ISOM 2700: Operations Management

Session 12: Inventory Management I

Dongwook Shin
Dept. ISOM, HKUST Business School
Course Roadmap
Bottleneck
Little’s law
Utilization

Control chart
Acceptance sampling
Six sigma

Maximize
Profits

Decision tree method

Linear programming

Economic order quantity

Newsvendor
Waiting time
Server utilization
1
Learning Objectives: Session 12

• Background of inventory management

• Newsvendor model

2
Inventory Management
Goal: Matching supply with demand

3
Why Hold Inventory?
• The transaction motive
• Economies of scale: production, transportation, discount,
replenishment, …
• Competition purpose

• The precautionary motive

• Demand uncertainty: unpredictable events
• Supply uncertainty: lead time, random yield, …

• The speculative motive

• Fluctuating value: ordering cost, selling price
• Demand increase: seasonality, promotion, …

4
Importance of Inventory
Evidence from the industry -- Based on an empirical
analysis over 353 publicly listed U.S. retailers for the
period 1985-2003
• Inventory is the largest asset on the balance sheet for
57% of publicly traded retailers in U.S
• Ratio inventory/total assets averages 35.1%
• Managers and analysts use inventory turnover, inventory
growth rate, and payables to inventory ratio to determine
how well a retailer is managing its inventory

Source: Gaur & Kesavan, 2009. Data obtained from Standard & Poor’s Compustat database

5
Inventory Costs
• Inventory carrying costs
• Insurance cost
• Maintenance cost
• Opportunity cost of alternative investment
• Underage costs
• Loss of profit
• Loss of good will or reputation (hard to quantify)
• Overage costs
• Leftover items are costly
• Fixed ordering (setup) costs
• Handling charges, preparing purchase order
• Supplier selection, negotiations
• Freight and insurance

6
Learning Objectives: Session 12

• Background of inventory management

• Newsvendor model

7
Newsvendor Problem
• Every morning, a newsboy
purchases newspapers to sell
in the day

• How many units to purchase?

• If purchases too many newspapers, money wasted on unsold
items
• If purchases too few newspapers, loses potential sales

• Not too much, not too little…

8
This is a Hard Problem…
• Uncertain demand
• Demand is not known at the time of order
• Pre-commitment of order quantity
• Order quantity must be chosen before observing demand
• Perishable product
• Leftovers are costly: After the demand is realized, they can lose
value, sometimes all of their value

These three characteristics are found

in the newsvendor model

9
Other Applications
• Fashion clothes
• Christmas trees and decorations
• Special event T-shirts
• Items with short life cycle
• Book publishing (traditional)…

Uncertain demand
Pre-commitment of order quantity
Perishable product
10
A Warm-up Example

• Demand for the newspaper is 50 with probability 0.3 and

100 with probability 0.7

• How many newspapers should the newsvendor order?

Cannot be determined precisely because of the lack of information
about underage & overage costs

11
Newsvendor Example 1: Pumpkin
• Demand information
• The retailer forecasts that the demand for pumpkins before
Halloween is
D ≤ 200 with probability 0.5
D ≤ 250 with probability 0.75
D ≤ 300 with probability 0.9
• Price & cost information
• The retailer buys pumpkins from a wholesaler at unit cost $2
• The retailer sells pumpkins at unit price $5
• Any unsold pumpkins has salvage value $1

• Question: How many pumpkins should the retailer buy so

that she could maximize her expected profit?

12
Marginal Analysis

Examination of the costs and benefits of a

marginal (small) change in the order quantity

13
Marginal Analysis: Incremental Profit
• Suppose the retailer has bought 200 pumpkins
• If the retailer buys one more pumpkin, what is the
expected value of this 201st pumpkin to the retailer?
• She needs to pay unit cost = $2
• Case 1: D ≤ 200 (prob. 0.5)
• The retailer cannot sell the 201st pumpkin, and gets $1
salvage value from it. Utility =$1 – $2= – $1
• Case 2: D > 200 (prob. 0.5)
• The retailer sells the 201st pumpkin at $5. Utility =$5 – $2=
$3
• Expected value of the 201st pumpkin
= 0.5 ´ (– $1) + 0.5 ´ $3 = $1 > 0

• Should the retailer buy the 201st pumpkin?

• Yes, because the expected value is positive!
14
Marginal Analysis: Incremental Profit
• Suppose the retailer has bought 250 pumpkins
• If the retailer buys one more pumpkin, what is the
expected value of this 251st pumpkin to the retailer?
• She needs to pay unit cost = $2
• Case 1: D ≤ 250 (prob. 0.75)
• The retailer cannot sell the 251st pumpkin, and gets $1
salvage value from it. Utility =$1 – $2= – $1
• Case 2: D > 250 (prob. 0.25)
• The retailer sells the 251st pumpkin at $5. Utility =$5 – $2=
$3
• Expected value of the 251st pumpkin
= 0.75 ´ (– $1) + 0.25 ´ $3 = 0

• Should the retailer buy the 251st pumpkin?

