Chapter VI.

Independent Demand Inventory Systems

Inventory is a store of goods held for future use. Types of
inventory include raw materials, work-in- process, finished
goods, and supplies.
Among some firms, inventory carrying costs represent from 10
to 40 percent of the cost of goods sold.
Why hold inventories? Benefits include:
1. Meet expected customer demand
2. Improve customer service
3. Uncertain sources of supply
4. Economies of scale (production, purchase, transportation)
5. Stable (smooth) rates of production
6. An inability to backlog demand
7. Decouple operations (internal and external)
8. Forward buying (hedge)
9. Display items
Costs to have inventory include:
1. Costs to: insure, finance, warehouse, manage or control,
security
2. Foregone return of alternative investments
3. Shop congestion
4. Tendency to hide quality related problems
5. Delays in meeting customer deliver dates
A. Independent demand inventory systems
Due to the nature of demand, an important distinction is made
between independent and dependent demand
1

 Independent demand inventory control procedures rely upon

unbiased forecasts of uncertain demand. Demand and lead
times may be treated as random variables. Forecasts are used
to develop production and purchase schedules for end items.
Dependent demand inventory control techniques utilize
material requirements planning (MRP) logic.
MRP converts a production schedule for end items that
experience independent demand into known requirements for
lower level items in order to determine production and
purchase schedules.
Regardless of the nature of demand (independent, dependent,
seasonal, terminal, lumpy, etc.) two fundamental issues
underlie all inventory planning:
1. How much to order (Q)
2. When to order
B. Methods of Stock Control
1. Push inventory control
A centralized form of inventory control whereby the
distribution network (points) is viewed collectively.

OEM

WH

Retl

WH

Retl

WH

Retl

Multiechelon Supply Chain

Available inventories are allocated to distribution points
according to some criterion such as projected needs,
historical use, or available space.
Push may afford greater economies of scale in production
and purchase orders.
2. Pull inventory control
A decentralized approach, whereby each point in the
distribution network is treated independently of all other
points.

Inventory policy decisions are made independently.

3. Benefits of centralized (push) inventory location strategies

a. Reduction of safety stocks
b. Greater opportunity to use private warehousing (volume)
c. Reduction of transshipment costs
d. Improvement of in-stock position in filling customer orders
4. Benefits of decentralized inventory location strategies
a. Improvement in order-cycle response time
b. Stimulation of sales
c. Reduction of warehouse-to-customer transport costs
d. Greater opportunity for public warehousing (larger
proportion of costs which vary with volume)
C. Inventory control objectives (2)

 Remember: two fundamental issues underlie all inventory

planning: (1) how much to order (Q), and (2) when to order.
1.Cost Objective: Minimize sum of relevant costs
2. Service objective: Maximize customer service
3. Four potential costs: holding (carrying), order, shortage, and
item acquisition (purchase)
a. Holding (Carrying) costs: H = C(i)
C is the unit cost and i is the annual holding cost
expressed as a percent of C
Determined for an average inventory level and divided
into 3 classes: (1) space costs, (2) capital costs, and (3)
risk costs
b. Order (Set-Up): S
Fixed costs associated with the acquisition of goods and
include costs for processing, transmitting, handling,
transportation, receipt, and inspecting orders.
c. Shortage Costs
Costs incurred when a customer order cannot be filled
from the inventory to which the order in assigned.
There are two classes: (1) lost sales, and (2) backorder
costs

d. Acquisition Cost (Manufacture or Purchase) - C represents

the unit cost
e. Total Cost = Holding Cost + Order Cost + Shortage Cost +
Acquisition Cost
4. Service Level (SL): defined several ways
a. The probability of satisfying all of demand during the leadtime period (the complement of a stock out)
b. The number of orders per year that were completely filled
c. The time between customer order transmittal and order
receipt
Note impact of customer service level on inventory levels.
D. Pull Strategy Inventory Control Models
Remember: 2 important issues in inventory control: order
quantity and order timing.
2 general classes of models: continuous review (fixed-order
quantity) and periodic review (fixed-order-period)
1. Continuous Review (Q-systems or fixed-order-quantity
systems)
a. Multi-period models
(1) Fixed order quantity, variable time between orders
(2) Perpetual inventory count
(3) On-hand inventory balance serves as order trigger (R)
(4) Example: 2-bin system
When inventory on-hand drops to a predetermined level, a
replenishment order is placed for a fixed amount of
5

inventory. EOQ in conjunction with reorder point (R) is

used to determine inventory policy (order quantity and
order release determination).
b. Single-period model
Places orders for a fixed amount of inventory, but does so
every period. Marginal analysis is used to determine
inventory policy (order quantity determination).
2. Periodic Review (P-systems or fixed-order-period systems)
a. Variable order quantity, fixed time between orders
b. Periodic inventory count
c. Time serves as order trigger
When a predetermined amount of time, P, has elapsed a
physical inventory count is taken. Based upon the number of
units in stock at that time (A) and an upper inventory unit
target (T), an order is placed for Q units, where Q = (T-A).
E. Pull strategy and Q-Systems: How much to order (Q)
Multi-period inventory model: order decisions for infinite
length planning period address how much to order as well as
when to place an order. Economic Order Quantity (EOQ)
model is commonly used as basis for determining Q.
Single-period model: order decisions for finite length (1
period) inventory planning address how much to order.
Marginal analysis commonly used as modeling basis.
1. Economic Order Quantity (EOQ)
Goal: to identify an order size (Qo) that will minimize the
sum of the relevant costs.
6

 A cost is only relevant if it can be avoided.

