Slide 1
Chapter 11 Structured Risk ManagementRisk ManagementExpected
Value of Perfect ControlSensitivity Analysis Robustness easy to do
with Precision TreeValue Added structured approach - Individual
random eventsRole of InformationExpected Value of Imperfect
InformationBayes Rule and EVIIOptimal Conditional
decisionSequential decisions with information delayReal Options
Chelst & Canbolat Value Added Decision Making
02/28/12#Chapter 11Risk Management ThemeYou cannot manage risk
if you do not admit there is uncertaintyManaging uncertainty also
includes unrealized upside potential and not just downside
lossesYou cannot allocate appropriate resources if you do not
quantify the risk or uncertaintyChelst & Canbolat Value Added
Decision Making
02/28/12#Chapter 112Figure 11.1: Decision tree Boss Controls
automation investment40.0%09.91.9FALSETake
Rate-85.8660.0%016.58.5How Much6.3240.0%0.413.80.8TRUETake
Rate-136.3260.0%0.62310Automation InvestmentLowHigh30% Take50%
Take50% Take30% TakeChelst & Canbolat Value Added Decision
Making
02/28/12#Chapter 11Investment in Automation -Question:
Robustness of Optimal SolutionThe High investment alternative
involves a new technology. Management is concerned that the capital
equipment estimate could be off by + 7%.There is even more concern
regarding the variable cost estimate that could be off by + 10%The
Low investment alternative is well tested and there is hope that
continuous improvement could reduce the variable cost by 5%.Because
they did not know, they set the take rate probabilities at 0.6 and
0.4 respectively. However, there is a lot of uncertainty regarding
this estimated probability.Chelst & Canbolat Value Added
Decision Making
02/28/12#Chapter 1145.8 5.2 = 0.6M50-50 chance of 30% and 50%
take rate averages out 400,000 sales per year. Each dollar decrease
is worth $400,00. A $1.50 decrease for Low investment variable cost
makes it equal to High alternativeIncreases in probability of Low
take rate decreases the expected volume. The higher variable cost
for Low Investment means it will get worse and worse. Increases in
probability every 10% probability shift causes expected sales to
decrease by 20,000. The difference in variable costs is (27-13) or
14. Thus a 10% shift reduces the HIGH advantage by $14(20,000)
$280,000 , 20% shift $560,000 and a 22% $616,000Price May be cut?
DOES NOT MAKE A DIFFERENCE in preference Every dollar affects both
equally. It will change the ROI Return on Investment.Volume OEM
forecast overly optimistic every 100,000 is 40,000 options Divide
$600,000/14 = 42,571 options 106427 sales.High Take rate too
conservative maybe 60% more sales make best alternative betterLow
Take rate too optimistic maybe 20% reduces expected sales of
options by 50,000 no longer optimal Investment in
AutomationRobustness of Optimal SolutionMagnitude of Difference
between two solutions ($6.32M-$5.86M) = $460,000Investment(s) How
much increase in HIGH Investment fixed cost results in change in
best decision?Variable Cost(s)How much would the variable cost for
Low Investment have to decline to make it preferred?Probability of
30% take rate: Increases? Decreases?What else and why?Chelst &
Canbolat Value Added Decision Making
02/28/12#Chapter 1155.8 5.2 = 0.6M50-50 chance of 30% and 50%
take rate averages out 400,000 sales per year. Each dollar decrease
is worth $400,00. A $1.50 decrease for Low investment variable cost
makes it equal to High alternativeIncreases in probability of high
take rate increases the expected volume. The lower variable cost
for High Investment means it will get better and better. Decreases
in probability every 10% probability shift causes expected sales to
decreases by 20,000. The difference in variable costs is (27-13) or
14. Thus a 10% shift reduces the HIGH advantage by $14(20,000)
$280,000 , 20% shift $560,000 and a 22% $616,000Price May be cut?
DOES NOT MAKE A DIFFERENCE in preference Every dollar affects both
equally. It will change the ROI Return on Investment.Volume OEM
forecast overly optimistic every 100,000 is 40,000 options Divide
$600,000/14 = 42,571 options 106427 sales.High Take rate too
conservative maybe 60% more sales make best alternative betterLow
Take rate too optimistic maybe 20% reduces expected sales of
options by 50,000 no longer optimal Activate: Precision Tree &
Sensitivity AnalysisOutput Separate WorksheetsSensitivity one
parameter at a timeOne line Objective function for optimal
strategy: A change in optimal decision is usually bend in line
Multiple lines Objective function for each decision. Crossing lines
change in optimal decisionTornado diagram more variables but less
infoSpider Plot more variables, more info, but limited to no more
than 3 or 4 variables too cluttered and confusing
Chelst & Canbolat Value Added Decision Making
02/28/12#Chapter 116Review: Figure 10.23: Sensitivity analysis
automation investment fixed cost of high investment$13.48 million X
axis Fixed cost input as negative value (-13) Axis would be
reversed if cost was stored as (13)Chelst & Canbolat Value
Added Decision Making
02/28/12#Chapter 11Review: Figure 10.24: Expected value of the
optimal decision for each value of fixed cost of high
investmentChelst & Canbolat Value Added Decision Making
