Post on 31-Dec-2015
description
transcript
1.040/1.4011.040/1.401Project ManagementProject Management
Spring 2006Spring 2006
Risk AnalysisRisk AnalysisDecision making under risk and Decision making under risk and
uncertaintyuncertainty
Department of Civil and Environmental Department of Civil and Environmental EngineeringEngineeringMassachusetts Institute of Technology Massachusetts Institute of Technology
PreliminariesPreliminaries
AnnouncementsAnnouncements RemainderRemainder
email Sharon Lin the team info by midnight, tonightemail Sharon Lin the team info by midnight, tonight Monday Feb 27 - Student Experience PresentationMonday Feb 27 - Student Experience Presentation Wed March 1st – Assignment 2 dueWed March 1st – Assignment 2 due
Today, recitation Joe Gifun, MIT facilityToday, recitation Joe Gifun, MIT facility Next Friday, March 3rd, Tour PDSI construction siteNext Friday, March 3rd, Tour PDSI construction site
1st group noon – 1:301st group noon – 1:30 2nd group 1:30 – 3:002nd group 1:30 – 3:00
Construction nightmares discussionConstruction nightmares discussion 16 - Psi Creativity Center, Design and Bidding 16 - Psi Creativity Center, Design and Bidding
phasesphases
Project Management PhaseProject Management Phase
FEASIBILITY
DESIGNPLANNING
CLOSEOUTDEVELOPMENT OPERATIONS
Financing&EvaluationRisk Analysis&Attitude
Risk Management PhaseRisk Management Phase
FEASIBILITY
DESIGNPLANNING
CLOSEOUTDEVELOPMENT OPERATIONS
RISK MNGRISK MNG
Risk management (guest seminar 1st wk April)Risk management (guest seminar 1st wk April) Assessment, tracking and controlAssessment, tracking and control Tools:Tools:
Risk Hierarchical modeling: Risk breakdown structuresRisk Hierarchical modeling: Risk breakdown structures Risk matrixesRisk matrixes Contingency plan: preventive measures, corrective actions, Contingency plan: preventive measures, corrective actions,
risk budget, etc.risk budget, etc.
Decision Making Under Decision Making Under Risk OutlineRisk Outline
Risk and UncertaintyRisk and Uncertainty Risk Preferences, Attitude and Risk Preferences, Attitude and
PremiumsPremiums Examples of simple decision treesExamples of simple decision trees Decision trees for analysisDecision trees for analysis Flexibility and real optionsFlexibility and real options
Uncertainty and RiskUncertainty and Risk
““risk” as risk” as uncertainty about a uncertainty about a consequenceconsequence
Preliminary questionsPreliminary questions What sort of risks are there and who What sort of risks are there and who
bears them in project management?bears them in project management? What practical ways do people use to What practical ways do people use to
cope with these risks?cope with these risks? Why is it that some people are willing to Why is it that some people are willing to
take on risks that others shun?take on risks that others shun?