• Indifferent, because the expected value is zero!
15
Marginal Analysis: Incremental Profit
• Suppose the retailer has bought 300 pumpkins
• If the retailer buys one more pumpkin, what is the
expected value of this 301st pumpkin to the retailer?
• She needs to pay unit cost = $2
• Case 1: D ≤ 300 (prob. 0.9)
• The retailer cannot sell the 301st pumpkin, and gets $1
salvage value from it. Utility =$1 – $2= – $1
• Case 2: D > 300 (prob. 0.1)
• The retailer sells the 301st pumpkin at $5. Utility =$5 – $2=
$3
• Expected value of the 301st pumpkin
= 0.9 ´ (– $1) + 0.1 ´ $3 = –$0.6 < 0

• Should the retailer buy the 301st pumpkin?

• No, because the expected value is negative!
16
Critical Fractile Analysis

• Cost of under-stocking: Cu = price − cost

• Cost of not having unit in stock when demand does materialize

• Cost of over-stocking: Co = cost − salvage value

• Cost of having unit in stock when demand does not materialize

17
Critical Fractile Analysis
• Suppose you have decided to stock/produce n units
• Consider the incremental decision to stock/produce one
more, namely an n+1st unit

Expected underage cost of n+1st unit

Expected
= underage cost × Pr(n+1st unit sells) gain of n+1st
= Cu × Pr (Demand > n) unit

Expected overage cost of n+1st unit

= overage cost × Pr(n+1st unit does not sell) Expected
= Co × Pr (Demand ≤ n) loss of n+1st
unit
18
Critical Fractile Analysis

• What is the expected incremental profit from stocking

the n+1st unit, i.e., Profit(n → n+1)?

• Profit(n → n+1)

= expected gain – expected loss

= Cu × Pr (Demand > n) – Co × Pr (Demand ≤ n)

= Cu – (Cu + Co ) × Pr (Demand ≤ n)

Note: Pr(Demand > n) = 1 – Pr(Demand ≤ n)

19
Critical Fractile Analysis: Profit Curve
Assumption: Demand is continuous

Profits Just right!

Too little!

Too much!

Inventory (n)

Key idea of the critical fractile analysis: At the optimal order

quantity, the expected value (profit) of ordering one
more unit must be close to zero.
20
Critical Fractile Analysis
• We want to find quantity n such that Profit(n → n+1) = 0
• So, Profit(n → n+1) = Cu – (Cu + Co ) × Pr(Demand ≤ n)
equals zero when

Cu
Pr Demand ≤ n =
Cu +Co

• This is how you find the optimal quantity n that

maximizes expected profits

21
Question for Students

• Suppose that the critical fraction is 0.4 for a newsvendor.

What is the correct answer below?

1. Underage cost is greater than the overage cost

2. To choose the optimal ordering quantity, we only need to
know the mean of the distribution
3. If the optimal quantity n is chosen, the newsvendor can
satisfy all customer demand with 40% probability
4. None of the above

22
Example: SnowTime Sporting Goods
Design Production Retailing

Feb 09 Sep 09 Feb 10 Sep 10

Production

• Properties of SnowTime Sporting Goods:

• Based on past sales, knowledge of the industry, and economic
conditions, the marketing department has a probabilistic
forecast
• One production opportunity
• Unsold items during the season can be sold at a discounted price
23
SnowTime Revenue and Cost Data
• For each unit
• c = production cost = $80
• r = selling price = $125
• s = salvage value of unsold unit = $20

• Recall
• Cu = underage cost = marginal gain of extra unit sold
• Co = overage cost = marginal cost of unsold unit

• Hence,
• Cu = r – c = $45 and
• Co = c – s = $60

24
SnowTime Demand Scenario
Cumulative probability
1
0.9
Demand Probability
0.8
8,000 0.11 0.7
10,000 0.11 0.6
12,000 0.28 0.5

14,000 0.22 0.4

0.3
16,000 0.18
0.2
18,000 0.10
0.1
0
8,000 10,000 12,000 14,000 16,000 18,000 20,000
Demand

25
Optimal Solution for SnowTime
• The optimal production quantity n* satisfies
∗
Cu
Pr Demand ≤ n = ≈0.429
Cu +Co
Cumulative 1
probability 0.9
0.8
0.7
0.6 Critical fractile = 0.429
0.5
0.4
0.3
0.2 n* = 12,000
0.1
0
8,000 10,000 12,000 14,000 16,000 18,000 20,000
Demand
26
Service Level
• What is the probability that SnowTime does not
experience a shortage (in other words, every customer
who comes for a jacket is able to get a jacket) ?

Cu
Pr Demand ≤ n∗ = ≈0.429
Cu +Co

27
Properties of the Optimal Solution
Cu
Pr Demand ≤ n∗ =
Cu +Co

Situation Cu ↑ Co ↑
Inventory Level Increase Decrease

Situation Cu<Co Cu>Co

Critical Fractile <0.5 >0.5

28
Normally Distributed Demand
• Suppose demand is a normal random variable with mean
𝜇 and standard deviation 𝜎; or Demand ~ N(𝜇, 𝜎 ")
• What is the optimal stocking level?

