accenture the rise of scm quants

8/13/2019 Accenture the Rise of SCM Quants

1/20

The Rise of SCM QuantsHow Finance and Risk Management

are Transforming Supply Chain

Strategy and ExecutionCarlos Alvarenga


2/20

1

Our business model is one ofvery high risk: We dig a very bighole in the ground, spend threebillion dollars to build a factory

in it, which takes three years, toproduce technology we haventinvented yet, to run products wehavent designed yet, for marketswhich dont exist. We do that

two or three times a year.Paul Otellini, CEO, Intel


3/20

Its difficult, though certainly notimpossible, to find a product companytoday that does not incorporate someelements of the discipline known as supplychain management (SCM) in its strategic

thinking and operational execution. Fromlogistics network optimization to strategicsourcing strategies, SCM has transformedthe way entire industries operate, alongthe way enabling operational models thatwere unthinkable a few decades earlier.

This transformation, for some SCMexecutives, seems to have run its course.Their companies use sophisticatedadvanced planning systems (APS)to coordinate global operations andextensive collaborative infrastructures to

make and deliver product worldwide. Formany people, this is the summit of whatSCM is, and they see the next decade ofwork as one focused on perfecting thesemodels. For others still catching up tothese leaders, the goal is to reach eitherbest-in-class capabilities or at leastindustry parity in SCM.

These executives do not see muchbeyond the current landscape andimagine that perfecting their currentmodels is the order of the day. But is thatconclusion correct? Is SCM in a periodof consolidation or refinement asmany executives and analysts suggest?This paper argues that this conclusion isincomplete. Why? Because we are seeingthe birth of the third generation of SCM-- one focused on expanding techniquesdeveloped in the fields of Financeand Financial Risk Management andintegrating them into what can be calledfinancial SCM (FSCM).

FSCM starts with a radical question: Whatif SCM had been invented in Finance and

not in Logistics?To the theoreticians andpractitioners of FSCM, the answer to thisquestion leads to a radically differentview of the goals, strategies and practicesthat great and good SCM companiesshould understand and use. In this paper,we discuss the most important of these

ideas, and investigate how they can andare being applied to traditional SCMchallenges.

We also explore the pros and cons ofthese ideas, and present a brief practicalframework for applying FSCM that mayserve to generate a new debate betweenSCM and Finance executives aboutthese concepts and their applicability toexisting strategies and operations. Lastly,the paper argues that FSCM is a welcomeand valuable evolution in the discipline

and that, far from entering a staticperiod in its history, SCM is about to berevolutionized yet again by new leaders,ideas and techniques that daily move fromthe theoretical into the very real worldin which SCM has had such profoundimpact.

2

Put all your eggs in one basket andthen WATCH THAT BASKET.

Andy Grove, Former CEO, Intel


4/20

3

Plans, Trains andAutomobilesIn any discussion of SCM, its usually agood thing to define its boundaries. Forthe purposes of the ensuing discussion,we assume a broad definition of the term:one spanning from the initial definition ofthe business case for a products creationto the termination of that product,perhaps years or decades later. For such

a discussion the common Supply ChainOperations Reference (SCOR) model(with a slight modification to Deliver toaccount for post-sale service activitiescommon to many firms), can be usefulto present these activities. To review,the SCOR model is a process referencemodel developed by the managementconsulting firms PRTM and AMRResearch, and endorsed by the Supply-Chain Council as the cross-industry defacto standard diagnostic tool for supplychain management. SCOR, as the model

is known, enables users to address,improve, and communicate supplychain management practices within andbetween all interested parties in theExtended Enterprise. The SCOR modelspans from the suppliers supplier to thecustomers customer. SCOR is based onfive distinct management processes: Plan,Source, Make, Deliver, and Return.

Plan- Processes that balance aggregatedemand and supply to develop a courseof action which best meets sourcing,production, and delivery requirements.

Source- Processes that procure goodsand services to meet planned or actualdemand.

Make- Processes that transformproduct to a finished state to meetplanned or actual demand.

Deliver- Processes that provide finished

goods and services to meet plannedor actual demand, typically includingorder management, transportationmanagement, and distributionmanagement.

Return- Processes associated withreturning or receiving returned productsfor any reason. These processes extendinto post-delivery customer support.

Figure 1 illustrates these five areas asrelates to supply chain management.

One can argue that the discipline ofSCM, which ties all of these areastogether, has progressed through twomajor phases since its creation in the18th century during the IndustrialRevolution and its formalization in HenryFords repetitive production lines at thestart of the 20th century.

The first of these phases was the PhysicsPhase. During this phase, the aim of thestrategists was to optimize the physicalflow of materials and goods through aphysical world. There were ships to berouted, production lines to be fine-tunedand, later, waste to be minimized.

All of these areas had one thing incommon: they were real in the sensethat they were and are, of course tangible parts of the physical world.Consequently, the tools and techniques

that this SCM phase valued from maps,to time and motion studies to Lean1techniques were those that werefocused on and improved real things. Onecould argue that a very common featureof improvement during this SCM phasewas that change was visual in the sensethat often before and after pictureswere the best explanations of what hadbeen improved in the supply chain. Thisphase eventually became the disciplinenow called SCM in the early 1980s.

Figure 1: Main SCOR Process Areas

Define and forecast market demand Define/engineer product Calculate Product Business Case

1: Plan

Management Process Relationship to Supply Chain Management

Define production strategy (make/buy)

Define/engineer Manufacturing strategy/process Define material, supplier and after-sales strategies

2: Source

Execute make/buy strategy Ramp production from prototype to launch Deliver product to customers and partners

3: Make

Establish and maintain after-sales support Provide and manage return and warranty services Execute product evolutions/versioning End-of-life product and services

4: Deliver, and

5: Return


5/20

4

The second evolution of SCM, whichbegan in the 1970s, was the InformationPhase. SCMs focus began to change,from strictly physical in nature, toincorporating the data used to managethe physical supply chain activities. Therise of companies such as i2 Technologiesand the invention of APS broughttwo profound changes in SCM: (a)technologists and computer scientistsbegan to enter the field in great numbersand (b) a formal level of abstraction was

introduced over the physical world.Let us examine each of these changes.

The first change introduction oftechnologists to the field -- meant thatpeople whose basic training had not beenin logistics or production were now notonly accepted into SCM but, in somecases, actually took over the intellectualleadership of the discipline. This occurredin software companies, academicdepartments and product companies alike.The Physicists did not go away, but they

did have to learn the language of IT andirreversibly incorporate IT systems intotheir operations.

The second, and more profound, changemeant that it became acceptable for SCMteams to focus on abstractions of physicalproblems. For example, the creation ofmultiple demand simulations and lifecyclemodels became routine. Many companiescreated teams to model non-existentsupply chains in their quest to optimizethe real one.

These two changes informationoptimization and abstraction ofthe physical supply chain laid thegroundwork for the third evolution ofSCM: the Finance Phase or financialsupply chain management (FSCM). Thesethree phases are illustrated in Figure 2.

Though admittedly early in its trajectory,the third phase can be defined by thefollowing three characteristics:

Intense focus on quantitative methodsand tools borrowed from Finance andFinancial Risk Management

Intense focus on financially optimalsolutions to real and virtual SCMproblems that may or may not correspond to the answers Physicsor Information-driven solutionswould yield

Increased abstraction away from thephysical SCM world toward whateventually will become two parallelSCM worlds the physical and the

abstract world -- similar to that whichexists in the banking and financeindustries today.

This final bullet highlights an analogoustransformation that illuminates thispapers hypothesis. In banking one findstwo parallel worlds: a physical worldwhere people still walk into a bank,apply for a home loan and then carry acheckbook to a settlement office to buya house, and a financial world in whichanalysts and traders are making multi-billion dollar decisions on the likelihoodthose same people and millions likethem will ever repay that loan. Indeed,

this parallel world creates value throughlevels of abstractions that, to a layperson, may seem totally removed fromthe physical world and in many cases arenothing more than instruments that allowone to speculate on the probabilities ofprobability itself. Curiously if ominouslyfor many people the virtual bankingworld creates, and in some cases destroysmuch more value than the physical world.That trend increases every day.

