Supply Chain Management

Managing Uncertainty in the

Supply Chain: Safety Inventory
 Outline
The role of safety inventory in a supply chain
Determining the appropriate level of safety inventory
Impact of supply uncertainty on safety inventory
Impact of aggregation on safety inventory
Impact of replenishment policies on safety inventory
Managing safety inventory in a multi-echelon supply
chain
Estimating and managing safety inventory in practice

Dr. Srikanta Routroy 2

 The Role of Safety Inventory
in a Supply Chain
Forecasts are rarely completely accurate.
If average demand is 1000 units per week, then half
the time actual demand will be greater than 1000, and
half the time actual demand will be less than 1000;
what happens when actual demand is greater than
1000?
If you kept only enough inventory in stock to satisfy
average demand, half the time you would run out.
Safety inventory: Inventory carried for the purpose of
satisfying demand that exceeds the amount forecasted
in a given period.
Dr. Srikanta Routroy 3
 Role of Safety Inventory
Average inventory is therefore cycle inventory plus
safety inventory.
There is a fundamental tradeoff:
Raising the level of safety inventory provides higher levels
of product availability and customer service.
Raising the level of safety inventory also raises the level of
average inventory and therefore increases holding costs.
Very important in high-tech or other industries where obsolescence
is a significant risk (where the value of inventory, such as PCs, can
drop in value)
Compaq and Dell in PCs
Dr. Srikanta Routroy 4
 Two Questions to Answer in
Planning Safety Inventory

What is the appropriate level of safety

inventory to carry?

What actions can be taken to improve

product availability while reducing
safety inventory?

Dr. Srikanta Routroy 5

 Determining the Appropriate
Level of Demand Uncertainty
Appropriate level of safety inventory determined by:
supply or demand uncertainty
desired level of product availability

Higher levels of uncertainty require higher levels of

safety inventory given a particular desired level of
product availability.
Higher levels of desired product availability require
higher levels of safety inventory given a particular
level of uncertainty
Dr. Srikanta Routroy 6
 Measuring Demand Uncertainty
Demand has a systematic component and a random
component
The estimate of the random component is the measure of
demand uncertainty.
Random component is usually estimated by the standard
deviation of demand.
Notation:
D = Average demand per period
sD = standard deviation of demand per period
L = lead time = time between when an order is placed and
when it is received
Uncertainty of demand during lead time is what is important.
Dr. Srikanta Routroy 7
 Measuring Demand Uncertainty
P = demand during k periods = kD

W = std dev of demand during k periods = sRSqrt(k)

(Demand between two periods is independent)

Coefficient of variation = cv = s/m =(std dev) /mean=

size of uncertainty relative to demand

Dr. Srikanta Routroy 8

 Measuring Product Availability
Product availability: a firms ability to fill a customers
order out of available inventory.
Stockout: a customer order arrives when product is not
available.
Product fill rate (fr): fraction of demand that is satisfied
from product in inventory.
Order fill rate: fraction of orders that are filled from
available inventory.
Cycle service level: fraction of replenishment cycles (RC)
that end with all customer demand met. RC is the interval
between two successive replenishment deliveries.
Dr. Srikanta Routroy 9
 Replenishment Policies
Replenishment policy: Decisions regarding when to
reorder and how much to reorder.
Continuous review: Inventory is continuously
monitored and an order of size Q is placed when the
inventory level reaches the reorder point ROP.
Periodic review: Inventory is checked at regular
(periodic) intervals and an order is placed to raise the
inventory to a specified threshold (the order-up-to
level).
Dr. Srikanta Routroy 10
 Continuous Review Policy: Safety
Inventory and Cycle Service Level
L: Lead time for replenishment
D L
DL
D: Average demand per unit

s sD
time
sD:Standard deviation of L
L
demand per period
ss F S (CSL) s L
1
DL: Mean demand during lead
time
sL: Standard deviation of ROP D L ss
demand during lead time
CSL: Cycle service level CSL F ( ROP, D L ,s L )
ss: Safety inventory
ROP: Reorder point Average Inventory = Q/2 + ss
Dr. Srikanta Routroy 11
 Estimating Safety Inventory
(Continuous Review Policy)
D = 2,500/week; sD = 500
L = 2 weeks; Q = 10,000; ROP = 6,000

