Content uploaded by Manmohan S. Sodhi
Author content
All content in this area was uploaded by Manmohan S. Sodhi on May 29, 2014
Content may be subject to copyright.
Content uploaded by Manmohan S. Sodhi
Author content
All content in this area was uploaded by Manmohan S. Sodhi on May 29, 2014
Content may be subject to copyright.
Managing Risk To Avoid
Supply-Chain Breakdown
FALL 2004 VOL.46 NO.1
REPRINT NUMBER 46109
Sunil Chopra and ManMohan S. Sodhi
Please note that gray areas reflect artwork that has
been intentionally removed. The substantive content
of the article appears as originally published.
FALL 2004 MIT SLOAN MANAGEMENT REVIEW 53
n March 17, 2000, lightning hit a power line in Albu-
querque, New Mexico. The strike caused a massive surge in the
surrounding electrical grid, which in turn started a fire at a
local plant owned by Royal Philips Electronics,N.V., damaging
millions of microchips. Scandinavian mobile-phone manufac-
turer Nokia Corp., a major customer of the plant, almost
immediately began switching its chip orders to other Philips
plants, as well as to other Japanese and American suppliers.
Thanks to its multiple-supplier strategy and responsiveness,
Nokia’s production suffered little during the crisis.
In contrast, Telefon AB L.M. Ericsson, another mobile-
phone customer of the Philips plant, employed a single-
sourcing policy.As a result, when the Philips plant shut down
after the fire, Ericsson had no other source of microchips,
which disrupted production for months. Ultimately, Ericsson
lost $400 million in sales.1(Ericsson has since implemented
new processes and tools for preventing such scenarios.2)
These two dramatically different outcomes from one event
demonstrate the importance of proactively managing supply-
chain risk. Supply-chain problems result from natural disasters,
labor disputes, supplier bankruptcy, acts of war and terrorism,
and other causes. They can seriously disrupt or delay material,
information and cash flows, any of which can damage sales,
increase costs — or both. Broadly categorized, potential supply-
chain risks include delays, disruptions, forecast inaccuracies, systems breakdowns, intellec-
tual property breaches, procurement failures,inventory problems and capacity issues. Each
category has its own drivers (see “Supply-Chain Risks and Their Drivers,”p. 54) and miti-
gation strategies (see “Assessing the Impact of Various Mitigation Strategies,”p. 55).
How a company fares against such threats depends on the type of disruption and the
organization’s level of preparedness. To prevent the kind of heavy sales losses suffered by
Ericsson after the Philips plant fire, managers must perform a delicate balancing act to
keep inventory, capacity and other elements at appropriate levels across the entire sup-
ply chain in a dynamic, fast-changing environment. Dell, Toyota, Motorola and other
leading manufacturers excel at identifying risks to their supply chains, and at creating
powerful mitigation strategies that neutralize potentially negative effects. With a clear
Managing Risk To Avoid
Supply-Chain Breakdown
Sunil Chopra is IBM Distinguished Professor of Operations Management and Information Sys-
tems at the Kellogg School of Management. Contact him at s-chopra@kellogg.northwestern.edu.
ManMohan S. Sodhi is associate professor of operations management at Cass Business School
in London. Contact him at m.sodhi@city.ac.uk.
O
By understanding the
variety and interconnect-
edness of supply-chain
risks, managers can tailor
balanced, effective risk-
reduction strategies for
their companies.
Sunil Chopra and
ManMohan S. Sodhi
54 MIT SLOAN MANAGEMENT REVIEW FALL 2004
understanding of the types of supply-chain risks, managers in
many types of industries can tailor effective risk-reduction
approaches to their own companies.
A Variety of Risks
Managing supply-chain risk is difficult because individual risks
are often interconnected. As a result, actions that mitigate one
risk can end up exacerbating another. Consider a lean supply
chain. While bare-bones inventory
levels decrease the impact of over-
forecasting demand, they simultane-
ously increase the impact of a supply
chain disruption. Similarly, actions
taken by any company in the supply-
chain can increase risk for any other
participating company.
Supply-chain risks can become
full-fledged supply-chain problems,
causing unanticipated changes in flow
due to disruptions or delays. Disrup-
tions can be frequent or infrequent;
short- or long-term; and cause prob-
lems for the affected organization(s),
ranging from minor to serious. A sim-
ple delay along the chain may create a
temporary risk, whereas a sole sup-
plier holding up a manufacturer to
force a price increase represents a
long-term risk. A machine breakdown
may have a relatively minor impact
on a manufacturing company with
redundant capacity, whereas a war
that disrupts shipping lanes can have a
major impact on a shipping company.
Most companies develop plans to
protect against recurrent, low-impact
risks in their supply chains. Many,
however, all but ignore high-impact,
low-likelihood risks. For instance, a
supplier with quality problems rep-
resents a common, recurrent dis-
ruption. Without much effort, the
customer can demand improvement
or find a substitute. In contrast, in
regions where earthquakes are rare,
preparedness to prevent major disrup-
tion may be weak or uneven.
