Chapter 2

Inventory
Management
and Risk Pooling

McGraw-Hill/Irwin Copyright © 2008 by The McGraw-Hill Companies, Inc. All rights reserved.
 Customers,
Field demand
Sources: Regional Warehouses: centers
plants Warehouses: stocking sinks
vendors stocking points
ports points

Supply

Inventory &
warehousing
costs
Production/
purchase Transportation Transportation
costs costs costs
Inventory &
warehousing
costs 2-2
 2.1 Introduction
Why Is Inventory Important?
Distribution and inventory (logistics) costs are
quite substantial
Total U.S. Manufacturing Inventories ($m):
 1992-01-31: $m 808,773
 1996-08-31: $m 1,000,774
 2006-05-31: $m 1,324,108

Inventory-Sales Ratio (U.S. Manufacturers):

 1992-01-01: 1.56
 2006-05-01: 1.25

2-3
 Why Is Inventory Important?
 GM’s production and distribution network
 20,000 supplier plants
 133 parts plants
 31 assembly plants
 11,000 dealers
 Freight transportation costs: $4.1 billion (60% for
material shipments)
 GM inventory valued at $7.4 billion (70%WIP; Rest
Finished Vehicles)
 Decision tool to reduce:
 combined corporate cost of inventory and transportation.
 26% annual cost reduction by adjusting:
 Shipment sizes (inventory policy)
 Routes (transportation strategy)

2-4
 Why Is Inventory Required?
 Uncertainty in customer demand
 Shorter product lifecycles
 More competing products

 Uncertainty in supplies
 Quality/Quantity/Costs/Delivery Times
 Delivery lead times
 Incentives for larger shipments

2-5
 Holding the right amount at the
right time is difficult!
 Dell Computer’s was sharply off in its
forecast of demand, resulting in inventory
write-downs
 1993 stock plunge
 IBM’s ineffective inventory management
 1994 shortages in the ThinkPad line
 Cisco’s declining sales
 2001 $ 2.25B excess inventory charge

2-6
 Inventory Management-Demand
Forecasts
 Uncertain demand makes demand
forecast critical for inventory related
decisions:
 What to order?
 When to order?
 How much is the optimal order quantity?

 Approach includes a set of techniques

 INVENTORY POLICY!!

2-7
 Supply Chain Factors in Inventory
Policy
 Estimation of customer demand
 Replenishment lead time
 The number of different products being considered
 The length of the planning horizon
 Costs
 Order cost:
 Product cost
 Transportation cost
 Inventory holding cost, or inventory carrying cost:
 Insurance on inventories
 Maintenance costs
 Obsolescence cost
 Opportunity costs
 Service level requirements
2-8
 2.2 Single Stage Inventory
Control
 Single supply chain stage
 Variety of techniques
 Economic Lot Size Model
 Demand Uncertainty
 Single Period Models
 Continuous Review Policy
 Variable Lead Times
 Periodic Review Policy
 Service Level Optimization

2-9
2.2.1. Economic Lot Size Model

FIGURE 2-3: Inventory level as a function of time

2-10
 Assumptions
 D items per day: Constant demand rate
 Q items per order: Order quantities are fixed, i.e., each
time the warehouse places an order, it is for Q items.
 K, fixed setup cost, incurred every time the warehouse
places an order.
 h, inventory carrying cost accrued per unit held in
inventory per day that the unit is held (also known as,
holding cost)
 Lead time = 0
(the time that elapses between the placement of an
order and its receipt)
 Initial inventory = 0
 Planning horizon is long (infinite).