DL = DL = (2500)(2) = 5000
ss = ROP - RL = 6000 - 5000 = 1000
Cycle inventory = Q/2 = 10000/2 = 5000
Average Inventory = cycle inventory + ss = 5000 + 1000
= 6000
Average Flow Time = Avg inventory / throughput =
6000/2500 = 2.4 weeks
Dr. Srikanta Routroy 12
 Estimating Cycle Service Level
(Continuous Review Policy)
D = 2,500/week; sD = 500
L = 2 weeks; Q = 10,000; ROP = 6,000

s s L R
L (500) 2 707
Cycle Service Level, CSL = Prob. (z less than equal to 1.414)=
0.92.
(This value is determined from a Normal probability
distribution table).

Dr. Srikanta Routroy 13

 Fill Rate
Proportion of customer demand
ESC
satisfied from stock fr 1
Stockout occurs when the Q
demand during lead time exceeds
ss
the reorder point ESC ss{1 F S }
ESC is the expected shortage per s L
cycle (average demand in excess
ss
of reorder point in each
s L f S
s L
replenishment cycle)
ss is the safety inventory
Q is the order quantity
ESC = -ss{1-NORMDIST(ss/sL, 0, 1, 1)} + sL NORMDIST(ss/sL, 0, 1, 0)
Dr. Srikanta Routroy 14
 Fill Rate
ESC
fr 1
Q
Expected Shortage per Cycle

ESC x ROP f x dx
x ROP
x D
x RL ss
L ss f x dx

x DL 2

x D ss
1 1 2s L 2
ESC * e dx
x RL ss
L
2 sL

Put
x DL
z
s dx s l dz

L

Dr. Srikanta Routroy 15

 Fill Rate
z2

zs ss
1 1
ESC * e 2
dz
ss
L
2 sL
z
sL

z2 z2
1 1
ss e 2
dz s z e 2
dz
2 2
L
ss ss
z z
sL sL

ss
1
z2
ss 1 FS s
2
s z e dz
2
L
L z
ss
sL

Put
z2
w dw zdz
2
Dr. Srikanta Routroy 16
 Fill Rate

ss
1 w
ESC ss 1 FS s L e dw
s L w
ss 2 2
2s L 2

ss ss
ESC ss 1 FS s L f s
sL sL

ESC Can be evaluated using EXCEL as given below:

ESC = -ss{1-NORMDIST(ss/sL, 0, 1, 1)} + sL

NORMDIST(ss/sL, 0, 1, 0)
Dr. Srikanta Routroy 17
 Evaluating Fill Rate
ss = 1,000, Q = 10,000, sL = 707, Fill Rate (fr) = ?

ESC = 10001 F 1.414 707
e w
dw
2 1

1,0001 0.9207
707
2 * e
79.3 103.776 24.5 25

fr = (Q - ESC)/Q = (10,000 - 25.13)/10,000 = 0.9975

Dr. Srikanta Routroy 18

 Factors Affecting Fill Rate
Safety inventory: Fill rate increases if safety
inventory is increased. This also increases the
cycle service level.
Lot size: Fill rate increases on increasing the lot
size even though cycle service level does not
change.

