Inventory Policy Decisions: "Every Management Mistake Ends Up in Inventory." Michael C. Bergerac

Inventory Policy
Decisions
“Every management mistake ends up in inventory.”
Michael C. Bergerac
Former Chief Executive
Revlon, Inc.
Chapter 9
CR (2004) Prentice Hall, Inc.
9-1
Inventory Decisions in
Strategy
Inventory Strategy
• Forecasting Transport Strategy
• Inventory decisions • Transport fundamentals
CONTROLLING
• Purchasing and supply
ORGANIZING
• Transport decisions
Customer
PLANNING
scheduling decisions
• Storage fundamentals service goals
• Storage decisions • The product
• Logistics service
• Ord. proc. & info. sys.
Location Strategy
• Location decisions
• The network planning process

9-2
What are Inventories?
•Finished product held for sale
•Goods in warehouses
•Work in process
•Goods in transit
•Staff hired to meet service needs
•Any owned or financially controlled
raw material, work in process, and/or
finished good or service held in
anticipation of a sale but not yet sold

9-3
Where are Inventories?
Material Inbound Production Outbound Finished goods Customers
sources transportation transportation warehousing
Receiving
Production
materials
Inventories
in-process
Shipping
Finished goods
Inventory
locations
CR (2004) Prentice Hall, Inc. 9-4

Reasons for Inventories
•Improve customer service
-Provides immediacy in product availability
•Encourage production, purchase, and transportation
economies
-Allows for long production runs
-Takes advantage of price-quantity discounts
-Allows for transport economies from larger shipment sizes
•Act as a hedge against price changes
-Allows purchasing to take place under most favorable price
terms
•Protect against uncertainties in demand and lead times
-Provides a measure of safety to keep operations
running when demand levels and lead times cannot be known
for sure
•Act as a hedge against contingencies
-Buffers against such events as strikes, fires, and
disruptions in supply
9-5
Reasons Against Inventories
•They consume capital resources that might be put to

better use elsewhere in the firm
•They too often mask quality problems that would more
immediately be solved without their presence
•They divert management’s attention away from careful
planning and control of the supply and distribution
channels by promoting an insular attitude about
channel management

9-6
Types of Inventories
•Pipeline
-Inventories in transit
•Speculative
-Goods purchased in anticipation of price increases
•Regular/Cyclical/Seasonal
-Inventories held to meet normal operating needs
•Safety
-Extra stocks held in anticipation of demand and
lead time uncertainties
•Obsolete/Dead Stock
-Inventories that are of little or no value due to being
out of date, spoiled, damaged, etc. 9-7
Nature of Demand
•Perpetual demand
-Continues well into the foreseeable future
•Seasonal demand
-Varies with regular peaks and valleys throughout
the year Accurately forecasting
•Lumpy demand demand is singly the
-Highly variable (3  Mean) most important factor
•Regular demand in good inventory
management
-Not highly variable (3 < Mean)
•Terminating demand
-Demand goes to 0 in foreseeable future
•Derived demand
-Demand is determined from the demand of another
item of which it is a part 9-8
Inventory Management
Philosophies
•Pull
-Draws inventory into the stocking location
-Each stocking location is considered independent
-Maximizes local control of inventories
•Push
-Allocates production to stocking locations based on
overall demand
-Encourages economies of scale in production
•Just-in-time
-Attempts to synchronize stock flows so as to just
meet demand as it occurs
-Minimizes the need for inventory
9-9
Inventory Management
Philosophies (Cont’d)
•Supply-Driven
-Supply quantities and timing are unknown
-All supply must be accepted and processed
-Inventories are controlled through demand
•Aggregate Control
-Classification of items:
›Groups items according to their sales level
based on the 80-20 principle
›Allows different control policies for 3 or more
broad product groups
9-10
Pull vs. Push Inventory Philosophies
PUSH - Allocate supply to each PULL - Replenish inventory with
warehouse based on the forecast order sizes based on specific needs
for each warehouse of each warehouse
Demand
forecast
Warehouse #1
Q1
A1
A2 Q2 Demand
Plant forecast
Warehouse #2
A3
Q3
A = Allocation quantity to each warehouse

Q = Requested replenishment quantity Demand
by each warehouse Warehouse #3 forecast
Costs Relevant to Inventory
Management
•Carrying costs
-Cost for holding the inventory over time
-The primary cost is the cost of money tied up in
inventory, but also includes obsolescence,
insurance, personal property taxes, and storage
costs
-Typically, costs range from the cost of short term
capital to about 40%/year. The average is about
25%/year of the item value in inventory.

9-12
Relevant Costs (Cont’d)
•Procurement costs
-Cost of preparing the order
-Cost of order transmission
-Cost of production setup if appropriate
-Cost of materials handling or processing at the
receiving dock
-Price of the goods

9-13
Relevant Costs (Cont’d)
•Out-of-stock costs
-Lost sales cost
›Profit immediately foregone
›Future profits foregone through loss of goodwill
-Backorder cost
›Costs of extra order handling
›Additional transportation and handling costs
›Possibly additional setup costs
9-14
Inventory Management Objectives
Good inventory management is a careful balancing act
between stock availability and the cost of holding
inventory.
Customer Service, Inventory Holding costs
i.e., Stock Availability
•Service objectives
-Setting stocking levels so that there is only a
specified probability of running out of stock
•Cost objectives
-Balancing conflicting costs to find the most
economical replenishment quantities and timing
9-15
Inventory’s Conflicting Cost Patterns
Total cost Minimum cost

reorder quantity
Cost
Procurement cost
Stockout cost

Replenishment quantity 9-16
Glossary of Terms
D = average annual demand, units
d = average period demand, units
S = procurement cost per order, $/order
I = carrying costs as a percent of product value, % per year
C = product value, $ per unit
sd = standard deviation of demand (d), units
k = out - of - stock cost, $ per unit
p = purchase price
s ' = standard deviation of compound demand distribution
E( z ) = partial expectation or unit normal loss integral
P = probability of being in - stock during lead time (Q - system)
or during lead time plus order cycle time (P - system
Q = order quantity
ROP = reorder point quantity, units
T = order interval, e.g., days
MAX = target inventory level, units
z = normal deviate or number of standard deviations from mean
on compound demand distribution
r = safety stock, or z x s' , units
TC = total relevant cost, $
SL = service level as a percent of total annual demand
LT , sLT = average and standard deviation of lead time
CPn = probability of n units being sold 9-17
Single Order Purchasing
Make a one-time purchase of an item. How much to order?
Procedure: Balance incremental profit against incremental loss.
Profit = Price per unit  Cost per unit