Though it may be hard to see today, thesame phenomenon the laying of thetheoretical and applied foundations fora parallel SCM world is beginning todevelop in SCM. The rest of this paperexamines some of the most importantelements of this evolution and suggestsa basic framework for applying FSCMby companies that want to start movingtoward this level of sophistication.

Figure 2: The Evolution of Supply Chain Management

Define optimal financial structures (profits, tax offsets, trade free zones, etc.) and align physical assets

F nance Phase 2000-Onwards

Implement advanced planning systems to optimize planning, decision making and collaboration

Information Phase (1980/90s)

Optimize physical nodes type, location and internal operations to minimize production costs and inefficiencies

Physics Phase (Pre-1980)

1: Demand Definition 2: Supply Definition 3: Product Execution 4: Deliver and 5: Return


6/20

5

Against The Gods andOther Strange IdeasTalk to almost anyone involved in the areaof FSCM and one book, above others,seems to be common reading. This isPeter L. Bernsteins insightful analysis ofthe history of risk, Against the Gods: The

Remarkable Story of Risk2

.Bernsteins basicthesis: that once one understands goodand bad outcomes as the result of risk andnot supernatural forces, ones view of theworld (and its gods) changes completely.People may take things like fire insuranceand credit cards for granted but theyexist only because brilliant people movedaway from divine causes for everydayevents and began to think of good andbad outcomes as the result of quantifiable(and non-quantifiable) uncertainties thatare inherent in our world.

Moreover, Bernstein describes how onebrilliant mathematician after anotherpushed the ideas of uncertainty,probability, and utility forward -- eachone adding a critical element to whatwould become the modern science of riskmanagement. For example, speaking ofthe Swiss mathematician Daniel Bernoulli,Bernstein writes:

Every aspect of life includes SCM. Takefor example when a large cosmeticscompany launches a new perfume and

in doing so does not inform its suppliersthat its sales forecast could be off by asmuch as +/- 70 percent not unheardof in this industry. This absolute forecasterror in a new product launch might be aminor worry for a big cosmetics companybut a life or death problem for a smallbox supplier to that same company. Inother words, as Bernoulli postulated, thesolution to the same SCM problem what is the absolute right forecast hasdifferent value to the cosmetics companyand to the box supplier. SCM strategists,

of course, have understood this problemfor decades, and the Physicists haveresponded by creating smart inventoryplans or just-in-time productiontechniques. The Informationists havefocused on collaborative forecastingsystems with the perfume suppliers ora new demand planning system. Bothare valid responses and help to varyingdegrees though it must be said that inpractice most often they ignore the utilityvariance Bernoulli discovered.

The term utility variance refers to thefact that the same outcome can havedifferent value for different agents. Inother words, as Bernoulli postulated, thesolution to the same SCM problem what is the absolute right forecast hasdifferent value to the cosmetics companyand to the box supplier. Similarly, a giventransport lane may have different value

to a carrier, a coal shipper and a cellphone shipper.

But the cosmetics company exampledescribed above begs the followingquestion: How would an FSCM specialistlook at this problem?

Before discussing the answers to thisquestion, its worth noting another criticalprinciple Bernstein illuminates in his bookand which is fundamental to anyone whohas ever worked as a securities trader:risk is neither good nor bad in and of itself.

In other words, risk is indifferent andamoral and should neither be embracednor rejected in ignorance. Indeed, everycollege Finance student knows that risk isa necessary pre-condition of reward.

This basic lesson sometimes seems lost onotherwise sophisticated SCM strategistsand analysts, however. This is no academicpoint, as will be shown later, but one ofthe more important errors many SCMPhysicists and Informationists make whenthey look at risk in their supply chains.

We deal with this critique in more detailbelow, but for now, it is suff icient to takeaway two key SCM lessons from

Against the Gods. First, our understandingof risk was and is an evolutionaryprocess that is quantitative in nature-- abstracting real problems intomathematical forms for analysis andsolution. Secondly, risk is both necessaryfor reward and indifferent per seto eithergood or bad outcomes to the risk taker.

Now, returning to the perfume problem,

and keeping in mind that the value theutility of the error correction is verydifferent for the big perfume vendor andfor the small box manufacturer: Whatwould an FSCM solution to this problemadd to the existing solution set? Howwould an approach founded not on thephysics of SCM or on the informationflows around them, be any different?

He suggests a systematicapproach for determining howmuch each individual desiresmore over less: the desireis inversely proportionateto the quantity of goodspossessed. For the first timein history Bernoulli is applying

measurement to somethingthat cannot be counted. Hehas acted as go-between inthe wedding of intuition andmeasurement. Cardano, Pascal,and Fermat provided a methodfor figuring the risks in eachthrow of the dice, but Bernoulliintroduces us to the risk-taker the player who chooses how

much to bet or whether tobet at all. Bernoulli laid theintellectual groundwork formuch of what was to follow,not just in economics, but intheories about how peoplemake decisions in every aspectof life.3


7/20

6

Before addressing these queries, let usfirst understand the tools and techniquesFSCM specialists are bringing to bear onthis kind of problem. Let us also look atfrom where these tools arise and howthey impact SCM problems.

The Birth of

SCM QuantsAs noted earlier, FSCM has three basictenets:

Quantitative methods and toolsborrowed from Finance

Increased abstraction away from thephysical SCM world toward the parallelabstract world

Intense focus on financially optimalsolutions that can sometimes overridetraditional SCM solutions.

The people who are developing FSCMapproach the forecasting problemdifferently from traditional SCMpractitioners because the latter groupview problem and their goals differently.Before returning to the perfume caseabove, lets examine some of the fieldsfrom which FSCM Quants4 (to borrowa term from Finance) are emerging andwhat each class of specialists is bringingto the discussion.

Real OptionsWhile the field of accounting hascontributed specialists to SCM fordecades, their role has been thetraditional one of calculators of cost andprofit. Finance itself has, until recently,rarely interacted with SCM. As far as thisauthor is aware, there are no professorsof SCM Finance teaching at any majorbusiness school, and most MBAs whospecialize in SCM take only the standard

corporate Finance class in order tograduate. However, this situation ischanging and the change is coming fromboth academics and Finance practitioners.

In the academic world, the change hascentered on fields such as Real Optionsand Valuation, both of which have majorrelevance to SCM. To summarize, a realoption, as distinct from a financial option,is an option related to the physical world.For example, a firm may or may not builda plant, invest in a new production lineor buy a truckload of plastic. The real is

not the new plant or truckload of plastic,but the possibilityof making either thingreal. In other words, when one party sellsa real option to another, they do not sellthe object of that option but the rightto make that option real or not. Theoptions objects to buy or not buy, tobuild or not build exist in the physicalworld but the option itself is an abstractcreation with value that is different fromthe option object. Thus, the goal of realoption research which has its originsin the 1970s is to create analyses and

tools that allow real options to be createdand priced correctly.

Real options provide flexibility in avariety of ways in the real world. Someallow deferral of an action, some allowfor alterations of consumption orproduction, while others allow switchingfrom one consumption or productionmethod to another. In theory, real optionsgive the SCM strategist an almost infinitevariety of possible solutions to real worldSCM problems.

To return to our perfume case, letsapply the theory of real options to thesituation. Today, the large cosmeticscompany can either buy or not buy someminimum number of boxes to supportits new product launch. If it sets thebuy quantity at zero, because it is tooconcerned about possible forecast error,the small box company has zero revenue.

A real option specialist, however, wouldlook at that same situation and say thatanother alternative exists i.e., the

creation of an option to buy the boxproduction without actually buying anyboxes. Such a real option, once created,would provide the perfume maker withoptional capacity to support its launchat a much lower cost than the traditionaland riskier minimum buy that wouldhave to be made today. It would also givethe box maker more revenue than it wouldget if the perfume maker decides that apossible 70 percent forecast error is toohigh to support a new product launch.