Dr. Srikanta Routroy 19

 Evaluating
Safety Inventory Given CSL
D = 2,500/week; sD = 500
L = 2 weeks; Q = 10,000; CSL = 0.90
DL = 5000, sL = 707
1.28

ss = FS-1(CSL)sL = FS-1(0.90)(707) = 906

ROP = DL + ss = 5000 + 906 = 5906

Dr. Srikanta Routroy 20

 Evaluating Safety Inventory
Given Desired Fill Rate
D = 2500, sD = 500, Q = 10000
If desired fill rate is fr = 0.975, how much safety
inventory should be held?
ESC = (1 - fr)Q = 250
ss ss
Solve ESC 250 ss 1 F S L f S
L L
ss
1
250 ESC ss 1 FS s L e w dw
s L w
ss 2 2
2s L 2

Dr. Srikanta Routroy

SS=67 21
Evaluating Safety Inventory Given
Fill Rate (try different values of ss)
Fill Rate Safety Inventory
97.5% 67
98.0% 183
98.5% 321
99.0% 499
99.5% 767

The required safety inventory grows rapidly with an increase in

the desired product availability.

Dr. Srikanta Routroy 22

Impact of Required Product Availability
and Uncertainty on Safety Inventory
Desired product availability (cycle service level or fill
rate) increases, required safety inventory increases
Demand uncertainty (sL) increases, required safety
inventory increases
Managerial levers to reduce safety inventory without
reducing product availability
reduce supplier lead time, L (better relationships with
suppliers)
reduce uncertainty in demand, sL (better forecasts, better
information collection and use)

Dr. Srikanta Routroy 23

Impact of Supply Uncertainty
D: Average demand per period
sD: Standard deviation of demand per period
L: Average lead time
sL: Standard deviation of lead time

DL DL

s s
2 2 2
L
L D D s L
Dr. Srikanta Routroy 24
 Impact of Supply Uncertainty
Demand information at Dell for PC is given as
D = 2,500/day; sD = 500. Target CSL = 0.90 and
Hard Drive supplier information as follows L = 7 days; sL = 7 days
Evaluate the safety inventory of hard drive, Dell has to carry.
DL = DL = (2500)(7) = 17500

s L
L s D sL
2
D
2 2

(7) 500 (2500) (7) 17500

2 2 2

ss = F-1s(CSL)sL = NORMSINV(0.90) x 17550= 22,491

Dr. Srikanta Routroy 25
Required safety inventory as a function of lead time
sL Std. dev. during SS(units) SS(days)
lead time
7 17,550 22,491 9
6 15,058 19,298 7.72
5 12,570 16,109 6.44
4 10,087 12,927 5.17
3 7,616 9,760 3.90
2 5,172 6,628 2.65
1 2,828 3,625 1.45
0 1,323 1,695 0.68

A reduction in supply chain uncertainty can help dramatically reduce

safety inventory required without hurting product availability.
 Techniques for reduction of
Safety Inventory
Models of aggregation

Information centralization

Specialization

Product substitution

Component commonality

Postponement

Dr. Srikanta Routroy 27

 Impact of Aggregation
n

D D
C
i
i 1
n

s s
C 2
D i
i 1

s Ls D
C C
L

ss F s (CSL) s L
1 C

Dr. Srikanta Routroy 28

 Impact of Aggregation
Car Dealer : 4 dealership locations (disaggregated)
D = 25 cars; sD = 5 cars; L = 2 weeks; desired CSL=0.90
What would the effect be on safety stock if the 4 outlets are
consolidated into 1 large outlet (aggregated)?

At each disaggregated outlet:

For L = 2 weeks, sL = 7.07 cars
ss = Fs-1(CSL) x sL = Fs-1(0.9) x 7.07 = 9.06
Each outlet must carry 9 cars as safety stock inventory, so safety
inventory for the 4 outlets in total is (4)(9) = 36 cars

Dr. Srikanta Routroy 29

 Impact of Aggregation
One outlet (aggregated option):
RC = D1 + D2 + D3 + D4 = 25+25+25+25 = 100 cars/wk
sRC = Sqrt(52 + 52 + 52 + 52) = 10
sLC = sDC Sqrt(L) = (10)Sqrt(2) = (10)(1.414) = 14.14
ss = Fs-1(CSL) x sLC = Fs-1(0.9) x 14.14 =18.12
or about 18 cars
If r does not equal 0 (demand is not completely
independent), the impact of aggregation is not as great.