Some RisksSome Risks Weather changesWeather changes Different productivityDifferent productivity (Sub)contractors are(Sub)contractors are
UnreliableUnreliable Lack capacity to do Lack capacity to do
workwork Lack availability to do Lack availability to do
workwork UnscrupulousUnscrupulous Financially unstableFinancially unstable
Late materials deliveryLate materials delivery LawsuitsLawsuits Labor difficultiesLabor difficulties Unexpected Unexpected
manufacturing costsmanufacturing costs Failure to find Failure to find
sufficient tenantssufficient tenants
Community oppositionCommunity opposition Infighting & Infighting &
acrimonious acrimonious relationshipsrelationships
Unrealistically low bidUnrealistically low bid Late-stage design Late-stage design
changes changes Unexpected Unexpected
subsurface conditionssubsurface conditions Soil typeSoil type GroundwaterGroundwater Unexpected ObstaclesUnexpected Obstacles
Settlement of adjacent Settlement of adjacent structuresstructures
High lifecycle costsHigh lifecycle costs Permitting problemsPermitting problems ……
Importance of RiskImportance of Risk Much time in construction Much time in construction
management is spent focusing on risksmanagement is spent focusing on risks Many practices in construction are Many practices in construction are
driven by riskdriven by risk Bonding requirementsBonding requirements Insurance Insurance LicensingLicensing Contract structureContract structure
General conditionsGeneral conditions Payment TermsPayment Terms Delivery MethodDelivery Method Selection mechanismSelection mechanism
OutlineOutline
Risk and UncertaintyRisk and Uncertainty Risk Preferences, Attitude and Risk Preferences, Attitude and
PremiumsPremiums Examples of simple decision treesExamples of simple decision trees Decision trees for analysisDecision trees for analysis Flexibility and real optionsFlexibility and real options
Decision making under riskDecision making under riskAvailable TechniquesAvailable Techniques
Decision modelingDecision modeling Decision making under uncertaintyDecision making under uncertainty Tool: Tool: Decision treeDecision tree
Strategic thinking and problem Strategic thinking and problem solving:solving: Dynamic modeling (end of course)Dynamic modeling (end of course)
Fault treesFault trees
Introduction to Decision Introduction to Decision TreesTrees
We will use decision trees both forWe will use decision trees both for Illustrating decision making with Illustrating decision making with
uncertaintyuncertainty Quantitative reasoningQuantitative reasoning
RepresentRepresent Flow of timeFlow of time DecisionsDecisions Uncertainties (via events)Uncertainties (via events) Consequences (deterministic or stochastic)Consequences (deterministic or stochastic)
Decision Tree NodesDecision Tree Nodes
Decision (choice) NodeDecision (choice) Node
Chance (event) NodeChance (event) Node
Terminal (consequence) nodeTerminal (consequence) node Outcome (cost or benefit)Outcome (cost or benefit)
Time
Risk PreferenceRisk Preference
People are not indifferent to uncertaintyPeople are not indifferent to uncertainty Lack of indifference from uncertainty arises Lack of indifference from uncertainty arises
from uneven preferences for different from uneven preferences for different outcomesoutcomes
E.g. someone may E.g. someone may dislike losing $x far more than gaining $xdislike losing $x far more than gaining $x value gaining $x far more than they disvalue value gaining $x far more than they disvalue
losing $x.losing $x.
Individuals differ in comfort with Individuals differ in comfort with uncertainty based on circumstances and uncertainty based on circumstances and preferencespreferences
Risk averse individuals will pay “risk Risk averse individuals will pay “risk premiums” to avoid uncertaintypremiums” to avoid uncertainty
Risk preferenceRisk preference
The preference depends on decision maker The preference depends on decision maker point of viewpoint of view
Categories of Risk Categories of Risk AttitudesAttitudes
Risk attitude is a general way of Risk attitude is a general way of classifying risk preferencesclassifying risk preferences
ClassificationsClassifications Risk averse fear loss and seek surenessRisk averse fear loss and seek sureness Risk neutral are indifferent to Risk neutral are indifferent to
uncertaintyuncertainty Risk lovers hope to “win big” and don’t Risk lovers hope to “win big” and don’t
mind losing as muchmind losing as much Risk attitudes change over Risk attitudes change over
TimeTime CircumstanceCircumstance
Decision RulesDecision Rules
The pessimistic rule (maximin = minimax)The pessimistic rule (maximin = minimax) The conservative decisionmaker seeks to:The conservative decisionmaker seeks to:
maximize the minimum gain (if outcome = payoff)maximize the minimum gain (if outcome = payoff) or minimize the maximum loss (if outcome = loss, risk)or minimize the maximum loss (if outcome = loss, risk)
The optimistic rule (maximax)The optimistic rule (maximax) The risklover seeks to maximize the maximum The risklover seeks to maximize the maximum
gaingain Compromise (the Hurwitz rule):Compromise (the Hurwitz rule):
Max (Max (αα min + (1- min + (1- αα) max) , 0 ) max) , 0 ≤ α ≤ ≤ α ≤ 11 αα = 1 pessimistic = 1 pessimistic αα = 0.5 neutral = 0.5 neutral αα = 0 optimistic = 0 optimistic
The bridge case – unknown The bridge case – unknown prob’tiesprob’ties
replace
repair
$ 1.09 million
Investment PV
$1.61 M
$0.55
$1.43
•Pessimistic rulePessimistic rule• min (1, 1.61) = 1 replace the bridgemin (1, 1.61) = 1 replace the bridge
•The optimistic rule (maximax)The optimistic rule (maximax)• max (1, 0.55) = 0.55 repair … and hope it max (1, 0.55) = 0.55 repair … and hope it works!works!
