Professional Documents
Culture Documents
Investment Decisions
• Measurement of Risk
• Method of Incorporating Risk into
Capital Budgeting
Measures of Risk
• Range
• Mean absolute Deviation
• Semi-Variance
• Coefficient of Variation
example
NPV Probability
200 0.3
600 0.5
900 0.2
For your Project
1 10 0.9
2 15 0.85
3 20 0.82
4 25 0.72
σ ( NPV ) = ∑ 2t
t =1 (1 + i )
Example
Year 1 Year 2 Year 3
n
At
NP V = ∑ − I
t =1 (1 + i ) t
n σt
σ ( NPV ) = ∑ t
t =1 (1 + i )
Example
• Work out previous Year Aver. Std Dev
one CF CF
• Next, see table; given
I=10,000 1 5000 1500
2 3000 1000
3 4000 2000
4 3000 1200
Moderately correlated cash flow
• Page 212
Using probability table
• Given average NPV=96,000
• Dtd devn NPV = 60,000
• What is the probability that NPV will be
less than 0?
steps
• Calculate z = (0-96000)/60000
• Z= -1.6
• So probability=
Advanced Techniques in RA
• Sensitivity Analysis
• Scenario Analysis
• Simulation approach
• Decision Tree Analysis
Sensitivity Analysis I
• Method
– Sensitivity analysis is a risk analysis technique that tells how
much NPV will change in response to given changes in one cash
flow factor with other factors held constant.
NPV ($)
NPV when unit sales price Unit sales price
goes up by 15%
$0
NPV when unit sales price
goes down by 20%
Original NPV
value Asset beta
$0 Utility price
Input price
Operating costs
NPV breakeven Construction costs
analysis
Input price growth
Based on: Financial Management, Eugene F. Brigham and Michael C. Ehrhardt, 2008
Sensitivity Analysis for your project
• Find what happens to your project’s NPV if
the cash flow affected by:
Increase/decrease in Initial Outlay
fall / rise in Sales
hike / dip in Variable Costs
up / down in Fixed Costs
Based on: Financial Management, Eugene F. Brigham and Michael C. Ehrhardt, 2008
Scenario Analysis II
Expected NPV
Excel (mean value)
Sheet Standard Deviation
Slot in all base-, best-, of NPV
Base-, best-, and worst- and worst-case scenarios
case scenarios of each cash of each cash
flow
Probability (%)
50% FCF from base-
40% case scenario
30%
FCF from worst- 20% FCF from best-
case scenario 10% case scenario
$0 NPV ($)
Mean value
Your Project’s Scenario Analysis
• Consider worst, normal, best scenario for
your project & work out scenario analysis
• All items in P&L undergo changes
Monte Carlo Simulation
Expected NPV
(mean value)
Standard Deviation
Probability distribution of each cash Randomly picking up of NPV
flow and correlations between them scenarios
10%
8%
6%
4%
2%
$0 NPV ($)
6 step process
• Define the problem
• Identify the fixed & variable factors
• Identify the alternatives available
• Construct a mathematical model
• Run the model & get results
• Decide best of the alternatives
For Your Project
• Assign same probabilities as in the book
with respect to the three variables Cost of
the Project, Life of the project and annual
cash flows
• Your mathematical model could be NPV,
IRR
example
Annual Cash Flow ######## Project life
Value Probability ######## 3 0.05
########
1000 0.02######## 4 0.10
1500 0.03######## 5 0.30
2000 0.15######## 6 0.25
########
2500 0.15######## 7 0.15
3000 0.30######## 8 0.10
3500 0.20######## 9 0.03
####
4000 0.15 10 0.02
Initial Decisions
1. You want to set 10 runs of simulations
2. Exogenous variables identified are
Annual cash flows and Life of project
3. You want to use two-digit random
numbers only
4. Model applied is NPV
n
AnnualCash Flow
NPV = ∑ − InitialInvestment
t =1 (1 + Riskfreerate )
Steps
1. assign two-digit random numbers range
from 0-99
2. Assign probabilities
Set up correspondence between
Values of Exogenous variables & 2
digit random numbers
Annual cash flow Project Life
Value PROB CUM 2 DIGIT Value Yrs Prob Cum Prob 2 Digi
Rs. PROB RN RN
Normal
Step Rectangular
Distbn
What to use when