Loss = Cost per unit  Salvage value per unit
If CPn is probability of n units being sold, then

Daily stocking of
CPn x Loss = (1  CPn) x Profit newspapers in
or vending
CPn = Profit/(Profit + Loss) machines is a
good example
Now, increase order quantity until CPn just matches cumulative
probability of selling additional units.
9-18
Single Order Purchasing (Cont’d)
Example A clothing item is purchased for a seasonal sale. It
costs $35, but it has a sale price of $50. After the season is
over, it is marked down by 50% to clear the merchandise.
The estimated quantities to be sold are:
Probability of
Number of selling exactly n Cumulative
items, n items probability
10 0.15 0.15
15 0.20 0.35
20 0.30 0.65
25 0.20 0.85
30 0.10 0.95
35 0.05 1.00
1.00
Single Order Purchasing (Cont’d)
Solution
Profit = $50 35 = $15

Loss = $35  (0.5)(50) = $10
CPn = 15/(15 + 10) = 0.60
CPn is between 15 and 20 items, round up and order 20

items.

9-20
Simple Two-Bin Pull Method
Given:
d = 50 units/week
Note: No uncertainty
I = 10%/year
in demand or lead
S = $10/order time—manage
C = $5/unit regular (cycle) stock
LT = 3 weeks only
Develop a simple control system by finding the

replenishment quantity (Q) and the reorder point
(ROP).
The relevant total cost is:
TC = ordering cost  carrying costs
= DS  ICQ
Q 2 9-21
Reorder Point Method Under Certainty
for a Single Item
Quantity on-hand
plus on-order
Q
Reorder
point, R
0 Lead Lead Time

time time
Order Order Order Order
Placed Received Placed Received

Two-Bin Method (Cont’d)
Using differential calculus, the optimal value for Q will

be:
Q* = 2DS/IC = 2(50x52)(10)/(0.10x5) = 322 units
The reorder point is: Famous EOQ

formula
ROP = d(LT) = 3(50) = 150 units
Rule When the inventory level drops to 150 units

(ROP) then reorder 322 units (Q*).
Reorder Point Control with
Demand Uncertainty
Given:
d = 50 units/week C = $5/unit
sd = 10 units/week LT = 3 weeks
I = 10%/year P = 99% during lead time
S = $10/order Good method for products:
Find Q* and ROP 1. Of high value
2. That are purchased from
From the EOQ formula one vendor or plant
3. Having few economies of
scale in production,
Q* = 2(50x52)(10) = 322 units purchasing, or
0.10(5)
transportation
Reorder Point Control for a Single Item
Quantity on hand
Q
Place
order
Q
DDLT
ROP
Receive
order P
0
Stockout
LT LT
Time
Reorder Point Control for a Single Item
Quantity on hand
Quantity for +on order
control backorders
Actual
Inventory level
on hand
ROP
Safety stock
0
LT Time LT
Reorder Point Control (Cont’d)
Finding the reorder point requires an understanding of
the demand-during-lead-time distribution
DDLT
P
Week 1 Week 2 Week 3
+ + =
sd=10 sd=10 sd=10 S’=17.3 z
d =100 d =100 d =100 X = 300 ROP
Weekly demand is normally distributed X = d  LT = 100(3) = 300
with a mean of d = 100 and a standard s ' = sd LT = 10 3 = 17.3
deviation of sd = 10
Lead time is 3 weeks 9-27
Now,
X = d(LT ) = 50(3) = 150 units
s' = sd LT = 10 3 = 17.32 units
Hence,
ROP = X  zs' = X  r
= 150  2.33(17.32) = 190 units
where 2.33 is the normal deviate at a probability of
0.01 taken from a normal distribution table.

9-28
Total relevant cost
The total relevant cost equation is now extended to
include the costs of safety stock as well as out-of-stock.
The out-of-stock cost (k) is $2/unit. The price term is
dropped. Hence,
TC = DS  IC Q  ICr  k D s'E(z)
Q 2 Q
= 2,600(10)  (0.1)(5) 322  (0.1)(5)(40)
322 2
 2 2,600(17.32)(0.0034)
322
= $182.20
where E(z) = 0.0034 from a unit normal loss table at a
z value of 2.33 9-29
With known stockout costs k
Setting Q involves balancing both costs and service
level at optimum. Since P and Q are interrelated,
an iterative approach is required.
1 Solve initially for Q
Q = 2DS
IC
2 Using Q, find
P =1  QIC If backordering is allowed
Dk
or
P =1  QIC
Dk QIC If sales are lost
9-30
3 Using P, find revised Q
2D[S  ks' E ]
Q= d (z)
IC
4 Repeat steps 2 and 3 until no further change
5 Compute ROP and other statistics

9-31
Example Given:
Monthly demand forecast, d 11,107 units

Std. error of forecast., sd 3,099 units
Replenishment lead-time, LT 1.5 months
Item value, C $0.11/unit
Cost for processing
vendor order, S $10/order
Carrying cost, I 20%/year
Stockout cost, k $0.01/unit
Backordering is allowed
Find optimal Q and P

9-32
Solution
Estimate Q
Q = 2DS = 2(11,107)(12)(10) =11,008 units
IC 0.20(0.11)
Estimate P
P =1  QIC =1 11,008(0.20)(0.11) = 0.82
Dk 11,107(12)(0.01)
Revise Q
Find App A, z@0.82=0.92 and from App B,
E(0.92)=0.0968
For these data, s'd was previously calculated as
3,795 units

9-33
 
2DS  ks' E 
Q=  d (z) 
IC
= 2(11,107)(12)[(10 0.01(3,795)(0.068)]
0.20(0.11)
= 12,872 units
Revise P
P =1 12,872(0.20)(0.11) = 0.79
11,107(12)(0.01)
Now, z@0.81=0.81 and E(0.81)=0.1181

