1

Inventory Optimization

Main reference for all models/concepts:

Vandeput, N. (2020). Inventory Optimization: Models and
Copyright: Nicolas Vandeput
Simulations. Berlin, Boston: De Gruyter.
 Introduction to Inventory
Optimization

Copyright
Nicolas Vandeput
 Why do we need inventory?

Copyright 5
Nicolas Vandeput
 Why do we need inventory?
💥Mitigate risks
Supplier delays
Forecast error
Production hazards
💶Reduce costs
Transaction costs through scale effect
Purchasing costs thanks to grouped orders and
volume discounts
Financial speculation
🚚Logistical constraints
Compensate for any bottleneck (production,
logistic)
🛍️Marketing
Merchandizing: you want to be seen.
Satisfy clients in time (MTS/MTO)
🏴‍☠️Shortage gaming

Copyright 6
Nicolas Vandeput
Why do we want to reduce inventory?

7
 Why do we limit inventory?
💥Mitigate risks
Obsolescence & Dead stocks
💶Reduce costs
Holding costs
Perishable goods
🚚Logistical constraints
Storage space
🛍️Marketing
Feeling of scarcity
BUT Hides bottlenecks: no incentive
to streamline operations.

Copyright 8
Nicolas Vandeput
 Too little 📦 Inventory Too much

🔜 Shortages 🕸️ Dead stocks

🚚 Transaction costs 📦 Holding costs

⚠️ Production flexibility ⚠️ Storage space

Copyright 9
Nicolas Vandeput
 Inventory Policies

Copyright
Nicolas Vandeput
 How do we control inventory?
Inventory policy: How much to order, When to order, (Where to store)

Supply Demand

How much? How much? How much?

When? When? When?
Forecasts

Copyright 11
Nicolas Vandeput
 Policy #1: (s,Q)
Continuous Review, Reorder-Point & Fixed Order
Quantity If you’re below this
threshold, make an order!

📆 When? Less than s in stock

When I have less than 10% ink in my printer,
Policy

Static
I order a new cartridge.
📦 How Much? A fixed quantity, Q
When we have 2 weeks of demand forecast left, we
Dynamic order 1000 units to our supplier.
🐇 You make an order when you need it:
limited risk of shortage (better when high
demand/supply variability)
Pro

🚚 Fixed, optimized order quantity (useful for

operations, batch size, full truckload)

⏱️ Need to be able to make orders at any point

in time to your internal/external supplier
Con

Cannot group orders of different products

toward the same supplier

Copyright 12
Nicolas Vandeput
 Policy #1: (s,Q)
Continuous Review & Fixed Order Quantity
📆 When? Less than s in stock
When I have less than 10% ink in my printer,
Policy

Static
I order a new cartridge.
📦 How Much? A fixed quantity, Q
When we have 2 weeks of demand forecast left, we
Dynamic order 1000 units to our supplier.
🐇 You make an order when you need it:
limited risk of shortage (better when high
demand/supply variability)
Pro

🚚 Fixed, optimized order quantity (useful for

operations, batch size, full truckload)

⏱️ Need to be able to make orders at any point

in time to your internal/external supplier
Con

Cannot group orders of different products

toward the same supplier

Copyright 13
Nicolas Vandeput
 Policy #2: (R,S)
Periodic Review & Up-To Level
📆 When? Every R periods
Static Every Friday,
Policy

I buy milk so that I have 3 litters.

📦 How Much? Up-to a level S
Every week, we buy enough raw materials to have 4
Dynamic weeks of forecast in stock
Simple to group different SKUs in a single order
toward one supplier
Pro

📆 Fixed order timings can allow smooth

operations (collaboration with supplier).

🐢 Less reactive: riskier to wait for a specific

period to make an order
Con

Order quantities are not aligned with batch size

📆 Fixed order timings can create bottlenecks

Copyright 14
Nicolas Vandeput
 Policy #2: (R,S)
Periodic Review & Up-To Level
📆 When? Every R periods
Static Every Friday,
Policy

I buy milk so that I have 3 litters.

📦 How Much? Up-to a level S
Every week, we buy enough raw materials to have 4
Dynamic weeks of forecast in stock
Simple to group different SKUs in a single order
toward one supplier
Pro

📆 Fixed order timings can allow smooth

operations (collaboration with supplier).

