Stochastic Inventory Models

Statistical Reorder Point Models

Single period reordering, Newsvendor model
Multiple period ordering
Product availability
Base stock model
(Q,r) model
Service level approach
Cost penalty approach
Managerial levers to reduce safety stocks

Outline
The role of safety inventory in a supply chain
Determining the appropriate level of
Reorder point
safety inventory

Replenishment Policies
Continuous
Periodic

Product availability measures

Cycle Service Level
Fill Rate

Impact of Lead Time Variability on Safety Stocks

EOQ Assumptions

1. Instantaneous production.
2. Immediate delivery.
3. Deterministic demand.
4. Constant demand.
5. Known fixed setup costs.

EPQ model relaxes this one

lags can be added to EOQ or other models
newsvendor and (Q,r) relax this one
WW model relaxes this one
can use constraint approach

6. Single product or separable products.

Chapter 17 extends (Q,r) to

multiple product cases
4

Modeling Philosophies for Handling Uncertainty

1. Use deterministic model adjust solution
- EOQ to compute order quantity, then add safety stock
- deterministic scheduling algorithm, then add safety lead time

2. Use stochastic model

- news vendor model
- base stock and (Q,r) models
- variance constrained investment models

The Newsvendor Approach

Assumptions:
1. single period
2. random demand with known distribution
3. linear overage/shortage costs
4. minimum expected cost criterion

Examples:
newspapers or other items with rapid obsolescence
Christmas trees or other seasonal items
capacity for short-life products

L.L. Bean Example

Demand Di
(in hundreds)

Probability pi

Cumulative Probability of
Demand Being Di or Less
(Pi)

Probability of Demand
Being Greater than Di
(1-Pi)

0.01

0.01

0.99

0.02

0.03

0.97

0.04

0.07

0.93

0.08

0.15

0.85

0.09

0.24

0.76

0.11

0.35

0.65

10

0.16

0.51

0.49

11

0.20

0.71

0.29

12

0.11

0.82

0.18

13

0.10

0.92

0.08

14

0.04

0.96

0.04

15

0.02

0.98

0.02

16

0.01

0.99

0.01

17

0.01

1.00

0.00

Newsvendor Example L.L Bean

Scenario:

Cost of shirts is $45.

Selling price is $100.
Unsold shirts(outlet) can be sold off at $50.
Holding parkas and transporting to outlet costs $10

Model Parameters:
cs =
co =

L.L. Bean Example

Expected demand Di pi 1,026
10

Expected profit Di ( p c) (1,000 Di )(c s) pi

i 4

17

1,000( p c) pi $49,900
i11

Expected profit
from extra 100 parkas

Expected profit from

ordering 1,300 parkas

= 5,500 x Prob(demand 1,100) 500

x Prob(demand < 1,100)
= $5,500 x 0.49 $500 x 0.51 = $2,440
= $49,900 + $2,440 + $1,240 +
$580
= $54,160

L.L. Bean Example

Additional
Hundreds

Expected Marginal
Benefit

Expected Marginal
Cost

Expected Marginal
Contribution

11th

5,500 x 0.49 = 2,695

500 x 0.51 = 255

2,695 255 = 2,440

12th

5,500 x 0.29 = 1,595

500 x 0.71 = 355

1,595 355 = 1,240

13th

5,500 x 0.18 = 990

500 x 0.82 = 410

990 410 = 580

14th

5,500 x 0.08 = 440

500 x 0.92 = 460

440 460 = 20

15th

5,500 x 0.04 = 220

500 x 0.96 = 480

220 480 = 260

16th

5,500 x 0.02 = 110

500 x 0.98 = 490

110 490 = 380

17th

5,500 x 0.01 = 55

500 x 0.99 = 495

55 495 = 440

Newsvendor Model Notation

X demand (in units), a random variable.

G ( x) P ( X x), cumulative distribution function of demand
(assumed continuous.)
g ( x)

d
G ( x) density function of demand.
dx

co cost (in dollars) per unit left over after demand is realized.
c s cost (in dollars) per unit of shortage.
Q production/order quantity (in units); this is the decision variable.
11

Newsvendor Model
Cost Function:
Note: for any given
Y (Q ) expected overage expected shortage cost day, we will be either
over or short, not both.
But in expectation,
co E units over cs E units short
overage and shortage
can both be positive.

co max Q x,0 g ( x)dx cs max x Q,0 g ( x)dx

Q

co (Q x) g ( x)dx cs ( x Q ) g ( x)dx

12

Newsvendor Model (cont.)

