Академический Документы
Профессиональный Документы
Культура Документы
1
6.1 Introduction to Decision Analysis
• Maximizing expected profit is a common
criterion when probabilities can be
assessed.
• Payoff Tables
– Payoff table analysis can be applied when:
• There is a finite set of discrete decision alternatives.
• The outcome of a decision is a function of a single future event.
– In a Payoff table -
• The rows correspond to the possible decision alternatives.
• The columns correspond to the possible future events.
• Events (states of nature) are mutually exclusive and collectively
exhaustive.
• The table entries are the payoffs.
3
TOM BROWN INVESTMENT
DECISION
• Tom Brown has inherited $1000.
• He has to decide how to invest the money for one
year.
• A broker has suggested five potential investments.
– Gold
– Junk Bond
– Growth Stock
– Certificate of Deposit
– Stock Option Hedge
4
TOM BROWN
5
TOM BROWN - Solution
7
The Payoff Table
8
The Payoff Table
10
Decision Making Under Uncertainty
11
Decision Making Under Uncertainty -
The Maximin Criterion
12
Decision Making Under Uncertainty -
The Maximin Criterion
• This criterion is based on the worst-case scenario.
– It fits both a pessimistic and a conservative decision
maker’s styles.
– A pessimistic decision maker believes that the worst
possible result will always occur.
– A conservative decision maker wishes to ensure a
guaranteed minimum possible payoff.
13
TOM BROWN - The Maximin Criterion
=MAX(H4:H7)
=MIN(B4:F4)
Drag to H7
* FALSE is the range lookup argument in
the VLOOKUP function in cell B11 since the =VLOOKUP(MAX(H4:H7),H4:I7,2,FALSE
values in column H are not in ascending )
order
15
The Maximin Criterion - spreadsheet
I4
Cell I4 (hidden)=A4
Drag to I7
17
Decision Making Under Uncertainty -
The Minimax Regret Criterion
18
Decision Making Under Uncertainty -
The Minimax Regret Criterion
• The Minimax Regret Criterion
– To find an optimal decision, for each state of nature:
• Determine the best payoff over all decisions.
• Calculate the regret for each decision alternative as the
difference between its payoff value and this best payoff value.
– For each decision find the maximum regret over all
states of nature.
– Select the decision alternative that has the minimum of
these “maximum regrets.”
19
Decision Making Under Uncertainty -
The Maximax Criterion
• This criterion is based on the best possible scenario.
It fits both an optimistic and an aggressive decision maker.
21
TOM BROWN - The Maximax Criterion
Th
eo
pti
al m
The Maximax Criterion Maximum
de
Decision Large rise Small rise No change Small fall Largecisfall Payoff
Gold -100 100 200 300 0 ion 300
Bond 250 200 150 -100 -150 200
Stock 500 250 100 -200 -600 500
C/D 60 60 60 60 60 60
22
Decision Making Under Uncertainty -
The Principle of Insufficient Reason
23
TOM BROWN - Insufficient Reason
• Sum of Payoffs
– Gold 600 Dollars
– Bond 350 Dollars
– Stock 50 Dollars
– C/D 300 Dollars
• Based on this criterion the optimal decision
alternative is to invest in gold.
24
Decision Making Under Uncertainty
– Spreadsheet template
Payoff Table
RESULTS
Criteria Decision Payoff
Maximin C/D Account 60
Minimax Regret Bond 400
Maximax Stock 500
Insufficient Reason Gold 100
EV Bond 130
EVPI 141
25
Decision Making Under Risk
• The probability estimate for the occurrence of
each state of nature (if available) can be
incorporated in the search for the optimal
decision.
• For each decision calculate its expected payoff.