Leading companies deal with this
range of supply-chain risks by hold-
ing reserves. Just as insurance compa-
nies hold cash reserves to meet claims, top manufacturers hold
supply-chain reserves that include excess inventory, excess capac-
ity and redundant suppliers. The big challenge for managers here:
Mitigate risk by intelligently positioning and sizing supply-chain
reserves without decreasing profits. So while stockpiling inven-
tory may shield a company against delivery delays by suppliers,
building reserves in an undisciplined fashion also drives up costs
and hurts the bottom line. The managers’role here is akin to that
Before companies can devise effective means of reducing supply-chain risks, managers
must first understand the universe of risk categories as well as the events and conditions
that drive them. Then, armed with clear, specific knowledge about these crucial risks,
companies can proceed to select and tailor mitigation strategies likely to be most effec-
tive. (See “Assessing the Impact of Various Mitigation Strategies.”)
Supply-Chain Risks and Their Drivers
Category of Risk
Disruptions
Delays
Systems
Forecast
Intellectual Property
Procurement
Receivables
Inventory
Capacity
Drivers of Risk
■Natural disaster
■Labor dispute
■Supplier bankruptcy
■War and terrorism
■Dependency on a single source of supply as well as the
capacity and responsiveness of alternative suppliers
■High capacity utilization at supply source
■Inflexibility of supply source
■Poor quality or yield at supply source
■Excessive handling due to border crossings or to change
in transportation modes
■Information infrastructure breakdown
■System integration or extensive systems networking
■E-commerce
■Inaccurate forecasts due to long lead times, seasonality,
product variety, short life cycles, small customer base
■“Bullwhip effect” or information distortion due to sales
promotions, incentives, lack of supply-chain visibility and
exaggeration of demand in times of product shortage
■Vertical integration of supply chain
■Global outsourcing and markets
■Exchange rate risk
■Percentage of a key component or raw material procured
from a single source
■Industrywide capacity utilization
■Long-term versus short-term contracts
■Number of customers
■Financial strength of customers
■Rate of product obsolescence
■Inventory holding cost
■Product value
■Demand and supply uncertainty
■Cost of capacity
■Capacity flexibility
FALL 2004 MIT SLOAN MANAGEMENT REVIEW 55
of a stock portfolio manager: Attain the highest achiev-
able profits (reward) for varying levels of supply-chain
risk and do so efficiently. (See “Choosing Supply-Chain
Risk/Reward Trade-Offs,”p. 56.) This means the manager
must seek additional profits for any level of risk protection
and preparedness or increase prevention and preparedness
without reducing profits. Success at this task requires a
good understanding of supply-chain risks and remedies,
both broad and tailored to the manager’s own company.
Delays Delays in material flows often occur when a sup-
plier, through high utilization or another cause of inflexi-
bility, cannot respond to changes in demand. Other
culprits include poor-quality output at supplier plants (or
at their suppliers’plants), high levels of handling or inspec-
tions during border crossings and changing transportation
modes during shipping. If material-flow delays are fre-
quent, however, companies can plan mitigation strategies
based on historical information.
Organizations can avoid delays, or at least prepare for
them, by appropriately and economically placing and sizing
their capacity and inventory reserves. One simple solution is
to maintain excess flexible capacity in existing plants. Toyota
Motor Corp.,for example, accomplishes this on its assembly
lines by employing team leaders who can work on any sta-
tion. Besides reducing the need for extra, station-specific workers
to cover absences, the strategy also ensures that daily production
goals are met, even when minor assembly-line problems occur.3
Another solution for delays is balancing capacity and inventory,
depending on the cost of the products. For example, telecommu-
nications equipment maker Cisco Systems Inc. has capacity to
assemble higher-value items in the United States. This enables the
company to respond quickly to orders from up-market domestic
customers. In contrast, Cisco holds an inventory of lower-value,
high-demand items produced in low-cost (but not very respon-
sive) locations overseas. Thus, by matching approach to product
value, Cisco reduces both supplier-delay risks and inventory costs.
Yet another solution is to combine inventory with different
transport modes. Dell Inc. holds very little inventory of high-value
components in the United States. Instead, the personal computer
manufacturer uses high-cost air transportation to deliver compo-
nents from the Far East as needed. For less expensive components,
however, Dell keeps some inventory that is shipped regularly at low
cost to the United States. In this way Dell, minimizes delay-related
risk as well as inventory-related costs.
Disruptions Disruptions to material flows anywhere in the supply
chain are unpredictable and rare but often quite damaging.
