Inventory Management and Risk Pooling

Dr Prashant Gupta

Inventory
Inventory can be broadly defined as the stock of
goods, commodities or other economic
resources that are stored or reserved at any
given period for future production or for meeting
future demand.
Inventory may also be defined as usable but
idle resource having economic value. If
resource is some physical and tangible object
such as materials, then it is generally termed as
stock.

Importance of Inventory
Inventory - Double-Edged Sword

Changes in inventory levels affect return on assets (ROA),

an important internal and external metric.
Ultimate challenge is to balance supply and demand as well
as cost burden and customer service.
Inventory turns is a measure of how well a companys
products are doing in the market and how well its inventory
is managed.

Inventory Management Decisions

Inventory management answers two questions
How much to order
When to order

Types of Inventory
Raw materials, Component parts (Production
Inventories)
Spares and other indirect materials ( MRO
Inventories))
In-process (partially completed) products
Finished goods
Tools, machinery, and equipment
Labor
Working capital

Reasons To Hold Inventory

Meet unexpected / variation in customer demand

Smoothen seasonal or cyclical demand
Hedge against price increases
Take advantage of price discounts
Quantity discounts
Maintain independence of operations
Allow flexibility in production scheduling
Safeguard for variation in raw material delivery time

Inventory
Where do we hold inventory?
Suppliers and manufacturers
warehouses and distribution centers
retailers

Types of Inventory
WIP( Work in process)
raw materials
finished goods

Why do we hold inventory?

Economies of scale
Uncertainty in supply and demand

Role of Inventory in the Supply Chain

Improve Matching of Supply
and Demand
Improve Forecasting
Reduce Material Flow Time
Reduce Waiting Time
Reduce Buffer Inventory

Economies of Scale

Supply / Demand
Variability

Seasonal
Variability

Cycle Inventory

Safety Inventory

Seasonal Inventory

Figure Error! No text of

Role of Cycle Inventory in a Supply Chain

Lot or batch size:: Quantity that a supply chain stage either
produces or orders at a given time.
Cycle inventory:: Average inventory that builds up in the supply
chain because a supply chain stage either produces or purchases
in lots that are larger than those demanded by the customer.
Q = lot or batch size of an order
d = demand per unit time= Average Flow Rate

Inventory profile:: Plot of the inventory level over time

Cycle inventory = Q / 2 (depends directly on lot size)
Average flow time = Average inventory / Average flow rate
Average flow time from cycle inventory = Q / (2d)

Inventory Costs
Holding (or carrying) costs
Costs for storage, handling, insurance, etc
Setup (or production change) costs
Costs for arranging specific equipment setups, etc
Ordering costs
Costs of placing an order,replenishing inventory, etc (Fixed
& Variable)
Shortage costs
Costs of canceling an order, temporary or permanent loss
of sales when demand cannot be met, etc

Inventory Carrying Costs

Inventory
carrying
costs

Capital
costs

Inventory investment

Inventory
service
costs

Insurance
Taxes
Plant warehouses

Storage
space costs

Public warehouses
Rented warehouses
Company-owned
warehouses
Obsolescence

Inventory
risk costs

Damage
Pilferage
Relocation costs

Inventory Systems
Single-Period Inventory Model:
One time purchasing decision (Example: vendor selling t-shirts
at a football game).
Seeks to balance the costs of inventory over-stock and
inventory under - stock.
Multi-Period Inventory Models:
Fixed-Order Quantity Model
Constant amount ordered when inventory declines to
predetermined level
Event triggered (Example: Below Pre-determined Level, running
out of stock)
Fixed-Time Period Model
Order placed for variable amount after fixed passage of time
Time triggered (Example: Monthly sales call by sales
representative)

Single-Period Inventory Model

Cuu
P
Coo Cuu

This
Thismodel
modelstates
statesthat
thatwe
we
should
shouldcontinue
continueto
toincrease
increase
the
thesize
sizeof
ofthe
theinventory
inventoryso
so
long
longas
asthe
theprobability
probabilityof
of
selling
sellingthe
thelast
lastunit
unitadded
addedisis
equal
equalto
toor
orless
lessthan
thanthe
theratio
ratio
of:
of:Cu
Cu//(Co
(Co++Cu)
Cu)

PCo
Where :
PCo(1-P)
(1-P)Cu
Cu
Co Cost per unit of demand over estimated

Cu Cost per unit of demand under estimated

P Probability that the unit will not be sold

Single Period Model Example

Our Institute purchases event specific TShirts for LBSIM Festival every year.
Based on our past experience, we sell on
average 2,400 shirts with a standard
deviation of 350. We make Rs 100 on
every shirt we sell at the event, but lose
Rs 50 on every shirt not sold. How many
shirts should we make for the event?