02/28/12#Chapter 11Review: Figure 10.25: Sensitivity analysis
automation investment low take rate probabilityDecision changes
when probability approaches 0.6 (a 50% increase)Chelst &
Canbolat Value Added Decision Making
02/28/12#Chapter 11List of Variable RangesFixed investment: High
Investment: 7% of basePrice: 0 to 10% of baseVariable Cost of Low
investment: 10 % of baseVariable Cost of High investment: 0 to 5 %
of baseProbability of Low Take rate: 0.2 absoluteLow take rate
(30%): 0 to 10% absoluteVolume: 0 to 15% of baseChelst &
Canbolat Value Added Decision Making
02/28/12#Chapter 1110 Precision Tree Sensitivity Analysis
Tornado Diagram Many parameters: unlimited Uses Min & Max
values specified in the range and calculates Objective
function.Ranks the analysis in order of their range of impact on
the objective looks like tornadoDoes NOT show changed
decisions!!Chelst & Canbolat Value Added Decision Making
02/28/12#Chapter 1111Figure 11.2: Tornado diagram Boss Controls
automation investmentRange of parameter
Prob of Low Take (0.2 to 0.6)
Vehicles (850 K to 1 million)
Price ($54 to $60)
Low take rate (20% to 30%)
High Invest. ($13 m + 910K)Chelst & Canbolat Value Added
Decision Making
02/28/12#Chapter 11 Precision Tree Sensitivity Analysis Spider
Diagram practical limit of 4 parametersMore detailed than Tornado
but harder to include many variables.X axis change input (percent)Y
axis change in expected valueAggregation of many one-way
sensitivity analyses but scaled to a common percentage.Shows the
slope of the impact on the objective function and
non-linearities.Shows changes in decisions bends in line
graphChelst & Canbolat Value Added Decision Making
02/28/12#Chapter 1113List of Variable RangesFixed investment:
High Investment: 7% of basePrice: 0 to 10% of base one sided (lower
value)Range of Change in input % from a negative % to 0%Probability
of Low Take rate: 0.2 absoluteDecision does not change except at
the very highest value slight bend in line at endVolume: 0 to 15%
of base one sided (lower value)Range of Change in input % from a
negative % to 0%
Chelst & Canbolat Value Added Decision Making
02/28/12#Chapter 1114Figure 11.3: Spider plot for Boss Controls
automation investment3.544.555.566.577.588.5-60%-40%-20%0%20%40%60%
Expected ValueChange in Input (%)Spider Graph of Decision Tree
'Automation Investment'Expected Value of Entire Model Prob.
(D13)Vehicles (Mil.) (C10)Price (C4)High_Investment (D6)Decision
changes: bendChelst & Canbolat Value Added Decision Making
02/28/12#Chapter 11Manage RiskImpact of Strategies to Change
Risk ProfileShift the risk profile to the right Figure 11-4b. add
value to all possible outcomes eliminate altogether an operating
cost in a project. Cut off the downside risk Figure 11-4cMove
outcomes to some guaranteed level. Minimum purchase quantity in a
contractincrease the mean and remove the most disastrous
possibilities. Insurance cuts off the downside risk (costs money)
leftward shift in the whole risk profile but reduce the overall
expected valueChelst & Canbolat Value Added Decision Making
02/28/12#Chapter 11Figure 11.4: Impact of risk management
actions on risk profileFigure a: BaselineF Figure b: Shift to right
by adding net value (cost elimination) Figure c: Chop off left
eliminate downside risk
Chelst & Canbolat Value Added Decision Making
02/28/12#Chapter 11Change Risk Profile Manage RiskCentrally
concentrate uncertainty: Figure 11-4d Risk sharing: sell half of a
risky opportunity for a price equal to half of its expected
value
Reduce but not eliminate extremely negative outcomes: Figure
11-4e Magnitude reduction consistent with the way managers view
riskProbability reduction not as well understood
Chelst & Canbolat Value Added Decision Making
02/28/12#Chapter 11Figure 11.4: Impact of risk management
actions on risk profileFigure a: Baseline Figure d: Centralize
through risk sharing
Figure e: reduce magnitude of negative outcome
Chelst & Canbolat Value Added Decision Making
02/28/12#Chapter 11Make or Buy Decision: Non-strategic (strictly
cost)Decision Context: Manufacture a component yourself or contract
with a supplier to manufacture it. There is a design for a
component but you are not sure when it comes time to manufacture,
that the design will be feasible as is. If not, there will need to
be a quick major redesign of the component. If you manufacture it,
you expect that with the redesign it will cost 8% more than the
original estimate. The decision to make or buy must be made now
before you have time to fully check out the design. The demand for
the product is also uncertain. If you sign a contract with the
supplier for a specific piece price, if the current design turns
out to be infeasible, you know the supplier will use the design
change as an excuse to increase the price 15%. Chelst &
Canbolat Value Added Decision Making
02/28/12#Chapter 1120Make or Buy Decision: Construct Influence
Diagram (Ignore data)
Make or Buy Data: Random EventsRandom Events1. Design
Feasibility Prob.Current Design will Work 0.4Need a Major Redesign
0.6
2. Demand Volume Prob. Low 1.0 million0.3 Medium 1.25 million0.5
High 1.5 million0.2 Chelst & Canbolat Value Added Decision
Making
02/28/12#Chapter 1121Make or Buy Data: Cost DataCosts: Make