Examples abound of how natural disasters, labor strikes, fires and
terrorism have halted the flow of materials. For instance, follow-
ing a February 1997 fire at a parts factory owned by Japanese
manufacturer Aisin Seiki Co. Ltd., a key supplier for Toyota, the
auto giant was forced to temporarily shut down production at
most of its Japanese plants.4The negative impact can be interna-
tional: The 1994 Kobe earthquake in Japan, to cite just one case,
left California-based sound card maker Kelly Micro Systems and
many other small companies without any supply of parts.5The
California dockworkers strike in 2002 produced shortages of
high-demand retail items.6The 2001 bankruptcy of U.K.-based
UPF-Thompson, sole chassis supplier to Ford Motor Co.’s Land
Rover unit, caused major problems for the automaker.7Immedi-
ately after the attacks of September 11, 2001, U.S. auto manufac-
turers ran short of parts because transport trucks had been
delayed at the Canadian border.8In addition, supply disruptions
also can increase prices, as the Midwest discovered painfully in
August 2001, when regional gasoline prices skyrocketed follow-
ing a refinery fire at the height of summer demand.9
Companies can counter disruptions in material flow by
building inventory, or by having redundant suppliers (since it is
unlikely that all suppliers would be disrupted simultaneously).
However, holding inventory in this situation can get very costly.
The reason is simple: While holding costs are incurred contin-
ually,the inventory would be used only in the rare event of a dis-
ruption. In essence, the company pays (and continues to pay)
for reserves that may never be tapped. Still, building inventory
Decreases Risk
Greatly Decreases Risk
Greatly Increases Risk
Increases Risk
Mitigation strategy
Add capacity
Add inventory
Have redundant suppliers
Increase responsiveness
Increase flexibility
Aggregate or pool demand
Increase capability
Have more customer accounts
Disruptions
Delays
Forecast risk
Procurement risk
Receivables risk
Capacity risk
Inventory risk
Unfortunately, there is no silver-bullet strategy for protecting organi-
zational supply chains. Instead, managers need to know which mitiga-
tion strategy works best against a given risk. (Systems risk and
intellectual property risks are not included here.)
Assessing the Impact of Various Mitigation Strategies
56 MIT SLOAN MANAGEMENT REVIEW FALL 2004
does make sense if the disruption can be predicted with reason-
able confidence. In 2002, for example, many retailers selectively
built up inventories after learning of the impending California
dockworker’s strike. As a result, when supply was disrupted,
as predicted, damage was minimal. Stockpiling inventory as a
hedge against disruption also makes sense for commodity prod-
ucts with low holding costs and no danger of obsolescence. The
large petroleum reserve kept by the United States is a perfect
example of this strategy.
For products with high holding costs and/or a high rate of
obsolescence, using redundant suppliers is a better strategy.
Motorola Inc., for example, buys many of its handset compo-
nents from multiple vendors. Doing so prepares the company for
disruptions without building up fast-depreciating inventory.
Motorola lowers the cost of redundancy by using multiple sup-
pliers for high-volume products and single sourcing for low-vol-
ume products. This approach helps the company lower the risk of
disruption while preserving economies of scale at its suppliers.
Systems Risk The more a company networks its information sys-
tems, the greater the threat that a failure anywhere can cause failure
everywhere. Although rare, a breakdown of information infrastruc-
ture can devastate today’s highly networked environments. Case in
point: the “Love Bug”computer virus. In 2002, the fast-spreading
infection shut down e-mail at the Pentagon, NASA and Ford,
among others, causing billions of dollars in estimated damages.10
The banking industry has long recognized systems risk as a
major threat. In 1988, the Basel Committee on Banking Supervi-
sion warned about the growing reliance on globally integrated
systems. “The greater use of more highly automated technology
has the potential to transform risks from manual processing
errors to system failure risks,”the committee wrote.11
The best defense against systems failure? Robust backup sys-
tems and well-designed, well-communicated recovery processes
that duplicate all data and transactions. Such approaches helped
securities firms recover quickly and convincingly following the
World Trade Center attacks in 2001.12
Forecast Risk Forecast risk results from a mismatch between a
company’s projections and actual demand. If forecasts are too
low, products might not be available to sell. Forecasts that are too
high result in excess inventories and, inevitably, price mark-
downs. Long lead times, seasonal demand, high product variety
and smaller product life cycles all increase forecast error. Errors
tend to be greater when a few customers make larger purchases
(as opposed to many customers making smaller purchases).
Forecast inaccuracies can also result from information distor-
tion within the supply chain. In late 2003, for example, product
shortages in Western Europe led Nokia customers to order more
than they needed so they would be able to meet demand if Nokia
began rationing or allocations. Unfortunately, the exaggerated
figures distorted Nokia’s reading of the market, causing the com-
pany to inaccurately forecast sales.13
Other causes of information distortion include promotions
and incentives that lead to forward buying; batching of pur-
chases, which leads to higher volatility in orders; and lack of
knowledge of end-customer demand at upstream locations. Dis-
tortion increases in the supply chain as you get farther away from
the end consumer,a phenomenon known as the bullwhip effect.14
Companies can reduce the sting of the bullwhip effect,
though, by adjusting pricing and incentives to decrease variation
in orders. Increasing the visibility of demand information across
the supply chain also helps. Continuous replenishment programs
(CRP), and collaborative planning, forecasting and replenish-
ment (CPFR), and other supply-chain initiatives also can soften
the bullwhip effect.