2-11
 Deriving EOQ
 Total cost at every cycle:
hTQ
K 
2
 Average inventory holding cost in a cycle: Q/2
 Cycle time T =Q/D
KD hQ
 Average total cost per unit time: 
Q 2

2 KD
Q 
*

2-12
 EOQ: Costs

FIGURE 2-4: Economic lot size model: total cost per unit time

2-13
 Sensitivity Analysis
Total inventory cost relatively insensitive to order quantities

Actual order quantity: Q

Q is a multiple b of the optimal order quantity Q*.
For a given b, the quantity ordered is Q = bQ*

b .5 .8 .9 1 1.1 1.2 1.5 2

Increase 25% 2.5% 0.5% 0 .4% 1.6% 8.9% 25%

in cost

2-14
 2.2.2. Demand Uncertainty
 The forecast is always wrong
 It is difficult to match supply and demand
 The longer the forecast horizon, the worse the
forecast
 It is even more difficult if one needs to predict
customer demand for a long period of time
 Aggregate forecasts are more accurate.
 More difficult to predict customer demand for
individual SKUs
 Much easier to predict demand across all SKUs
within one product family

2-15
 Single Period Models
 Using historical data
 identify a variety of demand scenarios
 determine probability each of these scenarios will occur
 Given a specific inventory policy
 determine the profit associated with a particular scenario
 given a specific order quantity
 weight each scenario’s profit by the likelihood that it will occur
 determine the average, or expected, profit for a particular ordering
quantity.
 Order the quantity that maximizes the average profit.

2-16
Single Period Model Example

FIGURE 2-5: Probabilistic forecast

2-17
 Additional Information
 Fixed production cost: $100,000
 Variable production cost per unit: $80.
 During the summer season, selling price:
$125 per unit.
 Salvage value: Any swimsuit not sold
during the summer season is sold to a
discount store for $20.

2-18
 Two Scenarios
 Manufacturer produces 10,000 units while
demand ends at 12,000 swimsuits
Profit
= 125(10,000) - 80(10,000) - 100,000
= $350,000
 Manufacturer produces 10,000 units while
demand ends at 8,000 swimsuits
Profit
= 125(8,000) + 20(2,000) - 80(10,000) - 100,000
= $140,000

2-19
Probability of Profitability Scenarios
with Production = 10,000 Units
 Probability of demand being 8000 units =
11%
 Probability of profit of $140,000 = 11%
 Probability of demand being 12000 units =
27%
 Probability of profit of $350,000 = 27%
 Total profit = Weighted average of profit
scenarios

2-20
Order Quantity that Maximizes
Expected Profit

FIGURE 2-6: Average profit as a function of production quantity

2-21
 Risk-Reward Tradeoffs
 Optimal production quantity maximizes
average profit is about 12,000
 Producing 9,000 units or producing 16,000
units will lead to about the same average
profit of $294,000.
 If we had to choose between producing
9,000 units and 16,000 units, which one
should we choose?

2-22
 Risk-Reward Tradeoffs

FIGURE 2-7: A frequency histogram of profit

2-23
 Risk-Reward Tradeoffs
 Production Quantity = 9000 units
 Profit is:
 either $200,000 with probability of about 11 %
 or $305,000 with probability of about 89 %
 Production quantity = 16,000 units.
 Distribution of profit is not symmetrical.
 Losses of $220,000 about 11% of the time
 Profits of at least $410,000 about 50% of the time
 With the same average profit, increasing the production
quantity:
 Increases the possible risk
 Increases the possible reward

2-24
 2.2.5. Multiple Order
Opportunities
REASONS
 To balance annual inventory holding costs and annual fixed order
costs.
 To satisfy demand occurring during lead time.
 To protect against uncertainty in demand.

TWO POLICIES
 Continuous review policy
 inventory is reviewed continuously
 an order is placed when the inventory reaches a particular level or reorder point.
 inventory can be continuously reviewed (computerized inventory systems are
used)

 Periodic review policy

 inventory is reviewed at regular intervals
 appropriate quantity is ordered after each review.
 it is impossible or inconvenient to frequently review inventory and place orders if
necessary.

2-25
2.2.6. Continuous Review Policy
 Daily demand is random and follows a normal distribution.
 Every time the distributor places an order from the
manufacturer, the distributor pays a fixed cost, K, plus an
amount proportional to the quantity ordered.
 Inventory holding cost is charged per item per unit time.
 Inventory level is continuously reviewed, and if an order is
placed, the order arrives after the appropriate lead time.
 If a customer order arrives when there is no inventory on
hand to fill the order (i.e., when the distributor is stocked
out), the order is lost.
 The distributor specifies a required service level.