Such a scenario may seem fanciful, butin reality it is not. Real options havealready entered the production world.A high tech company, for example,buys options on engineering hoursfrom small engineering companieswhen it launches certain new products,where the key constraint to meetingany higher than expected demand is

not physical production capacity butengineering hours spent adjusting itshit products for different countries.5

Operations ResearchUnlike Finance, Operations Research(commonly known as OR) is a f ieldthat has long had close ties with SCM.Many of the mathematical engines thatlie at the heart of advanced planningsystems, for example, were created byOR specialists. However, a group of OR

researchers have of late turned theirattention to the incorporation of realoptions and risk modeling techniques suchas Monte Carlo simulation to productionand SCM problems.6

In their work, OR specialists look at anSCM problem like the perfume case aboveand apply their techniques, for example,to the production planning aspects of thissituation. In this case, the Physicistswould do all they could to lean theperfume supply chain and prepare it to

stop as quickly as possible should theproduct be launched and then prove to bea failure 60 days later. An Informationistapproach would focus on improvingthe forecast error itself, perhaps by anincreased round of focus groups withpotential customers around the worldor a series of collaborative planningsessions with major department stores.


8/20

7

An OR specialists, however, would gobeyond the static statement that there isa possibility of a 70 percent forecast errorand attempt, perhaps using mathematicalmethods, to simulate the possible demandoutcomes and define an optimal capacitystrategy vis--vis two possible outcomes:that which is most likely to happen andthat which would be most catastrophic .In other words, the OR strategist wouldprepare the perfume maker for whatdemand should be, given a mathematical

combination not just of past sales historybut of the impact of randomness on thatsales history (the Monte Carlo part) aswell as for the possible impact of theworst case scenario (if the forecast is offby 100 or more percent, for example).Again, this may seem like an academicscenario but this technique, too, is alreadyin use. Several Manufacturing firmshave created integrated FinanceORteams working on SCM/manufacturingproblems. They have adopted Monte Carlosimulation for capacity and production

planning projects that can betteralign capacity decisions with demandprobabilities to a degree that has notbeen possible with traditional forecast-driven capacity calculation methods.

Risk ManagementThe third and most radical fielddriving the FSCM evolution isFinancial Risk Management. Italso is the most misunderstoodfield in SCM literature today.

There is no current shortage of articlesand seminars addressing the topic of SCMrisk. The great majority of them sharea major problems, namely that they allapproach SCM risk from the Physicistpoint of view and they treat risk as abad thing to be avoided. Because themajority of the authors and consultantsworking on SCM risk are Physicists bytraining or inclination, they are limitedin their view of SCM risk by the physicalworld. To them SCM risk managementis about real SCM elements such asboat in Shanghai that can sink or a plantthat can be destroyed by a tornado.7

This phenomenon is best illustrated inan interesting study conducted by ARC

Research8, in which SCM managerswere asked to rate the risk that mostconcerned them and only two non-physical risks (government regulationand currency fluctuation) were listed.Their responses are presented in Figure 3.

The authors note that SCM managerswould see any events that mightdisrupt reliable availability of materialsas serious threats. Poor forecasts ofnew products and promotions areprobably a similar issue. Forecasts can beunderestimated or poorly communicated,leading to product shortages that reflectpoorly on supply chain managers.10

If this view existed in modern Finance,then risk management departments onWall Street would all be outside making

sure no robbers get into the building.This is hardly the case because, as notedearlier, in modern Finance risk can be bothreal (bank collapse), directly derivative(securities backed by revenues from thatbank) and indirectly derivative (derivativeson the revenues held by parties with nodirect relationship to the bank or its directrevenues). The second problem, againrelated to their training and inclinations,leads them to talk consistently about riskin negative terms. As noted earlier, inFinance risk is a good thing as well as bad.Risk creates the opportunity for reward.

IT breakdown

Loss of important customer

Government regulations

Currency fluctuations

Terrorism

Emergence of new competitor

Labor strikes

Political unrest

Natural disaster

0% 5% 10% 15% 20% 25% 30%

29.2%

25%

20.8%

20.8%

20.8%

12.5%

8.3%

8.3%

4.2%

Figure 3: What Macro Level Events Would Pose a High Risk to Your

Financial Results?9


9/20

8

The job of the risk manager is not toeliminate risk completely, becausethat would also eliminate rewards,but to manage it. What does it meanto manage risk? It means that riskcan only be handled, as the ARCauthors correctly note at the start oftheir analysis, in one of four ways:

Assumed

Avoided

Transferred

Hedged

However, most of SCM risk literaturetoday deals with only the first andsecond options. What about the third andfourth? Can SCM risk really be transferredor hedged? Most top SCM companiesunderstand what risk avoidance andassumption are, and some are even adeptat a kind of basic transfer mechanism

of it through bi-lateral sourcing oroutsourcing arrangements, but very fewunderstand the last, and most powerful,possibility. This paper deals with thislast method in more detail below but itis suff icient to note that the third andfourth options, more than the first two,will transform SCM risk managementin the coming decades much as it didmodern Finance. Again, this is no purelyacademic discussion. Already, a handful ofleading SCM companies in more than oneindustry use commodity hedging to hedgeor transfer supply (and some elementsof production) risk, and while these arecomplex technical strategies, they willbecome more common as their applicationis simplified and their value is appreciated.

FinancialOptimizationA fourth field currently receiving alot of attention in FSCM is financialoptimization, which refers to theoptimization of tax or profitability-relatedfinancial flows in a given product supply

chain. While this idea is not new, whatis new is the arrival of computationaland modeling tools that finally enablea systematic and repeatable processfor FO. Several companies have createdsoftware tools that allow for the modelingand optimization of complex productflows and their associated financialperformance that aim to optimizenot SCM physical performance in thetraditional sense but FSCM performanceby reducing taxable exposures andmaximizing total SCM profitability.

SCM strategists using these tools havedeveloped the ability to rapidly buildmodels of SCM and product-levelfinancial performance and then optimizethose models against goals like taxminimization or Return on Net Assets(RONA) or Financial Contribution. Indeed,often working alongside traditionalaccounting-advisory firms, these teamsare at the vanguard of merging SCMand CFO functions in real, workingenvironments across the world.

Indeed, in the next section, thispaper will present in more detailsome of these specific techniquesof FSCM that are in use today andwhich will become more widespreadas the FSCM evolution continues.


10/20

9

FSCM In Action: Four ExamplesReal Options:How Boeing UsesReal Options in

Product ProductionDecision-makingThe commercial aircraft industry ischaracterized by several exceedinglycomplex factors that SCM managersmust consider when makingdecisions about product developmentschedules, technology investmentsand strategic sourcing. One majoraircraft manufacturer has been at the

forefront of incorporating the followingFinance techniques into these kindsof real world operations decisions:

Quantitative technical risk modeling

Investment and risk modeling metrics

Real options

Demand (price and quantity) modeling

Learning curve optimization(maximize net profits by balancingrecurring and non-recurring costs)

Multi-gate and stagedinvestment processes

Structured spreadsheetmodeling architecture

Portfolio analysis

Indeed, this company has recently becomea source of patent applications on certainmathematical techniques it developedduring the course of its early adoptionof FSCM. The most public example ofthis work has been their incorporation

of real option theory into decisionsabout product development and risk-sharing on new product launches.

In an interview, the leader of theFSCM team notes that his modelsfocus on the following goals:

Projecting the highest riskvariables price and quantity

Determining project profitabilityin an uncertain market

Capturing variability of future

production and cash flows

Providing forecasts even whenmarket data is sparse

Designing products and servicesthat help customers manage theirown market and operating risks

This companys approach impacts realworld decisions as follows. Take forexample, the decision to sub-contracta sub-assembly in an aircraft to amanufacturer in Germany a decision we

label Project 1. Traditional approaches tothis decision would, in essence, determinea quantity, unit price and deliveryschedule. Furthermore, investment andpayment terms would be set and executedon a temporal basis as production wasstarted and sub-assemblies delivered.