Here again, forecast risk can be lessened by selectively hold-
ing inventory and/or building responsive production and deliv-
ery capacity. Holding inventory is appropriate for commodity
products with relatively low holding costs; responsive delivery is
better for expensive products with short life cycles (and corre-
sponding large forecast errors). Motorola practices responsive
delivery each day when it flies in phones from China in response
to demand by customer Nextel Communications Inc. Instead of
stocking parts for uncertain demand, Dell also flies in high-value
items from Asian suppliers on an as-needed basis.
X
B
A
Risk
Reward
High
Low
Low High
A higher-
efficiency
risk-return
curve
Current
position
Current
efficiency
risk-return
curve
Perhaps the biggest challenge companies face is mitigating
supply-chain risks without eroding profits. The manager’s role
here is similar to that of a stock portfolio manager: Achieve
the highest possible profits for varying levels of risk and do
so efficiently. In practice, this entails either (A) moving to a
higher level of efficiency by reducing risk while increasing
rewards, or (B) remaining at the current level of efficiency by
accepting reduced risk and reduced rewards.
Choosing Supply-Chain Risk/Reward Trade-Offs
FALL 2004 MIT SLOAN MANAGEMENT REVIEW 57
Intellectual Property Risk Intellectual property risk has grown rap-
idly as supply chains become less vertically integrated and more
global, and as companies outsource to the same manufacturers
used by competitors. Because profitability — and often business
models as well — depend on keeping a competitive edge, intel-
lectual property risk has dramatic, long-term implications.
Companies can mitigate intellectual property risk by bring-
ing, or keeping, some production in-house, or at least under
direct company control. That is a major reason why Motorola
owns some of the testing equipment at supplier locations. Man-
agers also can decrease risk by limiting the flow of new intellec-
tual property into countries with weak legal protections.
Companies like Cisco, which outsources all manufacturing, also
lower risk by creating business processes that cannot be easily
replicated by a single manufacturer. Electronics manufacturer
Sharp Corp. even repairs equipment itself, thus preventing any
possibility, accidental or otherwise, that its vendors will share
proprietary information with Sharp’s competitors. The company
goes so far as to reprogram various computer-aided machines
used by its vendors without sharing the information.
Procurement Risk Procurement risk refers to unanticipated
increases in acquisition costs resulting from fluctuating exchange
rates or supplier price hikes. For example, the recent weakening
Exploring “what if” scenarios like those below can help groups identify, understand and prioritize risks, a key prerequisite to tailoring
effective risk-mitigation strategies.
Stress Testing Your Supply Chain
Disruptions
Delays
Systems
Information
Processing
Intellectual
Property
Procurement
Receivables
Supplier-Related
■Supplier of a key part shuts
down plant for a month or at a
key part of the production cycle
■Supplier capacity drops by 20%
overnight
■Purchase orders of key parts or
raw materials delayed by month
■Supplier’s order-entry system
goes down for a week
■Supplier rations supplies by 20%
■Supplier increases minimum
order size by 20% then 100%
■Key supplier redesigns parts and
creates own product
■Supplier delays in processing
returns by twice as long
■Supplier forced to increase price
of key components by 20%
■Transportation costs go up 20%
overnight
Internal
■Key plant shuts down unexpect-
edly for one month
■Capacity at a key plant drops by
20% overnight
■Distribution or production
orders delayed by a month
■Key customer’s procurement
system inside your company
goes down for a week
■Company’s inventory/accounts
system goes down for a week
■To take advantage of volume
discounts, company begins to
order in quantities twice as
large as usual, but half as fre-
quently, which impacts sup-
plier’s ability to forecast
■Unforeseen cash squeeze
requires month-long delays in
paying key suppliers
Customer-Related
■Demand goes up by 20%
… for all products
… for a key product
… across the board
■Demand goes
down
by 20%
under conditions above
■Customer orders delayed by a
month
■Order entry system not working
for a week
■Key customer’s procurement
system inside your company
goes down for a week
■Credit card information stolen
from hacked e-commerce system
■Key customer begins to order
in batches that are twice as
large as usual but less frequent
(the impact of forecasting)
■Key customer withholds pay-
ments one month longer than
usual
■20% of receivable payments
delayed by one month
58 MIT SLOAN MANAGEMENT REVIEW FALL 2004
of the U.S. dollar drove up costs for U.S. companies sourcing in
Europe. The devaluation also created the risk of a dollar collapse,
the result of Asian economies propping up the currency to main-
tain exports to the United States. Interestingly, hedging against
the dollar’s fall also has led to higher oil prices, creating problems
for petrochemical and energy-intensive industries. Needless to
say, price increases are more likely when a company uses only one
supply source. Fortunately, managers have several sharp tools at
their disposal for minimizing price increases.
Exchange-rate risk can be countered by creating financial
hedges, balancing cost and revenue flows by region and build-
ing flexible global capacity. Toyota’s manufacturing strategy, to
cite one example, allows each plant to serve the local market
and at least one other market across the world. This lets Toyota
shift production if exchange rates change appreciably.