The bridge case – known The bridge case – known prob’tiesprob’ties
replace
repair
$ 1.09 million
Investment PV
$1.61 M
$0.55
$1.43
Expected monetary valueExpected monetary valueE = (0.25)(1.61) + (0.5)(0.55) + (0.25)(1.43) = E = (0.25)(1.61) + (0.5)(0.55) + (0.25)(1.43) =
$ 1.04 M$ 1.04 M
0.25
0.5
0.25
Data link
The bridge case – The bridge case – decisiondecision
The pessimistic rule (maximin = The pessimistic rule (maximin = minimax)minimax) Min (Ei) = Min (1.09 , 1.04) = $ Min (Ei) = Min (1.09 , 1.04) = $
1.04 repair1.04 repair In this case = optimistic rule In this case = optimistic rule
(maximax)(maximax) Awareness of probabilities change Awareness of probabilities change
risk attituderisk attitude
Other criteriaOther criteria
Most likely valueMost likely value For each policy option we select the For each policy option we select the
outcome with the highest probabilityoutcome with the highest probability Expected value of Opportunity LossExpected value of Opportunity Loss
To buy soon or to buy laterTo buy soon or to buy later
Buy soon
Current price = 100S1 = + 30%S2 = no price variationS3 = - 30%
Actualization = 5
-100-30+5 = -125
-100+5 = -95
-100+5+30 = -65
Buy later
-100
To buy soon or to buy laterTo buy soon or to buy later
Buy soon
-125
-95
-65
Buy later
-100
0. 5
0.25
0.25
The Utility TheoryThe Utility Theory
When individuals are faced with uncertainty When individuals are faced with uncertainty they make choices as is they are maximizing a they make choices as is they are maximizing a given criterion: the given criterion: the expected utilityexpected utility..
Expected utility is a measure of the individual's Expected utility is a measure of the individual's implicit preference, for each policy in the risk implicit preference, for each policy in the risk environment.environment.
It is represented by a numerical value It is represented by a numerical value associated with each monetary gain or loss in associated with each monetary gain or loss in order to indicate the utility of these monetary order to indicate the utility of these monetary values to the decision-maker.values to the decision-maker.
Adding a Preference Adding a Preference functionfunction
Expected (mean) valueExpected (mean) valueE = (0.5)(125) + (0.25)(95) + (0.25)(65) = -E = (0.5)(125) + (0.25)(95) + (0.25)(65) = -
102.5102.5Utility value:Utility value:
f(E) = ∑ Pf(E) = ∑ Paa * f(a) = 0.5 f(125) + 0.25 f(95) * f(a) = 0.5 f(125) + 0.25 f(95) + .25 f(65) =+ .25 f(65) =
= .5*0.7 + .25*1.05 + .25*1.35 = ~0.95= .5*0.7 + .25*1.05 + .25*1.35 = ~0.95Certainty value = -102.5*0.975 = -97.38Certainty value = -102.5*0.975 = -97.38
100125 65
1.7
1.35
Defining the Preference Defining the Preference FunctionFunction
Suppose to be awarded a $100M Suppose to be awarded a $100M contract pricecontract price
Early estimated cost $70MEarly estimated cost $70M What is the preference function of What is the preference function of
cost?cost? Preference means utility or satisfactionPreference means utility or satisfaction
$
utility
70
Notion of a Risk Notion of a Risk PremiumPremium
A risk premium is the amount paid by a A risk premium is the amount paid by a (risk averse) individual to avoid risk(risk averse) individual to avoid risk
Risk premiums are very common – what Risk premiums are very common – what are some examples?are some examples? Insurance premiumsInsurance premiums Higher fees paid by owner to reputable Higher fees paid by owner to reputable
contractorscontractors Higher charges by contractor for risky workHigher charges by contractor for risky work Lower returns from less risky investmentsLower returns from less risky investments Money paid to ensure flexibility as guard Money paid to ensure flexibility as guard
against riskagainst risk
Conclusion: To buy or Conclusion: To buy or not to buynot to buy
The risk averter buys a “future” The risk averter buys a “future” contract that allow to buy at $ 97.38contract that allow to buy at $ 97.38
The trading company (risk lover) will The trading company (risk lover) will take advantage/disadvantage of take advantage/disadvantage of future benefit/loss future benefit/loss
