Chapter 12
Value of Information
Slide 1 of 29
Introduction
Often you pay for information you are asking for:
• Investment Advice
• Management Consultants
• Market Investigation
• Palm Reading
You need this information to make a decision in the
future:
• To invest in a particular stock or not
• To restructure the organization of your company or not
• To introduce a product or not
• Should I marry this person or not
Slide 2 of 29
Introduction
Problem at hand:
Given your decision problem, how much should
you be willing to pay for this information?
•To answer this questions you have to determine the
value (in dollars) of information.
•We will first discuss a method for determining the value
of perfect information and next for imperfect
information.
WHICH ONE DO YOU VALUE MORE?
Slide 3 of 29
Probability and Perfect Information
Definition: Clairvoyant Expert on event A
If event A is about to occur, the expert says, it will. If event
A is not to occur, the expert says, it will not. The expert is
NEVER wrong. His information is PERFECT.
A = { Dow Jones index goes up}
"A" = {Expert Says Dow Jones index goes up}
You are considering investing in a company, but before
you do you want to make sure that the Dow Jones
index will go up as this increases your chances of
making a good investment. Therefore, you decide to
consult a clairvoyant expert on the event A.
Slide 4 of 29
Probability and Perfect Information
What does it mean to be clairvoyant in probabilistic terms?
Pr( { Expert Says Dow Jones } | { Dow Jones } ) =
Pr( "A" | A ) = 1
Similarly:
Pr(" A"| A) 1 1 Pr(" A" | A) 1 Pr(" A" | A) 0
Pr(" A"| A) 0 1 Pr(" A" | A) 0 Pr(" A" | A) 1
Perhaps more importantly, what about?
Pr({ Dow Jones } | { Expert Says Dow Jones } ) =
Pr( "A" | A )
Slide 5 of 29
Probability and Perfect Information
Pr(" A" | A) Pr( A)
Pr( A | " A")
Pr(" A")
Pr(" A" | A) Pr( A)
Pr(" A" | A) Pr( A) Pr(" A" | A) Pr( A)
1 Pr( A)
1
1 Pr( A) 0 Pr(A)
Conclusion:
Pr(A|"A") equals 1 no matter what the value of Pr(A) is.
Slide 6 of 29
Probability and Perfect Information
What about the probability Pr( {Expert Says Dow Jones } )?
Pr("A") = Pr( {Expert Says Dow Jones } ) =
Pr(" A"| A) Pr( A) Pr(" A"| A) Pr( A)
1 Pr( A) 0 Pr( A) Pr( A) Pr({Dow Jones })
This is true in general: if we consult a clairvoyant expert
about an event a with possible outcomes {A1 ,⋯ , An } then:
Pr(" Ai ") Pr( Ai ), for all i 1,⋯ , n
After consulting the clairvoyant expert about event a, no
uncertainty remains about event a.
Slide 7 of 29
Expected Value of Perfect Information
STOCK MARKET EXAMPLE:
MAX. PROFIT
EMV= Up (0.5) 1500
580
High Risk Flat (0.3)
100
Stock
Down (0.2) -1000
EMV=
EMV= Up (0.5) 1000
580
540
Low Risk Flat (0.3)
200
Stock
Down (0.2) -100
Savings Account
500
Slide 8 of 29
Expected Value of Perfect Information
Consider first talking to a clairvoyant expert and then
making the investment decision:
MAX. PROFIT
EMV= High Risk Stock
1500
1500
“Up” (0.5) Low Risk Stock
1000
Savings Account 500
EMV=
1000
EMV= High Risk Stock
100
500
“Flat” (0.3) Low Risk Stock
200
Savings Account 500
EMV= High Risk Stock
-1000
500
“Down” (0.2) Low Risk Stock
-100
Savings Account
500
Slide 9 of 29
Expected Value of Perfect Information
Of course, the clairvoyant expert will charge a fee and
you would like to know how much you would be willing
to pay before using his services.
MAX. PROFIT
Do not Consult Clairvoyant EMV=
580
Consult Clairvoyant EMV=
1000 - X
- X = Consulting Fee
Conclusion:
You would be willing to consult the clairvoyant expert if:
1000 - X 580 X 1000 - 580 = 420 (=EVPI)
Slide 10 of 29
Expected Value of Perfect Information
EVPI = Expected Value of Perfect Information
Interpretation:
EVPI is the maximum amount of money you would be
willing to pay for the services of the clairvoyant expert.