This MFG company, however, approachesthe problem differently in that it seeksto (a) understand the risk inherent inthis outsourcing decision in a specific,mathematical model and (a) apply that

risk to issues such as contract andpayment terms, flexibility models forboth buyer and seller, and a continuousadaptation of the risk profile of boththe value of Project 1 and the options/flexibility applied to it throughout itslifecycle. The mathematics for suchan approach are no mean feat and.Indeed, the technique developed atthis company has allowed them tosimplify the valuation process so as tomake incorporating options techniqueseasier in operational problems.11

Furthermore, the team leader regularlygives multi-day seminars to productmanagers that explain real option theoryand the companys own techniques. Theteam leader, who is a robotics engineerby training, also notes that the modelsthat he creates are both engineering andfinancial models. They include variablesof aircraft design and financial design,and they can examine each variablesoption value in order to make better

strategic decisions about the finalproduct. For example, he notes, thecompany can reduce our unit costs byspending more money on productionfacilities, by say automation. Wherethey spend their money depends on

what option values they have at anystage of an aircraft design. They cantrade off certain design features thatmay be technically more challengingagainst the possibility of selling moreunits by lowering their unit price.

Perhaps the biggest benefit to thisaerospace manufacturer of this newapproach is that thinking about designand production problems from the pointof view of risk and probability has helpedBoeing overcome a natural conservativestreak in its corporate culture. In otherwords, understanding risk has helpedthem make more strategic decisions,incorporating into those decisions notjust deterministic scenarios i.e., thosethat do not account for randomnessin demand -- that over time lose theirrelationship to the real world but adynamic, constantly revalued perspectiveof the competitive landscape in which itsaircraft and customers must operate.

OperationsResearch: HowIntel Uses MonteCarlo Simulation inCapacity PlanningCapacity modeling is a complex andnecessary part of most SCM strategy.Traditionally, this was a purely Physicsproblem, in that the goal of thestrategist was to match as closely aspossible product demand and productioncapacity. Traditional methods calculateddemand for period X and mapped itagainst capacity for the same period,then adjusted capacity up or down toreach an efficient level of productioncapability. This approach has severalshortcomings, not the least of which isthat the forecasts that typically drivecapacity requirement analyses are usually


11/20

10

deterministic. Because they tend totreat capacity as static capability i.e.,production levels are averaged overa given time period, with temporaryvariations in capacity typically aresmoothed out in longer term models.

A leading global semi-conductormanufacturer has begun to addressthese problems by applying Monte Carlosimulation to the capacity planningquestions in what it calls its a NextGeneration Capacity Model (NGCM).Without diving into the mathematicsof their techniques, the companysapproach follows four steps:

Identify the critical capacity constraint(CCR) using techniques based onthe theory of constraints12

Introduce a stochastic, i.e., random,element in its demand forecasts using

statistical techniques to define peakdemand days and standard deviations ofhistorical demand by peak demand days

Apply the new, risk-adjusted demandpatterns to the CCR, which in thiscase turned out to be warehouseoperations throughput, to determinethe new risk-adjusted stress/demand-strength/capacity model

Conduct a safety factor analysis todetermine the probability of failurein the new stress-strength model at

the facility and processing line levels

The results of this approach, accordingto one of their senior productionengineers, are several. As he notes ina paper describing their approach13:

Until the development of NGCM, it wasnot possible to provide capacity forecastrecommendations with correspondinglevels of certainty. This made it difficult todetermine the risks associated with bothtactical and strategic capacity decisions.

With NGCM, such capability now exists.And while this model was created toaddress warehouse capacity forecastingchallenges, the methods articulatedhere are applicable to other situations,especially those where TOC/DBR14arein use. Operations with fluctuatingdemand and variable processes willbenefit from the application of thiscapacity model concept as they seek toimprove cycle-time performance andreduce work in process inventory.

In an interview, the engineer adds that theincreased level of sophistication of therisk techniques being adapted to SCM ischanging the role of SCM at his company.As he notes, manufacturing drives thecompany but SCM is really being seenas a driver of competitive advantage.

Risk Management:Commodity Hedgingat Southwest,Lufthansa, GE and HPPerhaps the most publicized FSCMexample is the use of hedging strategiesin commodity management. A hedge,simply put, is a cost borne in order toreduce the (negative) risk of anotherexisting position. There are manykinds of hedges but some of the mostcommon are financial options such ascalls (the right to buy something) andputs (the right to sell something). Manycommodities have markets where optionson those commodities are traded; onesuch commodity is aviation fuel, thesecond-most important direct cost (afterlabor) in the aviation supply chain.

As has been widely reported in thebusiness and popular press, the majorityof Procurement departments at U.S.airlines have, in recent years, not usedcommodity hedging strategies to lockin future price protection against thepossibility that aviation fuel prices wouldsignificantly increase. As a consequence,as with car manufacturers and steel,U.S. airlines have been devastated inthe last year as petroleum prices havedramatically increased and they havefound themselves forced to buy fuelon spot markets. The consequences ofthis exposure have been severe: huge

losses, layoffs, and route reductions allat a time when the industry were juststarting to show some profitability afterseveral years of terrible performance.

Not all airline Procurement teams leftthemselves un-hedged against this risk,for at least one major U.S. airline and afew European carriers adopted commodityhedging strategies, and in doing so,illustrate FSCM concepts in action. Forexample, as one article noted, Southwestwas 100 percent hedged for first-quarter

2007, capped at an average crude-equivalent price of about $50 a barrelThe locked-in cost is the reason whySouthwests fuel bill is projected to go uponly about $500 million this year [2007],one quarter that of US Airways increase,despite a larger fleet.15 A 2004 academicpaper that analyzed the Southwesthedging program contains an interesting

quote from Southwests Director ofCorporate Finance, Dave Carter, whoclaims that, If we dont hedge jet fuelprice risk, we are speculating. It is ourfiduciary duty to try and hedge this risk.1

Carters comment gets to heart of whatsome FSCM theorists and practitionersstress about their field: that it is not justa good thing to do but an actual fiduciaryresponsibility to examine and employ,to the degree possible, certain aspectsof FSCM. This is a statement not made

lightly, for it has huge ramifications forcorporate governance, which are outsidethe scope of this paper. Nonetheless, itshard to argue that if you are a shareholdeat a Southwest competitor, and you knowthat these techniques were available toyour Procurement and Finance teams,you certainly would like to know whythey were not exercised. This positionwould be made all the more urgent bythe fact that in Europe, several otherairlines deployed the same techniquesto more or less the same effect.

Lufthansa, for example, hedged 83percent of its fuel requirements throughthe end of this year [2008] and saidthat it saved 109 million Euros lastyear through the practice [of hedging],common with European carriers.17As the New York Times notes:

The ability to lock into fixed fuel pricesmonths ahead of time called hedging can help offset these rising prices.But with the exception of Southwest

Airlines, most United States airlines areless hedged than European ones. AirFrance-KLM has hedged 78 percent ofits fuel consumption through March2009, at $70 to $80 a barrel, Jean-Cyril Spinetta, the chairman and chiefexecutive of the airline, said last month.Through a policy of hedging fuel fouryears in advance, the company savedabout $35 a barrel when oil was at $120a barrel. Other European airlines havealso taken the hedging route. British


12/20

11

Airways hedged 72 percent of its fuelneeds for the first half of the financialyear and 60 percent for the second half.Lufthansa has hedged 83 percent of itsfuel requirements through the end of2008 and said that it saved 109 millionEuros ($169 million), last year by doing so.

Even low-cost carriers like Air Berlin,

EasyJet and Ryanair are hedging,with Ryanair recently reversing a

longstanding avowal never to do so.

The airlines are not alone in these

advanced procurement techniques;

companies such as GE have also adopted

complex strategies for managing sourcing

risk. In 2005, GEs Global Research and

Development Center worked with a small

team of business managers to create

an arithmetic simulator to analyze a

variety of different scenarios for natural

gas hedging. To hedge its exposure, GEemploys a portfolio approach, relying

on swaps18, call options and limit orders

of varying forms.19Similarly, since the

1990s Hewlett-Packard has used a

risk management team to analyze and

manage growing portions of commodity

purchasing. In 2002, Corey Billington,

Blake Johnson, Alex Triantis described the

HP approach in a prescient paper they

published in years before anyone grasped

the power of their pioneering approach:20

To begin with, notes a paper that

highlights HPs work, a systematic

analytic process was developed to

optimize HPs value-risk objectives for

the procurement portfolio. Proprietary

software (HPRisk) is now used to

examine the impact of different mixes

of fixed and flexible quantity contracts

(each of which may have caps, floors,

or other structured price features) on

expected cost and percentile cost levels.