• Uniform Distribution : safe limits
• Trapezoidal : data in a small range around its
best estimate – large class of subjective
judgments satisfactorily
• Step Rectangular : divides data into ranges &
assign different probability to each
• Normal – where no statistical errors or random
disturbances affect the data
Simulation results
Run RN Rs CF R N Yrs NPV
1 53 3000 97 9 4277
2 66 3500 99 10 8506
3 30 2500 81 7 (829)
4 19 2000 09 4 (7660)
5 31 2500 67 6 (2112)
6 81 3500 70 7 4039
7 38 3000 75 7 1605
8 48 3000 83 7 1605
9 90 4000 33 5 2163
10 58 3000 52 6 66
Advantages Vs. Disadvantages
• Versatility • Difficult o model &
• Compels the decision specify the probability
maker to explicitly distbn
consider • Imprecision – rpugh
interdependencies apprxmn
• Enormous complexity
• Rf is used as
discounting rate
World Bank
• Powerful technique
• Efficient medium of communication
• Not replacing skilled judgment; more
judgment than traditional analysis
• Treatment of correlations between
variables can be completely misleading if
correlations are not handled properly
Decision Tree Analysis
• Decision Tree analysis lets us approximate the NPV
distribution if we can estimate the probability of certain
events within the project
• A decision tree is an expanded time line which branches
into alternate paths whenever an event can turn out
more than one way
– The place at which branches separate is called a node
– Any number of branches can emanate from a node but the
probabilities must sum to 1.0 (or 100%)
– A path represents following the tree along a branch
• Evaluating a project involves calculating NPVs along all possible
paths and developing a probability distribution
Decision tree analysis
• Identify the problem
• Delineate the decision tree
• Specify probability & monetary outcomes
• Evaluate the decision alternatives
Decision Tree Analysis—
Example 1
Q: The Wing Foot Shoe Company is considering a three-year project
to market a running shoe based on new technology. Success
depends on how well consumers accept the new idea and demand
the product. Demand can vary from great to terrible, but for
planning purposes management has collapsed that variation into
just two possibilities, good and poor. A market study indicates a
Example
60% probability that demand will be good and a 40% chance that it
will be poor.
It will cost Rs5M to bring the new shoe to market. Cash flow
estimates indicate inflows of Rs3M per year for three years at full
manufacturing capacity if demand is good, but just Rs1.5M per year
if it’s poor. Wing Foot’s cost of capital is 10%. Analyze the project
and develop a rough probability distribution for NPV.
Decision Tree Analysis—
Example
A: First, draw a decision tree diagram for the project. Then calculate
the NPV along each path.
0 1 2 3 NPV
P = .6 Rs3M Rs3M Rs3M Rs2.461M
(Rs5M) Rs-
Example
Then calculate the weighted NPV for the tree. The decision
tree explicitly
Demand NPV Probability Product calls out the fact
Good 2.641M 60% Rs1.585M that a big loss is
quite possible,
Poor -1.270M 40% Rs-.508M although the
expected NPV
Expected Rs1.077M is positive.
NPV =
Example 2
• The scientists at Spectrum have come up
with an electric moped. The firm is ready
for pilot production and test marketing .
This will cost Rs. 20 million and take 6
months. Mgt believes that there is 70%
chance that the pilot production and test
marketing will be successful. In the case
of success , they can build the plant
costing Rs 150m.
- cont’d
• The plant will generate an annual cash
inflow of Rs.30m for 20 yrs if the demand
is high or an annual cash inflow of Rs.20m
if the demand is low. High demand has a
probability 0.6; low demand 0.4; pl advise
the co using decision tree analysis
Example 3- Expected Value &
Decision Tree
Decision
Expected Value
Decision Tree