Numerous modeling variations
Modeling assumptions: Saw-Tooth Diagram
a. 1 product
b. Annual demand (D) known, occurs at a constant rate (d)
c. Known and constant order lead time (LT)
d. Instantaneous replenishment (no split deliveries)
e. No stock outs allowed
f. No quantity discounts
g. All costs known and constant
There are only two relevant costs: annual carrying and
ordering costs.
Total annual cost: TC = (Q/2)H + (D/Q)S
____________

Basic model: Qo = 2(D)S /H

Realism?

Reorder point (R) quantity under certainty

Example
2. Economic Production Lot-Size EOQ
3. EOQ with Quantity Discounts
4. Joint Replenishment EOQ
F. Pull strategy and Q-Systems: When to order (R)
7

 When to place an order is determined relative to an on-hand

balance or a Re-order Point (R) quantity.
The R quantity is to be determined for a preferred customer
service level rather than estimating stock out costs.
1. Four determinants of R quantities
a. Rate of demand (d)
b. Length of order (manufacturing) LT
c. Variability of demand and LT (d, LT, or DLT)
d. Degree of stock-out risk (preferred service level, SL)
2. Alternative demand during lead time probability distributions
a. Discrete distribution of demand during lead time (DDLT)
May be used due to lack of demand history or item nature
For a desired minimum service level, R is determined
relative to the cumulative demand frequency.
DDLT
3
4
5
6
7

Number of
Frequency Cumulative
Observations (Probability) Probability
10
.10
.10
20
.20
.30
40
.40
.70
20
.20
.90
10
.10
1.00
100
1.00

b. Continuous distribution of demand during lead time

 May be used given sufficient demand history or due to

nature of the item (e.g., divisible units)
General Model Form: R = m + s
Where, m is expected (mean) demand during lead time
and s is safety stock. Safety stock is used to satisfy the
variability in demand and/or lead time and is a function
of: (1) rate of demand, (2) length of lead time, (3)
variability of demand and/or lead time, and (4) desired
service level. Safety stock is determined as:
s = Z(DLT)
4 Model Variations
(1) Constant Demand (d), Constant Lead Time (LT):
R = d(LT)
(2) Variable Demand (d), Constant Lead Time (LT):
__
R = d(LT) + Z(DLT) where DLT = dLT
(3) Constant Demand (d), Variable Lead Time (LT):
R = d(LT) + Z(DLT) where DLT = d(LT)
(4) Variable Demand (d), Variable Lead Time (LT):
_______________
R = d(LT) + Z(DLT) where, DLT = LT(d2) + d2(LT2)
Example

G. Pull strategy and P-Systems with Demand Uncertainty

Orders are placed after an elapsed fixed period of time, the
order interval (P), has passed for a variable quantity of units.
Other differences between Q- and P-Systems:
(a) Perpetual versus periodic inventory accounting systems
(b) Higher than normal demand during the order cycle leads to
a shorter time between orders for Q-Systems while in PSystems it leads to larger order sizes
(c) P-Systems typically require larger safety stocks in order to
provide the same level of customer service.
1. General model form with uncertain (variable) demand,
constant lead time
Expected Demand
Q = During Protection
Interval

+ Safety Stock - Amount on hand

(A)

2. Specific model form (uncertain demand, constant lead time)

____
Q = d(P + L) + Z(d)P + L - A
Where, d is rate of demand, P is the time between orders, L =
lead time, Z reflects desired service level, d is the standard
deviation of demand, and A = amount on hand at reorder time
A reasonable approximation method for determining model
parameter P is to use the EOQ model to solve for the order
size. The time between orders is then simply the order
10

quantity divided by annual demand. T, or the upper

inventory target, is determined for a specified service level.
The standard deviation of demand during the protection
interval is found by determining the daily standard
deviation of demand and multiplying by the square root of
the daily length of the protection interval.
Example
3. Advantages of P-Systems versus Q-Systems
a. On-hand accuracy resulting from periodic physical
inventory counts
b. Ability to group orders from a supplier resulting in
potentially lower transportation and ordering
4. Disadvantages of P-Systems versus Q-Systems
a. Larger amount of safety stock for a given level of customer
service
b. Periodic physical inventory review cost
H. Single period inventory model (Continuous Review, FixedOrder-Quantity System)
The approach used assumes units cannot be carried over to a
subsequent period without penalty. The difference of this
model is due to the finite length (1 period) planning horizon.
Marginal analysis is used to determine order quantity.
Shortage Costs (Cs) = revenue per unit cost per unit

11

 Excess Costs (Ce) = cost per unit - salvage value per unit
Service Level = Cs/(Cs+Ce)
1. Uniform Demand Distribution:
So = Min. Demand + SL (Max - Min Demand)
2. Normal Demand Distribution: So = d + Zd
3. Discrete Demand Distribution: compare SL with cumulative
frequency. So established so that SL achieved as a minimum
requirement.
I. ABC Inventory Classification
J. Cycle Counts

12