Transcript Inventory Management and Risk Pooling (1)

The Value of Information
Designing & Managing the Supply Chain
Chapter 4
Byung-Hyun Ha
[email protected]
Outline
 Barilla SpA
 Introduction
 The Bullwhip Effect
 Effective Forecast
 Information for the Coordination of Systems
Barilla SpA
 Introduction
 Barilla SpA is the world’s largest pasta manufacturer
 The company sells to a wide range of Italian retailers, primarily
through third party distributors
 During the late 1980s, Barilla suffered increasing operational
inefficiencies and cost penalties that resulted from large week-toweek variations in its distributors’ order patterns
Barilla SpA
 Distribution channels
Barilla SpA
 Weekly demand for Barilla dry products
Barilla SpA
 Demand fluctuations
 The extreme fluctuation is truly remarkable when one considers
the underlying aggregate demand for pasta in Italy
 Causes of demand fluctuations








Transportation discounts
Volume discount
Promotional activity
No minimum or maximum order quantities
Product proliferation
Long order lead times
Poor customer service rates
Poor communication
Barilla SpA
 Impact of demand fluctuation
 Inefficient production or excess finished goods inventory
 Utilization of central distribution is low
• Workers
• Equipment
 Transportation costs are higher than necessary
Barilla SpA
 Just-in-Time Distribution (JITD) proposal
 Decision-making authority for determining shipments from Barilla
to a distributor would transfer from the distributor to Barilla
 Rather than simply filling orders specified by the distributor,
Barilla would monitor the flow of its product through the
distributor’s warehouse, and then decide what to ship to the
distributor and when to ship it
 Evaluation of the proposal




JITD proposal as a mechanism for reducing these costs?
Why should this work?
How does it work?
What makes Barilla think that it can do a better job of
determining a good product/delivery sequence than its
distributors?
Barilla SpA
 Resistance from the Distributors
 “Managing stock is my job; I don’t need you to see my warehouse or my
figures.”
 “I could improve my inventory and service level myself if you would
deliver my orders more quickly; I would place my order and you would
deliver within 36 hours.”
 “We would be giving Barilla the power to push products into our
warehouse just so that Barilla can reduce its costs.”
 Resistance from Sales and Marketing
 “Our sales levels would flatten if we put this program in place.”
 “How can we get the trade to push Barilla product to retailers if we don’t
offer some sort of incentive?”
 “If space is freed up in our distributors’ warehouses…the distributors
would then push our competitors’ product more than ours.”
 “…the distribution organization is not yet ready to handle such a
sophisticated relationship.”
Introduction
 Value of Information
 “In modern supply chains, information replaces inventory”
• Why is this true?
• Why is this false?
 Information is always better than no information
 Information






Helps reduce variability
Helps improve forecasts
Enables coordination of systems and strategies
Improves customer service
Facilitates lead time reductions
Enables firms to react more quickly to changing market
conditions
Increasing Variability of Orders
 Lee, Padmanabhan, Wang (1997)
Bullwhip Effect
 Order variability is amplified up the supply chain;
upstream echelons face higher variability
 Main factors contributing to increase in variability
 Demand forecasting
 Lead time
 Promotional sales
• Forward buying
 Volume and transportation discounts
• Batching
 Inflated orders
• IBM Aptiva orders increased by 2-3 times when retailers thought
that IBM would be out of stock over Christmas
• Motorola cell phones
Impact of Promotional Sales
 Order pattern of a single color television model sold
by a large electronics manufacturer to one of its
accounts, a national retailer
order stream
Impact of Promotional Sales
POS Data After Removing Promotions
Point-of-sales Data-Original
Demand Forecasting & Lead Time
 Single retailer, single manufacturer
 Retailer observes customer demand, Dt
 Retailer orders qt from manufacturer
Dt
Retailer
qt
Manufacturer
L
 Suppose a P period moving average forecasting is
used
Var (q)
2 L 2 L2
 1
 2
Var ( D)
P P
Chen et al. 2000
Demand Forecasting & Lead Time
 Var(q)/Var(D) for various lead times
14
Var(q)
Var(D)
L=5
L=5
12
10
L=3
L=3
8
6
L=1
L=1
4
2
0
0
5
10
15
20
P
25
30
Demand Forecasting & Lead Time
 Multi-stage supply chains
 Stage i places order qi to stage i+1
 Li is lead time between stage i and i+1
qo=D
Retailer
Stage 1
q1
L1
q2
L2
Manufacturer
Stage 2
Supplier
Stage 3
 Centralized: each stage bases orders on retailer’s
2
k
k


forecast demand
2 L 2 L 
Var (q k )
 1
Var ( D)

i 1
P
i



i 1
i

P2
 Decentralized: each stage bases orders on previous
stage’s demand
2
k 
2 Li 2 Li 
Var (q k )
  1 
 2 
Var ( D ) i 1 
P
P 
Demand Forecasting & Lead Time
 Var(qk)/Var(D) with regard to stages
30
Var(qk) 25
Var(D)
Dec, k=5
20
15
10
5
0
Cen, k=5
Dec, k=3
Cen, k=3
k=1
0
5
15
10
P
20
25
The Bullwhip Effect
 Managerial insights
 Bullwhip effect exists, in part, due to the retailer’s need to
estimate the mean and variance of demand
 The increase in variability is an increasing function of the lead
time
 The more complicated the demand models and the forecasting
techniques, the greater the increase
 Centralized demand information can significantly reduce the
bullwhip effect, but will not eliminate it
Coping with the Bullwhip Effect
 Reduce uncertainty
 POS
 Sharing information
 Sharing forecasts and policies
 Reduce variability
 Eliminate promotions
 Year-round low pricing
 Reduce lead times
 EDI
 Cross docking
 Strategic partnerships
 Vendor managed inventory
 Data sharing
Information for Effective Forecasts
 Pricing, promotion, new products
 Different parties have this information
 Retailers may set pricing or promotion without telling distributor
 Distributor/Manufacturer might have new product or availability
information
 Collaborative Forecasting addresses these issues
 e.g. Wal-Mart’s Collaborative Planning, Forecasting, and
Replenishment (CPFR)
Information for Coordination of Systems
 Information is required to move from local to global
optimization
 Questions
 Who will optimize?
 How will savings be split?
 Information is needed





Production status and costs
Transportation availability and costs
Inventory information
Capacity information
Demand information
Locating Desired Products
 How can demand be met if products are not in
inventory?
 Locating products at other stores
 What about at other dealers?
 What level of customer service will be perceived?
Lead-Time Reduction
 Why?




Customer orders are filled quickly
Bullwhip effect is reduced
Forecasts are more accurate
Inventory levels are reduced
 How?
 EDI
 POS data leading to anticipating incoming orders
Information to Address Conflicts
 Lot size – inventory:
 Advanced manufacturing systems
 POS data for advance warnings
 Inventory – transportation:




Lead time reduction for batching
Information systems for combining shipments
Cross docking
Advanced DSS
 Lead time – transportation:
 Lower transportation costs
 Improved forecasting
 Lower order lead times
 Product variety – inventory:
 Delayed differentiation
 Cost – customer service:
 Transshipment