In-House Facility investment fixed Cost -$55M Variable Cost/ per
partIf current design works - $100/partIf new Design is needed -
$108/partCosts: Buy from SupplierFacility investment fixed Cost -
$0 Variable Cost/ per partIf current design works - $140/partIf new
Design is needed - $161/partChelst & Canbolat Value Added
Decision Making
02/28/12#Chapter 1122Figure 11.8: Western Co. make or buy
decisionDesign FeasibilityProbabilityMake CostsBuy
CostsWorks0.4100140Does
NOT0.6108161Premium8%15%DemandProb.30.0%0.1210.311551.250.540.0%Demand1.50.2100177.550.0%0.21.25180Make5520.0%0.08Buy01.5205TRUECurrent
Design55183.3830.0%0.18116360.0%Demand108187.350.0%0.31.2519020.0%0.121.5217Decision183.3830.0%0114040.0%Demand140171.550.0%01.2517520.0%01.5210FALSECurrent
Design0186.93530.0%0116160.0%Demand161197.22550.0%01.25201.2520.0%01.5241.5Fixed
CostsLowMediumHighHighMediumLowHighMediumLowHighMediumLowMake or
BuyMakeBuyWorksDoes NOT workDoes NOT workWorksMinimize CostE(X)
E(X) Complex calculation & NOT sum of values on branchesChelst
& Canbolat Value Added Decision Making
02/28/12#Chapter 11Structured Risk Management Step:
SummaryWithin optimal decisionIdentify random paths with large
downside riskLarge values that are negative or poor relative to the
best pathsProbability associated with this sequence is not
insignificantAssess impact of Increasing relative value of that
pathDecreasing the probability of that pathBrainstorm strategies
for making the above happenQuantify these alternativesRepeat for
2nd best decisionChelst & Canbolat Value Added Decision
Making
02/28/12#Chapter 1124Summary of Risk Management Alternatives:
Table 11.4
Chelst & Canbolat Value Added Decision Making
02/28/12#Chapter 1125Summary of Risk Management Alternatives
Table 11.4 Continued
Chelst & Canbolat Value Added Decision Making
02/28/12#Chapter 1126New topic: Information ValuePerfect
InformationImperfect InformationSample InformationExpert
InformationAccuracy of test (medical or engineering)Delay decision
until information unfolds Options
Chelst & Canbolat Value Added Decision Making
02/28/12#Chapter 11Information GatheringTraditional Approach
Gather information (surveys, tests, pilot plant, prototypes) until
time or the budget runs out. Most information is gathered to
validate already made decision.New Approach - Gather information if
the cost of gathering it is less than the gain in expected
value.Process Restructure the decision tree to determine the
expected value with the informationCounterintuitive How can you
determine the value of information before you have even gathered
the information?Chelst & Canbolat Value Added Decision
Making
02/28/12#Chapter 1128Expected Value of Perfect Information:
EVPIGoal: Determine the expected value of perfect information
regarding an Uncertainty or Risk Hire a Clairvoyant Prophet Isaiah
(Thomas)This provides an upper bound on the value of all
information including imperfect information.If the information
never changes the optimal decision then EVPI = 0.Decision Tree
Process: Move the random event in question to the front of the tree
before the first decision is to be made.Recalculate the overall
expected value.The NET Improvement is the EVPI.Chelst &
Canbolat Value Added Decision Making
02/28/12#Chapter 1129Original Decision Tree Automation
InvestmentBoss Controls Base Tree60.0%0.62310.0TRUETake Rate-136.32
(=10*0.6 + 0.8*0.4)40.0%0.413.80.8Decision6.32 (MAX{6.32,
5.86})60.0%016.58.5FALSETake Rate-85.86 (=8.5*0.6 +
1.9*0.4)40.0%09.91.9Automation InvestmentHighLow50% Take Rate30%
Take Rate50% Take Rate30% Take RateChelst & Canbolat Value
Added Decision Making
02/28/12#Chapter 11I have no idea how to remove the chapter 10
in footer30Figure 11.7: EVPI tree for Boss Controls InvestmentTake
rate event moved to before decisionTRUE0.41.91.9EVPI = 6.76 - 6.32
= 0.4440.0%How Much1.9FALSE00.80.8Take Rate6.76FALSE08.58.560.0%How
Much10TRUE0.61010Perfect InformationHighHighLowLow30% Take50%
TakeOptimal decision depends on outcome of random eventChelst &
Canbolat Value Added Decision Making
02/28/12#Chapter 11Expected Value of Perfect Information:
BossBase strategy High Investment & E(X) = $6.32MIf information
indicates 30% take rate then shift to Low Investment with profit =
$1.9MIf information indicates 50% take rate then stay with High
Investment with profit = $10MWhat is the probability the
information will indicate a 30% take rate? Answer 0.4E(X) with
perfect information = 1.9(.4) + 10 (.6) = 6.76EVPI = 6.76 6.32 =
$0.44M E(Perfect Control) = 10 6.32 = $3.86 M much more valuable to
exert control over uncertaintyChelst & Canbolat Value Added
Decision Making
02/28/12#Chapter 1132Review to contrast EVPC with EVPIMaximum
value of risk management Expected Value of Perfect Control: Not
about obtaining information but rather exerting control over
destiny Goal: Determine the value of eliminating Uncertainty or
RiskThis provides an upper bound on the value of risk management
with regard to that uncertainty.Process: Assign probability of 1 to
the best outcome of an uncertain event.Recalculate the overall
expected value.The NET Improvement in expected value is the
EVPC.Chelst & Canbolat Value Added Decision Making
02/28/12#Chapter 1133Review: Expected Value of Perfect Control:
Automation Investment Assign probability of 1 to best outcome Net
Change: $10 6.32 = $3.68 million100%1.02310.0TRUETake Rate-1310
=10*1 + 0.8*0)0%013.80.8Decision10 (MAX(10,
8.5))100%016.58.5FALSETake Rate-88.5=8.5*1.0 +
1.9*0)0%09.91.950%30%50%30%Automation InvestmentHighLowChelst &
Canbolat Value Added Decision Making
02/28/12#Chapter 112nd example: EVPI Western Make or BuyBase
strategy Make: E(X) = $183.38MUncertaintiesDesign works or not
Bound on Testing Design (Imperfect)EVPI = $2.41 M Demand Bound on
value of surveys (Imperfect)EVPI = $2.16 M Both uncertaintiesEVPI =
$3.16 M
Chelst & Canbolat Value Added Decision Making
02/28/12#Chapter 1135Make or BuyDesign
Feasibility30.0%0.12Prob.Make CostsBuy
Costs1155Works0.410014040.0%DemandDoes
NOT0.6108161100177.5Premium8%15%50.0%0.21.2518020.0%0.08DemandProb.1.520510.3TRUECurrent
Design1.250.555183.381.50.230.0%0.181163 Fixed
Costs60.0%DemandMake55108187.3Buy050.0%0.31.2519020.0%0.121.5217Decision183.3830.0%0114040.0%Demand140171.550.0%01.2517520.0%01.5210FALSECurrent
Design0186.93530.0%0116160.0%Demand161197.22550.0%01.25201.2520.0%01.5241.5Make
or BuyMakeBuyWorksDoes NOT workLowMediumHighWorksDoes NOT
workLowMediumHighLowMediumHighLowMediumHighChelst & Canbolat
Value Added Decision Making
02/28/12#Chapter
113630.0%0155155FALSEDemand177.550.0%018018020.0%020520540.0%Decision171.530.0%0.12140140TRUEDemand0171.550.0%0.217517520.0%0.08210210Current
Design180.9830.0%0.18163163TRUEDemand0187.350.0%0.319019020.0%0.1221721760.0%Decision0187.330.0%0161161FALSEDemand0197.22550.0%0201.25201.2520.0%0241.5241.5Info
DesignWorksDoes NOT
workMakeBuyLowMediumHighLowMediumHighMakeBuyLowMediumHighLowMediumHighFigure
11.9: Make-BuyEVPI: Design FeasibilityNet Improvement183.38-180.98
=$2.40MDesign uncertainty resolved before decisionChelst &
Canbolat Value Added Decision Making
02/28/12#Chapter 113740.0%0155155FALSECurrent
Design0159.860.0%016316330.0%Decision0152.640.0%0.12140140TRUECurrent
Design0152.660.0%0.18161161Demand181.2240.0%0.2180180TRUECurrent
Design018660.0%0.319019050.0%Decision018640.0%0175175FALSECurrent
Design0190.7560.0%0201201.2540.0%0.08205205TRUECurrent
Design0212.260.0%0.1221721720.0%Decision0212.240.0%0210210FALSECurrent
Design0228.960.0%0242241.5Info
DemandLowMediumHighMakeBuyMakeBuyMakeBuyWorksDoes NOT workWorksDoes
NOT workWorksDoes NOT workWorksDoes NOT workWorksDoes NOT
workWorksDoes NOT workFigure 11.10: Make-BuyEVPI on DemandNet
Improvement183.38-181.22 =$2.16MDemand uncertainty resolved before
decisionChelst & Canbolat Value Added Decision Making
02/28/12#Chapter 1138Figure 11.11: Make-Buy Decision EVPI on
Feasibility & Demand CombinedNet Improvement183.38-180.22 =
$3.16MLess than the SUM of $2.16 (Demand EVPI) + $2.40 (Feasibility
EVPI)
Next slide: Schematic TreesChelst & Canbolat Value Added
Decision Making
02/28/12#Chapter 1139Make or BuyDemandDesignMake or
BuyDemandDesignMake or BuyDemandDesignMake or
BuyDemandDesignOriginalEVPI Demand = $2.4 MEVPI Design = $2.16MEVPI
Combined: Design & Demand =$3.16MMake or Buy Schematic Trees:
EVPIChelst & Canbolat Value Added Decision Making
02/28/12#Chapter 11Imperfect Information Conditional Decision/
ProbabilitiesP (High | Positive)P(Positive)InvestDownstream values
and/or probabilities are affected by an upstream random event
Decision made AFTER resolution of random eventOptimal decision path
differs depending upon the outcome of a random eventChelst &
Canbolat Value Added Decision Making
02/28/12#Chapter 1141Expected Value of Imperfect
InformationImperfect info partial resolution of uncertaintyTest or
Sample InformationFew tests, experiments or surveys are perfect.
EVPI is an upper bound on the value of imperfect information.EVII
without well documented test reliability: Conditional probabilities
based on judgment EVII with Bayes Rule is used primarily in
environments with extensive data on the reliability of tests both
false positives and false negatives.Oil industry Seismographic
data. Test wellsMedical Applications Weather forecastsChelst &
Canbolat Value Added Decision Making
02/28/12#Chapter 1142EVII without well documented test
reliabilityConditional probabilities based on judgment Expert
understands the uncertain relationship between test data
(performance, throughput, etc.) or market surveys and subsequent
outcome.Can the expert provide a probabilistic range of outcomes
that have accompanied similar test results?Understand concept of
conditional probability experience with both possible outcomes.Need
stable process environment A priori probabilities are always in a
narrow range, for example, of 0.40 to 0.60.Not used to forecast
rare eventsProblem people have invalid intuition. Cannot factor in
a priori estimates that are updated with imperfect
information.Chelst & Canbolat Value Added Decision Making
02/28/12#Chapter 1143Boss Controls (BC) is gearing up to
manufacture an option to be made available on 1 million new cars
world-wide. Initial estimates are that the take rate for the option
could be as low as 30% or as high as 50%. Assume for simplicity
sake, these two take rates are equally likely. Experience with
focus groups indicates that for options such as the one BC is
considering, the results will either be Enthusiastic (E) or Good
(G). In the past if the focus groups were Enthusiastic, the take