Single Period Model Example

Our Institute purchases event specific T-Shirts for LBSIM
Festival every year. Based on our past experience, we
sell on average 2,400 shirts with a standard deviation of
350. We make Rs 100 on every shirt we sell at the event,
but lose Rs 50 on every shirt not sold. How many shirts
should we make for the event ?
Cu = Rs 100 and Co = Rs 50; P 100 / (100 + 50) = .667
Z.667 = .432 (use NORMSDIST(.667)
therefore we need 2,400 + .432(350) = 2,551 shirts

A Multi-Period Inventory Model

Often, there are multiple reorder opportunities
Consider a central distribution facility which
orders from a manufacturer and delivers to
retailers. The distributor periodically places
orders to replenish its inventory.

Comparison of Fixed-Order Quantity and Fixed-Time Period Reordering

Inventory Systems
( P)
Fixed-Time Period
Reordering System

(Q)
Fixed-Order
Quantity System

Idle State
Waiting for demand

Idle State
Waiting for demand
Demand Occurs
Units withdrawn from
inventory or back ordered
Compute inventory position
Position = on hand +
On order back order

No

Demand Occurs
Units withdrawn from
inventory or back ordered
No

Has review time

arrived?
Yes
Compute inventory position
Position = on hand +
On order back order

Is position___
Reorder position?
Yes
Issue an order for
exactly Q units

Compute order quantity to

bring inventory upto
required level
Issue an order for
the number of units needed

Multi-Period Models:
Fixed-Order Quantity Model Assumptions
Demand for the product is constant and uniform
throughout the period.
Lead time (time from ordering to receipt) is constant.
Price per unit of product is constant.

Multi-Period Models:
Fixed-Order Quantity Model Assumptions(Contd.)
Inventory holding cost is based on average inventory.
Ordering or setup costs are constant.
All demands for the product will be satisfied (No back
orders are allowed).

Basic Fixed-Order Quantity Model and Reorder

Point Behavior
4. The cycle then repeats.

1. You receive an order quantity Q.

Number
of units
on hand

2. Your start using

them up over time.

R = Reorder point
Q = Economic order quantity
L = Lead time

Time

L
3. When you reach down to
a level of inventory of R,
you place your next Q
sized order.

Cost Minimization Goal

By
Byadding
addingthe
theitem,
item,holding,
holding,and
andordering
orderingcosts
costs
together,
together,we
wedetermine
determinethe
thetotal
totalcost
costcurve,
curve,which
whichin
in
turn
inventory order point that
turnis
isused
usedto
tofind
findthe
theQ
Qopt
opt inventory order point that
minimizes
minimizestotal
totalcosts
costs
Total Cost
C
O
S
T

Holding
Costs
Annual Cost of
Items (DC)
Ordering Costs
QOPT
Order Quantity (Q)

Basic Fixed-Order Quantity (EOQ)

Model Formula
Total
Annual =
Cost

Annual
Annual
Annual
Purchase + Ordering + Holding
Cost
Cost
Cost

D
Q
D
Q
TC
TC == DC
DC ++ SS++ H
H
Q
22
Q

TC=Total
TC=Totalannual
annual
cost
cost
DD=Demand
=Demand
CC=Cost
=Costper
perunit
unit
QQ=Order
=Orderquantity
quantity
S(Co)
S(Co)=Cost
=Costof
of
placing
placingan
anorder
orderor
or
setup
setupcost
cost
RR=Reorder
=Reorderpoint
point
LL=Lead
=Leadtime
time
HH(Cc)=Annual
(Cc)=Annual
holding
holdingand
andstorage
storage
cost
costper
perunit
unitof
of
inventory
inventory

Deriving the EOQ

Using
Using calculus,
calculus, we
we take
take the
the first
first derivative
derivative of
of
the
the total
total cost
cost function
function with
with respect
respect to
to Q,
Q, and
and
set
set the
the derivative
derivative (slope)
(slope) equal
equal to
to zero,
zero, solving
solving
for
for the
the optimized
optimized (cost
(cost minimized)
minimized) value
value of
of QQopt
opt
QQOPT
=
OPT =

2DS
2DS =
=
HH

We
Wealso
alsoneed
needaa
reorder
reorderpoint
pointto
to
tell
tellus
uswhen
whento
to
place
placean
anorder
order

2(Annual
2(AnnualDemand)(Order
Demand)(Orderor
orSetup
SetupCost)
Cost)
Annual
AnnualHolding
HoldingCost
Cost
__