Optimal Solution: taking derivative of Y(Q) with respect to Q,
setting equal to zero, and solving yields:

G (Q * ) P X Q *

Notes:
Q* co
Q * cs

cs
co c s

Critical Ratio is
probability stock
covers demand

G(x)

cs
co c s

Q*
13

Marginal Analysis
Specific Cost Element P(Specific Cost Element
Is Incurred)
the Specific Cost
Element
Overage Co

P(Demand is less than Q)

Expected Value of

Co *P(D<Q)

= P(D<Q)
Shortage Cs

P(Demand is greater than Q)

Cs *[1- P(D<Q)]

= 1 - P(D<Q)

14

14

Newsvendor Example T Shirts

Scenario:
Demand for T-shirts is exponential with mean 1000 (i.e., G(x) =
P(X x) = 1- e-x/1000). (Note - this is an odd demand distribution;
Poisson or Normal would probably be better modeling choices.)
Cost of shirts is $10.
Selling price is $15.
Unsold shirts can be sold off at $8.

Model Parameters:
cs = 15 10 = $5
co = 10 8 = $2

15

Newsvendor Example T Shirts (cont.)

Solution:
G (Q ) 1 e
*

Q
1000

cs
5

0.714
co cs 2 5

Q * 1,253

Sensitivity: If co = $10 (i.e., shirts must be discarded) then

G (Q ) 1 e
*

Q
1000

cs
5

0.333
co cs 10 5

Q * 405

16

Example
Billys Bakery bakes fresh bagels each morning. The daily demand for bagels is a
random variable with a distribution estimated from historical data given. From
this historical data, Billys Bakery has also determined that the expected daily
demand is 18 units. The bagels cost $2 to make, and they are sold for $3.50 each.
Bagels unsold at the end of the day are purchased by a nearby charity soup kitchen
for $1.50 cents each
Quantity sold

Probability

0.05

0.10

10

0.10

15

0.20

20

0.25

25

0.15

30

0.10

35

0.05

Based on this distribution, how many bagels should be baked at the start of
each day?
What is the expected number of bagels sold to charity?
17

Newsvendor Model with Normal Dist. Demand

Suppose demand is normally distributed with mean and
standard deviation . Then the critical ratio formula
reduces to:

cs
Q *

co cs

3.00

G (Q* )

cs
Q *
z where ( z )

co cs
Q* z

(z)

0.00
1

13

19

25

31

37

43

49

55

61

Note: Q* increases in both

and if z is positive (i.e.,
if ratio is greater than 0.5).
18

67

73

79

85

91

97 103 109 115 121 127 133 139 145 151 157

Case of Normally Distributed Demand

Define

Ordering quantity
Expected profit

Q z

E[ P(Q)] cs cs ES co EO

Expected understock (shortage)

ES ( x Q ) g ( x)dx L( z )
Q

Expected overstock (overage)

Q

EO (Q x ) g ( x)dx L( z )
0

19

Loss function

Multiple Period Problems

Difficulty: Technically, Newsvendor model is for a single period.
Extensions: But Newsvendor model can be applied to multiple
period situations, provided:

demand during each period is iid, distributed according

to G(x)
there is no setup cost associated with placing an order
stockouts are either lost or backordered
Key: make sure co and cs appropriately represent overage and
shortage cost.
20

Example
Scenario:

GAP orders a particular clothing item every Friday

mean weekly demand is 100, std dev is 25
wholesale cost is $10, retail is $25
holding cost has been set at $0.5 per week (to reflect obsolescence,
damage, etc.)

Problem: how should they set order amounts?

21

Example (cont.)
Newsvendor Parameters:
c0 = $0.5
cs = $15
Solution:
15
0.9677
0.5 15
Q 100

0.9677
25

Q 100
1.85
25
Q 100 1.85 (25 ) 146
G (Q * )

Every Friday, they should

order-up-to 146, that is, if
there are x on hand, then
order 146-x.
22

Newsvendor Takeaways

Inventory is a hedge against demand uncertainty.

Amount of protection depends on overage and

shortage costs, as well as distribution of demand.