Price increases by suppliers can be blunted in several ways: by
signing long-term contracts, having redundant suppliers or, in
rare instances, holding inventory. But be warned that long-term
purchasing can badly damage profits if prices for the contracted
goods fall. For example, obligations signed by California during
the peak of its electricity crisis in 2001 forced the state to pay
800% more than the 2002 market price.15
Contracting with redundant suppliers can work, but only if
companies can maintain economies of scale. Global giants like
Toyota seek out local economies of scale by single-sourcing at the
plant level, but enlisting redundant suppliers globally. So even
though a company might be the sole supplier to a Toyota plant, it
must keep prices down to compete for business across the entire
Toyota network. Alternatively, some firms use multiple, redun-
dant suppliers, even if it means sacrificing some economies of
scale. Cisco, for one, claims to have four or five more suppliers
than it needs. The company keeps the resulting higher costs in
check by monitoring and benchmarking suppliers against each
other.16 A good example of using inventory to counter the threat
of price increases is the U.S. strategic oil reserve policy. Meant
primarily to prevent oil supply disruption, the reserve also has
been used on occasion to keep prices down. In another instance,
the International Paper Co. keeps prices of raw materials down
by sourcing them from independent forest owners, as well as
from its own forests.
Receivables Risk Receivables risk, the possibility of being unable
to collect on receivables, can torpedo the performance of any
company. In 2002, Sears Roebuck’s credit division reported
unexpected losses caused by delinquent cardholders.17 As a
result, Sears stock plummeted more than 30% in one day.
The company learned the hard way that filtering customers
for creditworthiness is a very prudent and powerful way to
reduce receivables risk.
Another approach is to spread the risk across
more customers. Elmhurst, Illinois-based
McMaster-Carr Supply Co., a maintenance-
materials supplier with hundreds of thousands
of customers, enjoys a much lower receivables
risk than a competitor selling to a single, large
customer.The Achilles heel here is a widespread
economic shock that harms the creditworthi-
ness of all customers, a fate that befell Cisco
during the dot-com bust.
Inventory Risk Excess inventory hurts financial
performance. That was the case in late 2000,
when the personal computer industry carried
roughly 12 weeks of inventory. The killer combi-
nation of excess inventory and falling prices hurt
many companies, notably Compaq Computer
Corp. Inventory risk hinges on three factors: the
value of the product, its rate of obsolescence and
uncertainty of demand and supply. As we have
seen, holding excess inventory for products with
high value or short life cycles can get expensive.
The strategy can work quite well, however, for
low-value commodity products that have low
obsolescence rates. To complicate matters even
Benefit of
Pooling
Reserve
Risk
Covered
High
Low
Low High
Cost of
Reserve
Risk
Covered
High
Low
Low High
Reserve
Required
for a Given
Level of Risk
Coverage
Extent
of Pooling
High
Low
Low High
Managers working to optimize
the cost of building a supply-chain
reserve against the level of risk pro-
tection must skillfully balance three
key relationships. The first relation-
ship shows the increasing cost of
risk reduction, which means that
using inventory to cover a
high
level
of demand risk proportionately
costs far more than doing so with
a
low
level of demand risk. The sec-
ond relationship shows that pooling
forecast risk, receivables risk or
other risk reduces the amount
of reserve required for a given level
of risk coverage. Thus, the required
level of inventory needed to miti-
gate forecast risk
decreases
as it is
pooled. The third relationship illus-
trates how the benefit of pooling
grows with the level of risk covered.
This means pooling inventory pro-
duces significant benefits only for
products with high forecast or
inventory risks.
Balancing Supply-Chain Risk/Reward Relationships
FALL 2004 MIT SLOAN MANAGEMENT REVIEW 59
further, inventory risk also increases as product variety grows.
Three time-tested approaches can help managers mitigate
inventory risk: (1) pooling inventory, (2) creating common com-
ponents across products and (3) postponing or delaying the last
stage of production until all orders are in hand. Online bookseller
Amazon.com serves all its customers in the United States with
inventory housed in a handful of warehouses. Book retailer Bor-
ders Books & Music supplies its customers with inventory in sev-
eral hundred stores. Each Amazon warehouse pools demand over
a large geographical area, leading to more stable forecasts and
lower total inventory. The strategy helps Amazon achieve 14
inventory turns per year, compared with two for Borders.
The paint industry illustrates well how to leverage component
commonality and “postponement”of variety. Traditionally, man-
ufacturers held paint inventory in a rainbow of different colors.
Today, paint inventory is held as a common base, which is then
mixed to exact color specifications after the customer orders.
This simple but powerful change has significantly lowered paint
inventory at retail stores. Apparel maker Benetton Group SpA
also practices pooling and postponement. An inventory of un-
dyed sweaters gets stockpiled in one location; coloring takes place
only after specific orders have been received. This pooling of
demand across geographical areas, and across colors, helps
Benetton greatly reduce inventory risk while more effectively
meeting customer demand.18
Companies can also minimize inventory risk by working
with a highly responsive supplier, especially for high-value,
short life-cycle products. Excess capacity can also lower the
amount of inventory required. For example, by running plants
at 80% utilization, Toyota can handle demand variation with-
out having to hold inventory.