🐢 Less reactive: riskier to wait for a specific

period to make an order
Con

Order quantities are not aligned with batch size

📆 Fixed order timings can create bottlenecks

Copyright 15
Nicolas Vandeput
 Policy #3: (R,s,Q)
Periodic Review, Threshold, and Fixed Order Quantity
Policy

📆 When? Every R periods, if less than s

Static Every Friday, if I have less than 3 bottles of
📦 How Much? Q
milk, I buy a pack of 6.

🚚 Fixed, optimized order quantity (useful for Dynamic Every Friday, if I have less than three weeks’
operations, batch size, full truckload) worth of expected milk consumption, I buy a pack of 6.
Simple to group different SKUs in a single order
Pro

toward one supplier

📆 Fixed order timings can allow smooth operations
(collaboration with supplier).

🐢 Riskier policy with two ordering conditions (can

even be less reactive than a (R,S) policy). You can set
short review periods, but at the expense of grouped
Con

orders.
🔬 More difficult to optimize.

Copyright 16
Nicolas Vandeput
 Inventory Policies – Recap
✅Fixed order quantities ➡️ Packaging optimization

Inventory policies tell you when ✅ Fixed review periods ➡️ Grouped orders (but riskier)
and how much to order ✅ Dynamic ➡️ Future-proof

In practice, we don’t see

❌ Multiple conditions ➡️ Riskier, complex policies
real continuous systems

Replenishment Threshold Order Quantity Policy

Continuous Yes Fixed (s,Q) If less than s, order Q
Continuous Yes Up-to level (s,S) If less than s, order up to S
Periodic Yes Fixed (R,s,Q) Every R, if less than s, order Q
Periodic Yes Up-to level (R,s,S) Every R, if less than s, order up to S
Periodic No Fixed (R,S) Every R, order up to S

Copyright 17
Nicolas Vandeput
18
19
20
21
 Inventory Segmentation

Copyright 22
Nicolas Vandeput
 Inventory Segmentation
📦 Cycle stock Cs
Stock needed to fulfill the expected demand (or forecast) over a
replenishment cycle.

🛡️ Safety stock Ss
Stock to protect the supply chain against demand variability (or
forecast error) and supply lack of reliability.

🚚 In-transit stock Is
Inventory in-transit in between different locations. Usually,
these are goods ordered from a supplier but not yet available in
our warehouse for our clients to buy.

⏳ Backorders
Clients' orders waiting to be fulfilled

Copyright 23
Nicolas Vandeput
 Inventory Segmentation
Periodic Review with Up-to Level (R,S) Continuous with Fixed Order Quantity (s,Q)

Copyright 24
Nicolas Vandeput
 Inventory Segmentation
🍻 Beer Game 🍊 Fresh Connection

⏳
📦 Cycle Backorders

🛡️ Safety

📦 Cycle

🛡️ Safety
🚚 Transit

Copyright 25
Nicolas Vandeput
 Inventory Segmentation
On-hold Prebuild
Inventory booked by clients or waiting to be Stock needed to cover a special (marketing) event,
shipped to another location. season, or a lack of production capacity.
Excess stock Merchandising
Excess inventory compared to what is required Stock needed to fill the shelves (or to take more
(inventory targets). space in a store).
Dead stock Work-In-Progress
Goods that won’t sell based on the current forecast The stock of products currently under production is
(even in the long-term). called work-in-progress (WIP).

Copyright 26
Nicolas Vandeput
 Order-up-to level (S): 30

Policies In Practice Net inventory level: 0 (on-hand) -2 (backorders) +18 (in-transit)

= 16
Order = 14 (= Net Inventory – Order-up-to- level = 30 - 16)

I am following a periodic up-to-level

policy (R=1, S=30).
How much should I order?
(you need to account for incoming orders and backlog)

27
 Reorder point (s): 20
Net inventory level: 0 (on-hand) - 4 (backorders) + 20 (in-transit) = 16

Policies In Practice Order = 20

I am following a policy with a fixed

quantity (Q=20), a periodic review
(R=1), and a reorder point (s=20):
How much should I order?
(You can order multiple batches at once)

28
29
30
 Inventory Costs

Copyright
Nicolas Vandeput
 Supply Chain Costs

Purchasing costs Transaction costs Holding costs Expiration costs Shortage costs

Transaction costs

Purchasing costs Holding costs Shortage costs

Expiration costs

Copyright 32
Nicolas Vandeput
 Purchasing Costs
• Purchasing/production costs
(COGS - Cost of Goods Sold)
• Can include volume discounts
(scale effect)
• Value can be consistently
decreasing over time (think high-
tech)

The more you buy at once,

33
the lower the purchasing costs Copyright
Nicolas Vandeput
 Transaction Costs
Costs triggered by making an order (or a transaction) with a Hint: Most of the employees in a warehouse work due to
supplier (either internal or external). transactions (in/out flows) and not because of inventory

34 Copyright
Nicolas Vandeput
 Transaction Costs
Hint: Most of the employees in a warehouse work due to
Costs triggered by making an order (or a transaction) transactions (in/out flows) and not because of inventory
with a supplier (either internal or external).