If shortage cost exceeds overage cost, optimal order

quantity generally increases in both the mean and
standard deviation of demand.

23

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

24

Role of Safety Inventory

Safety inventory is the average inventory remaining when the
replenishment lot arrives
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)
HP and Dell in PCs

25

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?

26

Determining the Appropriate

Level of Safety Inventory
Measuring demand uncertainty
Measuring product availability
Replenishment policies
Evaluating cycle service level and fill rate
Evaluating safety level given desired cycle service level or
fill rate
Impact of required product availability and uncertainty
on safety inventory

27

Determining the Appropriate

Level of Safety Inventory
Appropriate level of safety inventory determined by:
supply or demand uncertainty
desired level of product availability

28

Selecting the Reorder Point

IP

On-hand inventory

Order
received

IP

Order
received

Order
received

Q
OH

OH

IP

Order
received

Q
OH

r
Order
placed

Order
placed
L

TBO

Order
placed
L

TBO

L
TBO
29

Time

Continuous Review Systems

IP
Order
received

On-hand inventory

Order
received

IP

IP

Order
received

Order
received
Q

r
Order
placed

Order
placed

Order
placed

0
l1
TBO1

l2
TBO2

l3

Time

TBO3

Q
AverageInv entory Cycle Inventory Safety Stock s
30
2

Measuring Demand Uncertainty

Random component is usually estimated by the
standard deviation of demand
Notation:
D = Average demand per period

= Average demand during lead time

= standard deviation of demand per lead time

l = lead time = time between when an order is placed

and when it is received
X= Random variable, demand during lead time
Uncertainty of demand during lead time is what is
important
31

Demand During Lead Time

D = 15

D = 15

75
Demand for week 1

D = 15

75
Demand for week 2

75
Demand for week 3

= 25.98

E[X] E

i 1

nD

2
D

i
D
i 1

Var[X] Var
225
Demand for 3-week lead time

n D
32

Continuous Review Systems

IP
Order
received

On-hand inventory

Order
received

IP

IP

Order
received

Order
received
Q

r
Order
placed

Order
placed

Order
placed

0
l1
TBO1

l2
TBO2

l3

Time

TBO3

Q
AverageInv entory Cycle Inventory Safety Stock s
33
2

Inventory Level Definitions

On-hand Stock
Net Inventory = On-hand stock - Backorders
Inventory Position(IP) = On hand + On order
Backorders
Safety Stock (SS) = Average level of net inventory when
the replenishment arrives.

34
34

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
Cycle service level: fraction of replenishment cycles that
end with all customer demand met

35

Cycle Service Level (type I)

Q+r

s
0
K
L

Time
P[X>r]

r-J= minimum lead time demand

r+K= maximum lead time demand
r-s= expected lead time demand
36

Fill Rate Metric (type II)

Proportion of customer demand satisfied from on hand inventory
Stockout occurs when the demand during lead time exceeds the
reorder point
Need to find the expected number of items that will be short per
replenishment cycle (ESPRC)

ESPRC
ESPRC
Fill rate 1
1
Order
Q
37

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.

38

Finding Expected Units Short

Consider both continuous and discrete cases:

ESPRC ( x - r )P[X x]
x r

ESPRC ( x r ) g ( x)dx
r

39

Base Stock Model Assumptions

1. There is no fixed cost associated with placing an order.
2. There is no constraint on the number of orders that can
be placed per year.
That is, we can replenish
one at a time (Q=1).

40

Base Stock Notation

Q
r
R
l

p(x)
G(x)
h
b
S(R)
B(R)
I(R)

=
=
=
=
=
=
=
=
=
=
=
=
=

1, order quantity (fixed at one)

reorder point
r +1, base stock level
delivery lead time
mean demand during lead time l
standard deviation of demand during lead time l
Prob{demand X during lead time l is equal to x}
Prob{demand X during lead time l is less than or equal to x}
unit holding cost
unit backorder cost
average fill rate (service level)
average backorder level
average on-hand inventory level
41

Inventory Balance Equations

Balance Equation:
(inventory position) = on-hand inventory - backorders + orders

Under Base Stock Policy

inventory position = R= r +1

42

Inventory Profile for Base Stock System (R=5)

inventory position = on-hand inventory +on order - backorders
7

On Hand Inventory
Backorders
Orders
Inventory Position

0
0

10

15

20

25

30

35

Time

43

Service Level (Fill Rate)