2-26
 Continuous Review Policy

= Inventory Position (On Hand + On Order)

= Inventory Level (On Hand)

Reorder Point
Order
Arrives Order
Place Place Place Place Arrives
Safety Stock Level Order Order Order Order
Order
Arrives
Order
Arrives

Lead Lead Lead Lead

Time Time Time Time
 Continuous Review Policy
 AVG = Average daily demand faced by the
distributor
 STD = Standard deviation of daily demand faced
by the distributor
 L = Replenishment lead time from the supplier to
the distributor in days
 h = Cost of holding one unit of the product for
one day at the distributor
 α = service level. This implies that the probability
of stocking out is 1 - α

2-28
 Continuous Review Policy
 (Q,R) policy – whenever inventory level
falls to a reorder level R, place an order for
Q units
 What is the value of R?

2-29
 Continuous Review Policy
 Average demand during lead time: L x AVG
 Safety stock:

z  STD  L
 Reorder Level, R:
L  AVG  z  STD  L
 Order Quantity, Q:
2 K  AVG
Q
h

2-30
Service Level & Safety Factor, z

Service 90% 91% 92% 93% 94% 95% 96% 97% 98% 99% 99.9%
Level

z 1.29 1.34 1.41 1.48 1.56 1.65 1.75 1.88 2.05 2.33 3.08

z is chosen from statistical tables to ensure

that the probability of stockouts during lead time is exactly 1 - α

2-31
Inventory Level Over Time
FIGURE 2-9: Inventory level as a function of time in a (Q,R) policy

Inventory level before receiving an order = z  STD  L

Inventory level after receiving an order = Q  z  STD  L

Average Inventory = Q
2  z  STD  L
2-32
Continuous Review Policy Example
 A distributor of TV sets that orders from a
manufacturer and sells to retailers
 Fixed ordering cost = $4,500
 Cost of a TV set to the distributor = $250
 Annual inventory holding cost = 18% of
product cost
 Replenishment lead time = 2 weeks
 Expected service level = 97%

2-33
 Continuous Review Policy
Example
Month Sept Oct Nov. Dec. Jan. Feb. Mar. Apr. May June July Aug

Sales 200 152 100 221 287 176 151 198 246 309 98 156

Average monthly demand = 191.17

Standard deviation of monthly demand = 66.53

Average weekly demand = Average Monthly Demand/4.3

Standard deviation of weekly demand = Monthly standard deviation/√4.3

2-34
 Continuous Review Policy
Example

Parameter Average weekly Standard Average Safety Reorder

demand deviation of demand stock point
weekly demand during lead
time
Value 44.58 32.08 89.16 86.20 176

Weekly holding cost = 0.18  250

 0.87
52

2  4,500  44.58
Optimal order quantity = Q  679
.87

Average inventory level = 679/2 + 86.20 = 426

2-35
 2.2.7. Variable Lead Times
 Average lead time, AVGL
 Standard deviation, STDL.
 Reorder Level, R:

R  AVG  AVGL  z AVGL  STD 2  AVG 2  STDL2

Amount of safety stock= z AVGL  STD 2

 AVG 2
 STDL2

2 K  AVG
Order Quantity = Q
h

2-36
 2.2.8. Periodic Review Policy
 Inventory level is reviewed periodically at regular
intervals
 An appropriate quantity is ordered after each review
 Two Cases:
 Short Intervals (e.g. Daily)
 Define two inventory levels s and S
 During each inventory review, if the inventory position falls below s,
order enough to raise the inventory position to S.
 (s, S) policy
 Longer Intervals (e.g. Weekly or Monthly)
 May make sense to always order after an inventory level review.
 Determine a target inventory level, the base-stock level
 During each review period, the inventory position is reviewed
 Order enough to raise the inventory position to the base-stock level.
 Base-stock level policy

2-37
 (s,S) policy
 Calculate the Q and R values as if this
were a continuous review model
 Set s equal to R
 Set S equal to R+Q.