Certainty Equivalent Certainty Equivalent ExampleExample Consider a risk averse individual with Consider a risk averse individual with
preference fn preference fn ff faced with an faced with an investment c that providesinvestment c that provides 50% chance of earning $2000050% chance of earning $20000 50% chance of earning $050% chance of earning $0
Average Average moneymoney from investment = from investment = .5*$20,000+.5*$0=$10000.5*$20,000+.5*$0=$10000
Average Average satisfactionsatisfaction with the with the investment=investment= .5*f($20,000)+.5*f($0)=.25.5*f($20,000)+.5*f($0)=.25
This individual would be willing to This individual would be willing to trade for a trade for a suresure investment yielding investment yielding satisfaction>.25 insteadsatisfaction>.25 instead Can get .25 satisfaction for a sure fCan get .25 satisfaction for a sure f--
11(.25)=$5000(.25)=$5000 We call this the We call this the certainty equivalentcertainty equivalent to the to the
investmentinvestment Therefore this person should be willing to Therefore this person should be willing to
trade this investment for a sure amount of trade this investment for a sure amount of money>$5000money>$5000
.25
Mean valueOf investment
Mean satisfaction withinvestment
Certainty equivalentof investment
$50
00
.50
Example Cont’d (Risk Example Cont’d (Risk Premium)Premium) The risk averse individual would be willing The risk averse individual would be willing
to trade the uncertain investment c for any to trade the uncertain investment c for any certain return which is > $5000certain return which is > $5000
Equivalently, the risk averse individual Equivalently, the risk averse individual would be willing to pay another party an would be willing to pay another party an amount amount rr up to $5000 =$10000-$5000 for up to $5000 =$10000-$5000 for other less risk averse party to guarantee other less risk averse party to guarantee $10,000$10,000 Assuming the other party is not risk averse, Assuming the other party is not risk averse,
that party wins because gain that party wins because gain rr on average on average The risk averse individual wins b/c more The risk averse individual wins b/c more
satisfiedsatisfied
Certainty EquivalentCertainty Equivalent More generally, consider situation in which haveMore generally, consider situation in which have
Uncertainty with respect to consequence Uncertainty with respect to consequence cc Non-linear preference function Non-linear preference function ff
Note: E[X] is the mean (expected value) operatorNote: E[X] is the mean (expected value) operator The mean The mean outcomeoutcome of uncertain investment c is of uncertain investment c is
E[c]E[c] In example, this was .5*$In example, this was .5*$20,000+.5*$0=$10,00020,000+.5*$0=$10,000
The mean The mean satisfaction withsatisfaction with the investment is the investment is E[f(c)]E[f(c)] In example, this was .5*f($20,000)+.5*f($0)=.25In example, this was .5*f($20,000)+.5*f($0)=.25
We call fWe call f-1-1(E[f(c)]) the (E[f(c)]) the certainty equivalentcertainty equivalent of c of c Size of Size of sure sure return that would give the same satisfaction return that would give the same satisfaction
as as cc In example, was fIn example, was f-1-1(.25)=f(.25)=f-1-1(.5*20,000+.5*0)=$5,000 (.5*20,000+.5*0)=$5,000
Risk Attitude ReduxRisk Attitude Redux
The shapes of the preference functions The shapes of the preference functions means can classify risk attitude by means can classify risk attitude by comparing the certainty equivalent and comparing the certainty equivalent and expected valueexpected value For risk For risk loving loving individuals, individuals, ff-1-1(E[f(c)])(E[f(c)])>E[c]>E[c]
They want Certainty equivalent > mean outcomeThey want Certainty equivalent > mean outcome For risk For risk neutralneutral individuals, individuals, ff-1-1(E[f(c)])(E[f(c)])=E[c]=E[c] For risk For risk averseaverse individuals, individuals, ff-1-1(E[f(c)])(E[f(c)])<E[c]<E[c]
Motivations for a Risk Motivations for a Risk PremiumPremium
Consider Consider Risk averse individual A for whom Risk averse individual A for whom ff--
11(E[f(c)])(E[f(c)])<E[c]<E[c] Less risk averse party BLess risk averse party B