9-34
Revise Q
Q = 2(11,107)(12)[10 0.01(3,795)(0.1181)] =13,246 units

0.20(0.11)
Continue to revise Q and P until no further

change occurs. P=78% and Q=13,395 units.
Note Although the in-stock probability during the

lead time is 78%, the actual service level is
SL=96%

9-35
Pull Methods (Cont’d)
Noninstantaneous resupply
At times, production or supply continues while demand
is depleting inventories. This requires a slight
modification of the EOQ formula. That is,
Qp* = 2DS p Just add

IC p  d this term
where
p = output or supply rate
d = demand rate
and p > d. ROP remains unchanged.
Reorder point control with demand and lead time uncertainties
The combined effect of these two uncertainties is particularly
hard to estimate accurately. It is the standard deviation of the
demand-during-lead-time distribution that is the problem,
especially if the level of demand and the length of the lead time
are related to each other. Ideally, we would simply observe the
actual demand occurring over each lead time period. If the
demand and lead time are independent of each other and each
are represented by separate distributions, we may estimate the
standard deviation (s′) from
Caution: Can result in very
high safety stock levels when
s' = LT (sd2)  d 2(sLT
2 )
lead-time variability is high
After computing s’, calculation of the ordering policy would

be identical to that presented previously. 9-37
Periodic review control with demand uncertainty
The inventory is reviewed at the time interval (T) to
determine the quantity on hand. The replenishment
quantity (Q) to be ordered is the difference between a
target level called MAX and the quantity on hand. We
need to find MAX and T*. Good method for
products:
1. Of low value
Given: 2. That are purchased
d = 50 units/week C = $5/unit from the same vendor
sd = 10 units/week LT = 3 weeks 3. Having economies of
scale in production,
I = 10%/year P = 0.99 purchasing, and
S = $10/order k = $2/unit transportation
Periodic Control for a Single Item
M
Quantity on hand
Q2
Q1
~
q
Stock
level Order
reviewed received
0
LT LT Time
T T
M = maximum level T = review interval

M - q = replenishment quantity q = quantity on hand
Qi = order quantity 9-39
LT = lead time
Periodic Review (Cont’d)
Estimate Q* from the EOQ formula as if under demand

certainty conditions. Recall that this is Q* = 322 units.
Now,
T* = Q*/d = 322/50 = 6.4 weeks
Construct the demand-during-lead-time-plus-order-
cycle-time distribution.
T is order
review time

9-40
DD(T* + LT)
s′ Z(s′)
X MAX
s = sd T  LT
' *
= d(T* + LT)

9-41
where
X = d(T *  LT ) = 50(6.4  3) = 470
s' = s T *  LT = 10 6.4  3 = 30.66
d
Find MAX
MAX = d(T* + LT) + z(s’)
= 50(6.4 + 3) + 2.33(30.66)
= 470 + 71.44 = 541 units
Rule Review the inventory every 6.4 weeks and place
an order for the difference between the MAX level of 541
units and the quantity on hand + quantity on order –
backorders.
The total relevant cost for this design is:
TC = DS/Q + ICQ/2 + ICr + ks’(D/Q)E(z)
= 2600(10)/322 + (.10)(5)(322/2)
+ (.10)(5)(71) + 2 (30.66)(2600/322)(.0034)
= $198
Note Compare this cost with that of the reorder point

method to see that periodic review control carries a
slight premium in cost due to more safety stock.

9-43
In/ On In/ On In/ On
Date Customer Sales hand Date Customer Sales hand Date Customer Sales Hand
10/26 Bal Fwd 80500 2/2 Copies 50000 35000 3/30 Sup Meats 25000 20000
10/26 100M 180500 2/5 Bel-Gar 5000 30000 3/30 Copies 50 19950
10/30 Progression 20000 160500 2/6 Bel-Gar 15000 15000 3/30 Ptrs Dvl 5000 14950
10/30 Ogleby 25000 135500 2/6 Superior 25000 0* 3/30 Belmont 10000 4950
11/2 Mid Ross 15000 120500 2/6 Unt Sply 15000 0* 4/2 Berea Prtg 4950 0
11/9 Unt Sply 50000 70500 2/6 Berea Prtg 15000 0* 4/2 Berea Prtg 15050 0*
11/29 Berea Lit 25000 45500 2/8 Sagamore 5000 0* 4/9 REM 500 0*
12/1 Dol Fed 10000 35500 2/14 100M 100000 4/12 Mid Ross 5000 0*
12/13 Card Fed 20000 15500 2/15 50M 150000 5/7 Ohio Ost 5000 0*
12/14 Belmont 15000 500 2/16 Bel-Gar 5000 145000 5/8 Inkspots 5000 0*
12/15 Shkr Sav 5000 500* 2/21 Bel-Gar 15000 130000 5/8 Prts Dvl 2500 0*
1/8 BFK 500 0 2/26 Inkspot 5000 125000 5/11 100M 100000
1/8 100M 100000 2/27 Lcl 25UAW 50000 75000 5/14 BVR 5000 95000
1/8 Card Fed 30000 70000 2/28 Ptrs Dvl 2500 72500 5/15 Guswold 10000 85000
1/9 Pt of View 10000 60000 2/28 Shkr Sav 25000 47500 5/16 ESB 15000 70000
1/17 Am Safety 5000 55000 3/1 Copies 35000 12500 5/16 Superior 50000 20000
1/23 Foster 15000 40000 3/2 Untd Tor 10000 2500 5/16 J Stephen 5000 15000
1/24 Gib Prtg 5000 35000 3/8 Sagamore 2500 0 5/16 Am Aster 15000 0
1/26 Bel-Gar 5000 30000 3/8 Sagamore 12500 0* 5/16 Am Aster 10000 0*
1/26 Copies 20000 10000 3/12 150M 150000 5/22 Sagamore 15000 0*
1/29 Slvr Lake 5000 5000 3/12 Untd Tor 40000 110000 Coding 21200
1/29 100M 105000 3/12 Preston 50000 60000 M. Base Cost Date Min 125M
2/2 Sagamore 20000 85000 3/20 Midland 15000 45000 2.64 4/2 Max 250M
Size M/Wgt Basis Grain Color Finish Grade Location Ctn. Skid Cont. Att.
8½x14 12.72 20 L White RmSeal Advantage Bond F 14 5M
*
No stock or insufficient stock to meet demand 9-44
Supply chain example
Suppose that inventory is to be maintained on a

distributor’s shelf for an item whose demand is
forecasted to be d = 100 units per day and sd = 10 units
per day. A reorder point is the method of inventory
control. The supply channel is shown in the diagram.
Determine the average inventory to be held at the

distributor where we have:
I = 10%/year C = $5/unit
S = $10/order P = 0.99 during lead time
9-45
Supply Chain Example (Cont’d)
Supplier
Processing time
X p = 1, s 2p = 0 .1
Transport time
X i = 4 , si = 1.0
Inbound transport 2
Outbound transport
Transport time
Pool point Distributor
X o = 2 , s o2 = 0 .25
9-46
Solution The reorder point inventory theory applies.
However, determining the statistics of the demand-
during-lead-time distribution requires taking the lead-
time for the entire channel into account.
Recall,
s' = LT(sd2 )  d 2(sLT