If he charges more than $420 you would not consult the
expert.
A = { Dow Jones index goes up}
"A" = {Expert Says Dow Jones index goes up}
Consider an Expert about event A, who is not
clairvoyant, but is considered to be an expert. What does
it mean in for an expert not to be perfect in his
assessment about event A?
Slide 11 of 29
Expected Value of Imperfect Information
Pr( {Expert Says Dow Jones } | {Dow Jones } ) =
Pr(" A"| A) 1
Hopefully, the probability above is close to 1
(otherwise why consider him/her and Expert?)
Pr( {Expert Says Dow Jones } | {Dow Jones } ) =
Pr(" A" | A) 0
Hopefully, the probability above is close to 0
(otherwise why consider him/her and Expert?)
When an expert about an event is not clairvoyant you
need to express your trust in his assessment by for
example, checking his past performances and inter-
viewing references.
Slide 12 of 29
Expected Value of Imperfect Information
Based on your background-check of the expert you
assess your trust in terms of subjective probabilities.
True Market State
Expert UP FLAT DOWN
Prediction
"UP" Pr("UP"|UP) Pr("UP"|FLAT) Pr("UP"|DOWN)
"FLAT" Pr("FLAT"|UP) Pr("FLAT"| FLAT) Pr("FLAT"|"|DOWN)
"DOWN" Pr("DOWN"|UP) Pr("DOWN"| Pr("DOWN"|"|DOWN)
FLAT)
Total 1 1 1
Slide 13 of 29
Expected Value of Imperfect Information
Actual Assessment of the Expert:
True Market State
Expert UP FLAT DOWN
Prediction
"UP" 80% 15% 20%
"FLAT" 10% 70% 20%
"DOWN" 10% 15% 60%
Total 1 1 1
Slide 14 of 29
Expected Value of Imperfect Information
STOCK MARKET EXAMPLE:
MAX. PROFIT
EMV= Up (0.5) 1500
580
High Risk Flat (0.3)
100
Stock
Down (0.2) -1000
EMV=
EMV= Up (0.5) 1000
580
540
Low Risk Flat (0.3)
200
Stock
Down (0.2) -100
Savings Account
500
Slide 15 of 29
Expected Value of Imperfect Information
Consider first talking to an "Imperfect expert” and then
making the investment decision:
“Up” (?)
{ Original Decision Problem | “Up” }
“Flat” (?)
{Original Decision Problem | “Flat” }
“Down” (?)
{ Original Decision Problem | “Down” }
Slide 16 of 29
Expected Value of Imperfect Information
Suppose the Imperfect expert said Dow Jones will go
UP
Up (?)
1500
High Risk
Flat (?)
Stock 100
Down (?)
-1000
Up (?)
1000
“Up” (?) Low Risk Flat (?)
200
Stock Down (?)
-100
Savings Account
500
Slide 17 of 29
Expected Value of Imperfect Information
Suppose the Imperfect expert said Dow Jones will stay
FLAT
Up (?)
1500
High Risk
Flat (?)
Stock 100
Down (?)
-1000
Up (?)
1000
“Flat” (?) Low Risk Flat (?)
200
Stock Down (?)
-100
Savings Account
500
Slide 18 of 29
Expected Value of Imperfect Information
Suppose the Imperfect expert said Dow Jones will go
DOWN
Up (?)
1500
High Risk
Flat (?)
Stock 100
Down (?)
-1000
Up (?)
1000
Low Risk Flat (?)
200
Stock Down (?)
“Down” (?)
-100
Savings Account
500
Slide 19 of 29
Expected Value of Imperfect Information
Note That:
• After consulting the expert the uncertainty remains
•After consulting an imperfect expert, the original
decision problem still remains. The only difference is
that probabilities of the original decision problem have
changed to reflect the additional information, i.e the
expert's advise.
To calculate the EMV of the decision problem after
consulting the imperfect expert we have to solve for the
probabilities in the decision tree above. Calculating these
probabilities is equivalent with FLIPPING the order of the
uncertainty nodes.
Slide 20 of 29
Expected Value of Imperfect Information
Market Expert’s Expert’s Market
Behavior Forecast Forecast Behavior
“Up” (0.80) Up (?)
Up (0.5) “Flat” (0.10) “Up” (?) Flat (?)
“Down” (0.10) Down (?)
“Up” (0.15) Up (?)