A can lessen the effects of risk by paying a A can lessen the effects of risk by paying a risk premium risk premium rr of up to E[c]- of up to E[c]-ff-1-1(E[f(c)]) to B (E[f(c)]) to B in return for a in return for a guarantee guarantee of of E[c] incomeE[c] income The risk premium shifts the risk to BThe risk premium shifts the risk to B The net investment gain for A is E[c]-r, but A is The net investment gain for A is E[c]-r, but A is
more satisfied because E[c] – r > fmore satisfied because E[c] – r > f-1-1(E[f(c)])(E[f(c)]) B gets average monetary gain of B gets average monetary gain of rr
Gamble or not to GambleGamble or not to Gamble
EMV(0.5)(-1) + (0.5)(1) = 0
Preference function f(-1)=0, f(1)=100Certainty eq. f-1(E[f(c)]) = 0No help from risk analysis !!!!!
Multiple Attribute Multiple Attribute DecisionsDecisions
Frequently we care about multiple Frequently we care about multiple attributesattributes CostCost TimeTime QualityQuality Relationship with ownerRelationship with owner
Terminal nodes on decision trees can Terminal nodes on decision trees can capture these factors – but still need to capture these factors – but still need to make different attributes comparablemake different attributes comparable
The bridge case - Multiple The bridge case - Multiple tradeoffstradeoffs
MTTF = mean time to failure
Computation of Pareto-Optimal SetFor decision D2
Replace MTTF 10.0000 Cost 1.00
C3 MTTF 6.6667 Cost 0.30
C4 MTTF 5.7738 Cost 0.00
Aim: maximizing bridge duration, minimizing cost
Pareto OptimalityPareto Optimality
Even if we cannot directly weigh one attribute Even if we cannot directly weigh one attribute vs. another, we can rank some consequencesvs. another, we can rank some consequences
Can rule out decisions giving consequences Can rule out decisions giving consequences that are inferior with respect to that are inferior with respect to allall attributes attributes We say that these decisions are “dominated by” We say that these decisions are “dominated by”
other decisionsother decisions Key concept here: May not be able to identify Key concept here: May not be able to identify
best decisionsbest decisions, but we can rule out obviously , but we can rule out obviously badbad
A decision is “Pareto optimal” (or efficient A decision is “Pareto optimal” (or efficient solution) if it is not dominated by any other solution) if it is not dominated by any other decisiondecision
03/06/06 - Preliminaries03/06/06 - Preliminaries AnnouncementsAnnouncements
Due dates Stellar Schedule and not SyllabusDue dates Stellar Schedule and not Syllabus Term projectTerm project
Phase 2 due March 17thPhase 2 due March 17th Phase 3 detailed description posted on Stellar, due May Phase 3 detailed description posted on Stellar, due May
1111 Assignment PS3 posted on Stellar – due date Assignment PS3 posted on Stellar – due date
March 24March 24 Decision making under uncertaintyDecision making under uncertainty
Reading questions/comments?Reading questions/comments? Utility and risk attitudeUtility and risk attitude You can manage construction risksYou can manage construction risks Risk management and insurances - RecommendedRisk management and insurances - Recommended
Decision Making Under Decision Making Under RiskRisk
Risk and UncertaintyRisk and Uncertainty Risk Preferences, Attitude and Risk Preferences, Attitude and
PremiumsPremiums Examples of simple decision treesExamples of simple decision trees Decision trees for analysisDecision trees for analysis Flexibility and real optionsFlexibility and real options
Decision Making Under Decision Making Under RiskRisk
Risk and UncertaintyRisk and Uncertainty Risk Preferences, Attitude and Risk Preferences, Attitude and
PremiumsPremiums Examples of simple decision treesExamples of simple decision trees Decision trees for analysisDecision trees for analysis Flexibility and real optionsFlexibility and real options
BiddingBidding
What What choiceschoices do we have? do we have? How does the chance of winning How does the chance of winning
vary with our bidding price?vary with our bidding price? How does our profit vary with our How does our profit vary with our
bidding price if we win?bidding price if we win?