Handling, picking, moving, shipping

Variable part (logistic/transportation costs)
• Supplier: Transportation*, packaging, picking
• Client: Inspection (≠ inventory control), warehouse
interims*, ….
Fixed part (ordering costs)
• (external) Supplier: Fixed fee, transportation*, time to
prepare an order*,…
• (internal) Supplier: Change-over time
• Client: Reception, working time (stock analysis, buying
process, negotiations, …), warehouse employees*, …

The more you buy at once,

35
the lower the transaction costs Copyright
Nicolas Vandeput
 Holding Costs
Costs related to storing (or possessing) products

Hint: costs related to leaving stock at night in a warehouse.

Hint: Most of the employees in a warehouse work due to

transactions (in/out flows) and not because of inventory
36 Copyright
Nicolas Vandeput
 Holding Costs
Costs related to storing (or possessing) products
Variable part
(more related to the products)
• Products’ cost of capital (financing cost)
• Products’ insurance
• Obsolescence (technology!) (≠ expiration costs)
• Damage, loss, thefts
• Inventory control (≠ inspections)
Fixed part
(more related to the warehouse)
• Warehouse’s cost of capital
• Warehouse’s insurance
• Storing equipment/infrastructure
• Software/IT (WMS)
Hint: costs related to leaving stock at night in a warehouse.
• Security
• Lighting, heating/cooling
Hint: Most of the employees in a warehouse work due to
transactions (in/out flows) and not because of inventory
The more you buy at once,
37
the higher the holding costs Copyright
Nicolas Vandeput
 Expiration Costs
• Costs related to shelf life
(expiration dates)

The more you buy at once,

38
the higher the expiration costs Copyright
Nicolas Vandeput
Shortage Costs
No more chocolate
ice cream.
Vanilla.
I’m out.

I’ll be back.

Oh, a waffle
truck.

39 Copyright
Nicolas Vandeput
 Shortage Costs
Costs incurred when demand exceeds No more chocolate
supply (inventory shortage) ice cream.
How clients react to shortages? Vanilla.
• Choose another product instead
• Come back later I’m out.
⚠️ Shortage
• Not buy anything
• Turn to the competition duration
• Cancel a bigger order
Cost of missing inventory I’ll be back.
• Goodwill & reputation (empty shelves!)
Rep
Backorder

• Penalty (based on Service Level Agreement) Oh, a waffle

Lost Sales

• Production stops truck.

• Emergency/expedition costs
• Loss of information/data?
• Margin/profits

? 40 Copyright
Nicolas Vandeput
 Costs
🍻 Beer Game 🍊 Fresh Connection

Copyright 41
Nicolas Vandeput
 Costs & Inventory

Purchasing costs Transaction costs Holding costs Expiration costs Shortage costs Every month, if I have less
than 1 week of stock,
I buy 1 month of stock.
(R,s,Q) =
Long review
(1 month, 1 week, 1 month)
period
?
Short review
period
Every day, if I have less
than 1 week of stock,
I buy 1 month of stock.

(R,s,Q) =
(1 day, 1 week, 1 month)

Copyright 42
Nicolas Vandeput
43
44
45
46
 Inventory Optimization

Copyright
Nicolas Vandeput
 Inventory Optimization
Science of optimizing inventory Logistic constraints
policies (when, how much,
Inventory Policy Objective
where) to maximize profits when, how much, where profits, costs, (ROI)
(optimal service level is a
byproduct). Supply Demand

Costs
Optimal inventory policy depends
on the supply chain environment
(supply, costs, and demand) and
logistic constraints*.
*Constraints: supply capacity (batch size), logistic capacity (number of trucks, warehouse 48
Blog: A Framework for Inventory Optimization volume), product constraints (shelf-life), financial constraints (budget, cash).
 Inventory Optimization
Average Demand
How much? How much? Demand Distribution How much?
When? When? When?
Average lead time Forecast
Supplier reliability Forecast error
Demand Distribution

Transaction costs
Supply Demand
Purchasing costs Holding costs Shortage costs
Expiration costs

Profit maximization

49
Blog: A Framework for Inventory Optimization