Dr. Srikanta Routroy 30

Safety Inventory in the Disaggregate and
Aggregate options

r Disaggregate Safety Inventory Aggregate Safety

Inventory
0 36.24 18.12
0.2 36.24 22.92
0.4 36.24 26.88
0.6 36.24 30.32
0.8 36.24 33.41
1.0 36.24 36.24

Dr. Srikanta Routroy 31

 Impact of Aggregation
If number of independent stocking locations decreases by n, the
expected level of safety inventory will be reduced by square root
of n (square root law).

Many e-commerce retailers attempt to take advantage of

aggregation (Amazon) compared to bricks and mortar retailers
(Borders).
Aggregation has two major disadvantages:
Increase in response time to customer order
Increase in transportation cost to customer
Some e-commerce firms (such as Amazon) have reduced aggregation to
mitigate these disadvantages

Dr. Srikanta Routroy 32

 Information Centralization
Virtual aggregation.

Information system that allows access to current inventory

records in all warehouses from each warehouse.

Most orders are filled from closest warehouse.

In case of a stock-out, another warehouse can fill the order.

Better responsiveness, lower transportation cost, higher product

availability, but reduced safety inventory.

Examples: McMaster-Carr, Gap, Wal-Mart.

Dr. Srikanta Routroy 33

 Specialization
Stock all items in each location or stock different
items at different locations?
Different products may have different demands in different
locations
There can be benefits from aggregation

Benefits of aggregation can be affected by:

coefficient of variation of demand (higher cv yields greater
reduction in safety inventory from centralization)
value of item (high value items provide more benefits from
centralization).
Dr. Srikanta Routroy 34
 Impact of coefficient of variation on
value of aggregation
Assume that WW Grainger, a supplier of MRO products has
1600 stores throughout United States.
Consider two products-large electric motors and industrial
cleaners.
The cost and demand information (mean and standard
deviation) in each store for two products are given as: electric
motor: $500/unit, 20, 40 and cleaner:$30, 1000, 100.
Demand information in each store is independent and supply
lead time is 4 weeks in both cases. Holding cost: 20% and
desired CSL:95%.
For each of the two products, evaluate the reduction in safety
inventories that will result if they moved from retail stores and
carried only in a centralized DC.
Dr. Srikanta Routroy 35
Value of Aggregation at Grainger

Inventory is stocked in each store case

Motors Cleaner
Disagg. cv 2 0.1
SS/store 3.29*40=132 3.29*100=329
TSI 211,200 526,400
Value TSI 211,200*500= 526,400*30=
$105,600,000 $15,792,000
Dr. Srikanta Routroy 36
 Value of Aggregation at Grainger

Inventory is aggregated at the DC

Motors Cleaner
Mean weekly demand 20*1,600= 32,000 1,600,000
Std. Agg. demand 40*40=1,600 40*100=4,000
CV 0.05 0.0025
Agg. SI 3.29*1,600=5,264 13,159
Value of SI $2,632,00 $394,770

Dr. Srikanta Routroy 37

 Value of Aggregation at Grainger

Savings
Motors Cleaner
Total inventory saving $102,968,000 $15,397,230
Total holding cost on agg. $25,742,000 $3,849,308
Holding cost saving/unit $15.47 $0.0046
Saving as a percentage of 3.09% 0.15%
product cost

Dr. Srikanta Routroy 38

 Product Substitution
Substitution: use of one product to satisfy the demand
for another product.
Manufacturer-driven one-way substitution
It is influenced by the cost differential between high and
low value item.
The value substitution increases as demand uncertainty
increases.
It is also influenced by the correlation between products.

Dr. Srikanta Routroy 39

 Product Substitution
Customer-driven two-way substitution.

Recognition of customer-driven two-way substitution

and joint management across substitutable products
allows a SC to reduce the required safety inventory
While ensuring a high level of product availability.

Dr. Srikanta Routroy 40

 Component Commonality
Using common components in a variety of
different products.