2
)
where
2
sLT = sp2  si2  so2
= 0.1  1.0  0.25
= 1.35 days

9-47
Average lead time
LT = X p  X i  X o = 1 4  2 = 7 days
Now
s' = 7x102 1002 x1.35 = 14,200 = 119.16 days
and
Q* = 2(100)(10) = 63 units
0.1(5)
*
Q
AIL =  z(s' ) = 63  2.33(199.16) = 309 units
2 2

9-48
Joint ordering
Perpetual inventory control for most firms is the problem of
managing items jointly rather than singly. This occurs since
more than one item is typically purchased from the same
vendor. The approach to joint ordering is to find a common
order review interval (T) and then to set separate target levels
(MAX) based on specific item costs and service levels.
A common review time may be specified, or it may be
computed based on appropriate economics.
2(O  S )
T* = i where
(I C D ) O = common procurement cost, $/order
i i

Note: Q* = T*xd 9-49
Joint Ordering Example
Given
Item
A B
Average daily demand (d) 30 75 units
Demand std. dev. ( sd) 8 10 units
Average lead time (LT) 14 14 days
Annual carrying cost (I) 25 25 %
Procurement cost (S) 30 20 $/order
with common cost (O) 80 $/order
In-stock probability (P) 80 92 %
Product value (C) 170 200 $/unit
Out-of-stock cost (k) 25 45 $/unit
Selling days per year 365 365 days

9-50
Joint Ordering Example (Cont’d)
Find common review time
T* = 2[80  (30  20)] = 4.35 days

[0.25/365][170(30) 200(75)]
Find target quantity (MAX) for item A

s'A = sd T *  LT = 8 4.35 14 = 34.3 units
A
A
then z@80%=0.84
MAX A = X  z(sA' ) = 30(4.35 14)  0.84(34.3) = 579 units

9-51
Joint Ordering Example (Cont’d)
which has an average inventory of
Avg. Inventoryi =T *(di / 2)  zi (si' )
Avg. InventoryA = 4.35(30 / 2)  0.84(34.3) = 94.1units
Find target quantity (MAX) for item B

sB' = sd T *  LT = 10 4.35 14 = 42.8 units
B
B
then for z@90%=1.41

MAX B = 75(4.35 14) 1.41(42.8) = 1437 units
which has an average inventory of

Avg. inventoryB = 4.35(75 / 2) 1.41(42.8) = 223 units
9-52
The Min-Max variant

This is basically a reorder point system, but the order
quantity is incremented by the amount of the difference
between the reorder point quantity and the quantity on
hand + quantity on order  backorders. This takes into
account that demand does not decrement inventory
levels evenly. Therefore, inventory levels may fall
below the reorder point at the time that it is reached.

9-53
Min-Max Inventory Control
Add increment ROPq to order size
M
Quantity on hand
Q1 Q2
~ Q*
ROP
q
LT LT Time
The T, R, M variant
This is a combination of the min-max and the periodic
review systems. The stock levels are reviewed
periodically, but control the release of the replenishment
order by whether the reorder point is reached. This
method is useful where demand is low, such that small
quantities might be released under a periodic review
method.

9-55
Inventory not
T,R,M variant
below R, so don’t
Inventory level
place an order
Q1
Q2
R
q
LT LT Time
T T
T = review time
R = reorder point M – Q = replenishment quantity
Stock to demand (a periodic review method)

This is an important periodic review method, not so much
because of its accuracy but because of its popularity in
practice. The method is synchronized with the period of
the forecast. The target quantity (MAX) is developed as
follows. An example
•Set the period of the forecast, say 4 weeks
•Add time for lead time, say 1 week
•Add an increment of time for safety stock, say 1 week
9-57
Stock to Demand (Cont’d)
Therefore, MAX is 6/4 times the monthly forecast. The
replenishment quantity is determined as follows.
At the time (T) of the monthly stock-level review, make a
forecast and determine the MAX level.
Units
MAX = Forecast x 6/4 12,500
Plus: Backorders 0
Less: Quantity on hand* -5,342
Less: Quantity on order -4,000
Order quantity (Q) 3,158
*Quantity on hand = actual quantity on hand +
quantity on order – backorders

9-58
Multiple item, multiple-location control

The theory that has been discussed previously is
useful when designing inventory control systems for
the practical problem of controlling many items at
many locations. Consider how a specialty chemical
company designed such a practical system. TASO is
the time to accumulate a stock order (truckload) for all
items in warehouse.

9-59
Multiple-Item, Multiple-Location Control
M
Quantity on hand
Q2 Q3
Q1
~
Stock
Order
order
received
0
LT LT Time
~
TASO TASO TASO

M = maximum level Qi = order quantity
TASO = time to accumulate stock order LT = lead time
9-60
Customer Service Level
For individual items
The service level (stock availability) actually achieved by
inventory control methods is not best represented by the
probability (P) of a stockout during the lead time. It is
more accurate to compute it as follows.
s 'E D /Q s 'E
SL = 1 (z) = 1 (z)
D Q
Using data from the reorder point under uncertainty
example, the service level would be:
Note: Higher
SL = 1 17.32(.0034)(2,600 / 322) = 0.999 than P
2,600 9-61
Customer Service Level (Cont’d)
This actual level is higher than P = 0.99 that was
used to set the inventory level. The reason is that
there are periods of time when the stock level is
above the reorder point and there is no risk of being
out of stock.
Methods for defining stock availability include:
•Probability of filling all item demand
•Probability of filling an order completely
•Probability of filling a percent of all item demand
•Weighted average of items filled on an order (fill
rate)
9-62
Customer Service Level (Cont’d)
For multiple items on the same order
If all items on an order have the same service level,
what is the probability of filling the order complete?
The service level for multiple items is the combination
of the individual item service levels as follows:
SL = SL1 x SL2 x SL3 …x SLn
Suppose 3 items have the following service levels—
0.95, 0.89, and 0.92. The probability of filling the order
complete is:
SL = 0.95 x 0.89 x 0.92 = 0.78