A scenario forecasting module feeds intothe optimization program to provide

distributions of demand, availability,

and price, and these are updated

through time based on newly collected

data. HPs web-enabled software also

supports the valuation, monitoring,

and management of its contracts.21

When airlines hedge their principal

raw material-- aviation fuel and

companies such as GE and HP do the

same with gas, copper, steel and other

basic commodities they demonstrate

the FSCMs tenet of solving a physical

problem availability of a raw material

with a non-physical f inancial solution.

These examples are simple in concept

yet complex in execution, of course; andhad fuel not increased so dramatically,

Southwest, Lufthansa and the others

might have been criticized for making

unnecessary expenditures in exotic

risk instruments. After all, fuel hedging

has existed without much attention in

one way or another in the U.S. aviation

industry since it was deregulated in 1978.

Financial

Optimization inHigh-Tech ConsumerElectronicsThe typical consumer electronics

manufacturing company has contract

manufacturers making products for it

(mostly in Asia) and sells those products

through a variety of channels all over

the world. When these companies launch

a new product, they typically rely onPhysics-based SCM modeling to map

out physical flows and distribution

nodes. The tax implications of these

product flows were a secondary or

tertiary consideration at best.

This changed for one major consumer

electronics manufacturer when it

purchased software that allowed it to

model its products flows from creation to

sales, and also model the cost, financial

and tax flows of its supply chain. In amulti-year effort that should serve as a

benchmark for others in the industry, this

company saved over US $2 billion through

tax optimization and cost avoidance.

This case offers two key lessons. First, it

took collaboration between the companys

SCM and Finance teams to create the

data sources and implementation methods

to achieve the results noted. Second, the

optimized FSCM flow is not always the

same as the optimized SCM flow. Indeed,

as one of the founders of the software

company explained in an interview, that

everyone understands the Physics of theseproblems. But getting the financial flows

right requires a real philosophical change:

one has to separate the two problems, be

willing to over-ride the physical answer,

and trust that the financial models are

accurate optimization environments.

This raises the issue as to what

consideration ultimately will drive

the final SCM model adopted by

companies. When that model is

the FSCM model when a Financesolution overrides a Physics solution

then, we see in action the evolution

that this paper has described.

It is not too far a stretch for SCM

executives, with their typically global

perspectives, to understand how

taxation and product contribution

could change their physically optimized

SCM models. Most will admit that,

if tax and profitability were fully

understood, their supply chains wouldlook a lot different than they do today.

There are other finance-based

techniques, though more complex

and more revolutionary, that will

have an even greater impact on SCM.

While still theoretical, they are worth

noting because they suggest what

the third evolution of SCM could

look like in the coming decades.


13/20

12

At the EfficientFrontier: Derivatives,Synthetics and theFuture of FSCMIn order to move beyond the currentlyapplied level of FSCM, its necessary toleave behind the constraints of traditionalphysical and even information-based SCM

techniques and look at what is possiblefrom a completely different perspective.Consider for example, a company thatmakes grass seed. In an interview withthe CFO of this company, he causallypoints out that their grass seed demand isusually pretty stable. Grass seed is grassseed, he jokes and its hardly an impulsebuy after all. Further conversation,however, about the factors that drivethe demand for his products leads tothe observation that demand for grassseed happens to correlate strongly with

temperature. When the mean temperaturein his customers cities is within the rangeof 55 to 85, the company sells lots ofproduct. When temperature is not withinthat range either above or below salesdrop of linearly until a certain pointin both directions, at which time theydisappear completely. That demand curveis illustrated in Figure 4.

This makes sense intuitively of course.Few people buy a lot of grass seed whenthe ground is so cold the seed cantpenetrate the soil. Nor will they buy alot if seed when the temperature is veryhot. Clearly, the grass seed company ishurt equally by temperatures that aretoo high and too low. A fatalistic SCMmanager either accepts that his businessis governed by the random luck ofweather (in this case temperature), whileanother more sophisticated manager may

try to lessen the impact of this weatherrisk by better forecasting or outsourcingproduction.

In contrast, an FSCM strategist wouldgo far beyond these common responsesand could among other possibilities decide to use what are known as weatherderivatives to tackle this problem.22Afterall, lots of industries from farming totractors to heating oil -- are affectedand there is exist markets where one canbuy weather hedges, which when usedcorrectly can act as a kind of weatherinsurance. The scope of this article doesnot allow a discussion of the mathematicsinvolved, but suffice it to say that aheating degree hedge could protect theproducer against the cold temperaturescenario and a cooling degree hedgecould protect against the oppositeextreme event. This hedging strategy isillustrated in Figure 5.

Theoretically, by buying the same kindof risk instruments agriculture and otherindustries use routinely, the grass seedCFO would in effect hedge his weatherdriven demand risk, something that thevast majority of SCM executives do notconsider doing today, even theoretically.23

Now consider a more complex problemfor a large farm equipment manufacturerwhich sells tractors and other vehicles.Like the seed CFO, the head of supplychain of this industrial company faces

demand volatility that derives from avariety of sources: weather, commoditymarkets, political decisions made inforeign countries. However, an analysisof his demand indicates that a significantportion of the volatility in his demandis most closely correlated with certainspecific factors: the price of specificcommodities, fuel costs, and rainfall.Since these factors are unpredictable thisexecutive carries massive amounts oftractors in lots, much the Big Three U.S.automotive companies are accustomedto doing. The pressure to maintain suchhigh finished goods inventories comesfrom the corporate VP of sales, whowill complain vigorously to his CEO iffarmers suddenly want tractors and thereare none available. It would seem anintractable problem: the executive hasno choice but to carry a vast inventoryin huge lots hoping demand will show upand is seemingly unable to further reducehis inventory costs as a result.

Grass seed demand curve

45o 55o 85o 95o

Figure 4: Grass Seed Demand Curve


14/20

13

Now, consider the FSCM approach. Giventhe fact that this company in effectself-insures against demand risk andthat because its inventory is big andexpensive, the head of supply chaindecides that his ideal strategy would beto transfer some of the risk of a suddendemand spike to an external entity. Inother words, the supply chain executivecan lower his inventory if he can buydemand insurance an instrumentthat, for a fee. will pay him for any lost

sales he may incur for having too littleinventory in the unlikely case of a suddenand unexpected surge in tractor demand.

The head of SCM contacts various majorinsurance companies and, in 2010 at least,has no luck finding an insurance companywilling to write such a policy. The storydoes not end there, however. Becausethere presently is no financial firmthat will accept general market risk,,the executive develops an artificialor synthetic demand curve that (a)follows his real demand curve and (b) isbuilt from items that can be individually,though not collectively, hedged. Thatstrategy is presented in Figure 6.

In this scenario, the executive uses asynthetic demand curve built on pre-existing risk instruments such as hedgesand swaps that collectively mirror hisreal demand curve. These risk instrumentseffectively protect his company againstthe scenario where demand spikes forany given set of reasons and he has no

inventory to sell. The natural evolutionof this strategy is that over time thesupply chain executive will come tosee the hedge instruments as anothervariable along with the traditionalphysical one of spare inventory to beadjusted against fluctuations in demand.