Capacity Risk Unlike inventory, capacity can only be increased or
decreased over a period of time. Thus, building excess capacity
usually becomes a strategic choice. Excess (and thus, underuti-
lized) capacity hurts financial performance. That was the case in
2002–2003, when many semiconductor firms had to operate at
about 50% capacity because of soft demand.
Managers can lower excess capacity risks by making existing
capacity more flexible. Flexibility is a form of pooling that allows
use of the same capacity for a variety of products. For example,
plants owned by Japanese truck manufacturer Hino Motors Ltd.
employ multiple assembly lines on which the number of workers
determines line speed. This flexibility lets Hino change produc-
tion on any line by moving workers (capacity) to meet fluctuat-
ing demand. It also greatly reduces the excess capacity of workers
Hino would have to carry if each worked only on a specific line.
Toyota decreases risks from idle capacity by ensuring that each
plant is flexible enough to supply more than one market.
Demand fluctuations can be satisfied from a variety of plants,
which decreases the total capacity required. As noted, Toyota car-
ries the idea of flexibility down to the shop floor, where team
leaders can work on any station in the assembly line, reducing the
need for spare station-specific workers to cover absences.19
Lastly, a company can minimize excess capacity by serving
geographically scattered customers from the same location. Ital-
ian automaker Ferrari SpA, for example, minimizes total pro-
duction capacity by centralizing all car production in a single
plant. The arrangement also gives Ferrari economies of scale,
even though the company procures and produces much less
than the big auto companies.
What Managers Should Do
With so many related risks and risk-mitigation approaches to
consider, managers must do two things when they begin to con-
struct a supply-chain risk management strategy. First, they must
create a shared, organizationwide understanding of supply-chain
risk. Then they must determine how to adapt general risk-miti-
gation approaches to the circumstances of their particular com-
pany. Managers can achieve the former through stress testing and
the latter through tailoring.
Stress Testing Stress testing is a group exercise that helps man-
agers and their companies understand and prioritize supply-
chain risks. “What if”scenarios help key players focus on the
supply chain one link at a time. This strategy offers an especially
effective way to gain buy-in and shared ownership in project
teams tackling supply-chain risk.
The first step in stress testing is to identify key suppliers, cus-
tomers, plant capacity, distribution centers and shipping lanes.
Next, the team surveys locations and amounts of inventory repre-
sented by components, work-in-process and finished goods. Man-
Reduce
costs
Reduce
costs
Cost of Risk
Mitigating
Reserve
Pool
reserves
Mitigate
risk
Mitigate
risk
Level of Risk
High
Low
Low High
Decentralize
reserves
Pool
reserves
Decentralize
reserves
The application of three key risk/reward relationships in the
supply chain (see “Balancing Supply-Chain Risk/Reward Rela-
tionships”) suggests four general strategic approaches, depend-
ing upon the relative level of risk and the cost of mitigating it.
Rules of Thumb for Tailored Risk Management
60 MIT SLOAN MANAGEMENT REVIEW FALL 2004
agers then probe each potential source of risk,
which helps assess possible supply-chain
impacts as well as the company’s level of pre-
paredness. Facilitators ask questions such as,
“What might happen if a particular supplier
could not deliver for a month?”or “What if a
supplier raised prices by 20% at the termination
of a contract?”Questions pertaining to key cus-
tomers might include: “What if demand went
up or down by 20%?”or “What if a customer
delayed cash payment by a month?”(See “Stress
Testing Your Supply Chain,”p. 57.) When con-
sidering questions during stress testing, man-
agers should realize that figures such as
“20%”or “one month”are not sacred, but sim-
ply represent numbers large enough to be sig-
nificant and small enough to be realistic.
It is wise to position stress testing as a
“thought experiment”to help the company
prepare for unforeseen events, rather than
focusing on the debatable likelihood of such
events. Similarly, it is useful to frequently
remind people of the goal: preparing the sup-
ply chain for unforeseen events and greatly
lowering risk, both at the lowest possible cost.
Through stress testing, managers should be
able to identify risk-mitigation priorities for
the near, medium and long term. They will
have identified product families at risk, as well
as individual plants, shipping lanes, suppliers
or customers that could pose risks. Managers
will also have a clear idea of what risks might
have an impact on sales, procurement costs,
revenues, prices or even reputation.
Tailoring Risk Management Approaches Leading
companies mitigate risk by building various
forms of reserves, including inventory, capacity, redundant suppli-
ers and responsiveness. Managers must keep a vigilant eye on the
trade-off between the risk and the cost of building a reserve to mit-
igate it.20 Three key relationships influence this optimal balance.
(See “Balancing Supply-Chain Risk/Reward Relationships,”p. 58.)
The first relationship is the increasing cost of risk reduction.