9-63
Push Inventory Control
Example
Three warehouses are used to supply 900 retail
drugstores. Each warehouse serves approximately 300
stores. A large purchase of clock radios is made, where
radios were to be a promotional item in the next forecast
period. The special buy will result in more stock than
needed, but the company expects to sell all stock
eventually. Warehouses are to have a 92% in-stock
probability. All of the purchased radios are to be allocated
to the warehouses based on the anticipated demand
levels at each warehouse. Account is taken of the
inventory already on hand. A total of 5000 radios is
purchased. The next purchase will be made in one
month. Further information is given below.
Push Inventory Control (Cont’d)
Current Forecast
stock Forecasted error (std.
Ware- level, demand, dev.), units
house units units
1 400 2,300 100
2 350 1,400 55
3 0 900 20
4,600
How should the allocation to the warehouses be

made?
9-65
Solution
Ware- Total
house requirements
a
1 2,428
2 1,470
3 926
4,824
a2,428 = 2,300 + 1.28(100)
Total requirements = Forecast + z(Forecast error)

where z@90% = 1.28. Therefore,
9-66
(1) (2)
(3)= (4) (5)=
(1)(2) Pro- (4)+(3)
Total On- Net ration
require- hand require- of Allo-
Ware- ments, stock, ments, excess, cation,
house units units units units units
a b
1 2,428 400 2,028 463 2,491
2 1,470 350 1,120 282 1,402
3 926 0 926 181 1,107
c
Total 4,824 4,074 926 5,000
aTotalrequirements less (quantity on hand + quantity on order – backorders)
bExcess purchase quantity times forecast for warehouse divided by total
forecast quantity. For example, (5,000 – 4,074) x 2,300/4,600 = 463

c5,000 – 4,074 = 926

Multi-Echelon Inventories
Control the entire channel inventory levels, not just a
single echelon.
How much stock here when
retailers also carry stock?
Warehouse
echelon
ad- R1
End customer demand

le
e tail LT R d 1 , s d1
R e,
Warehouse tim
lead-time, LTw R2
S W d 2 , sd 2
Supplier Warehouse
R3
d 3 , sd3
Retailer

9-68
Multi-Echelon Inventories (Cont’d)
Example
An item has the following cost characteristics. Item
values are CR=$10/unit and CW=$5/unit. Carrying cost
is I = 20%/year. Ordering costs are SR=$40/order and
SW=$75/order. Lead times are LTR=0.25 month and
LTW=0.5 months. In-stock probability for retailers and
warehouses is 90%. Monthly demand statistics are:
Monthly Std. dev.,

avg., units units
Retailer 1 202.5 16.8
Retailer 2 100.5 15.6
Retailer 3 302.5 18.0
Combined 605.5 32.4
9-69
Solution
Based on reorder point inventory control, the retailers’
inventory statistics are
Retailer 1 Retailer 2 Retailer 3
Reorder qty, Q 312 220 381
Reorder point, ROP 61 35 87
Avg. inv., AIL 167 120 202
The warehouse echelon order quantity is

2DW SW
QW = = 2(605.5x12)(75) =1,043.98, or 1,044 units
ICW 0.20(5)
ROPW = dW xLTW  zsW LTW = 605.5(.5) 1.28(32.4) .5
= 332 units
9-70
The warehouse echelon inventory is
Q
AILW = W  zsW LTW
2
= 1,043.98 1.28(32.4) 0.5
2
= 551.32. or 551units
The average warehouse inventory is the warehouse
echelon inventory less the retailers’ inventory, or 551
– 167 –120 – 202 = 62 units.
Rule When the total warehouse inventory (sum of
retailers’ inventory, inventory at the warehouse and on
order, and retailers’ orders less any inventory committed
to customers drops below 332 units, order 1,044 units.
9-71
Aggregate Inventory Control
Product items can be grouped according to 80-20
curve, each with different stocking policies
100
90
80
Total sales (%)
70
60
50
40
30
A items B items C items
20
10
0
0 20 40 60 80 100
Total items (%)
9-72
Inventory Consolidation
(“Risk Pooling”)
Illustration of risk pooling
Suppose there is a product stocked in two warehouses.

The replenishment quantities are determined by the
economic order quantity formula. The replenishment
lead-time is 0.5 months, the cost for a replenishment
order is $50, the inventory carrying cost is 2% per
month, and the item value is $75 per unit. The
probability of an out of stock during the lead-time period
is 5%. The demand is normally distributed with typical
demand over six months as follows.

9-73
Risk Pooling (Cont’d)
Combined
Demand Demand Demand in a
in Whse in Whse Central
Month A B Whse
1 35 67 102
2 62 83 145
3 46 71 117
4 25 62 87
5 37 55 92
6 43 66 109
Avg. (D) 41.33 67.33 108.66
6
Std. Dev. (sd) 11.38 8.58 19.07
Estimate the average inventory levels for two-

warehouse and one-warehouse supply channels.
9-74
Regular stock
2DS
RS = Q = IC
2 2
2(41.33)(50)
RSA = 0.02(75) = 52.49 = 26.25 units
2 2
2(67.33)(50)
RSB = 0.02(75) = 67.00 = 33.50 units
2 2
Regular stock in system is
RSs = RSA  RSB = 26.25  33.50 = 59.75 units
9-75
Regular stock if item is entirely in one warehouse
2(108.66)(50)
RSC = 0.02(75) = 85.11= 42.56 units
2 2
Safety stock
SS = z(sd ) LT
SSA = 1.96(11.38) 0.5 = 15.77 units