26
Decision Making Under Risk –
The Expected Value Criterion
• For each decision calculate the expected payoff
as follows:
27
TOM BROWN - The Expected Value
Criterion
The
opt
im a
The Expected Value Criterion l deci Expected
sion
Decision Large rise Small rise No change Small fall Large fall Value
Gold -100 100 200 300 0 100
Bond 250 200 150 -100 -150 130
Stock 500 250 100 -200 -600 125
C/D 60 60 60 60 60 60
Prior Prob. 0.2 0.3 0.3 0.1 0.1
28
When to use the expected value
approach
29
The Expected Value Criterion -
spreadsheet Cell H4 (hidden) = A4
Drag to H7
=SUMPRODUCT(B4:F4,$B$8:$F$8)
Drag to G7
=MAX(G4:G7)
=VLOOKUP(MAX(G4:G7),G4:H7,2,FALSE)
30
6.4 Expected Value of Perfect
Information
• The gain in expected return obtained from knowing
with certainty the future state of nature is called:
31
TOM BROWN - EVPI
If it were known with certainty that there will be a “Large Rise” in the market
32
TOM BROWN - EVPI
42
Posterior (revised) Probabilities
spreadsheet template
Bayesian
BayesianAnalysis
Analysis
Indicator
Indicator11 Indicator
Indicator22
States
States Prior
Prior Conditional
Conditional Joint
Joint Posterior
Posterior States
States Prior
Prior Conditional
Conditional Joint
Joint Posterior
Posterior
ofofNature Probabilities Probabilities Probabilities Probabilites
Nature Probabilities Probabilities Probabilities Probabilites ofofNature Probabilities Probabilities Probabilities Probabilites
Nature Probabilities Probabilities Probabilities Probabilites
Large
LargeRise
Rise 0.2
0.2 0.8
0.8 0.16
0.16 0.286
0.286 Large
LargeRise
Rise 0.2
0.2 0.2
0.2 0.04
0.04 0.091
0.091
Small Rise
Small Rise 0.3
0.3 0.7
0.7 0.21
0.21 0.375
0.375 Small Rise
Small Rise 0.3
0.3 0.3
0.3 0.09
0.09 0.205
0.205
No Change
No Change 0.3
0.3 0.5
0.5 0.15
0.15 0.268
0.268 No Change
No Change 0.3
0.3 0.5
0.5 0.15
0.15 0.341
0.341
Small Fall
Small Fall 0.1
0.1 0.4
0.4 0.04
0.04 0.071
0.071 Small Fall
Small Fall 0.1
0.1 0.6
0.6 0.06
0.06 0.136
0.136
Large Fall
Large Fall 0.1
0.1 00 00 0.000
0.000 Large Fall
Large Fall 0.1
0.1 11 0.1
0.1 0.227
0.227
s6
s6 00 00 0.000
0.000 s6
s6 00 00 0.000
0.000
s7
s7 00 00 0.000
0.000 s7
s7 00 00 0.000
0.000
s8
s8 00 00 0.000
0.000 s8
s8 00 00 0.000
0.000
P(Indicator 1)
P(Indicator 1) 0.56
0.56 P(Indicator 2)
P(Indicator 2) 0.44
0.44
43
Expected Value of Sample
Information
EVSI
• This is the expected gain from making decisions
based on Sample Information.
• Revise the expected return for each decision using
the posterior probabilities as follows:
44
TOM BROWN – Conditional Expected
Values
The revised probabilities payoff table
Decision Large rise Small rise No change Small fall Large fall
Gold -100 100 200 300 0
Bond 250 200 150 -100 -150
Stock 500 250 100 -200 -600
C/D 60 60 60 60 60
P(State|Positive) 0.286 0.375 0.268 0.071 0
P(State|negative) 0.091 0.205 0.341 0.136 0.227
GOLD|“Positive” forecast) =
EV(Invest in…….
=.286(-100)+.375( 100 )+.268( 200)+.071( 300)+0( 0 ) = $84
……. | “Negative” forecast) =
EV(Invest in GOLD
=.091(-100 )+.205( 100 )+.341( 200 )+.136( 300 )+.227( 0 ) = $120
45
TOM BROWN – Conditional Expected
Values
• The revised expected values for each decision:
Positive forecast Negative forecast
EV(Gold|Positive) = 84 EV(Gold|Negative) = 120
EV(Bond|Positive) = 180 EV(Bond|Negative) = 65
EV(Stock|Positive) = 250 EV(Stock|Negative) = -37
EV(C/D|Positive) = 60 EV(C/D|Negative) = 60
46
TOM BROWN – Conditional Expected
Values
• Since the forecast is unknown before it is
purchased, Tom can only calculate the expected
return from purchasing it.
• Expected return when buying the forecast = ERSI =
P(Forecast is positive)(EV(Stock|Forecast is positive)) +
P(Forecast is negative”)(EV(Gold|Forecast is negative))
= (.56)(250) + (.44)(120) = $192.5
47
Expected Value of Sampling
Information (EVSI)
• The expected gain from buying the forecast is:
EVSI = ERSI – EREV = 192.5 – 130 = $62.5
48
TOM BROWN – Solution
EVSI spreadsheet template
Payoff Table
Large Rise Small Rise No Change Small Fall Large Fall s6 s7 s8 EV(prior) EV(ind. 1) EV(ind. 2)
Gold -100 100 200 300 0 100 83.93 120.45
Bond 250 200 150 -100 -150 130 179.46 67.05
Stock 500 250 100 -200 -600 125 249.11 -32.95
C/D Account 60 60 60 60 60 60 60.00 60.00
d5
d6
d7
d8
Prior Prob. 0.2 0.3 0.3 0.1 0.1
Ind. 1 Prob. 0.286 0.375 0.268 0.071 0.000 #### ### ## 0.56
Ind 2. Prob. 0.091 0.205 0.341 0.136 0.227 #### ### ## 0.44
Ind. 3 Prob.