Reorder
Reorder point,
point, R
R == ddLL

d = average daily demand (constant)

L = Lead time (constant)

EOQ Cost Model

CO - cost of placing order

D - annual demand

CC - annual per-unit carrying cost

Q - order quantity

Annual ordering cost = COD/Q

Annual carrying cost = CCQ/2

Total cost = COD/Q + CCQ/2

Co D
Cc Q

Q
2
2Co D
Q2
Cc
Qopt
TC min

2Co D
Cc
Cc Qopt
Co D

Qopt
2

CoD
CcQ
TC

Q
2
TC
CoD
Cc

2
Q
2
Q
0

C oD
Cc

2
Q2

Q opt

2CoD
Cc

EOQ With Non- instantaneous Receipt

Inventory
level
Q(1-d/p)
Q
(1-d/p)
2

0
Order
receipt period

Maximum
inventory level
Begin
Order
receipt

End
Order
receipt

Average
inventory level

Time

EOQ With Non - instantaneous

Receipt

Q
d
Max inv level = Q - d Q 1 -
p
p

p = production rate
d = demand rate

1
d
Q
d
Avg inv level = Q 1 - 1 -
2
p
2
p
Cc Q
d
1 -
Total carrying cost =
2
p
C o D CcQ
d

TC

1 -

Q
2
p
Qopt

2C o D

d
C c 1 -
p

Price-Break Model Formula

Based on the same assumptions as the EOQ model,
the price-break model has a similar Qopt formula:

Q OPT

2DS
2(Annual Demand)(Order or Setup Cost)
=
=
iC
Annual Holding Cost

i = percentage of unit cost attributed to carrying inventory

C = cost per unit
Since C changes for each price-break, the formula
above will have to be used with each price-break cost
value

Price-Break Example Problem Data (Part 1)

A
Acompany
company has
has aa chance
chance to
to reduce
reduce their
theirinventory
inventory
ordering
ordering costs
costs by
by placing
placing larger
largerquantity
quantity orders
orders using
using
the
the price-break
price-breakorder
orderquantity
quantity schedule
schedule below.
below. What
What
should
should their
theiroptimal
optimal order
orderquantity
quantity be
be ifif this
this company
company
purchases
purchases this
this single
single inventory
inventoryitem
item with
with an
an e-mail
e-mail
ordering
ordering cost
cost of
of $4,
$4, aa carrying
carrying cost
cost rate
rate of
of 2%
2%of
of the
the
inventory
inventory cost
cost of
of the
the item,
item, and
and an
an annual
annual demand
demand of
of
10,000
10,000 units?
units?
Order Quantity(units) Price/unit($)
0 to 2,499
$1.20
2,500 to 3,999 1.00
4,000 or more .98

Price-Break Example Solution (Part 2)

First, plug data into formula for each price-break value of C
Annual Demand (D)= 10,000 units
Cost to place an order (S)= $4

Carrying cost % of total cost (i)= 2%

Cost per unit (C) = $1.20, $1.00, $0.98

Next, determine if the computed Qopt values are feasible or not

Interval from 0 to 2499, the
Qopt value is feasible

Q OPT =

2DS
=
iC

2(10,000)(4)
= 1,826 units
0.02(1.20)

Q OPT =

2DS
=
iC

2(10,000)(4)
= 2,000 units
0.02(1.00)

Interval from 4000 & more, the

Q OPT =
Qopt value is not feasible

2DS
=
iC

2(10,000)(4)
= 2,020 units
0.02(0.98)

Interval from 2500-3999, the

Qopt value is not feasible

Price-Break Example Solution (Part 3)

Since
Sincethe
thefeasible
feasiblesolution
solution occurred
occurredin
inthe
thefirst
firstpricepricebreak,
values occur
break,itit means
means that
thatall
all the
theother
othertrue
trueQ
Qopt
opt values occur
at
at the
thebeginnings
beginningsof
ofeach
eachprice-break
price-breakinterval.
interval. Why?
Why?
Because
Becausethe
thetotal
total annual
annualcost
costfunction
functionis
is
aau
ushaped
shapedfunction
function

Total
annual
costs

So
So the
thecandidates
candidates
for
forthe
thepricepricebreaks
breaks are
are1826,
1826,
2500,
2500,and
and4000
4000
units
units
0

1826

2500

4000

Order Quantity

Price-Break Example Solution (Part 4)

Next,
values into the total cost
Next,we
weplug
plugthe
thetrue
trueQ
Qopt
opt values into the total cost
annual
annualcost
costfunction
functionto
todetermine
determinethe
thetotal
total cost
costunder
under
each
eachprice-break
price-break