SSB = 1.96(8.58) 0.5 = 11.89 units
System safety stock in 2 warehouses
SSA  SSB = 15.77 11.89 = 27.66 units
9-76
Safety stock in 1 warehouse
SSc = 1.96(19.07) 0.5 = 26.43 units
Total inventory
Two
AIL = Regular stock + Safety stock warehouses
AIL = 59.75 + 27.66 = 87.41 units

In a one-warehouse channel
AIL = 42.56 + 26.43 = 68.99 units
Conclusion There is a reduction in the average
inventory level of an item as the number of stocking
points in the supply channel is decreased. In this
example, both regular stock and safety stock decline.
9-77
System-wide Inventory as a Fraction of the
Demand Divided Between Two Warehouses
100
System-Wide Inventory
95
Percent of Peak
90
85
80
75
70
0 0.2 0.4 0.6 0.8 1
One warehouse's demand as a fraction
of the total
9-78
Virtual Inventories
•Stockouts are filled from other stocking locations in
the distribution network
• Customers assigned to a primary stocking location
• Backup locations are usually determined by
“zoning” rules
• Expectation is that lower system-wide inventories
can be achieved while maintaining or improving
stock availability levels
• Total distribution costs should be lower to support
the cross filling of customer demand

9-79
Cross Filling Among 2
Stocking Locations
Stock Stock
location A
assignment
location B
assignment
Primary
Primary
Secondary
assignment
Demand 1 Demand 2
Virtual Inventories 9-80
Potential Benefit
of Cross Filling
Suppose that an item is stocked at a fill rate of 80% in
4 stocking locations. If cross filling is used, what is
the effective fill rate for the customer?
Fill rate = [1 – (.20)(.20)(.20)(.20)] x 100 = 99.8%
Customer service levels can be quite high

even if the item fill rate is low!
But are inventory costs lower?
Virtual Inventories
9-81
Regular Stock in 2 Locations
•Meaning of regular stock Stock Stock
•How it varies with:

location A location B
assignment
assignment
Primary
Primary
Demand dispersion Secondary
assignment
Demand 1 Demand 2
Method of stock control
Fill rate
Virtual Inventories
9-82
Stock Control Methods
and Regular Stock
If control is EOQ-based, average inventory level (AIL) is
EOQ
formula 
0.5

2S D 



AIL = Q = IC
 
  = kD0.5
2 2
AIL is a function of
demand with
If stock-to-demand control exponents ranging
from 0.5 to 1.0
AIL = kD1.0
Virtual Inventories
Regular Stock as a Percent of Demand Divided
Between Two Warehouses
D1.0
Percent of peak sytem-w ide D0.9 Stock-to-
100 demand
control
95
D0.5
90
inventory
D0.7
85
80
75
EOQ-based
70
control
0 20 40 60 80 100
O n e w a re h o u s e 's d e m a n d
a s a p e rc e n t o f to ta l d e m a n d
Virtual Inventories
Observation about
Regular Stock
A system of multiple stocking locations
will carry its maximum regular stock
when demand is balanced among them
Virtual Inventories
9-85
Fill Rate and Regular Stock
Cross filling increases regular stock as lower fill rates
are specified
Example
•2 locations
•Demand is dispersed 50 and 150
•Fill rate is 90%
•Stocking policy is D0.5 with k=1
Virtual Inventories
9-86
Example (Cont’d)
No cross filling Cross filling
Location A Location B Location A Location B

Demand 1 50 0 45a 5b
Demand 2 0 150 15 135
Total 50 150 60 140
Regular stock 7.1 12.2 7.7 11.8
System inv. 19.3 19.5
a50x.90=45
b[50x(1-0.90)]x0.905
Regular stock increases
with cross filling
Virtual Inventories
Regular Stock Penalty for Cross Filling Under
Various Stocking Policies
EOQ-based
150
% of inventory compared
w ith no cross-filling control
140
D0.5
130
D0.7
120
110 D0.9
100 D1.0
50 60 70 80 90 100
Stock-to-demand A v e ra g e fill ra te o n w a re h o u s e s
control
CR (2004) Prentice Hall, Inc. Virtual Inventories 9-88
Safety Stock in 2 Locations Stock Stock
locationA
location A locationB
locationB
assignment
assignment
Primary
•Meaning of safety stock
Primary
Secondary
•Safety stock depends on

assignment
Demand 1 Demand 2
Demand dispersion (variance is proportional

to (demand)
Fill rate
Observation
A system of multiple stocking
locations will carry its minimum
safety stock when demand is
balanced among them
Virtual Inventories
Safety Stock Estimation
Safety stock is estimated by
ss = zs * LT
where s* is the demand standard deviation at
location N
When cross filling,

s * = FR 2sd2
where sd is the demand standard deviation at the
primary location
At any location N
sN* = [FR(1  FR )N 1]2 sd2
Safety Stock in 2 Locations
Example
•2 locations
•Weekly demand and std. dev. are (50,5) and
(150,15)
•Lead time is 1 week
•Fill rate (FR) is 95%
•z is 1.65 for 95% stocking level (demand normally
distributed)
•Inventory control is EOQ based
Safety Stock for 2 Locations
No cross filling Cross filling
Location A Location B Location A Location B

Std. Dev. 1 5 0 4.7500 0.2375
Std. Dev. 2 0 15 0.7125 14.2500
Combined 5 15 4.8 14.3
Safety stock 8.3 24.8 7.9 23.5
System inv. 33.1 31.4
Safety stock decreases

with cross filling
Virtual Inventories
Safety Stock Reduction Due to Cross Filling as a
Percent of Demand Divided Between Two Warehouses
Lower safety
25 stocks from
Percent reduction
FR=70% lower fill rates

20
15
10
FR=90%
5
No cross-
0 filling
0 5 15 25 35 45 55 65 75 85 95 100
One warehouse's demand as a percent of
the system-wide demand
Virtual Inventories
Simplifying the Decision
Problem
An item potentially can be cross filled from 1 backup
warehouse. The item has a value of $200/unit, a carrying
cost of 25%/per year, a stocking level of 6-weeks demand, a
replenishment lead-time of 8 weeks, and a target fill rate of
95%. To cross haul the item from the secondary warehouse
incurs an extra $10/ unit in transportation. The stock control
policy is not known. Demand statistics are as follows:
Mean demand, Std. Dev.,

Location units units
1 300 138
2 100 80
System 400 160
Should the item be cross filled?

