backorders - Solution

!
! D$
# Qb
TRC(Q,b) = ct # & + ce #
# 2Q
"Q%
"
*
PBO
2ct D
ce
b* =
$
! b2 $
&
&& + cs # 2Q &
"
%
%
*
ceQPBO
(c
+ ce
(c
+ ce
cs
"
cs
$
= 1
$
cs + ce
#
Inventory Policy
Order Q*PBO when IOH = - b*
Order Q*PBO every T*PBO time periods
) = Q (c
+ ce
cs
%
' Q*
' PBO
&
Critical Ratio
CR =
cs
(c
+ ce

Q*(PBO) as percent of Q* (w/o backorders)
1000%
Critical Ratio
900%
If cs is very small,
then Q*PBO>>Q*
800%
CR =
700%
cs
(c
+ ce
600%
500%
400%
If cs is very big,
then Q*PBO=Q*
300%
200%
100%
0%
0.0
0.1
0.2
0.3
0.4
0.5
0.6
Critical Ratio Cs/(Cs+Ce)

0.7
0.8
0.9
1.0
Probabilistic Demand:
Single Period Models
Assumptions: Single Period Models

Demand
n
n
n
Discounts
Known vs Random
Continuous vs Discrete
n
n
Excess Demand
Lead Time
n
n
n
n
Instantaneous
Deterministic vs Stochastic
Internally Replenished
n
n
n
n
n
Independent
Correlated
Indentured
n
n
n
n
One vs Multi vs Multi-Echelon
Capacity / Resources
n
Continuous vs Periodic
Number of Locations
Unlimited
Limited / Constrained
Single Period
Finite Period
Infinite
Number of Items
n
One vs Many
Form of Product
n
n
None
Uniform with time
Non-linear with time
Planning Horizon
Review Time
n
None
All orders are backordered
Lost orders
Substitution
Perishability
Dependence of Items
n
None
All Units vs Incremental vs One Time
Single Stage
Multi-Stage
11
Example: NFL Replica Jerseys

Situation:
n
n
n
n
In 2002 Reebok had sole rights to sell

replica NFL football jerseys
Jerseys have unique names & numbers
Peak sales last about 8 weeks
Lead time from contract manufacturer is 12-16 weeks
Main Issue:
n
Reebok had to commit to an order in advance while the actual

demand was uncertain
Question:
n
How many Jerseys of each player should they order?

Case adapted from Parsons, J. (2004) Using A Newsvendor Model for Demand Planning of NFL Replica Jerseys, MIT
Supply Chain Management Program Thesis.
Image Source: http://commons.wikimedia.org/wiki/File:Tom_Brady_%28cropped%29.jpg
12

Data:
n
n
n
Unit cost = c = 10.90 $/jersey

Unit selling price = p = 24 $/jersey
Forecast demand = 32,000 jerseys ( = 11,000)
w History showed demand to be Normally distributed
Select Q* that maximizes profit where X = actual demand:
Profit = p MIN !" x,Q#$ cQ

How do I determine the best policy?
1. Data table
2. Marginal analysis
13
Solving Single Period Model:

Data Table
Sample spreadsheets in MS Excel and LibreOffice are available in this unit.
Data Table
Profit = pMIN (x,Q) cQ

Potential order sizes (Q)
=$D$4*MIN($B8,E$5)-$F$4*E$5
Probability of demand P[x]

=NORMDIST(B10,$B$2,$B$3,1)
Potential demand (x)

15
=SUMPRODUCT($C$7:$C$48,E7:E48)
Expected Profits
Expected Profit
$350.00
$325.00
$300.00
$275.00
Expected Total Profit ($k)
$250.00
$225.00
$200.00
$175.00
$150.00
$125.00
$100.00
$75.00
$50.00
$25.00
$0
10
15
20
25
30
Order Quantity
35
40
45
50
Solving Single Period Model:

Marginal Analysis
18
Marginal Analysis
For single-period problems we have two costs:

ce = Excess cost when D<Q ($/unit) i.e. having too much product
cs = Shortage cost when D>Q ($/unit) i.e. having too little product
Assuming a continuous distribution of demand , we get

ce P[XQ] = expected excess cost of the Qth unit ordered
cs (1-P[XQ]) = expected shortage cost of the Qth unit ordered
If E[Excess Cost] < E[Shortage Cost] then increase Q

We are at Q* when E[Shortage Cost] = E[Excess Cost]
Marginal Analysis
Marginal Shortage and Excess Costs
$14.00
cs 1 P #$ x Q%&
$13.00
$12.00
Marginal Cost per Jersey
$11.00
$10.00
$9.00
$8.00
$7.00
$6.00
$5.00
$4.00
$3.00
ce P "# x Q$%
$2.00
$1.00
$0
10
15
20
25
30
35
40
45
50
55
60
Order Quantity
Excess Cost
Shortage Cost
20
Marginal Analysis
ce P "# x Q$% = cs 1 P "# x Q$%
ce P "# x Q$% = cs cs P "# x Q$%

ce P "# x Q$% + cs P "# x Q$% = cs
P "# x Q$% ce + cs = cs
P "# x Q$% =
cs
(c
+ cs
The Critical Ratio

NFL Jersey Example
- solved

Data:
n
n
n
Total cost = c = 10.90 $/jersey

Selling price = p = 24 $/jersey
Forecast demand ~N(32000, 11000)
Solution:
n
n
n
n
cs = p c = 24 -10.90 = $13.10
ce= c = $10.90
CR = (13.10)/( 10.9 + 13.10) =0.546
Select Q where P[xQ] = 0.546
w Normal Table or use spreadsheet:
23
Standard Normal Table

P[xQ] = 0.546
Find k= 0.115
Recall k=(Q-)/
So, Q = +k
= 32000+(0.115)(11000)
Q = 33,267 units
24

Data:
n
n
n

Solution:
n
n
n
n
cs = p c = 24 -10.90 = $13.10
ce= c = $10.90
CR = (13.10)/( 10.9 + 13.10) =0.546
w =NORMINV(CR, Mean, StdDev)
w =NORMINV(0.546, 32000, 11000)
Q* = 33,267 - the profit maximizing quantity
But what if I can sell the left overs at a discount?