Bidding Decision Tree with Bidding Decision Tree with Stochastic Costs, Stochastic Costs, Competing BidsCompeting Bids
Decision Tree Example: Decision Tree Example: Procurement TimingProcurement Timing
DecisionsDecisions Choice of order time (Order early, Choice of order time (Order early,
Order late)Order late) EventsEvents
Arrival time (On time, early, late)Arrival time (On time, early, late) Theft or damage (only if arrive early)Theft or damage (only if arrive early)
Consequences: CostConsequences: Cost Components: Delay cost, storage cost, Components: Delay cost, storage cost,
cost of reorder (including delay)cost of reorder (including delay)
Decision Making Under Decision Making Under RiskRisk
Risk and UncertaintyRisk and Uncertainty Risk Preferences, Attitude and Risk Preferences, Attitude and
PremiumsPremiums Decision trees for representing Decision trees for representing
uncertaintyuncertainty Decision trees for analysisDecision trees for analysis Flexibility and real optionsFlexibility and real options
Analysis Using Decision Analysis Using Decision TreesTrees
Decision trees are a powerful Decision trees are a powerful analysis toolanalysis tool
Example analytic techniquesExample analytic techniques Strategy selection (Monte Carlo Strategy selection (Monte Carlo
simulation)simulation) One-way and multi-way sensitivity One-way and multi-way sensitivity
analysesanalyses Value of informationValue of information
Monte Carlo simulationMonte Carlo simulation Monte Carlo simulation randomly generates values for Monte Carlo simulation randomly generates values for
uncertain variables over and over to simulate a model.uncertain variables over and over to simulate a model. It's used with the variables that have a known range It's used with the variables that have a known range
of values but an uncertain value for any particular of values but an uncertain value for any particular time or event. time or event.
For each uncertain variable, you define the possible For each uncertain variable, you define the possible values with a probability distribution.values with a probability distribution.
Distribution types include:Distribution types include:
A simulation calculates multiple scenarios of a model A simulation calculates multiple scenarios of a model by repeatedly sampling values from the probability by repeatedly sampling values from the probability distributionsdistributions
Computer software tools can perform as many trials Computer software tools can perform as many trials (or scenarios) as you want and allow to select the (or scenarios) as you want and allow to select the optimal strategyoptimal strategy
…
With Risk Preferences for With Risk Preferences for Large Uncertainties at Large Uncertainties at
lower bidlower bid
Decision Making Under Decision Making Under RiskRisk
Risk and UncertaintyRisk and Uncertainty Risk Preferences, Attitude and Risk Preferences, Attitude and
PremiumsPremiums Decision trees for representing Decision trees for representing
uncertaintyuncertainty Examples of simple decision treesExamples of simple decision trees Decision trees for analysisDecision trees for analysis Flexibility and real optionsFlexibility and real options
Flexibility and Real Flexibility and Real OptionsOptions
Flexibility isFlexibility is providing additional providing additional choiceschoices
Flexibility typically has Flexibility typically has Value by acting as a way to lessen the Value by acting as a way to lessen the
negative impacts of uncertaintynegative impacts of uncertainty Cost Cost
Delaying decisionDelaying decision Extra timeExtra time Cost to pay for extra “fat” to allow for Cost to pay for extra “fat” to allow for
flexibilityflexibility
Ways to Ensure of Ways to Ensure of Flexibility Flexibility
in Constructionin Construction Alternative Delivery Alternative Delivery Clear spanning (to Clear spanning (to
allow movable walls) allow movable walls) Extra utility Extra utility
conduits (electricity, conduits (electricity, phone,…)phone,…)
Larger footings & Larger footings & columns columns
Broader foundationBroader foundation Alternative Alternative
heating/electrical heating/electrical