rate ended up being at the HIGH end 70% time. However, if the focus
groups reactions were just good, then 80% of the time the take rate
was at the LOW end. Focus groups have an optimistic bias and tend
to be enthusiastic 80% of the time. Boss Controls: Focus Groups
& Imperfect Information based on ExperienceChelst &
Canbolat Value Added Decision Making
02/28/12#Chapter 1144EVII & Decision Trees - ExperienceAdd
an uncertain node at the front of tree to represent uncertain
outcome of focus groupInsert the probabilities that reflect the
likelihood of different responses: Here P(E) = .8 and P(G) =
.2Probability of outcomes (Take rates) are now Conditional
probabilities based on past experience (or Bayes Rule)Insert the
conditional probabilities into tree and calculate expected
value.Chelst & Canbolat Value Added Decision Making
02/28/12#Chapter 1145
Figure 11.14: Decision tree of EVII for BC automation investment
Expert estimates conditional probabilitiesEVII = 6.436 6.320 =
0.116Less than one third ofEVPI was $400,000Conditional
ProbabilitiesChelst & Canbolat Value Added Decision Making
02/28/12#Chapter 11Conditional Probabilities Consistent with
Original EstimatesA Priori Probability that Take Rate is 30% - Use
Partition Formula P(A) = P(A|B)P(B) + P(A|B)P(B) P(T=30%) = P(T=30%
| G) P(G) + P(T=30% | E) P(E)P(T=30%) = (3/4)(.4) + (1/3) (.6) = .5
original estimateP(T=50%) = P(T=50% | G) P(G) + P(T=50% | E)
P(E)P(T=50%) = (1/4)(.4) + (2/3) (.6) = .5 original estimateChelst
& Canbolat Value Added Decision Making
02/28/12#Chapter 1147Boss Control: Conditional DecisionIf focus
groups reaction is ENTHUSIASTIC then HIGH investment in
automationIf focus groups reaction is GOOD then Low investment in
automation
Chelst & Canbolat Value Added Decision Making
02/28/12#Chapter 1148INTUITION?Bayes Rule & Reliable
TestRare Disease How Rare: 1 in 1,000Probability of positive
reading for a person with the disease test is very reliable P(Pos.|
Disease) = P(P|D) = .99Probability of negative reading for a person
without the disease 4% false positives P(Neg. | No Disease) =
P(N|Dc) = .96Key Question: P(Disease | Pos) = P(D|P) = ??Let Dc = D
complement, or D , or No diseaseChelst & Canbolat Value Added
Decision Making
02/28/12#Chapter 1149Next slide provides intuitive
explanation.Bayes Rule & Reliable Test - ResultsBayes Rule
(General Formula): with Bc = B complement or NOT BDenominator uses
partitioning (all ways that A can occur) to determine P(A)Bayes
Rule (Reliable Test): (Pos = Positive test result)
Intuitive 1000 tested yields 40 false positives (4% error rate)
and 1 true positive
Chelst & Canbolat Value Added Decision Making
02/28/12#Chapter 1150Have them guess with no calculations Most
answers will be above .90Intuitive explanation1,000 take test and 1
has disease1 Disease 1 positive999 No disease but 5% false positive
(1-.96) 40 false positivesP(Disease | Pos.) = 1/(40+1) = .002
Probability MisunderstandingPeople do NOT know how to integrate
prior knowledge and data accuracy.Especially problematic withLow
probability events and highly accurate testWeakly reliable
testsChelst & Canbolat Value Added Decision Making
02/28/12#Chapter 1151Bayesian Posterior (after positive result)
ProbabilitiesInitial Probability of
Success.7.8.9.95.10.210.310.500.68.30.500.630.790.89.40.610.730.86
0.93.450.660.770.880.94.50.700.800.900.95.60.780.860.930.97.70.840.900.950.98Test
Accuracy Assume Positive = NegativeFor 0.5, .45, and even 0.40, the
final estimates are close to test accuracy.Column heading close to
cell value. For initial low probability events, test accuracy and
final probability are far apart.Chelst & Canbolat Value Added
Decision Making
02/28/12#Chapter 1152Assume, for example, all of the experience
with imperfect data involved predicting events with an initial
probability of approximately 0.3.If in the past when the results of
a survey indicated success, success followed 80% of the time, you
would not use BAYES Rule. The post survey results reflect the
actual conditional probabilities.
Bayes rule is appropriate for a standardized testing procedure
that is used over a wide range of initial (a priori) probabilities.
The testing procedures accuracy is known.EVII: Make or Buy
DecisionDecision Context: Manufacture a component yourself or
contract with a supplier to manufacture it.Design Reliability is a
key concern. Experts initially estimate that the current design
will work with probability of only 0.4.However there is a complex
test that can be used to ALMOST validate or invalidate the design.
This testing procedure is used in a wide range of
situations.Looking back at past data over a wide range of initial
success estimatesIf the design worked, how often were the test
results GOOD?Test results GOOD almost validates P(Test results Good
| Design Works) = 0.98If the design failed, how often were the test
results BAD?Test Results BAD almost Invalidates P(Test results Bad
| Design Fails) = 0.94Chelst & Canbolat Value Added Decision
Making
02/28/12#Chapter 1153EVII - Bayes Rule & Decision TreesAdd
an uncertain node at front of tree to represent uncertain outcome
of testUse Bayes rule to calculate conditional probabilities.Use
partition rule to calculate the probabilities of the test results.