Just the same as

what we did for
discrete distribution.
So, what is special
about normal
distribution case?
n*

Cu
Pr (Demand ≤ n∗ ) = F(n∗ ) =
Cu +Co
29
Normally Distributed Demand

• Optimal ordering quantity n* can be written as

n∗ = 𝜇 + z∗ ×𝜎
where z∗ is given by

Cu
z∗ = normsinv
Cu +Co

A function in Excel; inverse

of the standard normal
• Easy to calculate n∗ ! distribution function with
mean 0 and std. dev. 1

30
 The Standard Normal Distribution
z 0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09
0.0 0.5000 0.5040 0.5080 0.5120 0.5160 0.5199 0.5239 0.5279 0.5319 0.5359
0.1 0.5398 0.5438 0.5478 0.5517 0.5557 0.5596 0.5636 0.5675 0.5714 0.5753
0.2 0.5793 0.5832 0.5871 0.5910 0.5948 0.5987 0.6026 0.6064 0.6103 0.6141
0.3 0.6179 0.6217 0.6255 0.6293 0.6331 0.6368 0.6406 0.6443 0.6480 0.6517
0.4 0.6554 0.6591 0.6628 0.6664 0.6700 0.6736 0.6772 0.6808 0.6844 0.6879
Φ(z) = Prob(N(0,1) < z) 0.5 0.6915 0.6950 0.6985 0.7019 0.7054 0.7088 0.7123 0.7157 0.7190 0.7224
0.6 0.7257 0.7291 0.7324 0.7357 0.7389 0.7422 0.7454 0.7486 0.7517 0.7549
0.7 0.7580 0.7611 0.7642 0.7673 0.7704 0.7734 0.7764 0.7794 0.7823 0.7852
0.8 0.7881 0.7910 0.7939 0.7967 0.7995 0.8023 0.8051 0.8078 0.8106 0.8133
0.9 0.8159 0.8186 0.8212 0.8238 0.8264 0.8289 0.8315 0.8340 0.8365 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
Φ(z) = 98.7% implies z* = 1.5
1.6
0.9332
0.9452
0.9345
0.9463
0.9357
0.9474
0.9370
0.9484
0.9382
0.9495
0.9394
0.9505
0.9406
0.9515
0.9418
0.9525
0.9429
0.9535
0.9441
0.9545
2.23 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

31
Reading Standard Normal Distribution
Function Table
• Q: For what z is there a 70.19% chance that the outcome
of a standard normal will be that z or smaller?
• Find the probability inside the table
• Add the row and column headers for that probability
• A: the answer is z = 0.53
• Round-up rule: Whenever you are looking up a target value in a
table and the target value falls between two entries, choose the
entry that leads to the larger order quantity
Standard Normal Distribution Function Table (continued), F (z )

z 0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09
0.0 0.5000 0.5040 0.5080 0.5120 0.5160 0.5199 0.5239 0.5279 0.5319 0.5359
0.1 0.5398 0.5438 0.5478 0.5517 0.5557 0.5596 0.5636 0.5675 0.5714 0.5753
0.2 0.5793 0.5832 0.5871 0.5910 0.5948 0.5987 0.6026 0.6064 0.6103 0.6141
0.3 0.6179 0.6217 0.6255 0.6293 0.6331 0.6368 0.6406 0.6443 0.6480 0.6517
0.4 0.6554 0.6591 0.6628 0.6664 0.6700 0.6736 0.6772 0.6808 0.6844 0.6879
0.5 0.6915 0.6950 0.6985 0.7019 0.7054 0.7088 0.7123 0.7157 0.7190 0.7224
32
Reading Standard Normal Distribution
Function Table
• Q: What is the probability the outcome of a standard
normal will be z = 0.28 or smaller?
• Look for the intersection of the fourth row (with the header 0.2)
and the ninth column (with the header 0.08) because 0.2+0.08 =
0.28, which is the z we are looking for.
• A: The answer is 0.6103, which can also be written as
61.03%
Standard Normal Distribution Function Table (continued), F (z )

z 0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09
0.0 0.5000 0.5040 0.5080 0.5120 0.5160 0.5199 0.5239 0.5279 0.5319 0.5359
0.1 0.5398 0.5438 0.5478 0.5517 0.5557 0.5596 0.5636 0.5675 0.5714 0.5753
0.2 0.5793 0.5832 0.5871 0.5910 0.5948 0.5987 0.6026 0.6064 0.6103 0.6141
0.3 0.6179 0.6217 0.6255 0.6293 0.6331 0.6368 0.6406 0.6443 0.6480 0.6517
0.4 0.6554 0.6591 0.6628 0.6664 0.6700 0.6736 0.6772 0.6808 0.6844 0.6879
0.5 0.6915 0.6950 0.6985 0.7019 0.7054 0.7088 0.7123 0.7157 0.7190 0.7224

33
Takeaways
• Critical fractile analysis to derive formula for optimal
ordering quantity of the newsvendor problem

Cu
Pr Demand ≤ n =
Cu +Co

• Next class: Inventory Management II

• Economic Order Quantity (EOQ) model

34