To FSCM strategists, this companyhas simply bought a kind of demandinsurance, which, when all is saidand done, fulfills the same riskmanagement function as hundreds of

millions of dollars of unsold inventorywaiting for a demand spike.24

As unlikely as the above scenariomay seem, such techniques are beingdeveloped today by FSCM strategistsin industry, consultancies and a smallnumber of financial services firms.What all these theoretical but soonpossible strategies have in common isthat they are built on a concept calledthe Efficient Frontier, first laid out byHarry Markowitz in his 1952 paper

that launched portfolio theory.25

In essence, the Efficient Frontier (seeFigure 7) is a curve that consists ofthe optimal risk-reward portfolios fora given level of risk. This concept fromFinance is important for FSCM becauseFCSM aims not just at managing risk,which is the subject of so much SCMrisk management thinking today, but ofincreasing the value of the company,another idea borrowed from Finance.Be it by using real option methods to

make decisions about airplane features,or reducing profitability by commodityhedging, or by deploying syntheticdemand curves to create at least someincome in situations where before therewere just lost sales the aim of FSCM isto expand the role of SCM strategy intothe world of value maximization, not justdelivering products or creating forecasts.If one thinks of a given products supplychain as really a set of present andfuture cash f lows (again, analogous

to a Project in Finance terms), thenas one paper puts it, by choosing theproject to invest in, managers search forefficiency, that is attaining the frontierof possibilities, as well as optimality, thatis reaching the point on that frontierthat maximizes firm value given themarket prices of risk factors..26FSCM, bytrying at all times to optimize the valueof each products supply chain relativeto market risk something that rarelyhappens explicitly today, supply chainas a discipline can attains its highest

conceptual evolution to date and actuallybegin an evolution towards becomingcloser to a modern Corporate Financefunction, which is what will happen inthe coming decades. As the events of2008 and 2009 have shown, however,such a evolution will not be not withoutits own risks, however, and there aremany lessons to be drawn from Financethat FSCM strategists must consider asthey evolve their thinking and practices.

Grass seed demand curve

45o 55o 85o 95o

TempDemand volatility coverage

Option

value

Heating

degree day

hedge

Cooling

degree day

hedge

Figure 5: Hedging Out Weather-Driven Demand Risk


15/20

14

Time

Demand peak Demand volatilitycoverage

Syntheticdemand curve

Mirror indexhedge

Farmequipment

demand curve

Hedge limit

Demand/option value

Figure 6: Hedging Out Complex Demand Risk

Risk % (Standard deviation)

Return%

A portfolio above thiscurve is impossible

Low risk/low return

Medium risk/medium return

High risk/high return

Portfolios below the curve are notefficient, because for the same riskone could achieve a greater return.

Optional portfoliosshould lie on thiscurve (know as theEfficient Frontier)

Figure 7: The Efficient Frontier Portfolio Theory


16/20

An FSCM Framework:A Practical ProposalOne of the aims of writing this paper wasto determine if FSCM is at a stage where

a basic framework can be discerned fromthe case studies presented earlier. Theanswer to this question is a cautious yes.The examples presented in this paperare a first attempt at developing a basicFSCM framework for Finance and SCMstrategists to consider. It is the hope ofthis author that a decade from now theFSCM discipline will have evolved in waysunimaginable today.

From an organizational perspective, thefirst step is to create an FSCM team.

The FSCM Framework delineated belowsuggests a four-step process to createand prepare this team:

1) Learn. Create an integrated FSCMteam that combines both Financeand SCM expertise. Send this teamthrough a comprehensive set of courseson advanced OR, real option, riskmanagement and financial optimizationtools and techniques. They will not findthis course set in any one institution;rather, they will have to design a bespoke

curriculum from a variety of academicand professional sources. Full course ofpreparation could take as long as a yearor two, though certain industries (such asEnergy) already boast professionals whounderstand complex SCM and financialmethods. A great product company coulddo worse than to recruit some of theseexperts as it builds its first FSCM teams.

2) Rethink. As the team evolves, askthem to consider the companys SCM notonly as set of physical flows but in themore abstract sense presented above.In other words, instead of thinking onlyabout things such as safety stockand production agility, ask them tothink about external drivers of demandvolatility, what a synthetic demand curvesmight look like, sources of financial risk aswell as ways in which those risks can betransferred or hedged.

3) Question. Ask the team to becomefamiliar with the vendors of the emerging

tools in FSCM. All have their pros andcons. In addition, meet with and questionthought leaders in FSCM areas and beginto understand the theories behind thetools and techniques, which will maketool evaluation and adoption faster andmore effective.

4) Experiment. FSCM is a nascentdiscipline. It is early in its evolutionbut that process has started. All of theleading companies in this field have atleast one thing in common: they are

experimenters. Indeed, one should seethis exercise as a radical search for SCMinnovation that has the potential tocompletely change the way a companydoes business and sooner ratherthan later. Indeed, what most of thesecompanies also have in common is thatonce they begin this investigation, theydo not abandon it. Despite the rigor of themethods and the intellectual challenge ofthe perspectives required, the best SCMstrategists are stimulated by taking thisnext step in their thinking and careers.This energy should be harnessed by theircompanies, with strategists given room toexperiment and fail if need be.

From a technical perspective, theFSCM Framework suggests a three stepapproach:

1) Understand the financial flows that runparallel to the physical and informationflows in the companys supply chain. Inthis step, the FSCM team creates f inancialflow and cost models that become theFSCM world in which it will strategizeand manage.

2) Understand the way risk is createdin this parallel world and how, today, itis either being ignored or managed (i.e.,assumed, ignored, transferred or hedged).This should be done qualitatively at firstbut then quantitatively as the FSCM teambuilds its level of sophisticationand expertise.

3) Develop two roadmaps. The firstroadmap addresses creation of a fully

parallel FSCM model that can be usedto adjust and, if need be, over-ride thephysical SCM model. The covers creationof a risk management toolkit that canbe used to (a) evaluate SCM (financial)risk and (b) manage that risk in complex,multi-faceted strategies that fully exploitthe innovations of the global financialsystem in the last three decades.

4) Lastly, deploy the roadmaps and keepa constant qualitative and quantitativealignment between the physical and

financial supply chains. Redesignprocesses such as Sales and OperationsPlanning to incorporate both worldsconcurrently. Work with financialpartners to develop new risk managementmechanisms for FSCM, where those dontexist today. In the end, transform theSCM team from its current focus to onethat is much more externally focusedand one which, in a few years, is just athome on a trading floor as a shop floor.It is crucial to note that FSCM in no wayeliminates or diminishes the importanceof the physical SC and physical SCM; itjust shines a new light on those aspectsthat may reveal heretofore unseenproblems and solutions.

Achieving the state described bythe four steps above will necessarilytransform from a function focused onlyon real product flows and facilities toone that is concerned with corporate-level risk, financial performance andshareholder risk and returns in the mostdirect way possible by controllingthe full scope of the products andservices that, in the end, are thedrivers and determinants of corporateperformance in the marketplace.

15


17/20

16

When Genius FailsPerhaps the most interesting publicdebate in Finance at the start of thenew century was over the use of riskinstruments such as derivatives27inthe global financial system itself. Adecade before the global financial crisisof 2008, far-sighted critics attacked

these instruments both from a technicalperspective i.e., that they concentraterisk too narrowly or that they actuallydont work when they are most needed and from a social perspective i.e.,that they are unfair to the poor or thatthey are a means by which a plutocraticelite manipulate an unwitting public intobearing the cost of poor decision. Forexample, Tim Weithers, of the Universityof Chicago, summarized the most seriousworries back in 2007 as follows:

The sheer size of the derivatives andother instruments versus the underlyingassets on which those derivatives arebased. In other words, for any $1 ofmortgage debt, there might be $10 ofbets on whether that mortgage will everbe paid off.

The increasing involvement of the hedgefund community in this market, asthese institutions are opaque and lightlyregulated.

The operational backlogs and issues

surrounding confirmations, clearing,and settlement.28

The millions of bankruptcies andforeclosures caused by the collapse ofthe Western financial system in late2008 only confirmed Weithers fearsin the minds of most people. Given theeconomic roller coaster of the last twoyears, its worth elaborating on each ofthese worries and how they might applyto FSCM. The first issue Weithers notes,though couched in arcane Finance terms,refers to the parallel universe issuediscussed earlier in this paper. The worryis the imbalance between the values ofthe real sources of risk and values ofthe parallel derivative instruments. In anFSCM future, the equivalent might be thatApple could hedge the risk of its next iPodfailing in a demand risk marketplace runby, say NYSE Euronext29. In this scenario,a large parallel market would exist wherethousands of side bets were placed onthe outcome of the product launch.