(These appear in the denominator of the Bayes Rule equation.)Green
on the next page highlights the test result probabilitiesYellow on
the next page highlights the conditional probabilities. These vary
because they depend upon the results of the tests.Chelst &
Canbolat Value Added Decision Making
02/28/12#Chapter 1154EVII - Bayes Rule & Decision
TreesActivity: calculate conditional probabilitiesDataP(Design
Works) = P(W)= 0.4P(Design Fails)= P(F) = 0.6P(Test Results Good |
Design Works) = P(G|W)= 0.98P(Test Results Bad | Design
Fails)=P(B/F) =0.94Activity: Use Bayes Rule to calculateP(Design
Works | Test Results Good) = P(W|G)= ??P(G) = ??P(F/ G)=
??Precision Tree Calculate Bayesian Probabilities by hand and
Insert all of the initial probabilities upfrontInsert conditional
probabilities downstream. Chelst & Canbolat Value Added
Decision Making
02/28/12#Chapter
115591.6%Demand100177.5FALSEDesign55178.328.4%Demand108187.342.8%Decision0173.660991.6%Demand140171.5TRUEDesign0173.668.4%Demand161197.23Test
Design181.381.4%Demand100177.5TRUEDesign55187.1698.6%Demand108187.357.2%Decision0187.161.4%Demand140171.5FALSEDesign0196.8798.6%Demand161197.23EVII
Make-BuyGoodBadMakeBuyWorksFailsWorksFailsMakeBuyWorksFailsWorksFails++++++++Figure
11.12EVII for Make/Buy: Test Design183.38 181.38=2.0 EVII = $2M and
EVPI =$2.4M+ means collapsed nodeRed Demand values are expected
valuesChelst & Canbolat Value Added Decision Making
02/28/12#Chapter 1156Conditional DecisionIf test results are
GOOD then buy from supplierLess fear of 15% price increaseIf test
results are BAD then make it yourselfConcerned over suppliers
opportunity for significant price increaseChelst & Canbolat
Value Added Decision Making
02/28/12#Chapter 1157Activity: Concrete examples of IMPERFECT
InformationDescribe a context in which a decision can be made after
gathering imperfect information and there is still related
uncertainty.Product
DevelopmentExample____________________________________________Imperfect
Information ________________________Decision AFTER
____________________________Updated future uncertainties
_________________________Can you quantify accuracy?
_______________________ManufacturingExample____________________________________________Imperfect
Information ________________________Decision AFTER
____________________________Updated future uncertainties
_________________________Can you quantify accuracy?
_____________________________Chelst & Canbolat Value Added
Decision Making
02/28/12#Chapter 1158Sequential Decisions with Information in
betweenInability to predict future accurately Must make decisions
under uncertaintyA firm unable to determine level of demand in
future or predict rivals reactionsUnderstate some perceived risks
in order to obtain approvalCan management delay high cost PART of
decision until more knowledge is availablePartial Investment Gain
Information Broader scope of subsequent investmentChelst &
Canbolat Value Added Decision Making
02/28/12#Chapter 1159Investment DecisionsThree important
characteristics of Investment DecisionsPartially or completely
irreversibleUncertainty over future rewards from the
investmentAssess the probabilities of alternative outcomesLeeway
about timing of your investmentPostpone action to get more
information about the futureHow should a firm decide on an
investing on a project or a new facility?Chelst & Canbolat
Value Added Decision Making
02/28/12#Chapter 1160Real OptionsAn option represents a Right,
but not an Obligation, to do something under predefined
arrangementsBuy Option (expand or substitute) Put Option (contract
or cut back)Flexibility to adapt in response to new information
enhances the investment opportunitys value by improving its upside
potentialAn approach that offers a positive and radical
reassessment of risk and explorationThe opportunities to acquire
real assetsReal OptionsReal Options term coined by Stewart Myers
(1977)Chelst & Canbolat Value Added Decision Making
02/28/12#Chapter 1161Options AnalysisFinancial options Data
Availability Precise modelsTechnical optionsData are less
accurateOne time decisionsEstimates of values are approximates
within bands described by sensitivity analysisAnalytical niceties
that might lead to greater precision might be a waste of effortTo
decide whether to do the R&D that will lead to a real option on
a launch of a new product, managers only need to know if the value
of option is greater than the cost to acquire itChelst &
Canbolat Value Added Decision Making
02/28/12#Chapter 1162Real Options UncertaintyConventional
ApproachMinimize RiskReact to uncertaintiesWhat is the best choice
under the given circumstances?Work with predetermined set of
decisionsReal OptionsProactive towards uncertainties Prepare plans
to manage the risksIdentify parts of the system that have most
uncertainty, and try to see how these situations can be
exploitedIdentify new possible paths: change decision tree by
adding flexibilityChelst & Canbolat Value Added Decision
Making
02/28/12#Chapter 1163Table 11.10 Common Real
OptionsOptionDescriptionRelevant Application IndustriesDeferProject
that can be postponed allows learning more about project outcomes
before making a commitment. Real estate development, farming, paper
products, offshore oil leaseStageA multi-stage project whose
construction involves a series of cost outlays could be delayed or
killed in a midstream. R&D intensive industry such as
pharmaceuticals or other long development capital intensive
projectsAlter Operating ScaleA project whose operating scale can be
expanded or contracted according to market conditions. Mining,
facilities planning, fashion apparel, consumer goodsAbandonProject
can be abandoned permanently when market conditions are worsen
severely and project resources could be sold or put to other more
valuable uses.Capital intensive industries (airline, railroad), new
product introduction, financial servicesSwitchThe project permits
changing its output mix or producing the same outputs using
different inputs in response to changes in the price of inputs and
outputs.Any good sought in small batches or subject to volatile
demand (e.g., consumer electronics, toys, machine
parts)ExploreStart with a pilot or prototype project and follow-up
with a full-scale project if the pilot or prototype succeeds. High
production cost areasChelst & Canbolat Value Added Decision
Making
02/28/12#Chapter 11Options Manufacturing and Product
ExamplesDesign all vehicles to facilitate pricey add-ons for
specific market segments. (Vehicle personalization) Design a truck
such that four-wheel steering is a later option that can be
designed into it. Production system that can change easily Inputs:
Dual fuel burners (oil and gas)Production lines designed to switch
equipment so that they can produce different productsFlexible
machines rapid tool changeoverModular Design: Option to upgrade a
computer systemEngines? _______________Labor Contract pay premium
for option to reduce workforce or close plants if necessary (Put
Option)Chelst & Canbolat Value Added Decision Making
02/28/12#Chapter 1165Remaining Text Examples and FiguresEVII and
Oil DrillingTechnology ChoiceSchematic treeDecision treeRisk
ProfileContingent ContractNegotiations 2 perspectivesMercks
options
Chelst & Canbolat Value Added Decision Making
02/28/12#Chapter 11
Figure 11.13: Decision tree for oil drilling case with imperfect
informationChelst & Canbolat Value Added Decision Making
02/28/12#Chapter 11Figure 11.15: Schematic tree for technology
development example
Chelst & Canbolat Value Added Decision Making
02/28/12#Chapter 11
Figure 11.16: Decision Tree for Omega case Chelst & Canbolat
Value Added Decision Making
02/28/12#Chapter 11Figure 11.17: Cumulative risk profile for
technology development case - Omega
Chelst & Canbolat Value Added Decision Making
02/28/12#Chapter 11Figure 11.18: Contingent Contract Total sales
from perspectives of Biotech and BSG
a) BioTech perspectiveb) BSG perspectiveChelst & Canbolat
Value Added Decision Making
02/28/12#Chapter 11Figure 11.19: Mercks options and major
uncertainties in Project Gama
Chelst & Canbolat Value Added Decision Making
02/28/12#Chapter 11FactorChangeOptimalComments
Reduce Cost Increase Linked to RedesignFrom $8 to $7
From $8 to $3$730,000
$3.65 MIf redesign is needed try to contain added cost of
manufacturing.