Let:
X = (random) demand during lead time l
so E[X] = . Consider a specific replenishment order.
Since inventory position is always R,
the only way this item can stock out is if X R=(r+1).
Expected Service Level:
S (r ) P ( X r 1)
G ( R 1) G (r ) if X is discrete

if X is continuous
G (r 1)
44

Backorder Level
Note: At any point in time, number of orders equals number of
demands during last l time units (X) so from our previous balance
equation:
R = on-hand inventory backorders + orders
on-hand inventory backorders = R X
Note: on-hand inventory and backorders are never positive at the same
time, so if X=x, then
if x R
0
backorders
x R if x R
Expected Backorder Level (also
called First-Order Loss Function):
( x R ) p ( x) if X is discrete
xR
B ( R ) E max x R, 0
( x R ) g ( x)dx if X is continuous
R
45

Backorder Level

poisson

B(r ) ( x R) p ( x) p ( R) ( R)[1 G ( R)]

x R

normal distribution B (r ) ( r )[1 ( z )] ( z )

B (r ) L( z ) where z

simpler version for

spreadsheet computing

r 1

46

Inventory Level
Observe:
on-hand inventory = R X + backorders
E[X] = from data
E[backorders] = B(R) from previous slide
Result:
I ( R) R B( R)
if X : Poisson
p ( R) ( R )G ( R)
2
g R ( R )G ( R) if X : Normal

48

Base Stock Example (Poisson)

l = one month
= 10 units (per month)
Assume Poisson demand, so

10 k e 10

G ( x) p(k )
k!
k 0
k 0
x

Note: Poisson
demand is a good
choice when no
variability data
is available.
49

Base Stock Example Calculations

Poisson Dis. With mean = 10

50

Base Stock Example Results

Service Level: For fill rate of 90%, we must set R-1= r =14, so
R=15 and safety stock s = r- = 4. Resulting service is 91.7%.
Backorder Level:
B(R) = B(15) = 0.103
Inventory Level:
I(R) = R - + B(R) = 15 - 10 + 0.103 = 5.103

51

Optimal Base Stock Levels

Objective Function:
Y(R) = hI(R) + bB(R)
= h(R-+B(R)) + bB(R)
= h(R- ) + (h+b)B(R)

holding plus backorder cost

Solution: if we assume G is continuous, we can use calculus

to get
b
G( R * )
hb

Implication: set base stock

level so fill rate is b/(h+b).
Note: R* increases in b and
decreases in h.

52

Optimal Base Stock Levels

Objective Function:
continuous,
Y ( R ) hI ( R ) bB R

h R B ( R ) bB R
h R h b B R

holding plus backorder cost

Implication: set base stock

level so fill rate is b/(h+b).
Note: R* increases in b
and decreases in h.

Solution: if we assume X is
d
0
Y (R )
dR
d

h R h b B R
dR
d
h h b
B R
dR

d
h h b
( x R ) g ( x )dx

dR R

h h b g ( x)dx
R

h h b G R 1
b
*
G R
h 53 b

Base Stock Normal Approximation

If G is normal(,), then
G ( R*)

b
hb

R *
z

where (z)=b/(h+b). So
R* = + z
Note: R* increases in
and also increases in
provided z>0.

54

Optimal Base Stock Example

Data: Approximate Poisson with mean 10 by normal with mean

10 units/month and standard deviation 10 = 3.16 units/month.

Set h=$15, b=$25.

Calculations:

b
25

0.625
h b 15 25

since (0.32) = 0.625, z=0.32 and hence

R* = + z = 10 + 0.32(3.16) = 11.01 11

Observation: from previous table fill rate is G(11) = 0.583, so

maybe backorder cost is too low.
55

Base Stock Insights

1.

Reorder points control prob of stockouts by establishing safety stock.

2.

To achieve a given fill rate, the required base stock level (and hence
safety stock) is an increasing function of mean and (provided
backorder cost exceeds shortage cost) std dev of demand during
replenishment lead time.

3.

The optimal fill rate is an increasing in the backorder cost and a

decreasing in the holding cost. We can use either a service constraint
or a backorder cost to determine the appropriate base stock level.

4.

Base stock levels in multi-stage production systems are very similar to

kanban systems and therefore the above insights apply.

56