Join
Sidebar
Close
Business Analytics: Data Analysis & Decision Making 5 Chapter 16
Helpful
Unhelpful
April 29, 2025
Related documents
Business Analytics: Data Analysis & Decision Making 5 Chapter 8
Business Analytics: Data Analysis & Decision Making 5 Chapter 6
Business Analytics: Data Analysis & Decision Making 5 Chapter 16
Managerial Economics: Theory Applications And Cases 8 Chapter 6
Managerial Economics: Theory Applications And Cases 8 Chapter 11
Managerial Economics: Theory Applications And Cases 8 Chapter 5
Managerial Economics: Theory Applications And Cases 8 Chapter 12
Managerial Economics: Theory Applications And Cases 8 Chapter 17
Managerial Economics: Theory Applications And Cases 8 Appendix E
Managerial Economics: Theory Applications And Cases 8 Chapter 15
of 18
Unlock access to all the studying documents.
View Full Document
Name:
Class:
Date:
chapter 16
Answer Key
Name:
Class:
Date:
chapter 16
26.
c
27.
c
28.
a
29.
c
30.
c
31.
d
32.
d
33.
b
34.
a
35.
b
36.
c
37.
d
38.
d
39.
d
40.
e
41.
d
42.
a
43.
a
44.
e
45.
c
46.
c
47.
e
48.
b
49.
b
50.
c
51.
The
standard deviation has dropped considerab
ly,
to
about $370,000, and
the
distribution
is
not
nearly
as
chapter 16
skewed. The reduction
is
expected value
is
clearly offset
by
an
accompanying reduction
in
risk.
52.
53.
54.
It
might
not
match
our
intuition, based
on
the
answer
to
Question 71. Ho
wever,
we
must simulate
to
find
the
55.
We
are given a mean and standard devi
ation, which means market growth rate
is
normally distribu
ted.
56.
chapter 16
The change
is
clearly worth
the
cost.
59.
Yes. Both simulations 4
and
5 (Q=35,000 units
and
Q=40,000 units) have a minimum that
is
negative,
so
there
is
at
least some chance
of
getting a negative number w
ith these alternatives (although
the
smallest p
ercentile, 5
th
percentile,
60.
The
output
plot
now
shows a mean
of
approximately $900,000.
Name:
Class:
Date:
chapter 16
62.
Name:
Class:
Date:
chapter 16
63.
(A)
chapter 16
It
appears that a capacity level
of
500,000
achieves the largest mean NPV.
(B)
(C)
(D)
From
the
above @Risk output,
it
is
clear t
hat the variance (or standard deviation)
increases with larger capacity levels.
Because a risk-averse person doesn't like
large variance,
the
optimal capacity
level would probably decrease.
66.
Raising customer satisfaction
to
90% translate
s
to
an
NPV
of
$9270.
68.
chapter 16
69.
Raising customer satisfaction
to
85%,
at
a cost
of
$300
per customer, translates
to
an
NPV
of
$815
0. Since
71.
In
the first case the value drops
to
$14,500
and
in
the
second
it
increases
to
$44,000.
Name:
Class:
Date:
chapter 16
72.
The standard deviation
of
Simulation 5 (Q=40
000)
is
the
largest,
so
it
has
the most
“spread
or
dispersio
n”.
73.
75.
76.
The
output
graph below indicates about a 41% chance
of
a N
PV below zero.
Name:
Class:
Date:
chapter 16
77.
79.
We
can see from
the
plot
in
the answer
to
Question
78
that
the
NPV distribution
is
skewed
to
the
left (i.e., includes
80.
The
mean time
in
this
case
this occurs
is
in
year
25,
o
nly a slight increase from before.
81.
The mean discounted value
of
the opt
ions
is
approximately $22,600.
Name:
Class:
Date:
chapter 16
82.
There
is
a 22.34% chance
of
a negative
return
83.
The
extra
5%
customer satisfaction translates
to
an
NPV
of
$8770.
84.
In
each
period where a competitor h
as
not
yet
entered
the
market,
we
can model competitor ent
ry with a 0-1
86.
It
appears that
an
extra
5%
is
worth approximately $500
to
GM
(based
on
$9270-$8770, and $8770-$8295)
87.
Clearly,
on
average, brand A ends
up
wit
h the largest market share
by
far.
chapter 16
88.
Simulation 4 (Q=35,000 units)
has
a median (
50
th
percentile)
of
$100,063 which
is
the largest
value.
89.
We
are given a minimum, maximum and
most likely value, which indicates a triangular
distribution.
91.
(A)
Note that
the
median
is
obtained from
@RISK Detailed Statistics window.
(B)
The ending value after 100 years
is
approximately
“zero”
a large fraction
of
t
he time
--
hence
the
median
is
approximately
0
--
but
the mean
is
fairly large because
of
the
occasional times where the ending
value
is
really large.
92.
93.
chapter 16
Since there
is
a 56.4% chance
of
being
less than 0.2, then there
is
a
43.6%
chance
of
exceeding
0.2
94.
We
calculate
the
profit
in
an
output
cell
the
same way
as
before, except
now
only when
the
conservative
96.
chapter 16
Simulation 4 (Q=35,000 units) yields the largest
mean NPV
of
$93,653.
97.
Including the required investment
of
$0.9
m
illion, the resulting NPV
is
$155,000.
98.
The
output
below shows
an
expected NPV
of
$49,474.
Name:
Class:
Date:
chapter 16
100.
Although this project has a positive NPV, there
is
also substantial risk involved. A firm with
a relatively low
tolerance
for
risk may indeed decide
not
to
pursue this project.
Name:
Class:
Date:
chapter 16
101.
102.
The plant should
be
built
for
a capacity
of
3
5,000 units. This provides the highest mean
and
median NPV, although
104.
The
random variable inputs are
the
market size growth
rate
in
each year,
the
initial market size, and
the
competitor entry time
for
each
of
the
three potential competitors.
The
output
is
the
project NPV.
Name:
Class:
Date:
chapter 16