Simplifying the Decision
Answer
Solve for K in AIL=KD. K=D1-/TO, where TO is 52
weeks/6 weeks of demand or 8.67. This assumes a
control policy based on  of 0.7. Hence,
K=(400x52)1-.7/8.67=2.28. The demand ratio r between
the 2 locations is 100/400=0.25. Now solve for two
parameters of the cross-filling curve.
1 1.7
X= tD = 10([400x52]) =1.73
ICK 0.25(200)(2.28)
and
Y = zs LT = 1.96(160) 8 = 0.4
 0.7
KD 2.28(400x52)
Now, from the decision curve, don’t cross fill this item.
Decision Curve for FR=.95 and =0.7
LEGEND
D = annual system demand
s = system demand std.
dev.
Ratio of minimum demand to total

Below
C = item unit cost decision
I = annual carrying cost curve—don’t
rate (%) cross fill

K, = inventory
control parameters
t = transportation rate
FR = item fill rate
r = ratio of minimum
demand, r
demand to system
demand
LT = lead-time in demand
std. dev. time units
z = normal deviate at FR %

tD 1 zs LT
X = Y = 
X
ICK KD
Virtual Inventories
9-96
Square Root Law
of Inventory Consolidation
The amount of inventory (regular stock) at multiple stocking
points can be estimated by the square root law when
•Inventory control at each point is based on EOQ
principles
•There is an equal amount on inventory at each point
The square root law is:
IT = Ii n
where
IT = amount of inventory at one location
Ii = amount of inventory at each of n locations
n = number of stocking points
Square Root Law (Cont’d)
Example Suppose that there is $1,000,000 of inventory at 3
stocking points for a total of $3,000,000. If it were all
consolidated into 1 location, we can expect:
IT = 1,000,000 3 = $1,732,051
If we wish to consolidate from 3 to 2 warehouses, the

level of inventory in each warehouse would be:
I
Ii = T = $1,732,051 = $1,224,745
2 1.41
For a total system inventory of n x I = 2 x $1,224,745 =
$2,449,490.
Square Root Law (Cont’d)
More simply
n2
I2 = I1 n
1
where
I2 = system inventory in n2 locations
I1 = system inventory in n1 locations
n2 = no. of new locations
n1 = no. of previous locations
So,
I2 = 3,000,000 2 = 2,449,490
3
9-99
Warehouse average inventory, Ii ($000s)
Inventory-Throughput
3000
Curve
2500
2000
Ii = 1.57Di0.72
R = 0.85
1500
1000
500
0
0 10000 20000 30000 40000 50000
CR (2004) Prentice Hall, Inc. Annual warehouse throughput, Dj ($000s) 9-100
Inventory-Throughput Curve
Example
Suppose two warehouses with 390,000 lb. and
770,000 lb. of throughput respectively are to be
consolidated into a single warehouse with 390,000 +
770,000 = 1,160,000 lb. of annual throughput. How
much inventory is likely to be in the single warehouse?
The inventory-throughput curve developed from the
company’s multiple warehouse stocks is shown in the
figure below.
Note Reading from the plot, the inventory in the

consolidated warehouse has dropped to 262,000 lb.
from 132,000 + 203,000 = 335,000 lb. in the two
warehouses.
9-101
Inventory-Throughput Curve
450
400
350
Average inventory level, AIL i (000 lb.)
AILi = 3.12Di0.628
R2 = 0.77
300
262
250
203
200
150
132
100
Current
warehouse Consolidated
50 warehouse
0
0 200 400 600 800 1000 1200 1400 1600 1800
Annual warehouse throughput,D i (000 lb.)

9-102
Turnover Ratio
Annual sales
Turnover ratio =
Average inventory
A fruit grower stocks its dried fruit products in 12 warehouses
around the country. What is the turnover ratio for the
distribution system?
Ware- Annual Average Ware- Annual Average

house warehouse inventory house warehouse inventory
no. throughput, $ level, $ no. throughput, $ level, $
1 21,136,032 2,217,790 7 43,105,917 6,542,079
2 16,174,988 2,196,364 8 47,136,632 5,722,640
3 78,559,012 9,510,027 9 24,745,328 2,641,138
4 17,102,486 2,085,246 10 57,789,509 6,403,076
5 88,228,672 11,443,489 11 16,483,970 1,991,016
6 40,884,400 5,293,539 12 26,368,290 2,719,330
Totals 425,295,236 43,701,344
$425,295,236
TO ratio = = 9.7 $ are at cost
$43,701,344 9-103
A table entry is the proportion of the
Areas under area under the curve from a z of 0 to a