Can be an effective approach to exploit

aggregation and reduce component inventories.

Dr. Srikanta Routroy 41

 Problem on Component
Commonality
Evaluate the safety stock requirement for the following example.
Suppose Dell is to produce 27 different PCs with three distinct
components: processor, memory and hard drive.

Monthly demands for each computer is independent and normally

distributed variable with mean 5000, and standard deviation 3000.
Suppose Dell is targeting 95% CSL and replenishment lead time
for each component is one month.
Case 1: If Dell designs specific components for each PC resulting:
3*27=81 components.
Case 2: Common components: 3 processors, 3 memory and 3 hard
drives to create 27 kinds of computers.
Dr. Srikanta Routroy 42
 Problem on Component
Commonality
Case 1:
Safety inventory per component= F-1(0.95)* 3000=4,935 units
implies that TSI= 4,935*81= 399,699 units

Case 2:
Each component ends up in nine different finished products.

Aggregating across the nine products, the monthly demand for each
component be normally distributed with the following:
Mean demand of common component (CC) across nine products=4,500
Standard deviation of CC across nine products= (9)0.5*3000=9000
Safety inventory required for each CC= F-1(0.95)* 9000=14,804
Total safety inventory=14,804*9= 133,236

Dr. Srikanta Routroy 43

 Problem on Component
Commonality
Case 1:
Safety inventory per component= F-1(0.95)* 3000=4,935 units
implies that TSI= 4,935*81= 399,699 units

Case 2:
Each component ends up in nine different finished products.

Aggregating across the nine products, the monthly demand for each
component be normally distributed with the following:
Mean demand of common component (CC) across nine products=4,500
Standard deviation of CC across nine products= (9)0.5*3000=9000
Safety inventory required for each CC= F-1(0.95)* 9000=14,804
Total safety inventory=14,804*9= 133,236

Dr. Srikanta Routroy 44

 Value of Component Commonality
Component commonality decreases
450000
Safety Inventory. Marginal benefit, however
400000
350000
decreases with increasing commonality
300000
Safety
250000
Inventory SS
200000
150000
100000
50000
0
1 2 3 4 5 6 7 8 9

Number of finished Products

Per component
Dr. Srikanta Routroy 45
 Postponement
The ability of a supply chain to delay product
differentiation or customization until closer to the
time the product is sold.
Goal is to have common components in the supply
chain for most of the push phase and move
product differentiation as close to the pull phase
as possible.
Examples: Dell, Benetton

Dr. Srikanta Routroy 46

 Impact of Replenishment
Policies on Safety Inventory
Continuous review policies
Periodic review policies 1
SS FS (CSL) * s L

SS FS (CSL) * s T L
Average Lot Size=Q
OUL DT L SS =DT=D*T

OUL= Order up to level Review Lead Time

Period
Dr. Srikanta Routroy 47
 Example Problem

Weekly demand for Logo at a Wal-Mart store is

normally distributed with a mean of 2,500 boxes and a
standard deviation of 500. The replenishment lead
time is two weeks and the store manager has decided
to review inventory every four weeks. Assuming a
periodic review replenishment policy, evaluate the
safety inventory that the store should carry to provide
a CSL of 90 percent. Evaluate the OUL for such a
policy.

Dr. Srikanta Routroy 48

 Solution to Example Problem 11.10
Mean demand during (T+L) periods= 15,000 units
Standard deviation of demand during (T+L) periods

s T L T L * s D 6 * 500 1,225
The required safety stock for a CSL of 0.90=

F-1(CSL)*s T L 1.28*1225=1,570

OUL= 15,000+1,570=16,570

Dr. Srikanta Routroy 49

 Estimating and Managing
Safety Inventory in Practice
Account for the fact that supply chain demand is
lumpy.
Adjust inventory policies if demand is seasonal.
Use simulation to test inventory policies.
Start with a pilot.
Monitor service levels.
Focus on reducing safety inventories

Dr. Srikanta Routroy 50