The second issue is the involvementof powerful, secretive and lightlyregulated agents using and abusingrisk mechanisms, especially as regardsleverage or the multiplicative impactthat leveraged derivatives positionscan provide an agent. An FSCM parallelmight be if a hedge fund begins to takerisk positions on the outcome of certain

product cash flows e.g., iPod sales forthe next five years and by doing sodistorts the natural course that launchwould have taken.

Direct SCM speculation does not existtoday formally but thats only because noone has worked out a mechanism to doit something that todays sophisticatedFinancial Services firms could do themoment they set their mind to it. The lastissue refers to the mechanics of disposingof derivative instruments, especially intimes when the global financial system

is stressed. Given what bankers havedone to the world in the recent past,it would be tempting to dismiss FSCMas an idea that, like mortgage-backedsecurities, is pointless or even dangerous.However, mortgages and hedges did notcause the recent Financial crisis greedypeople did. And to dismiss the possibilitiesand flexibility that FSCM and, say, afunctioning demand risk marketplacewould provide SCM strategists, is to missthe view of the beneficial forest for the

sight of the current mass of rotten trees.Even if and this is a very big if indeed the birth of true SCM risk marketsexperiences some of the same problemsseen in the global financial systemrecently (and that is almost impossibleto imagine) the benefits will greatlyoutweigh the cost. A walk around theacres of unsold inventory in the autobusiness or a look at the mounds ofunsold cosmetics or piles of unsoldcomputers all a crude form of self-insurance against demand volatility

shows just how basic a level at whichSCM risk management operates today.Moreover, where lessons can be learnedfrom Finance in the regulation of SCMrisk instruments and markets say inaccounting treatments or in Value at Risk(VaR)30methodologies such lessonsshould also be adopted in FSCM.

As in Finance, genius in FSCM will failwhen overpowered by greed. But the samecan be said of any of the SCM techniquestaken for granted today. Indeed, as oneleader at Boeing put it when describingthe manufacturing challenges in its latestaircraft launch, nobody inside Boeingthought building the 787 would be easy.After all, the company decided to bet on

pushing the boundaries of the possibleIf everything was going perfect, it [wouldmean] you werent trying hard enough.31

CFO as Chief SCMOfficerThe revolution in FSCM is somethingthe very best SCM companies32shouldembrace, but doing so has manyimplications. Traditional roles will haveto change, as will power over SCM

decisions. In the end, what FSCM willdo is move SCM into the realm ofFinance, into the world of the CFO. Thismeans that SCM strategists need torethink their training and expand theirview and knowledge of finance andrisk management in the coming years.Just as the original physical evolutionand subsequent information evolutionschanged SCM forever, so will this nextwave of thinkers, tools and techniques.

Just how long a road this will be is

illustrated, again, by the ARC survey citedearlier. When asked to list their top SCMrisk management techniques, ARC surveyrespondents noted the results presentedin Figure 8.

Not one of the top five solutions is relatedto FSCM as presented in this paper. Bynow, that should be of no surprise. Whatalso will be of no surprise is how manyreaders will react negatively to thispaper. Many will dismiss FSCM as somefantastic, theoretical view of the futurethat, if it ever exists at all, will only comeinto being decades from now. Hopefully,the brief case studies cited here point outthe error of that reaction.

It is worth noting that the first outlinesfor this paper were laid out in 2005. Atthat time, only a few people were talkingabout SCM risk; today it is perhaps themost talked about subject in SCM. A fewyears ago, real options in contractingwere a revolutionary idea; today, they arein use and demonstrating their value.


18/20

17

A few years ago, the idea of incorporatingMonte Carlo simulation was purelytheoretical; today, several softwaresolutions allow you to do just that.

Similarly, the tools for hedging demandvolatility may seem fanciful now, butthey wont take long to create once theinsurance companies, banks and financialfirms that created SCM teams in thelast couple of years realize that they siton edge of a multi-billion dollar marketthat is largely ignored today. Indeed, ina recent interview, the head of SCM riskat a major insurance company notedthat supply chain as a concept exists ather company and that they are coveringcertain specific supply chain risks todaysuch as:

Supplier default/bankruptcy

Supply quality (i.e., supplier delivermaterials but they are unusable becauseof spoilage or some other defect)

Foreign exchange risk on supply contracts

Gaps between marine and property risk

Demand concentration risk (i.e., all of acompanys demand comes from only afew customers; should one of them fail,

the company is severely exposed)

Weather risk in agricultural industries

Carbon credit risk (i.e., an enterprisedepends on a carbon credit and wantsto ensure that, if the credit disappears,it hedges against incremental cost ofreplacement)

Risk of government seizure of plantand property

The insurance executive also noted that in2008 the company even insured againsta few product-income related risks.She added that this is new, and actually

a kind of securitization (a structuredfinance process that distributes riskby aggregating debt instruments in apool, then issues new securities backedby the pool). However, in the end theissue is pricing the peril and this drivesthe business case to insure or not. Theinsurance executive also noted that hergroup is very entrepreneurial and isused to basically creating custom risksolutions from scratch so that whatwill be possible in five years may not beimaginable today. This sentiment echoes

the thoughts of that group of prescientauthors writing in 2002 about themerging of Finance and SCM:34

What they wrote is still accurate and itstime for the wider SCM community to

begin to understand the coming FSCMevolution and see it as the next majorstep in a discipline that has moved fromthe background to the forefront ofglobalization.

In closing, think back to the perfumemaker and its supplier. In the future, whois to say the supplier will not be ableto buy a hedge instrument that wouldextract that 70 percent forecast error riskand spread it around channel partners orrisk-seekers in other industries lookingfor a high return? Who is to say that theperfume manufacturer will not be ableto buy a real option for box productioncapacity instead of forcing the supplierto accept a risky production commitmentthat could put it out of business? Who isto say the average investor might not findthe ability to invest in specific productcash flows more efficiently than in thecompanies that bundle them into onecorporate stock?

How different SCM would look in thatworld a world in which Finance and

SCM have merged to form an entirelynew discipline whose focus is not justwarehouses, pallets and trucks but themost fundamental elements of corporatevalue creation and preservation.

0% 20% 40% 60% 80%

Engage in S&OP

Source from multiple vendors, carriers etc.

Collaborative forecasting

Add redundant logistics or manufacturing

capacity

Implement supplier charge-backs

70.8%

70.8%

54.2%

29.2%

12.5%

Figure 8: What Are the Best Ways to Avoid Supply Chain Risks?33

The tools and lessons of financialengineering need to be adaptedto these new environments, andembodied in new products andservices that enable these emergingcapabilities. Similar transformationshave already occurred in more

traditional commodity-basedindustries, such as energy, metals,chemicals, or agriculture, wherecompanies purchase primaryinputs or sell primary outputs inactive markets.

As the technology industry, alongwith other industries, movestowards being more commodity-based, expertise in manufacturingfacilities investment and processescritical competencies of technologyfirms in the pastwill becomemore peripheral, while expertise

in market analysis, contracting,trading, and risk management willbecome increasingly central to thesuccess of these firms. Companiesthat develop those competencieswill be well positioned tocreate significant value for theirshareholders over the long run.


19/20

Notes

18

1. Lean manufacturing or lean production, whichis often known simply as "Lean", is a productionpractice that considers the expenditure ofresources for any goal other than the creation ofvalue for the end customer to be wasteful, andthus a target for elimination.

2. Peter L. Bernstein, Against the Gods: TheRemarkable Story of Risk, (New York: John Wiley &Sons Inc., 1996, 1998.

3. Bernstein, 106.

4. The term quant was first coined to refer to themathematicians/physicists who got PhDs in the1980s and 1990s and rather than go into science,chose much more lucrative careers in f inance atrend that continued until 2008 and will pick upagain in 2010.

5. For an example of this field, see BardiaKamrad and Keith Ord, Market Risk and ProcessUncertainty in Production Operations, (Paperdelivered at the Real Options Conference, ColumbiaBusiness School), 2006.