Reduce Risk that Design will not WorkFrom 0.6 to 0.5
From 0.6 to 0.3
From 0.6 to 0.0$980,000
$4.16 M
$11.9 MModify design quickly to reduce need for major redesign
later.
New Optimal: Use Supplier
Value of Perfect Control
Manage Uncertainty of DemandNot appropriateDoes not make sense
to reduce total demand to lower total cost.
FactorChangeOptimalComments
Percentage Price increase by Supplier if design does not
workFrom 15% to 14%
From 15% to 8%$0
$3.5 MObtain commitment from supplier not to take advantage of
redesign to raise prices disproportionately.
Supplier Price Reduction if Volumes are HighUp to $8 reduction
in priceNo ImpactNegotiate major price reduction for high
volumes.
EVPI of Design Feasibility$2.4 MTest feasibility of current
design
treeCalc_1NameAutomation InvestmentPtree1 Compatibility3Output
LabelR-Value Ref.6SheetRef0Eval.
Function592147GenInfo0,1,1,0,0,Exponential,
0,0,-1,0,-1,0,.0001Creation Version1.0.?Output Value NFDef.
Link=Required Version5.0.0Output Prob NFEXT REFS0Recommended
Version5.0.0Input Value NFDef. FormLast Modified By
Version5.7.0Input Prob NFCalc MacroHighest#7Anchor CellBranch
NamebformtypevalformulapbformuladistributioncumPayoffFunctionlinkENDNODEFORMULAVALPBGenInfoIntRefsRefRefsNodeNamesCollapsed6.32Automation
Investment0002,0,0,2,2,3,0,0,00How
Much5.86Low0DEFAULT-81,0,0,2,4,5,1,0,00Take
Rate6.32High0DEFAULT-131,0,0,2,7,6,1,0,00Take Rate1.930%
Take0DEFAULTDEFAULT9.90.44,0,0,0,2,0,008.550%
Take0DEFAULTDEFAULT16.50.64,0,0,0,2,0,001050%
Take0DEFAULTDEFAULT230.64,0,0,0,3,0,000.830%
Take0DEFAULTDEFAULT13.80.44,0,0,0,3,0,00
PTModuleNamePerfect ControlPtree1 Compatibility3Output
LabelR-Value Ref.0SheetRef0Eval.
Function514964GenInfo0,2,1,0,0,Exponential,
0,0,-1,0,-1,0,.0001Creation Version1.0.?Output Value NFDef.
Link=Required Version5.0.0Output Prob NFEXT REFS0Recommended
Version5.0.0Input Value NFDef. FormLast Modified By
Version5.7.0Input Prob NFCalc MacroHighest#7Anchor CellBranch
NamebformtypevalformulapbformuladistributioncumPayoffFunctionlinkENDNODEFORMULAVALPBGenInfoIntRefsRefRefsNodeNamesCollapsed10Perfect
Control0002,0,0,2,5,2,0,0,00How
Much10High0DEFAULT-131,0,0,2,4,3,1,0,00Take Rate1050%
Take0DEFAULTDEFAULT2314,0,0,0,2,0,000.830%
Take0DEFAULTDEFAULT13.804,0,0,0,2,0,008.5Low0DEFAULT-81,0,0,2,7,6,1,0,00Take
Rate8.550% Take0DEFAULTDEFAULT16.514,0,0,0,5,0,001.930%
Take0DEFAULTDEFAULT9.904,0,0,0,5,0,00
treeCalc_2NamePerfect InformationPtree1 Compatibility3Output
LabelR-Value Ref.0SheetRef0Eval.
Function566421GenInfo0,3,1,0,0,Exponential,
0,0,-1,0,-1,0,.0001Creation Version1.0.?Output Value NFDef.