Flat (0.3) “Flat” (0.70) “Flat” (?) Flat (?)
“Down” (0.15) Down (?)
“Up” (0.20) Up (?)
“Flat” (0.20) Flat (?)
Down (0.2) “Down” (0.60) “Down” (?) Down (?)
How can we solve for these probabilities?
Via Bayes theorem using a probability table
Slide 21 of 29
Expected Value of Imperfect Information
STEP 1: Construct a probability table
Pr(Up) Pr(Flat) Pr(Down)
0.500 0.300 0.200
"A" Pr("A"|Up) Pr("A"|Flat) Pr("A"|Down)
"Up" 0.800 0.150 0.200
"Flat" 0.100 0.700 0.200
"Down" 0.100 0.150 0.600
Check 1.000 1.000 1.000
Pr("A" Up) Pr("A" Flat) Pr("A" Down) Pr("A") Pr(Up|"A") Pr(Flat|"A") Pr(Down|"A") Check
0.400 0.045 0.040 0.485 0.825 0.093 0.082 1.000
0.050 0.210 0.040 0.300 0.167 0.700 0.133 1.000
0.050 0.045 0.120 0.215 0.233 0.209 0.558 1.000
Check 0.500 0.300 0.200 1.000
Slide 22 of 29
Expected Value of Imperfect Information
STEP 2: Insert the probabilities in the probability tree
Market Expert’s Expert’s Market
Behavior Forecast Forecast Behavior
“Up” (0.80) Up (0.825)
Up (0.5) “Flat” (0.10) “Up” (0.485) Flat (0.093)
“Down” (0.10) Down (0.082)
“Up” (0.15) Up (0.167)
Flat (0.3) “Flat” (0.70) “Flat” (0.300) Flat (0.700)
“Down” (0.15) Down (0.133)
“Up” (0.20) Up (0.233)
“Flat” (0.20) Flat (0.209)
Down (0.2) “Down” (0.60) “Down” (0.215) Down (0.558)
Slide 23 of 29
Expected Value of Imperfect Information
STEP 3: Calculate EMV after consulting the expert
Up (0.825)
1500
High Risk 1164 Flat (0.093)
100
Stock
Down (0.082)
-1000
1164 Up (0.825)
835 1000
“Up” (0.485) Low Risk Flat (0.093)
200
Stock Down (0.082)
-100
Savings Account
500
Slide 24 of 29
Expected Value of Imperfect Information
STEP 3B: Calculate EMV after consulting the expert
Up (0.167)
1500
High Risk 187
Flat (0.700)
Stock 100
Down (0.133)
-1000
500 Up (0.167)
293 1000
“Flat” (0.300) Low Risk Flat (0.700)
200
Stock Down (0.133)
-100
Savings Account
500
Slide 25 of 29
Expected Value of Imperfect Information
STEP 3C: Calculate EMV after consulting the expert
Up (0.233)
1500
High Risk -188
Flat (0.209)
Stock 100
Down (0.558)
-1000
Up (0.233)
500 219 1000
Low Risk Flat (0.209)
200
Stock Down (0.558)
“Down” (0.215) -100
Savings Account
500
Slide 26 of 29
Expected Value of Imperfect Information
STEP 3D: Calculate EMV after consulting the expert
“Up” (0.485) 1164
{ Original Decision Problem | “Up” }
822 500
“Flat” (0.300)
{Original Decision Problem | “Flat” }
500
{ Original Decision Problem | “Down” }
“Down” (0.215)
Slide 27 of 29
Expected Value of Imperfect Information
STEP 4: Calculate EVII for consulting the expert
MAX. PROFIT
Do not Consult Imperfect Expert EMV=
580
Consult Imperfect Expert EMV=
822 - X
- X = Consulting Fee
Conclusion:
You would be willing to consult the clairvoyant expert if:
822 - X 580 X 822 - 580 = 242 (=EVII)
EVII = Expected Value of Imperfect Information
Slide 28 of 29
Expected Value of Imperfect Information
Interpretation:
EVII is the maximum amount of money you would be
willing to pay for the services of the imperfect expert.
If he charges more than $242 you would not consult the
expert.
Note:
•EVPI EVII. Interpretation: Perfect Information is
always better than imperfect information.
•When performing sensitivity analysis EVPI calculation
of every uncertain event should be considered. When
EVPI is high for a particular uncertain event, investment
to reduce uncertainty may be warranted.
Slide 29 of 29