6. A Monte Carlo method is a technique thatinvolves using random numbers and probabilityto solve problems. The term Monte Carlo Methodwas coined by S . Ulam and Nicholas Metropolisin reference to games of chance, a popularattraction in Monte Carlo, Monaco (Hoffman, 1998;Metropolis and Ulam, 1949), (http://www.vertex42.com/ExcelArticles/mc/MonteCarloSimulation.html).

7. See, for example, Best Prac tices: SuccessfullyManaging Security And Risk In A Global SupplyChain, Patrick Connaughton, Forrester Research,August 3, 2007.

8. ARC Strategies Series: Risk Management, SteveBanker, Sid Snitkin and Adrian Gonzalez, 2005. Its

also notable that the number one risk they fearedwas a breakdown of their IT systems, which speaksvolumes about the impact the Information Phasehas had on supply chain management.

9. Source: ARC Strategies Series: Risk Management,Steve Banker, Sid Snitkin, & Adrian Gonzalez, 2005.

10. Ibid, 6.

11. European Real Options: An Intuitive Algorithmfor the Black-Scholes Formula, Vinay Datar andScott Matthews, Journal of Applied Finance, Spring/Summer, 2004.

12. Theory of Constraints (TOC) is an overallmanagement philosophy introduced by Dr. Eliyahu

M. Goldratt in his 1984 book titled The Goal, thatis geared to help organizations continually achievetheir goal. The title comes from the contention thatany manageable system is limited in achieving moreof its goal by a very small number of constraints,and that there is always at least one constraint. TheTOC process seeks to identify the constraint andrestructure the rest of the organization around it.

13. Capacity Modeling with Monte CarloSimulation for Finished Goods Warehouses, by ScottEdwards, Proceedings of the 2006 Crystal Ball UserConference, 6.

14. Drum-buffer-rope is the Theory of Constraintsproduction application. It is named after thethree essential elements of the solution; thedrum or constraint or weakest link, the buffer ormaterial release duration, and the rope or releasetiming. The aim of the solution is to protect theweakest link in the system, and therefore the

system as a whole, against process dependencyand variation and thus maximize the systemsoverall effectiveness. (http://www.dbrmfg.co.nz/Production%20DBR.htm).

15. Southwest net, including fuel-hedgingcosts, dips: Results match average of analysts'forecasts, Padraic Cassidy, MarketWatch, Jan.17, 2007. http://www.marketwatch.com/news/story/southwest-air-net-income-including/story.aspx?guid=%7B557D6875-2B3C-41F8-AA67-E8D63C1038BA%7D (accessed: March 19, 2008).

16. Fuel Hedging in the Airline Industry: TheCase of Southwest Airlines, Dave Carter, DanRogers, and Betty Simkins, Oklahoma StateUniversity, 2004.

17. European Airlines Reap Benefits of Oil Hedging,Caroline Brothers, N.Y. Times, June 12, 2008. http://www.nytimes.com/2008/06/12/business/12air.html?_r=1&ref=business&oref=slogin (accessed:June 12, 2008).

18. Swap: A custom-made and negotiatedtransaction designed to manage financial risk over aperiod of 1 to 12 years. Two individuals can create aswap, or a swap may be made through a third partysuch as a brokerage firm or a bank. Swaps are usedto manage risk and often settlements occur in cash,not in delivery of the actual product or financialinstruments. Examples of swap transactions includecurrency swaps, interest rate swaps, and priceswaps for a variety of commodities. (http://www.yourdictionary.com/finance/swap).

19. Treasury & Risk Magazine Best PracticesSummit Awards, 2006. http://www.treasuryandrisk.com/events/06_aha_fr_win.php (accessed: June30, 2008)

20. A Real Options Perspective on SupplyChain Management in High Technology, CoreyBillington, Blake Johnson, Alex Triantis, Journalof Applied Corporate Finance, Summer 2002,42-43. This paper, though overlooked by mostSCM strategists and consultants, was a visionarywork in its analysis and statements about thefuture of SCM. Indeed, Johnson (Stanford) andTriantis (Maryland) continue to be thought

leaders in the FSCM world, in the areas of riskmanagement and real options re spectively.

21. Ibid, p.41.

22. A derivative simply refers to a financialinstrument that derives its value/price from someother indicator, in this case temperature.

23. There is a lot of literature on weather riskhedging. For an excellent introduction to the topicsee, Weather Forecasting for Weather Derivatives,Sean D. Campbell and Francis X. Diebold, Journal ofthe American Statistical Association, March 2005,Vol. 100, No. 469.

24. In 2005, the author filed for a patent for amethod to construct synthetic demand curves forthe purposes of hedging out demand volatility.

25. Riskglossary.com, http://www.riskglossary.com/link/efficient_frontier.htm. The efficientfrontier was first defined by Harry Markowitz inhis groundbreaking (1952) paper that launchedportfolio theory. That theory considers a universeof risky investments and explores what might bean optimal portfolio based upon those possibleinvestments.

26. The Value of Real and Financial RiskManagement, Boyer, Boyer and Garcia, SerieScientifique, Centre interuniversitaire de recherceen analyse des organizations, 2005, 21.

27. A derivative is a financial instrument that isderived from some other asset, index, event, valueor condition (known as the underlying asset).Rather than trade or exchange the underlying assetitself, derivative traders enter into an agreementto exchange cash or assets over time based on the

underlying asset. A simple example is a futurescontract: an agreement to exchange the underlyingasset at a future date.

28. Credit Derivatives: Macro-Risk Issues, TimWeithers, University of Chicago, April 20, 2007.

29. NYSE Euronext (NYX) is a leading globaloperator of financial markets and provider ofinnovative trading technologies. The company'sexchanges in Europe and the United States tradeequities, futures, options, fixed-income andexchange-traded products.

30. Value at Risk (VaR): A technique which uses thestatistical analysis of historical market trends andvolatilities to estimate the likelihood that a givenportfolio's losses will exceed a certain amount,InvestorWords.com, http://www.investorwords.com/5217/VAR.html (accessed: July 5, 2008).

31. The 787 Encounters Turbulence:Technical glitches and manufacturingwoes could delay Boeing's breakthrough,BusinessWeek, June 18, 2006.

32. And, paradoxically, in some ways the very worstas well.

33. Source: Investopedia. http://www.investopedia.com/terms/e/efficientfrontier.asp (accessed June29, 2008).

34. Billington et al., pp.42-43.


20/20

AcknowledgementsThe author wishes to express his gratitudeto the following individuals for their timeand insights on FSCM: Scott Edwards(Intel), Scott Matthews (Boeing), AlexTriantis and Sandor Boyson (The SmithSchool, University of Maryland), ScottKoerwer (The Moore School, University

of South Carolina), Francis X. Diebold(The Wharton School, University ofPennsylvania), Erwin Hermans (Accenture),Alen Yakar (Garanti Bank International N.V.,Istanbul), Herbert Wolf (Alix Partners), andOliver Scutt (Jonova Inc.).

About The AuthorCarlos A. Alvarenga is the Global Lead forOperations Finance and Risk at Accenture,as well as a Senior Research Fellow at theRobert H. Smith School of Business at The

University of Maryland. He is also author ofthe Operations Finance and Risk blog,SC Quant (http://scquant.com).

This material originally appeared in thebook, X-SCM, The New Science of X-tremeSupply Chain Management, (Routledge,2010).

About AccentureAccenture is a global managementconsulting, technology services andoutsourcing company, with approximately236,000 people serving clients in morethan 120 countries. Combining unparalleledexperience, comprehensive capabilitiesacross all industries and business functions,

and extensive research on the worldsmost successful companies, Accenturecollaborates with clients to help thembecome high-performance businesses andgovernments. The company generated netrevenues of US$25.5 billion for the fiscalyear ended Aug. 31, 2011. Its home page iswww.accenture.com.

Copyright 2011 AccentureAll rights reserved.

Accenture, its logo, andHigh Performance Delivered

accenture the rise of scm quants

Documents