This simply means that using inventory to cover a high level of
demand risk costs much more than covering a low level of risk.
The second relationship shows that pooling forecast risk, receiv-
ables risk or some other risk reduces the amount of reserve
required for a given level of risk coverage. Thus,the required level
of inventory needed to mitigate forecast risk decreases as it is
pooled. The third relationship shows how the benefit of pooling
grows with the level of risk covered: The benefit of pooling inven-
tory is great only if the product has high forecast or inventory risk.
Managers can balance these relationships to tailor their
response to risk with a surer grasp of extent and cost of reserve.
The following rules of thumb can be applied to tailor risk-miti-
gation strategies: When the cost of building a reserve is low,
reserves should be decentralized. When the cost is high, reserves
should be pooled. If the level of risk is low, focus on reducing
costs. If the risk is high, focus on risk mitigation. (See “Rules of
Thumb for Tailored Risk Management,”p. 59.) By tailoring
reserves for all risk-mitigation strategies, companies can maxi-
mize rewards for the same level of risk, or lower risks for the same
reward. (See “Tailoring Reserves for Risk Mitigation.”)
Once a company clearly understands its supply-chain risk, it can select the appro-
priate general mitigation approach and specific tailored strategy.
Tailoring Reserves for Risk Mitigation
Mitigation Approach
Increase Capacity
Acquire Redundant
Suppliers
Increase
Responsiveness
Increase Inventory
Increase Flexibility
Pool or Aggregate
Demand
Increase Capability
Tailored Strategies
■Focus on low-cost, decentralized capacity
for predictable demand.
■Build centralized capacity for unpredictable
demand. Increase decentralization as cost of
capacity drops.
■Favor more redundant supply for high-volume
products, less redundancy for low-volume
products.
■Centralize redundancy for low-volume prod-
ucts in a few flexible suppliers.
■Favor cost over responsiveness for commodity
products.
■Favor responsiveness over cost for short life-
cycle products.
■Decentralize inventory of predictable, lower-
value products.
■Centralize inventory of less predictable, higher-
value products.
■Favor cost over flexibility for predictable, high-
volume products.
■Favor flexibility for low-volume unpredictable
products.
■Centralize flexibility in a few locations if it is
expensive.
■Increase aggregation as unpredictability
grows.
■Prefer capability over cost for high-value,
high-risk products.
■Favor cost over capability for low-value com-
modity products.
■Centralize high capability in flexible source
if possible.
FALL 2004 MIT SLOAN MANAGEMENT REVIEW 61
Beyond reserve costs, companies must also consider product
volumes. Fast-moving standard items, with low margins and low
forecast risk, call for different reserves than slow-moving special
products with high margins and greater forecast risk. When plan-
ning capacity, managers should select an efficient, low-cost sup-
plier for fast-moving (low-risk) items. In contrast, a more
responsive supplier better suits slow-moving (high-risk and
high-value) items. For example, Cisco tailors its response by
manufacturing fast-moving products in specialized, inexpensive
but not-so-responsive Chinese plants. High-value, slow-moving
items are assembled in responsive, flexible (and more expensive)
U.S. plants. Sony Corp. also exploits this strategy, utilizing flexi-
ble but high-cost plants in Japan and low-cost but specialized
plants in Malaysia and China.
As much as possible, a specialized, decentralized approach
offers the best way to keep capacity for fast-moving, commodity
items with low forecast risk. Doing so should produce greater
responsiveness and lower transportation costs — but only if
doing so maintains adequate economies of scale. In contrast,
capacity for slow-moving, short life-cycle products with high fore-
cast risk should be made more flexible and centralized to pool
demand. This helps explain why automakers, for example, often
build specialized plants for fast-moving products in each major
market, but centralize the flexible production of high-end,
slower-selling models.
When capacity is expensive, managers can reduce supply-
chain costs by centralizing capacity to pool risk. As costs
decline, capacity must be decentralized further. Consider the
personal computer industry. PCs can be assembled to order in
two different ways. In one, the Dell model, capacity is central-
ized. In the other model, widely used in India, several compa-
nies sell component kits to local assemblers for assembly on
demand. Given the low cost of assembly capacity in India, it is
economical to decentralize capacity, even though this action
reduces pooling and increases the overall size of assembly
capacity across the supply chain. In contrast, given the higher
cost of capacity in the United States, centralizing buffer capac-
ity is more effective.
In addition to separating products with different risk charac-
teristics, managers must also consider separating capacity for the
low-risk and high-risk aspects of each product. Utility companies
use this strategy by employing low-cost, coal-fired power plants
to handle predictable base demand, and utilizing responsive but
high-cost gas- and oil-fired power plants to handle uncertain
peak demand. Similarly, Benetton produces the predictable base
load of its knit garments using a cheaper process that starts with
dyed thread and only produces the uncertain portion using the
more expensive process of dyeing knit garments.21
By continually stress testing their supply chains and tailoring
reserves, managers can protect and improve the bottom line in the
face of many types of supply-chain risks. Like Ericsson, smart com-
panies do not wait for lightning to strike twice before taking action.