2-38
 Periodic Review (P) Policy
r+L
r
L
Target

Place
Order Order
Order Place Arrives
Arrives Order
Order
Arrives
Place
Order

Place = Inventory Position (On Hand + On Order)

Order Order
Arrives

Lead Lead Lead Lead = Inventory Level (On Hand)

Time Time Time Time

Review Period Review Period Review Period

 Base-Stock Level Policy
 Determine a target inventory level, the base-
stock level
 Each review period, review the inventory
position is reviewed and order enough to raise
the inventory position to the base-stock level
 Assume:
r = length of the review period
L = lead time
AVG = average daily demand
STD = standard deviation of this daily demand.

2-40
 Base-Stock Level Policy
 Average demand during an interval of r +
L days= ( r  L)  AVG

 Safety Stock= z  STD  r  L

2-41
 Base-Stock Level Policy
Example
 Assume:
 distributor places an order for TVs every 3 weeks
 Lead time is 2 weeks
 Base-stock level needs to cover 5 weeks
 Average demand = 44.58 x 5 = 222.9
 Safety stock = 1.9  32.8  5
 Base-stock level = 223 + 139 = 362
 Average inventory level = 3442 .58  1.9  32.08  5  203.17

 Distributor keeps 5 (= 203.17/44.58) weeks of supply.

2-42
 2.3 Risk Pooling
 Demand variability is reduced if one
aggregates demand across locations.
 More likely that high demand from one
customer will be offset by low demand
from another.
 Reduction in variability allows a decrease
in safety stock and therefore reduces
average inventory.

2-44
 Demand Variation
 Standard deviation measures how much
demand tends to vary around the average
 Gives an absolute measure of the variability
 Coefficient of variation is the ratio of
standard deviation to average demand
 Gives a relative measure of the variability,
relative to the average demand

2-45
 Acme Risk Pooling Case
 Electronic equipment manufacturer and distributor
 2 warehouses for distribution in New York and New
Jersey (partitioning the northeast market into two
regions)
 Customers (that is, retailers) receiving items from
warehouses (each retailer is assigned a warehouse)
 Warehouses receive material from Chicago
 Current rule: 97 % service level
 Each warehouse operate to satisfy 97 % of demand
(3 % probability of stock-out)

2-46
 New Idea
 Replace the 2 warehouses with a single
warehouse (located some suitable place) and
try to implement the same service level 97 %
 Delivery lead times may increase
 But may decrease total inventory investment
considerably.

2-47
 Historical Data
PRODUCT A
Week 1 2 3 4 5 6 7 8
Massachusetts 33 45 37 38 55 30 18 58

New Jersey 46 35 41 40 26 48 18 55

Total 79 80 78 78 81 78 36 113

PRODUCT B
Week 1 2 3 4 5 6 7 8
Massachusetts 0 3 3 0 0 1 3 0
New Jersey 2 4 3 0 3 1 0 0
Total 2 6 3 0 3 2 3 0

2-48
 Summary of Historical Data
Statistics Product Average Demand Standard Coefficient of
Deviation of Variation
Demand
Massachusetts A 39.3 13.2 0.34

Massachusetts B 1.125 1.36 1.21

New Jersey A 38.6 12.0 0.31

New Jersey B 1.25 1.58 1.26

Total A 77.9 20.71 0.27

Total B 2.375 1.9 0.81

2-49
 Inventory Levels
Product Average Safety Stock Reorder Q
Demand Point
During Lead
Time
Massachusetts A 39.3 25.08 65 132

Massachusetts B 1.125 2.58 4 25

New Jersey A 38.6 22.8 62 131

New Jersey B 1.25 3 5 24

Total A 77.9 39.35 118 186

Total B 2.375 3.61 6 33

2-50
 Savings in Inventory
 Average inventory for Product A:
 At NJ warehouse is about 88 units
 At MA warehouse is about 91 units
 In the centralized warehouse is about 132 units
 Average inventory reduced by about 36 percent
 Average inventory for Product B:
 At NJ warehouse is about 15 units
 At MA warehouse is about 14 units
 In the centralized warehouse is about 20 units
 Average inventory reduced by about 43 percent

2-51
 Critical Points
 The higher the coefficient of variation, the greater the
benefit from risk pooling
 The higher the variability, the higher the safety stocks
kept by the warehouses. The variability of the demand
aggregated by the single warehouse is lower
 The benefits from risk pooling depend on the behavior of
the demand from one market relative to demand from
another
 risk pooling benefits are higher in situations where
demands observed at warehouses are negatively
correlated
 Reallocation of items from one market to another
easily accomplished in centralized systems. Not
possible to do in decentralized systems where
they serve different markets

2-52
 2.4 Centralized vs.
Decentralized Systems
 Safety stock: lower with centralization
 Service level: higher service level for the same
inventory investment with centralization
 Overhead costs: higher in decentralized system
 Customer lead time: response times lower in the
decentralized system
 Transportation costs: not clear. Consider
outbound and inbound costs.