D
Q
D
Q iC
TC
=
DC
+
S
+
TC = DC +
S+
iC
Q
2
Q
2
TC(0-2499)=(10000*1.20)+(10000/1826)*4+(1826/2)(0.02*1.20)
TC(0-2499)=(10000*1.20)+(10000/1826)*4+(1826/2)(0.02*1.20)
==$12,043.82
$12,043.82
TC(2500-3999)=
TC(2500-3999)=$10,041
$10,041
TC(4000&more)=
TC(4000&more)=$9,949.20
$9,949.20

Finally,
, which is this
Finally,we
weselect
select the
theleast
least costly
costlyQ
Qopt
opt, which is this
problem
problem occurs
occursin
in the
the4000
4000 &&more
more interval.
interval. In
In summary,
summary,
our
ouroptimal
optimal order
orderquantity
quantityis
is4000
4000units
units

When to Order
Reorder Point : level of inventory at which to place
a new order
R=dL
where
d = demand rate per period
L = lead time

Safety Stocks
Safety Stock
Buffer added to on hand inventory during lead time

Stock - Out
An inventory shortage

Service Level
Probability that the inventory available during lead
time will meet demand

Inventory level

Reorder Point With A Safety Stock

Reorder
point, R

Safety stock

LT

Time

LT

Reorder Point With Variable Demand

R dL z d L
where
d = average daily demand
L = lead time
d standard deviation of daily demand
z = number of standard deviations for desired service level
z d L safety stock
Variance (daily variances) x (number of days of lead time)
= 2
dL
2L
d
d L

S tan dard deviation

Reorder Point For A Service Level

Probability of
meeting demand during
lead time = service level
Probability of
a stockout

Safety stock

dL
Demand

Reorder Point For Variable Demand

Carpet store wants a reorder point with a 95%
service level ( 5% stock -out probability)

d = 30 m per day
L = 10 days
d 5 m per day
R d L z d

Determining Z Value For Service Level

..
.

Z
1.6

0.00

..
.
0.4452

0.01

..
.
0.4463

...

Service level =
area to left of Z value or 95%
0.5000

0.4505

0.05

..
.
0.4505

Probability of
a stockout = 5%

Z = 1.65

Reorder Point For Variable Demand

Carpet store wants a reorder point with a 95% service level
(5% stock -out probability)
d = 30 m per day
L = 10 days

d 5 m per day
R d L z d L
(30)(10) (1.65)(5)( 10 ) 300 26.1 326.1m
Safetystock z d L (1.65)(5)( 10 ) 26.1m

Example Safety Stock Calculation

Key parameters
SL = 99 percent availability (Z=2.33)
D = 5 units/day, DD = 2.54
LT = 10 days, LT = 2
OQ = 300 units

Combining Demand and Lead Time Uncertainty

+
Lead Time
LT = 10
LT = 2.00

Demand
D=5
DD = 2.54

Average LTD (Lead time demand) = LT * D = 50

Combined Uncertainty
C =

LT * DD2 + D2 * LT2

C =

10 * 2.542 + 52 * 2.002

C = 12.83

Example Safety Stock Calculation

Key parameters
SL = 99 percent availability (Z=2.33)
D = 5 units/day, DD = 2.54
LT = 10 days, LT = 2
OQ = 300 units
Safety Stock= ZC =2.33 x 12.83= 29.89 or 30 Units
Reorder Point= D x LT + Safety Stock= 5x10+30=80 Units

Order Quantity For A Periodic

Inventory System
Q d t b L z d

tb L I

where
d = average demand rate
t b fixed time between orders
L = lead time

d standard deviation of demand

z d

t b L safety stock
I = inventory in stock

Fixed-Period Model Example

d = 6 bottles per day
d 1.2 bottles
t b 60 days
L = 5 days
I = 8 bottles
z = 1.65 (for 95% service level)

Fixed-Period Model Example

d = 6 bottles per day
d 1.2 bottles
tb 60 days
L = 5 days
I = 8 bottles
z = 1.65 (for 95% service level)
Q d tb L z d tb L I
(6)(60 5) (165
. )(12
. ) 60 5 8 397.96 bottles

The (s,S) Policy

(s, S) Policy: Whenever the inventory position drops
below a certain level, s, we order to raise the inventory
position to level S.
s is the reorder point, and S is the order-up-to level. The
difference between the two levels is driven by the fixed
costs associated with ordering, transportation, or
manufacturing
The reorder point (s) is a function of:

Lead Time
Average demand
Demand variability
Service level

A View of (s, S) Policy

Inventory Level

Inventory Position

Lead
Time

s
0
Time

Optional Replenishment System

Maximum Inventory Level, S

Q=S-I
Actual Inventory Level, I

S
I

If Q > q, order Q, otherwise do not order.