positive value of z. To find the area
from a z of 0 to a negative z, subtract the
tabled value from 1.
Standardized z
0.0
0.1
.00
0.5000
0.5398
.01
0.5040
0.5438
.02
0.5080
0.5478
.03
0.5120
0.5517
.04
0.5160
0.5557
.05
0.5199
0.5596
.06
0.5239
0.5636
.07
0.5279
0.5675
.08
0.5319
0.5714
.09
0.5359
0.5753
Normal 0.2
0.3
0.4
0.5793
0.6179
0.6554
0.5832
0.6217
0.6591
0.5871
0.6255
0.6628
0.5910
0.6293
0.6664
0.5948
0.6331
0.6700
0.5987
0.6368
0.6736
0.6026
0.6406
0.6772
0.6064
0.6443
0.6808
0.6103
0.6480
0.6844
0.6141
0.6517
0.6879
0.5 0.6915 0.6950 0.6985 0.7019 0.7054 0.7088 0.7123 0.7157 0.7190 0.7224
Distribution 0.6
0.7
0.8
0.9
0.7257
0.7580
0.7881
0.8159
0.7291
0.7611
0.7910
0.8186
0.7324
0.7642
0.7939
0.8212
0.7357
0.7673
0.7967
0.8238
0.7389
0.7704
0.7995
0.8264
0.7422
0.7734
0.8023
0.8289
0.7454
0.7764
0.8051
0.8315
0.7486
0.7794
0.8078
0.8340
0.7517
0.7823
0.8106
0.8365
0.7549
0.7852
0.8133
0.8389
1.0 0.8413 0.8438 0.8461 0.8485 0.8508 0.8531 0.8554 0.8577 0.8599 0.8621
1.1 0.8643 0.8665 0.8686 0.8708 0.8729 0.8749 0.8770 0.8790 0.8810 0.8830
1.2 0.8849 0.8869 0.8888 0.8907 0.8925 0.8944 0.8962 0.8980 0.8997 0.9015
1.3 0.9032 0.9049 0.9066 0.9082 0.9099 0.9115 0.9131 0.9147 0.9162 0.9177
1.4 0.9192 0.9207 0.9222 0.9236 0.9251 0.9265 0.9279 0.9292 0.9306 0.9319
1.5 0.9332 0.9345 0.9357 0.9370 0.9382 0.9394 0.9406 0.9418 0.9429 0.9441
1.6 0.9452 0.9463 0.9474 0.9484 0.9495 0.9505 0.9515 0.9525 0.9535 0.9545
1.7 0.9554 0.9564 0.9573 0.9582 0.9591 0.9599 0.9608 0.9616 0.9625 0.9633
1.8 0.9641 0.9649 0.9656 0.9664 0.9671 0.9678 0.9686 0.9693 0.9699 0.9706
1.9 0.9713 0.9719 0.9726 0.9732 0.9738 0.9744 0.9750 0.9756 0.9761 0.9767
2.0 0.9772 0.9778 0.9783 0.9788 0.9793 0.9798 0.9803 0.9808 0.9812 0.9817
2.1 0.9821 0.9826 0.9830 0.9834 0.9838 0.9842 0.9846 0.9850 0.9854 0.9857
2.2 0.9861 0.9864 0.9868 0.9871 0.9875 0.9878 0.9881 0.9884 0.9887 0.9890
2.3 0.9893 0.9896 0.9898 0.9901 0.9904 0.9906 0.9909 0.9911 0.9913 0.9916
2.4 0.9918 0.9920 0.9922 0.9925 0.9927 0.9929 0.9931 0.9932 0.9934 0.9936
2.5 0.9938 0.9940 0.9941 0.9943 0.9945 0.9946 0.9948 0.9949 0.9951 0.9952
2.6 0.9953 0.9955 0.9956 0.9957 0.9959 0.9960 0.9961 0.9962 0.9963 0.9964
2.7 0.9965 0.9966 0.9967 0.9968 0.9969 0.9970 0.9971 0.9972 0.9973 0.9974
2.8 0.9974 0.9975 0.9976 0.9977 0.9977 0.9978 0.9979 0.9979 0.9980 0.9981
2.9 0.9981 0.9982 0.9982 0.9983 0.9984 0.9984 0.9985 0.9985 0.9986 0.9986
3.0 0.9987 0.9987 0.9987 0.9988 0.9988 0.9989 0.9989 0.9989 0.9990 0.9990
3.1 0.9990 0.9991 0.9991 0.9991 0.9992 0.9992 0.9992 0.9992 0.9993 0.9993
3.2 0.9993 0.9993 0.9994 0.9994 0.9994 0.9994 0.9994 0.9995 0.9995 0.9995
3.3 0.9995 0.9995 0.9995 0.9996 0.9996 0.9996 0.9996 0.9996 0.9996 0.9997
CR (2004) Prentice Hall, Inc. 3.4 0.9997 0.9997 0.9997 0.9997 0.9997 0.9997 0.9997 0.9997 0.9997 0.9998
9-104
Unit Normal
Loss
Integrals

Inventory Policy Decisions: "Every Management Mistake Ends Up in Inventory." Michael C. Bergerac

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

Inventory Policy Decisions: "Every Management Mistake Ends Up in Inventory." Michael C. Bergerac

Transféré par

Droits d'auteur :

Formats disponibles

Inventory Policy

CR (2004) Prentice Hall, Inc.

CR (2004) Prentice Hall, Inc.

CR (2004) Prentice Hall, Inc. 9-4

•They consume capital resources that might be put to

CR (2004) Prentice Hall, Inc.

A = Allocation quantity to each warehouse

CR (2004) Prentice Hall, Inc.

CR (2004) Prentice Hall, Inc.

Total cost Minimum cost

CR (2004) Prentice Hall, Inc.

Profit = Price per unit  Cost per unit

If CPn is probability of n units being sold, then

Profit = $50 35 = $15

CPn is between 15 and 20 items, round up and order 20

CR (2004) Prentice Hall, Inc.

Develop a simple control system by finding the

0 Lead Lead Time

CR (2004) Prentice Hall, Inc. 9-22

Using differential calculus, the optimal value for Q will

The reorder point is: Famous EOQ

Rule When the inventory level drops to 150 units

CR (2004) Prentice Hall, Inc.

5 Compute ROP and other statistics

CR (2004) Prentice Hall, Inc.

Monthly demand forecast, d 11,107 units

Find optimal Q and P

CR (2004) Prentice Hall, Inc.

Now, z@0.81=0.81 and E(0.81)=0.1181

CR (2004) Prentice Hall, Inc.

Q = 2(11,107)(12)[10 0.01(3,795)(0.1181)] =13,246 units

Continue to revise Q and P until no further

Note Although the in-stock probability during the

CR (2004) Prentice Hall, Inc.

Qp* = 2DS p Just add

After computing s’, calculation of the ordering policy would

M = maximum level T = review interval

Estimate Q* from the EOQ formula as if under demand

CR (2004) Prentice Hall, Inc.

CR (2004) Prentice Hall, Inc.

Note Compare this cost with that of the reorder point

CR (2004) Prentice Hall, Inc.

Suppose that inventory is to be maintained on a

Determine the average inventory to be held at the

s' = LT(sd2 )  d 2(sLT

CR (2004) Prentice Hall, Inc.

CR (2004) Prentice Hall, Inc.

CR (2004) Prentice Hall, Inc.

CR (2004) Prentice Hall, Inc.

T* = 2[80  (30  20)] = 4.35 days

Find target quantity (MAX) for item A

CR (2004) Prentice Hall, Inc.

Find target quantity (MAX) for item B

then for z@90%=1.41

which has an average inventory of

The Min-Max variant

CR (2004) Prentice Hall, Inc.

CR (2004) Prentice Hall, Inc.

Stock to demand (a periodic review method)

CR (2004) Prentice Hall, Inc.

Multiple item, multiple-location control

CR (2004) Prentice Hall, Inc.

TASO TASO TASO