2-53
2.5 Managing Inventory in the
Supply Chain
 Inventory decisions are given by a single decision maker
whose objective is to minimize the system-wide cost
 The decision maker has access to inventory information
at each of the retailers and at the warehouse
 Echelons and echelon inventory
 Echelon inventory at any stage or level of the system
equals the inventory on hand at the echelon, plus all
downstream inventory (downstream means closer to
the customer)

2-54
Echelon Inventory

FIGURE 2-13: A serial supply chain

2-55
 Reorder Point with Echelon
Inventory
 Le = echelon lead time,
 lead time between the retailer and the
distributor plus the lead time between the
distributor and its supplier, the wholesaler.
 AVG = average demand at the retailer
 STD = standard deviation of demand at
the retailer
 Reorder point R  Le  AVG  z  STD  Le

2-56
4-Stage Supply Chain Example
 Average weekly demand faced by the
retailer is 45
 Standard deviation of demand is 32
 At each stage, management is attempting
to maintain a service level of 97% (z=1.88)
 Lead time between each of the stages,
and between the manufacturer and its
suppliers is 1 week

2-57
 Costs and Order Quantities
K D H Q

retailer 250 45 1.2 137

distributor 200 45 .9 141

wholesaler 205 45 .8 152

manufacturer 500 45 .7 255

2-58
 Reorder Points at Each Stage
 For the retailer, R=1*45+1.88*32*√1 = 105
 For the distributor, R=2*45+1.88*32*√2 =
175
 For the wholesaler, R=3*45+1.88*32*√3 =
239
 For the manufacturer, R=4*45+1.88*32*√4
= 300

2-59
 More than One Facility at Each
Stage
 Follow the same approach
 Echelon inventory at the warehouse is the
inventory at the warehouse, plus all of the
inventory in transit to and in stock at each of the
retailers.
 Similarly, the echelon inventory position at the
warehouse is the echelon inventory at the
warehouse, plus those items ordered by the
warehouse that have not yet arrived minus all
items that are backordered.

2-60
Warehouse Echelon Inventory

FIGURE 2-14: The warehouse echelon inventory

2-61
 2.6 Practical Issues
 Periodic inventory review.
 Tight management of usage rates, lead times, and
safety stock.
 Reduce safety stock levels (LT reduction,
Aggregation).
 Introduce or enhance cycle counting practice.
 ABC approach.
 Shift more inventory or inventory ownership to
suppliers.
 Quantitative approaches.
FOCUS: not reducing costs but reducing inventory levels.
Significant effort in industry to increase inventory turnover
Annual _ Sales
Inventory _ Turnover _ Ratio 
Average _ Inventory _ Level
2-62
 2.7 Forecasting
RULES OF FORECASTING
 The forecast is always wrong.
 The longer the forecast horizon, the
worse the forecast.
 Aggregate forecasts are more accurate.

2-63
 Utility of Forecasting
 Part of the available tools for a manager
 Despite difficulties with forecasts, it can be
used for a variety of decisions
 Number of techniques allow prudent use
of forecasts as needed

2-64
 Techniques
 Judgment Methods
 Sales-force composite
 Experts panel
 Delphi method
 Market research/survey
 Time Series
 Moving Averages
 Exponential Smoothing
 Trends
 Regression
 Holt’s method
 Seasonal patterns
 Trend + Seasonality
 Causal Methods

2-65
 SUMMARY
 Matching supply with demand a major challenge
 Forecast demand is always wrong
 Longer the forecast horizon, less accurate the
forecast
 Aggregate demand more accurate than
disaggregated demand
 Need the most appropriate technique
 Need the most appropriate inventory policy

2-66