25
Considering Other Costs

Other costs:
g = salvage value, $/unit

B = Penalty for not satisfying demand
(beyond lost profit), $/unit
The excess and shortage costs change:
cs = p c + B
ce = c - g
Critical Ratio = cs/(cs+ce)
= (p-c+B)/(p-c+B+c-g)
=(p-c+B)/(p+B-g)

Data:
n
n
n
n

Salvage value = g = 7 $/jersey
Solution:
n
n
n
n
cs = p c = 24 - 10.90 = $13.10
ce= c - g = 10.90 7.00 = $3.90
CR = (13.10)/( 3.9 + 13.10) =0.771
w =NORMINV(CR, Mean, StdDev)=NORMINV(.771, 32000, 11000)
Q* = 40,149- the profit maximizing quantity
But, how do I determine the profitability?

27
Key Points from Lesson
28
Key Points
Newsvendor problems are everywhere
Fashion items, perishable goods, fleet sizing,

contracting, space missions, etc.
Whenever you have to make a firm bet in the face of
uncertain demand in a single period
Classic trade off between:
Having too much (excess cost ce)

Having too little (shortage cost cs)
Critical Ratio captures this trade-off
CR = Cs / (Cs + Ce)
CR = Pct of demand distribution to cover
= P[xQ]
CTL.SC1x -Supply Chain & Logistics Fundamentals
Questions, Comments, Suggestions?

Use the Discussion!
MIT Center for

Transportation & Logistics
caplice@mit.edu

w6l1 ShortageCosts ANNOTATED v20 PDF

Transféré par

Informations du document

Titre original

Copyright

Formats disponibles

Partager ce document

Partager ou intégrer le document

Options de partage

Avez-vous trouvé ce document utile ?

Ce contenu est-il inapproprié ?

Droits d'auteur :

Formats disponibles

w6l1 ShortageCosts ANNOTATED v20 PDF

Transféré par

Droits d'auteur :

Formats disponibles

CTL.

SC1x -Supply Chain & Logistics Fundamentals

Single Period Inventory Models:

MIT Center for

Assumptions: EOQ with Planned Backorders

One vs Multi vs Multi-Echelon

CTL.SC1x - Supply Chain and Logistics Fundamentals

Lesson: Single Period Inventory Models

EOQ with Planned Backorders

Lesson: Single Period Inventory Models

EOQ with Planned Back Orders

CTL.SC1x - Supply Chain and Logistics Fundamentals

Lesson: Single Period Inventory Models

Lesson: Single Period Inventory Models

EOQ with Planned Backorders

From similar triangles:

CTL.SC1x - Supply Chain and Logistics Fundamentals

Planned Backorders - Solution

CTL.SC1x - Supply Chain and Logistics Fundamentals

Lesson: Single Period Inventory Models

EOQ with Planned Backorders

Lesson: Single Period Inventory Models

EOQ with Planned Backorders

Critical Ratio Cs/(Cs+Ce)

Lesson: Single Period Inventory Models

Lesson: Single Period Inventory Models

Assumptions: Single Period Models

One vs Multi vs Multi-Echelon

CTL.SC1x - Supply Chain and Logistics Fundamentals

Lesson: Single Period Inventory Models

Example: NFL Replica Jerseys

In 2002 Reebok had sole rights to sell

Reebok had to commit to an order in advance while the actual

How many Jerseys of each player should they order?

CTL.SC1x - Supply Chain and Logistics Fundamentals

Lesson: Single Period Inventory Models

Example: NFL Replica Jerseys

Unit cost = c = 10.90 $/jersey

Select Q* that maximizes profit where X = actual demand:

Profit = p MIN !" x,Q#$ cQ

Lesson: Single Period Inventory Models

Solving Single Period Model:

Lesson: Single Period Inventory Models

Sample spreadsheets in MS Excel and LibreOffice are available in this unit.

Profit = pMIN (x,Q) cQ

Probability of demand P[x]

Potential demand (x)

CTL.SC1x - Supply Chain and Logistics Fundamentals

Lesson: Single Period Inventory Models

Expected Total Profit ($k)

CTL.SC1x - Supply Chain and Logistics Fundamentals

Lesson: Single Period Inventory Models

Solving Single Period Model:

Lesson: Single Period Inventory Models

For single-period problems we have two costs:

Assuming a continuous distribution of demand , we get

If E[Excess Cost] < E[Shortage Cost] then increase Q

Lesson: Single Period Inventory Models

Marginal Cost per Jersey

CTL.SC1x - Supply Chain and Logistics Fundamentals

Lesson: Single Period Inventory Models

ce P "# x Q$% = cs 1 P "# x Q$%

ce P "# x Q$% = cs cs P "# x Q$%