(Where q = minimum acceptable order quantity)

Selective inventory management

Selective inventory management is resorted so
that we can put our limited control efforts more
judiciously to the more significant group of items.
An appropriate inventory policy is chosen
depending upon value, criticality and usage
frequency.

Summary of techniques of selective inventory

control

S. NO. TITLE

BASIS

MAIN USES

A-B-C

Value of consumption

To control raw material, components,

and work-in-progress inventory

H-M-L

Unit price of material

Mainly to control purchases

(High, Medium, Low)

F-S-N-D

Consumption pattern of the material

To control obsolescence

(Fast Moving, Slow , Non Moving, Dead)

4
5

S-D-E
G-O-L-F

Problems faced in procurement

(Scarce, Difficult, Easy to obtain)

Lead time analysis and purchasing

strategies

Source of material

Procurement strategies

(Govt., Ordinary, Local, Foreign Sources)

V-E-D

Criticality of the material

(Vital, Essential, Desirable)

S-OS

Nature of suppliers
(Seasonal, Off-seasonal)

X-Y-Z

Value of items in storage

V-I-R

Priority ( Vital, Important, Routine)

To determine stocking levels of spare

parts
Procurement / holding strategies for
seasonal items like agricultural products
To review the inventories and their uses
at scheduled intervals

ABC Classification System

Items kept in inventory are not of equal importance in terms of:

Money invested

profit potential

sales or usage volume

stock-out penalties

60

% of
Rs Value 30
0

% of
Use

30

A
B

60

So, identify inventory items based on percentage of total

Rupee value, where A items are roughly top 15 %, B items
as next 35 %, and the lower 65% are the C items.

ABC Classification System

Demand volume & value of items vary
Classify inventory into 3 categories
Class
A
B
C

% of Units

% of Dollars
5 - 15 70 - 80
30 15
50 - 60 5 - 10

ABC Classification Example

Unit
Cost
60
350
30
80
30
20
10
320
510
20

Anual
Usage
90
40
130
60
100
180
170
50
60
120

Total
% of Total % of Total
%
Part
Value
%Value %QuantityCumulative %
9
30,600
35.8
6.0
6.0
8
16,000
18.7
5.0
11.0
2
14,000
16.4
4.0
15.0
1
5,400
6.3
9.0
24.0
4
4,800
5.6
6.0
30.0
3
3,900
4.6
13.0
43.0
6
3,600
4.2
18.0
61.0
5
3,000
3.5
10.0
71.0
10
2,400
2.8
12.0
83.0
7
1,700
2.0
17.0
100.0
Rs 85,400
Class
A
B
C

Items
9,8,2
1, 4, 3
6, 5, 10, 7

% Value % Units
71
15
16.5
28
12.5
57

Risk Pooling
Consider these two systems:
Warehouse One

Market One

Warehouse Two

Market Two

Supplier

Market One
Supplier

Warehouse
Market Two

Risk Pooling Effect

Centralized

2++
2

Relative Reduction = 1 - 1 / N
(under certain conditions)

++
2

N = 25

80%

Risk Pooling:Important Observations

Centralizing inventory control reduces both
safety stock and average inventory level for the
same service level.
This works best for
High coefficient of variation, which reduces required
safety stock.
Negatively correlated demand.

Risk Pooling Example

Compare the two systems:

two products
maintain 97% service level
$60 order cost
$0.27 weekly holding cost
$1.05 transportation cost per unit in decentralized
system, $1.10 in centralized system
1 week lead time

Risk Pooling Example

Risk Pooling Example

Risk Pooling Example

Warehouse

Product

Average
Demand
during Lead
Time

Standard
Deviation of
Demand

Safety
Stock

Reorder
Point (s)

Order
Quantity
(Q)

Market 1

39.3

13.2

25.08

65

132

Market 2

38.6

12.0

22.8

62

131

Market 1

1.125

1.36

2.58

24

Market 2

1.25

1.58

24

Total

77.9

20.71

39.35

118

186

Total

2.375

1.9

3.61

33

Risk Pooling Example

Risk Pooling:Important Observations

Centralizing inventory control reduces both
safety stock and average inventory level for the
same service level.
This works best for
High coefficient of variation, which reduces required
safety stock.
Negatively correlated demand.

To Centralize or not to Centralize

What is the effect on:

Safety Stock?
Service Level?
Overhead Costs?
Customer Lead Time?
Transportation Costs?