Contingent plans forContingent plans for Value engineeringValue engineering Geotechnical conditionsGeotechnical conditions Procurement strategyProcurement strategy
Additional elevatorAdditional elevator Larger electrical Larger electrical
panelspanels Property for expansionProperty for expansion Sequential Sequential
constructionconstruction Wiring to roomsWiring to rooms
Adaptive StrategiesAdaptive Strategies
An adaptive strategy is one that An adaptive strategy is one that changes the course of action based changes the course of action based on what is observed – i.e. one that on what is observed – i.e. one that has flexibilityhas flexibility Rather than planning statically up front, Rather than planning statically up front,
explicitly plan to adapt as events unfoldexplicitly plan to adapt as events unfold Typically we delay a decision into the Typically we delay a decision into the
futurefuture
Real OptionsReal Options Real Options theory provides a means of Real Options theory provides a means of
estimating financial estimating financial valuevalue of flexibility of flexibility E.g. option to abandon a plant, expand bldgE.g. option to abandon a plant, expand bldg
Key insight: NPV does not work well with Key insight: NPV does not work well with uncertain costs/revenuesuncertain costs/revenues E.g. difficult to model option of abandoning E.g. difficult to model option of abandoning
invest.invest. Model events using stochastic diff. Model events using stochastic diff.
equationsequations Numerical or analytic solutionsNumerical or analytic solutions Can derive from decision-tree based frameworkCan derive from decision-tree based framework
ConsiderationsConsiderations
TradeoffsTradeoffs Short-term speed and flexibilityShort-term speed and flexibility
Overlapping design & construction and different Overlapping design & construction and different construction activities limits changesconstruction activities limits changes
Short-term cost and flexibilityShort-term cost and flexibility E.g. value engineering away flexibilityE.g. value engineering away flexibility Selection of low bidderSelection of low bidder Late decisions can mean greater costsLate decisions can mean greater costs
NB: both budget & schedule may ultimately be NB: both budget & schedule may ultimately be better off w/greater flexibility!better off w/greater flexibility!
Frequently retrofitting $ > up-front $Frequently retrofitting $ > up-front $
Decision Making Under Decision Making Under RiskRisk
Risk and UncertaintyRisk and Uncertainty Risk Preferences, Attitude and Risk Preferences, Attitude and
PremiumsPremiums Decision trees for representing Decision trees for representing
uncertaintyuncertainty Examples of simple decision treesExamples of simple decision trees Decision trees for analysisDecision trees for analysis Flexibility and real optionsFlexibility and real options
ReadingsReadings RequiredRequired
More information:More information: Utility and risk attitude – Stellar Readings Utility and risk attitude – Stellar Readings
sectionsection Get prepared for next class:Get prepared for next class:
You can manage construction risks – StellarYou can manage construction risks – Stellar On-line textbook, from 2.4 to 2.12On-line textbook, from 2.4 to 2.12
Recommended:Recommended: Meredith Textbook, Chapter 4 Prj Meredith Textbook, Chapter 4 Prj
OrganizationOrganization Risk management and insurances – Risk management and insurances –
StellarStellar
Risk - MIT librariesRisk - MIT libraries Haimes, Risk modeling, assessment, and managementHaimes, Risk modeling, assessment, and management
Mun, Mun, Applied risk analysis : moving beyond Applied risk analysis : moving beyond uncertaintyuncertainty
Flyvbjerg, Mega-projects and riskFlyvbjerg, Mega-projects and risk
Chapman, Managing project risk and uncertainty : a Chapman, Managing project risk and uncertainty : a constructively simple approach to decision makingconstructively simple approach to decision making
Bedford, Probabilistic risk analysis: foundations and Bedford, Probabilistic risk analysis: foundations and methodsmethods
… … and a lot more!and a lot more!