Ind 4 Prob.
RESULTS
Prior Ind. 1 Ind. 2 Ind. 3 Ind. 4
optimal payoff 130.00 249.11 120.45 0.00 0.00
optimal decision Bond Stock Gold
EVSI = 62.5
EVPI = 141
Efficiency= 0.44 49
6.6 Decision Trees
50
Characteristics of a decision tree
• A Decision Tree is a chronological representation of the
decision process.
• The tree is composed of nodes and branches.
Chance ( S 1) A branch emanating from a
P
node decision node corresponds to a
P(S 2)
Decision decision alternative. It includes a
s i o n1 P(S
node D ec i t 1 )
cost or benefit value.
Cos
3
)
Dec
isio P(S 1 A branch emanating from a state of
Cos n 2 P(S 2) nature (chance) node corresponds to a
t2
P ( S particular state of nature, and includes
3) the probability of this state of nature.
51
BILL GALLEN DEVELOPMENT
COMPANY
– BGD plans to do a commercial development on a
property.
– Relevant data
• Asking price for the property is 300,000 dollars.
• Construction cost is 500,000 dollars.
• Selling price is approximated at 950,000 dollars.
• Variance application costs 30,000 dollars in fees and expenses
– There is only 40% chance that the variance will be approved.
– If BGD purchases the property and the variance is denied, the property
can be sold for a net return of 260,000 dollars.
– A three month option on the property costs 20,000 dollars, which will
allow BGD to apply for the variance. 52
BILL GALLEN DEVELOPMENT
COMPANY
– A consultant can be hired for 5000 dollars.
– The consultant will provide an opinion about the
approval of the application
• P (Consultant predicts approval | approval granted) = 0.70
• P (Consultant predicts denial | approval denied) = 0.80
• BGD wishes to determine the optimal strategy
– Hire/ not hire the consultant now,
– Other decisions that follow sequentially.
53
BILL GALLEN - Solution
54
BILL GALLEN - The Decision Tree 0
ing
3
oth
Do n 0
Buy land Apply for variance
nt -300,000 -30,000
lu ta Pu
rc h
ons ase
c -20
ih re t = 0 ,00 op
tion
t s 0
no Co
Do
Le
to t us Apply for variance
Hir no co -30,000
ec
on t h ns
Co su ire ide
st
= - ant
lt a c r th
50
00 on e d
su
lta ecis
nt ion
55
BILL GALLEN - The Decision Tree
Buy land and -300000 – 30000 – 500000 + 950000 = 120,000
apply for variance Build Sell
ved -500,000 950,000
ro
App .4
0
Den
ied
-300000 – 30000 + 260000 = -70,000
0.6 Sell
260,000
Buy land Build Sell
d
p rove -300,000 -500,000 950,000
Ap 100,000
0.4
12 Den
ied
0.6
0
Buy land and -300000 – 30000 – 500000 + 950000 = 120,000
ng apply for variance Build Sell
othi
Do n 0 -500,000 950,000
Buy land Apply for variance
a nt Pu -300,000 -30,000
lt -300000 – 30000 + 260000 = -70,000
n su r ch
ase Sell
co
e 0 -20
op
hi r s t = ,00
0 tion 260,000
n ot Co
Do Buy land Build Sell
-300,000 -500,000 950,000
100,000
Apply 12
for variance
Hir -30,000
ec
on
Co su
st lta
=- nt
50 Purchase option and
00 -50,000
apply for variance 61
0.4
Apply for variance
-30,000
Let us consider the Pre nial
decision to hire a
De
dic -5000
consultant othing
0.6
N
t
Do
Buy land Apply for variance
-300,000 -30,000
Purc
hase
optio
-20,0 n
BILL GALLEN – 00
Apply for variance
59
BILL GALLEN - The Decision Tree
115,000
Build Sell
ved -500,000 950,000
r o
App
Den ?
ied
-75,000
Sell
? 260,000
60
BILL GALLEN - The Decision Tree
115,000
Build Sell
ved -500,000 950,000
r o
App
Den ?
ied
-75,000
Sell
? 260,000
61
BILL GALLEN - The Decision Tree
115,000
Build Sell
d
23 -500,000 24 950,000
25
rove
App
22 Den ?
.7
ied
-75,000
Sell
?