REFERENCES
1. R. Eglin, “Can Suppliers Bring Down Your Firm?” Sunday Times
(London), Nov. 23, 2003, appointments sec., p. 6.
2. A. Norrman and U. Jansson, “Ericsson’s Proactive Supply Chain
Risk Management Approach After a Serious Sub-Supplier Accident,”
International Journal of Physical Distribution & Logistics Management
34, no. 5 (2004): 434-456.
3. K. Mishina, “Toyota Motor Manufacturing, U.S.A. Inc.,” Harvard
Business School case no. 9-693-019 (Boston: Harvard Business
School Publishing, Sept. 8, 1992).
4. M. Nakamoto, “Fire Hits Parts Supply Network at Toyota,” Financial
Times, Feb. 4, 1997, p. 34.
5. B. Johns, “Damage to Chip Makers Puts Sourcing in Spotlight,”
Journal of Commerce, Jan. 30, 1995, p. 1A.
6. A. Zimmerman, A. Merrick and O. Sook, “Retailers Scramble To
Keep Stores Stocked,” Wall Street Journal, Oct. 21, 2002, sec. B, p. 1.
7. T. Lester, “Inside Track: Making It Safe To Rely on a Single Partner,”
Financial Times, Apr. 1, 2002, p. 7.
8. M. Kane, “Inventory Controls Re-examined: Attacks Reveal Vulnerabil-
ity of Just-in-Time,” Cleveland Plain Dealer, Oct. 29, 2001, sec. C, p. 4.
9. J. Kohler, “Prices Jump at Gasoline Pumps in the Area,” St. Louis
Post-Dispatch, Aug. 29, 2001, sec. A, p. 1.
10. K. Maney and M.J. Zuckerman, “FBI Hunts ‘Love Bug’Source:
Damage From E-mail Source Cuts Across USA and Worldwide,” USA
Today, May 5, 2000, sec. A, p. 1.
11. Basel Committee on Banking Supervision, “Sound Practices for
the Management and Supervision of Operational Risk” (Basel,
Switzerland: Basel Committee Publications, Feb. 2003), 1.
12. J. Gillis, “Backup Systems Passed Trying Test; Despite Scale of
Destruction, Wall St. Data Largely Saved,” Washington Post, Sept. 27,
2001, sec. E, p. 1.
13. “Nokia Feels the Squeeze From Shortage,” Off the Record
Research, Nov. 13, 2003.
14. H.L. Lee, V. Padmanabhan and S. Whang, “The Bullwhip Effect in
Supply Chains,” Sloan Management Review 38 (spring 1997): 93-102.
15. “California May Have New Energy Deals,” New York Times, Aug.
8, 2002, sec. C, p. 12.
16. Author’s interview with Randy Pond, senior vice president, Cisco
Systems Inc., Nov. 2002.
17. C.L. Hays, “Sears Earnings Will Be Hurt by Credit Unit,” New York
Times, Oct. 18, 2002, sec. C, p. 1.
18. J.L. Heskett and S. Signorelli, Benetton (A), Harvard Business
School case no.9-685-014 (Boston: Harvard Business School Publish-
ing, 1989).
19. K. Mishina, “Toyota Motor Manufacturing, U.S.A. Inc.,” 4-5.
20. See the newsvendor problem in S. Chopra and P. Miendl, “Supply-
Chain Management: Strategy, Planning, and Operations,” 2nd ed.,
(Upper Saddle River, New Jersey, Prentice-Hall, 2003), 346-352.
21. J.L. Heskett and S. Signorelli, Benetton (A), Harvard Business School
case no. 9-685-014 (Boston: Harvard Business School Publishing, 1984).
Reprint 46109. For ordering information, see page 1.
Copyright Massachusetts Institute of Technology, 2004. All rights reserved.
PDFs ■Reprints ■Permission to Copy ■Back Issues
Electronic copies of MIT Sloan Management Review
articles as well as traditional reprints can be purchased on
our Web site: www.sloanreview.mit.edu or you may order
through our Business Service Center (9 a.m.-5 p.m. ET)
at the phone numbers listed below.
To reproduce or transmit one or more MIT Sloan
Management Review articles by electronic or mechanical
means
(including photocopying or archiving in any
information storage or retrieval system)
requires written
permission.
To request permission, use our Web site
(www.sloanreview.mit.edu), call or e-mail:
Toll-free
in U.S. and Canada: 877-727-7170
International: 617-253-7170
e-mail: smrpermissions@mit.edu
To request a free copy of our reprint catalog or order
a back issue of MIT Sloan Management Review,
please contact:
MIT Sloan Management Review
77 Massachusetts Ave., E60-100
Cambridge, MA 02139-4307
Toll-free
in U.S. and Canada: 877-727-7170
International: 617-253-7170
Fax: 617-258-9739
e-mail: smr-orders@mit.edu
Reproducedwithpermission of the copyright owner.Furtherreproductionprohibitedwithoutpermission.