Link=Required Version5.0.0Output Prob NFEXT REFS0Recommended
Version5.0.0Input Value NFDef. FormLast Modified By
Version5.7.0Input Prob NFCalc MacroHighest#9Anchor CellBranch
NamebformtypevalformulapbformuladistributioncumPayoffFunctionlinkENDNODEFORMULAVALPBGenInfoIntRefsRefRefsNodeNamesCollapsed6.76Perfect
Information0DEFAULT001,0,0,2,3,2,0,0,00Take Rate1050%
Take000.62,0,0,2,6,5,1,0,00How Much1.930%
Take000.42,0,0,2,7,4,1,0,00How
Much0.8High0DEFAULT0.84,0,0,0,3,0,0010High0DEFAULT104,0,0,0,2,0,008.5Low0DEFAULT8.54,0,0,0,2,0,001.9Low0DEFAULT1.94,0,0,0,3,0,00
treeCalc_3NameImperfect InfoPtree1 Compatibility3Output
LabelR-Value Ref.0SheetRef0Eval.
Function489238GenInfo0,4,1,0,0,Exponential,
0,0,-1,0,-1,0,.0001Creation Version1.0.?Output Value NFDef.
Link=Required Version5.0.0Output Prob NFEXT REFS0Recommended
Version5.0.0Input Value NFDef. FormLast Modified By
Version5.7.0Input Prob NFCalc MacroHighest#15Anchor CellBranch
NamebformtypevalformulapbformuladistributioncumPayoffFunctionlinkENDNODEFORMULAVALPBGenInfoIntRefsRefRefsNodeNamesCollapsed6.436Imperfect
Info0001,0,0,2,2,3,0,0,00Focus
Group7.24Enthusiastic000.82,0,0,2,7,4,1,0,00How
Much3.22Good000.22,0,0,2,13,10,1,0,00How
Much7.24High0DEFAULT-131,0,0,2,6,5,2,0,00Take Rate1050%
Take0DEFAULTDEFAULT230.74,0,0,0,4,0,000.830%
Take0DEFAULTDEFAULT13.80.34,0,0,0,4,0,006.52Low0DEFAULT-81,0,0,2,9,8,2,0,00Take
Rate8.550% Take0DEFAULTDEFAULT16.50.74,0,0,0,7,0,001.930%
Take0DEFAULTDEFAULT9.90.34,0,0,0,7,0,002.64High0DEFAULT-131,0,0,2,12,11,3,0,00Take
Rate1050% Take0DEFAULTDEFAULT230.24,0,0,0,10,0,000.830%
Take0DEFAULTDEFAULT13.80.84,0,0,0,10,0,003.22Low0DEFAULT-81,0,0,2,15,14,3,0,00Take
Rate8.550% Take0DEFAULTDEFAULT16.50.24,0,0,0,13,0,001.930%
Take0DEFAULTDEFAULT9.90.84,0,0,0,13,0,00
treeCalc_4NameReturn on InvestmentPtree1 Compatibility3Output
LabelR-Value Ref.0SheetRef0Eval.
Function617879GenInfo1,5,1,0,0,Exponential,
0,0,-1,0,-1,0,.0001Creation Version1.0.?Output Value NFDef.
Link=Required Version5.0.0Output Prob NFEXT REFS0Recommended
Version5.0.0Input Value NFDef. Form-2Last Modified By
Version5.7.0Input Prob NFCalc MacroHighest#7Anchor CellBranch
NamebformtypevalformulapbformuladistributioncumPayoffFunctionlinkENDNODEFORMULAVALPBGenInfoIntRefsRefRefsNodeNamesCollapsed0.7325Return
on Investment0002,0,0,2,5,2,0,0,00How
Much0.4861538462High0DEFAULT-131,0,0,2,4,3,1,0,00Take
Rate0.769230769250%
Take0DEFAULTDEFAULT230.64,0,0,0,2,0,000.061538461530%
Take0DEFAULTDEFAULT13.80.44,0,0,0,2,0,000.7325Low0DEFAULT-81,0,0,2,7,6,1,0,00Take
Rate1.062550% Take0DEFAULTDEFAULT16.50.64,0,0,0,5,0,000.237530%
Take0DEFAULTDEFAULT9.90.44,0,0,0,5,0,00
treeCalc_5Automation Invesment Decision
TreesPrice60LowHigh40.0%0Investment8139.91.9Variable Cost27140Take
Rate-85.8660.0%0Vehicles (Mil.)116.58.5How MuchTake
RateProb.6.32Low30%0.440.0%0.4High50%0.613.80.80Take
Rate-136.3260.0%0.62310To see utility scores click on "Automation
Investment."In the upper right hand corner, check "use utility
function."Change the display to "expected utility" or "certainty
equivalent."0.0%0EVPC = 10 - 6.32 = 3.689.91.90Take
Rate-88.5100.0%016.58.5How Much100.0%013.80.80Take
Rate-1310100.0%1231000.41.91.9EVPI = 6.76 - 6.32 = 0.4440.0%How
Much1.9000.80.8Take Rate6.76008.58.560.0%How
Much1000.6101030.0%09.91.90Take Rate-86.5270.0%016.58.580.0%How
Much07.2430.0%0.2413.80.8EVII = 6.436 - 6.320= 0.1160Take
Rate-137.2470.0%0.562310Focus Group6.43680.0%0.169.91.90Take
Rate-83.2220.0%0.0416.58.520.0%How Much03.2280.0%013.80.80Take
Rate-132.6420.0%0231040.0%0.49.90.23750Take
Rate-80.732560.0%0.616.51.0625How
Much0.732540.0%013.80.06153846150Take
Rate-130.486153846260.0%0230.7692307692
Automation InvestmentLowHigh30% Take50% Take50% Take30%
TakePerfect ControlHigh50% Take30% TakeLow50% Take30% TakePerfect
Information50% Take30% TakeHighHighLowLowImperfect
InfoEnthusiasticGoodHigh50% Take30% TakeLow50% Take30% TakeHigh50%
Take30% TakeLow50% Take30% TakeReturn on InvestmentHigh50% Take30%
TakeLow50% Take30% Take
Sheet1
Sheet2