.3 26 27
260,000
62
The Decision Tree
Determining the Optimal Strategy
63
BILL GALLEN - The Decision Tree
Determining the Optimal Strategy
0 500 115,000 115,000 115,000
.7 )=8 00 115,000 115,000 115,000
0 0 )(0 805 115,000
15 ,0 500 Build Sell
(1 0800 23 -500,000 24 950,000
25
805 r oved
p
58,000 Ap
22 D enie
?
0.70
-75,000 -75,000 -75,000 -75,000
-22 -75,000
50 d -75,000
0 -75,000
(-7 -225 0.30 Sell
5,0 00 ? 26 27
00 260,000
)(0 -22
.3) 500
=-
22
50
0
With 58,000 as the chance node value,
we continue backward to evaluate
the previous nodes. 64
BILL GALLEN - The Decision Tree
Determining the Optimal Strategy
$115,000
Build,
$10,000 Sell
o t
d
$20,000 Do n
r ov e
.7
i r e $58,000
App
h
l Buy land; Apply
Hi p rova for variance
re $20,000 i c t s ap
P red
.4
.3
De
Pre
nie
di c t
sd $-5,000
d
eni
al
.6 Do nothing
Sell
land
$-75,000
65
BILL GALLEN - The Decision Tree
Excel add-in: Tree Plan
66
BILL GALLEN - The Decision Tree
Excel add-in: Tree Plan
67
6.7 Decision Making and Utility
• Introduction
– The expected value criterion may not be appropriate if
the decision is a one-time opportunity with substantial
risks.
– Decision makers do not always choose decisions
based on the expected value criterion.
• A lottery ticket has a negative net expected return.
• Insurance policies cost more than the present value of the
expected loss the insurance company pays to cover
insured losses.
68
The Utility Approach
69
Determining Utility Values
70
Determining Utility Values
Indifference approach for assigning utility values
71
Determining Utility Values
Indifference approach for assigning utility values
72
Determining Utility Values
Indifference approach for assigning utility values
1-p Rmax
Rij
Rmin
1-p Rmax
Rij
Rmin
d 1 150 100
d 2 -50 140
d 1 150 100
d 2 -50 140
U(100)=.7U(150)+.3U(- $150
$100
50)
1-p
= .7(1) + .3(0) = .7
p -50
76
TOM BROWN - Determining Utility Values
• Data
– The highest payoff was $500. Lowest payoff was -$600.
– The indifference probabilities provided by Tom are
Payoff -600 -200 -150 -100 0 60 100 150 200 250 300 500
Prob. 0 0.25 0.3 0.36 0.5 0.6 0.65 0.7 0.75 0.85 0.9 1
77
TOM BROWN – Optimal decision (utility)
RESULTS
Criteria Decision Value
Exp. Utility Stock 0.675
78
Three types of Decision Makers
• Risk Averse -Prefers a certain outcome to a chance
outcome having the same expected value.
79
The Utility Curve for a
Utility Risk Averse Decision Maker
U(200)
U(150)
EU(Game)
The
Theutility
utilityofofhaving
having$150
$150on
onhand…
hand…
U(100)
…is
…islarger
largerthan
thanthe
theexpected
expectedutility
utility
ofofaagame
gamewhose
whoseexpected
expectedvalue
value
isisalso
also$150.
$150.
U(200)
U(150)
EU(Game) AArisk
riskaverse
aversedecision
decisionmaker
makeravoids
avoids
the
thethrill
thrillofofaagame-of-chance,
game-of-chance,
whose
whoseexpected
expectedvalue
valueisisEV,
EV,ififhe
he
U(100) can
canhave
haveEV EVononhand
handfor
forsure.
sure.
Furthermore,
Furthermore,aariskriskaverse
aversedecision
decision
maker
makerisiswilling
willingtotopay
payaapremium…
premium…
…to
…tobuy
buyhimself
himself(herself)
(herself)out
outofofthe
the
game-of-chance.
game-of-chance.
100 CE 150 200 Payoff
0.5 0.5 81
Utility
Risk Averse Decision Maker
ker
n Ma
s io
e ci
lD
utra
k Ne
Ris
Risk Taking Decision Maker
Payoff
82